diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.13995801259622112,
+  "epoch": 0.5598320503848845,
   "eval_steps": 500,
-  "global_step": 1000,
+  "global_step": 4000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -7008,6 +7008,21006 @@
       "learning_rate": 4.8437575681263656e-05,
       "loss": 0.628,
       "step": 1000
+    },
+    {
+      "epoch": 0.14009797060881735,
+      "grad_norm": 0.4506290084182641,
+      "learning_rate": 4.843362954163273e-05,
+      "loss": 0.5924,
+      "step": 1001
+    },
+    {
+      "epoch": 0.14023792862141357,
+      "grad_norm": 0.4476783073678505,
+      "learning_rate": 4.842967858615316e-05,
+      "loss": 0.6682,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1403778866340098,
+      "grad_norm": 0.4357549411702438,
+      "learning_rate": 4.842572281563691e-05,
+      "loss": 0.6332,
+      "step": 1003
+    },
+    {
+      "epoch": 0.14051784464660602,
+      "grad_norm": 0.4883894311666662,
+      "learning_rate": 4.842176223089694e-05,
+      "loss": 0.642,
+      "step": 1004
+    },
+    {
+      "epoch": 0.14065780265920225,
+      "grad_norm": 0.4432669792413136,
+      "learning_rate": 4.8417796832747186e-05,
+      "loss": 0.6049,
+      "step": 1005
+    },
+    {
+      "epoch": 0.14079776067179847,
+      "grad_norm": 0.46556766571285496,
+      "learning_rate": 4.841382662200257e-05,
+      "loss": 0.646,
+      "step": 1006
+    },
+    {
+      "epoch": 0.14093771868439467,
+      "grad_norm": 0.5157114366864044,
+      "learning_rate": 4.8409851599479015e-05,
+      "loss": 0.657,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1410776766969909,
+      "grad_norm": 0.4366811539636766,
+      "learning_rate": 4.8405871765993433e-05,
+      "loss": 0.6326,
+      "step": 1008
+    },
+    {
+      "epoch": 0.14121763470958712,
+      "grad_norm": 0.43049502553637836,
+      "learning_rate": 4.8401887122363714e-05,
+      "loss": 0.616,
+      "step": 1009
+    },
+    {
+      "epoch": 0.14135759272218335,
+      "grad_norm": 0.48197017641080875,
+      "learning_rate": 4.839789766940875e-05,
+      "loss": 0.6432,
+      "step": 1010
+    },
+    {
+      "epoch": 0.14149755073477957,
+      "grad_norm": 0.45845061514071844,
+      "learning_rate": 4.839390340794841e-05,
+      "loss": 0.6262,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1416375087473758,
+      "grad_norm": 0.457307052580857,
+      "learning_rate": 4.838990433880355e-05,
+      "loss": 0.6317,
+      "step": 1012
+    },
+    {
+      "epoch": 0.14177746675997202,
+      "grad_norm": 0.44363047689877544,
+      "learning_rate": 4.838590046279602e-05,
+      "loss": 0.599,
+      "step": 1013
+    },
+    {
+      "epoch": 0.14191742477256822,
+      "grad_norm": 0.45048773354087446,
+      "learning_rate": 4.838189178074867e-05,
+      "loss": 0.626,
+      "step": 1014
+    },
+    {
+      "epoch": 0.14205738278516444,
+      "grad_norm": 0.44588282768255694,
+      "learning_rate": 4.8377878293485305e-05,
+      "loss": 0.5985,
+      "step": 1015
+    },
+    {
+      "epoch": 0.14219734079776067,
+      "grad_norm": 0.4579523725156681,
+      "learning_rate": 4.8373860001830755e-05,
+      "loss": 0.5844,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1423372988103569,
+      "grad_norm": 0.46436658695739996,
+      "learning_rate": 4.8369836906610816e-05,
+      "loss": 0.5891,
+      "step": 1017
+    },
+    {
+      "epoch": 0.14247725682295312,
+      "grad_norm": 0.46569223528579046,
+      "learning_rate": 4.836580900865227e-05,
+      "loss": 0.6164,
+      "step": 1018
+    },
+    {
+      "epoch": 0.14261721483554934,
+      "grad_norm": 0.4543042451718399,
+      "learning_rate": 4.836177630878289e-05,
+      "loss": 0.588,
+      "step": 1019
+    },
+    {
+      "epoch": 0.14275717284814557,
+      "grad_norm": 0.44767680240836294,
+      "learning_rate": 4.8357738807831446e-05,
+      "loss": 0.6551,
+      "step": 1020
+    },
+    {
+      "epoch": 0.14289713086074177,
+      "grad_norm": 0.44934339880943197,
+      "learning_rate": 4.835369650662767e-05,
+      "loss": 0.6167,
+      "step": 1021
+    },
+    {
+      "epoch": 0.143037088873338,
+      "grad_norm": 0.45729799528169507,
+      "learning_rate": 4.834964940600231e-05,
+      "loss": 0.5913,
+      "step": 1022
+    },
+    {
+      "epoch": 0.14317704688593422,
+      "grad_norm": 0.46783307870591107,
+      "learning_rate": 4.8345597506787075e-05,
+      "loss": 0.6267,
+      "step": 1023
+    },
+    {
+      "epoch": 0.14331700489853044,
+      "grad_norm": 0.4863803445979481,
+      "learning_rate": 4.8341540809814686e-05,
+      "loss": 0.62,
+      "step": 1024
+    },
+    {
+      "epoch": 0.14345696291112667,
+      "grad_norm": 0.48368481434141886,
+      "learning_rate": 4.8337479315918825e-05,
+      "loss": 0.6458,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1435969209237229,
+      "grad_norm": 0.44240686647941085,
+      "learning_rate": 4.833341302593417e-05,
+      "loss": 0.5951,
+      "step": 1026
+    },
+    {
+      "epoch": 0.14373687893631912,
+      "grad_norm": 0.46334653267689374,
+      "learning_rate": 4.832934194069639e-05,
+      "loss": 0.6634,
+      "step": 1027
+    },
+    {
+      "epoch": 0.14387683694891532,
+      "grad_norm": 0.4566293728675077,
+      "learning_rate": 4.832526606104213e-05,
+      "loss": 0.6574,
+      "step": 1028
+    },
+    {
+      "epoch": 0.14401679496151154,
+      "grad_norm": 0.4482024873976331,
+      "learning_rate": 4.832118538780902e-05,
+      "loss": 0.5956,
+      "step": 1029
+    },
+    {
+      "epoch": 0.14415675297410777,
+      "grad_norm": 0.45091056986953776,
+      "learning_rate": 4.8317099921835697e-05,
+      "loss": 0.638,
+      "step": 1030
+    },
+    {
+      "epoch": 0.144296710986704,
+      "grad_norm": 0.4546329207907057,
+      "learning_rate": 4.8313009663961746e-05,
+      "loss": 0.6386,
+      "step": 1031
+    },
+    {
+      "epoch": 0.14443666899930022,
+      "grad_norm": 0.4279917091758518,
+      "learning_rate": 4.830891461502777e-05,
+      "loss": 0.6225,
+      "step": 1032
+    },
+    {
+      "epoch": 0.14457662701189644,
+      "grad_norm": 0.4653479806373598,
+      "learning_rate": 4.8304814775875326e-05,
+      "loss": 0.6273,
+      "step": 1033
+    },
+    {
+      "epoch": 0.14471658502449264,
+      "grad_norm": 0.44000485919639887,
+      "learning_rate": 4.8300710147346996e-05,
+      "loss": 0.6049,
+      "step": 1034
+    },
+    {
+      "epoch": 0.14485654303708886,
+      "grad_norm": 0.4481538097684572,
+      "learning_rate": 4.829660073028631e-05,
+      "loss": 0.6526,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1449965010496851,
+      "grad_norm": 0.4392800745551569,
+      "learning_rate": 4.829248652553779e-05,
+      "loss": 0.6282,
+      "step": 1036
+    },
+    {
+      "epoch": 0.14513645906228131,
+      "grad_norm": 0.43714289037219956,
+      "learning_rate": 4.8288367533946964e-05,
+      "loss": 0.6162,
+      "step": 1037
+    },
+    {
+      "epoch": 0.14527641707487754,
+      "grad_norm": 0.47850026452151156,
+      "learning_rate": 4.828424375636031e-05,
+      "loss": 0.653,
+      "step": 1038
+    },
+    {
+      "epoch": 0.14541637508747376,
+      "grad_norm": 0.4387192439390933,
+      "learning_rate": 4.828011519362531e-05,
+      "loss": 0.6186,
+      "step": 1039
+    },
+    {
+      "epoch": 0.14555633310007,
+      "grad_norm": 0.438096448983694,
+      "learning_rate": 4.827598184659043e-05,
+      "loss": 0.5983,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1456962911126662,
+      "grad_norm": 0.4545201024028468,
+      "learning_rate": 4.827184371610511e-05,
+      "loss": 0.6423,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1458362491252624,
+      "grad_norm": 0.45563857615067394,
+      "learning_rate": 4.826770080301978e-05,
+      "loss": 0.5828,
+      "step": 1042
+    },
+    {
+      "epoch": 0.14597620713785864,
+      "grad_norm": 0.45222712800980513,
+      "learning_rate": 4.826355310818585e-05,
+      "loss": 0.5934,
+      "step": 1043
+    },
+    {
+      "epoch": 0.14611616515045486,
+      "grad_norm": 0.4831310051395056,
+      "learning_rate": 4.825940063245572e-05,
+      "loss": 0.6841,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1462561231630511,
+      "grad_norm": 0.46283974537294126,
+      "learning_rate": 4.8255243376682744e-05,
+      "loss": 0.6435,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1463960811756473,
+      "grad_norm": 0.4735013090808921,
+      "learning_rate": 4.825108134172131e-05,
+      "loss": 0.6406,
+      "step": 1046
+    },
+    {
+      "epoch": 0.14653603918824354,
+      "grad_norm": 0.4506208707883696,
+      "learning_rate": 4.824691452842675e-05,
+      "loss": 0.6372,
+      "step": 1047
+    },
+    {
+      "epoch": 0.14667599720083974,
+      "grad_norm": 0.4364931698009048,
+      "learning_rate": 4.824274293765536e-05,
+      "loss": 0.6026,
+      "step": 1048
+    },
+    {
+      "epoch": 0.14681595521343596,
+      "grad_norm": 0.44459989160312974,
+      "learning_rate": 4.823856657026448e-05,
+      "loss": 0.6316,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1469559132260322,
+      "grad_norm": 0.44933496657307287,
+      "learning_rate": 4.8234385427112385e-05,
+      "loss": 0.5813,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1470958712386284,
+      "grad_norm": 0.4594564123055747,
+      "learning_rate": 4.8230199509058326e-05,
+      "loss": 0.6263,
+      "step": 1051
+    },
+    {
+      "epoch": 0.14723582925122464,
+      "grad_norm": 0.4597297128739321,
+      "learning_rate": 4.822600881696256e-05,
+      "loss": 0.6166,
+      "step": 1052
+    },
+    {
+      "epoch": 0.14737578726382086,
+      "grad_norm": 0.4797057798768006,
+      "learning_rate": 4.822181335168634e-05,
+      "loss": 0.6791,
+      "step": 1053
+    },
+    {
+      "epoch": 0.1475157452764171,
+      "grad_norm": 0.463031471453022,
+      "learning_rate": 4.821761311409184e-05,
+      "loss": 0.6442,
+      "step": 1054
+    },
+    {
+      "epoch": 0.14765570328901328,
+      "grad_norm": 0.46690698533496683,
+      "learning_rate": 4.821340810504228e-05,
+      "loss": 0.6209,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1477956613016095,
+      "grad_norm": 0.47419444368207214,
+      "learning_rate": 4.8209198325401815e-05,
+      "loss": 0.6504,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14793561931420574,
+      "grad_norm": 0.4959825016958297,
+      "learning_rate": 4.8204983776035605e-05,
+      "loss": 0.615,
+      "step": 1057
+    },
+    {
+      "epoch": 0.14807557732680196,
+      "grad_norm": 0.4481720138656539,
+      "learning_rate": 4.8200764457809784e-05,
+      "loss": 0.6137,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14821553533939819,
+      "grad_norm": 0.45342457377099943,
+      "learning_rate": 4.819654037159146e-05,
+      "loss": 0.6605,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1483554933519944,
+      "grad_norm": 0.4383482872844384,
+      "learning_rate": 4.8192311518248726e-05,
+      "loss": 0.6383,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14849545136459064,
+      "grad_norm": 0.4558068303448877,
+      "learning_rate": 4.818807789865065e-05,
+      "loss": 0.6221,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14863540937718683,
+      "grad_norm": 0.44912539719700584,
+      "learning_rate": 4.818383951366729e-05,
+      "loss": 0.6413,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14877536738978306,
+      "grad_norm": 0.4355408755527561,
+      "learning_rate": 4.817959636416969e-05,
+      "loss": 0.6451,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14891532540237928,
+      "grad_norm": 0.46358234313926816,
+      "learning_rate": 4.8175348451029836e-05,
+      "loss": 0.6249,
+      "step": 1064
+    },
+    {
+      "epoch": 0.1490552834149755,
+      "grad_norm": 0.47591686796466215,
+      "learning_rate": 4.817109577512073e-05,
+      "loss": 0.6829,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14919524142757173,
+      "grad_norm": 0.4231210119222125,
+      "learning_rate": 4.8166838337316334e-05,
+      "loss": 0.5962,
+      "step": 1066
+    },
+    {
+      "epoch": 0.14933519944016796,
+      "grad_norm": 0.47903022082147834,
+      "learning_rate": 4.81625761384916e-05,
+      "loss": 0.6666,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14947515745276416,
+      "grad_norm": 0.4356774379772111,
+      "learning_rate": 4.8158309179522454e-05,
+      "loss": 0.584,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14961511546536038,
+      "grad_norm": 0.43493168432918317,
+      "learning_rate": 4.8154037461285796e-05,
+      "loss": 0.6003,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1497550734779566,
+      "grad_norm": 0.4152139582969938,
+      "learning_rate": 4.8149760984659506e-05,
+      "loss": 0.6396,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14989503149055283,
+      "grad_norm": 0.4578712882918065,
+      "learning_rate": 4.814547975052245e-05,
+      "loss": 0.6249,
+      "step": 1071
+    },
+    {
+      "epoch": 0.15003498950314906,
+      "grad_norm": 0.46527522340987304,
+      "learning_rate": 4.814119375975447e-05,
+      "loss": 0.6691,
+      "step": 1072
+    },
+    {
+      "epoch": 0.15017494751574528,
+      "grad_norm": 0.4551938793084905,
+      "learning_rate": 4.813690301323636e-05,
+      "loss": 0.6347,
+      "step": 1073
+    },
+    {
+      "epoch": 0.1503149055283415,
+      "grad_norm": 0.472894226480286,
+      "learning_rate": 4.813260751184992e-05,
+      "loss": 0.6558,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1504548635409377,
+      "grad_norm": 0.45206474588868106,
+      "learning_rate": 4.812830725647793e-05,
+      "loss": 0.5915,
+      "step": 1075
+    },
+    {
+      "epoch": 0.15059482155353393,
+      "grad_norm": 0.46083474871884716,
+      "learning_rate": 4.8124002248004126e-05,
+      "loss": 0.6336,
+      "step": 1076
+    },
+    {
+      "epoch": 0.15073477956613016,
+      "grad_norm": 0.44095545335189645,
+      "learning_rate": 4.811969248731323e-05,
+      "loss": 0.6307,
+      "step": 1077
+    },
+    {
+      "epoch": 0.15087473757872638,
+      "grad_norm": 0.46628362437250337,
+      "learning_rate": 4.8115377975290955e-05,
+      "loss": 0.6267,
+      "step": 1078
+    },
+    {
+      "epoch": 0.1510146955913226,
+      "grad_norm": 0.4570685734330465,
+      "learning_rate": 4.811105871282395e-05,
+      "loss": 0.6891,
+      "step": 1079
+    },
+    {
+      "epoch": 0.15115465360391883,
+      "grad_norm": 0.4494515219252805,
+      "learning_rate": 4.81067347007999e-05,
+      "loss": 0.6147,
+      "step": 1080
+    },
+    {
+      "epoch": 0.15129461161651506,
+      "grad_norm": 0.44047992997082713,
+      "learning_rate": 4.810240594010742e-05,
+      "loss": 0.6443,
+      "step": 1081
+    },
+    {
+      "epoch": 0.15143456962911125,
+      "grad_norm": 0.4423964257933292,
+      "learning_rate": 4.80980724316361e-05,
+      "loss": 0.6316,
+      "step": 1082
+    },
+    {
+      "epoch": 0.15157452764170748,
+      "grad_norm": 0.43454276433469047,
+      "learning_rate": 4.809373417627654e-05,
+      "loss": 0.6244,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1517144856543037,
+      "grad_norm": 0.4576456453029119,
+      "learning_rate": 4.8089391174920275e-05,
+      "loss": 0.6158,
+      "step": 1084
+    },
+    {
+      "epoch": 0.15185444366689993,
+      "grad_norm": 0.4376862180207936,
+      "learning_rate": 4.808504342845986e-05,
+      "loss": 0.6287,
+      "step": 1085
+    },
+    {
+      "epoch": 0.15199440167949615,
+      "grad_norm": 0.4272520839128296,
+      "learning_rate": 4.808069093778879e-05,
+      "loss": 0.5709,
+      "step": 1086
+    },
+    {
+      "epoch": 0.15213435969209238,
+      "grad_norm": 0.4562369557196528,
+      "learning_rate": 4.807633370380155e-05,
+      "loss": 0.651,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1522743177046886,
+      "grad_norm": 0.4304596270726819,
+      "learning_rate": 4.807197172739357e-05,
+      "loss": 0.6338,
+      "step": 1088
+    },
+    {
+      "epoch": 0.1524142757172848,
+      "grad_norm": 0.4427035165037056,
+      "learning_rate": 4.806760500946132e-05,
+      "loss": 0.6196,
+      "step": 1089
+    },
+    {
+      "epoch": 0.15255423372988103,
+      "grad_norm": 0.4603650971833003,
+      "learning_rate": 4.806323355090218e-05,
+      "loss": 0.6526,
+      "step": 1090
+    },
+    {
+      "epoch": 0.15269419174247725,
+      "grad_norm": 0.4348487042293239,
+      "learning_rate": 4.805885735261454e-05,
+      "loss": 0.6375,
+      "step": 1091
+    },
+    {
+      "epoch": 0.15283414975507348,
+      "grad_norm": 0.4515390501287558,
+      "learning_rate": 4.805447641549774e-05,
+      "loss": 0.5708,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1529741077676697,
+      "grad_norm": 0.4715753314020402,
+      "learning_rate": 4.805009074045213e-05,
+      "loss": 0.6477,
+      "step": 1093
+    },
+    {
+      "epoch": 0.15311406578026593,
+      "grad_norm": 0.4434658029205517,
+      "learning_rate": 4.8045700328378986e-05,
+      "loss": 0.6013,
+      "step": 1094
+    },
+    {
+      "epoch": 0.15325402379286215,
+      "grad_norm": 0.44663741205752033,
+      "learning_rate": 4.804130518018058e-05,
+      "loss": 0.6081,
+      "step": 1095
+    },
+    {
+      "epoch": 0.15339398180545835,
+      "grad_norm": 0.41670655291144576,
+      "learning_rate": 4.803690529676019e-05,
+      "loss": 0.6378,
+      "step": 1096
+    },
+    {
+      "epoch": 0.15353393981805458,
+      "grad_norm": 0.49686928110177025,
+      "learning_rate": 4.803250067902202e-05,
+      "loss": 0.6445,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1536738978306508,
+      "grad_norm": 0.4685195748194762,
+      "learning_rate": 4.802809132787125e-05,
+      "loss": 0.6066,
+      "step": 1098
+    },
+    {
+      "epoch": 0.15381385584324703,
+      "grad_norm": 0.46286655961754153,
+      "learning_rate": 4.802367724421407e-05,
+      "loss": 0.6144,
+      "step": 1099
+    },
+    {
+      "epoch": 0.15395381385584325,
+      "grad_norm": 0.43665548007096733,
+      "learning_rate": 4.8019258428957605e-05,
+      "loss": 0.6404,
+      "step": 1100
+    },
+    {
+      "epoch": 0.15409377186843948,
+      "grad_norm": 0.4700705212693585,
+      "learning_rate": 4.8014834883009966e-05,
+      "loss": 0.6297,
+      "step": 1101
+    },
+    {
+      "epoch": 0.1542337298810357,
+      "grad_norm": 0.46140020925769776,
+      "learning_rate": 4.8010406607280244e-05,
+      "loss": 0.63,
+      "step": 1102
+    },
+    {
+      "epoch": 0.1543736878936319,
+      "grad_norm": 0.432596752729328,
+      "learning_rate": 4.800597360267849e-05,
+      "loss": 0.6014,
+      "step": 1103
+    },
+    {
+      "epoch": 0.15451364590622813,
+      "grad_norm": 0.4087158619129061,
+      "learning_rate": 4.800153587011573e-05,
+      "loss": 0.6101,
+      "step": 1104
+    },
+    {
+      "epoch": 0.15465360391882435,
+      "grad_norm": 0.4306421416237517,
+      "learning_rate": 4.799709341050396e-05,
+      "loss": 0.6088,
+      "step": 1105
+    },
+    {
+      "epoch": 0.15479356193142058,
+      "grad_norm": 0.439174570947343,
+      "learning_rate": 4.799264622475616e-05,
+      "loss": 0.6271,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1549335199440168,
+      "grad_norm": 0.45230787848804416,
+      "learning_rate": 4.7988194313786275e-05,
+      "loss": 0.6314,
+      "step": 1107
+    },
+    {
+      "epoch": 0.15507347795661303,
+      "grad_norm": 0.4452846746911816,
+      "learning_rate": 4.79837376785092e-05,
+      "loss": 0.6189,
+      "step": 1108
+    },
+    {
+      "epoch": 0.15521343596920922,
+      "grad_norm": 0.43395577576303523,
+      "learning_rate": 4.7979276319840824e-05,
+      "loss": 0.5867,
+      "step": 1109
+    },
+    {
+      "epoch": 0.15535339398180545,
+      "grad_norm": 0.4198939865361689,
+      "learning_rate": 4.797481023869801e-05,
+      "loss": 0.6033,
+      "step": 1110
+    },
+    {
+      "epoch": 0.15549335199440167,
+      "grad_norm": 0.4115113980248625,
+      "learning_rate": 4.797033943599859e-05,
+      "loss": 0.5792,
+      "step": 1111
+    },
+    {
+      "epoch": 0.1556333100069979,
+      "grad_norm": 0.42305720865767826,
+      "learning_rate": 4.796586391266134e-05,
+      "loss": 0.5721,
+      "step": 1112
+    },
+    {
+      "epoch": 0.15577326801959412,
+      "grad_norm": 0.4594490270870182,
+      "learning_rate": 4.796138366960603e-05,
+      "loss": 0.5923,
+      "step": 1113
+    },
+    {
+      "epoch": 0.15591322603219035,
+      "grad_norm": 0.4660997680749236,
+      "learning_rate": 4.7956898707753405e-05,
+      "loss": 0.6168,
+      "step": 1114
+    },
+    {
+      "epoch": 0.15605318404478657,
+      "grad_norm": 0.4300722165746083,
+      "learning_rate": 4.795240902802517e-05,
+      "loss": 0.5975,
+      "step": 1115
+    },
+    {
+      "epoch": 0.15619314205738277,
+      "grad_norm": 0.4593249810088033,
+      "learning_rate": 4.794791463134399e-05,
+      "loss": 0.598,
+      "step": 1116
+    },
+    {
+      "epoch": 0.156333100069979,
+      "grad_norm": 0.4615991027920945,
+      "learning_rate": 4.79434155186335e-05,
+      "loss": 0.6267,
+      "step": 1117
+    },
+    {
+      "epoch": 0.15647305808257522,
+      "grad_norm": 0.441601529655813,
+      "learning_rate": 4.7938911690818347e-05,
+      "loss": 0.6607,
+      "step": 1118
+    },
+    {
+      "epoch": 0.15661301609517145,
+      "grad_norm": 0.45355246625005374,
+      "learning_rate": 4.793440314882408e-05,
+      "loss": 0.6071,
+      "step": 1119
+    },
+    {
+      "epoch": 0.15675297410776767,
+      "grad_norm": 0.44036948902223627,
+      "learning_rate": 4.792988989357727e-05,
+      "loss": 0.6154,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1568929321203639,
+      "grad_norm": 0.4734397090177979,
+      "learning_rate": 4.7925371926005435e-05,
+      "loss": 0.6102,
+      "step": 1121
+    },
+    {
+      "epoch": 0.15703289013296012,
+      "grad_norm": 0.43617965348321686,
+      "learning_rate": 4.792084924703705e-05,
+      "loss": 0.6104,
+      "step": 1122
+    },
+    {
+      "epoch": 0.15717284814555632,
+      "grad_norm": 0.4478534547540619,
+      "learning_rate": 4.791632185760158e-05,
+      "loss": 0.6235,
+      "step": 1123
+    },
+    {
+      "epoch": 0.15731280615815255,
+      "grad_norm": 0.44766677539513156,
+      "learning_rate": 4.791178975862945e-05,
+      "loss": 0.6439,
+      "step": 1124
+    },
+    {
+      "epoch": 0.15745276417074877,
+      "grad_norm": 0.44554264209996225,
+      "learning_rate": 4.790725295105205e-05,
+      "loss": 0.5959,
+      "step": 1125
+    },
+    {
+      "epoch": 0.157592722183345,
+      "grad_norm": 0.4296700348089923,
+      "learning_rate": 4.790271143580174e-05,
+      "loss": 0.6539,
+      "step": 1126
+    },
+    {
+      "epoch": 0.15773268019594122,
+      "grad_norm": 0.4412703085902547,
+      "learning_rate": 4.789816521381185e-05,
+      "loss": 0.6375,
+      "step": 1127
+    },
+    {
+      "epoch": 0.15787263820853745,
+      "grad_norm": 0.4373118306763795,
+      "learning_rate": 4.7893614286016684e-05,
+      "loss": 0.6112,
+      "step": 1128
+    },
+    {
+      "epoch": 0.15801259622113367,
+      "grad_norm": 0.44070716127461335,
+      "learning_rate": 4.7889058653351485e-05,
+      "loss": 0.6195,
+      "step": 1129
+    },
+    {
+      "epoch": 0.15815255423372987,
+      "grad_norm": 0.4052645175970106,
+      "learning_rate": 4.788449831675248e-05,
+      "loss": 0.6118,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1582925122463261,
+      "grad_norm": 0.4205527476901149,
+      "learning_rate": 4.7879933277156884e-05,
+      "loss": 0.6098,
+      "step": 1131
+    },
+    {
+      "epoch": 0.15843247025892232,
+      "grad_norm": 0.4232443251461348,
+      "learning_rate": 4.787536353550285e-05,
+      "loss": 0.5962,
+      "step": 1132
+    },
+    {
+      "epoch": 0.15857242827151855,
+      "grad_norm": 0.4555075573119837,
+      "learning_rate": 4.787078909272951e-05,
+      "loss": 0.6099,
+      "step": 1133
+    },
+    {
+      "epoch": 0.15871238628411477,
+      "grad_norm": 0.45597750842746004,
+      "learning_rate": 4.786620994977695e-05,
+      "loss": 0.6403,
+      "step": 1134
+    },
+    {
+      "epoch": 0.158852344296711,
+      "grad_norm": 0.4523579064101052,
+      "learning_rate": 4.7861626107586236e-05,
+      "loss": 0.6276,
+      "step": 1135
+    },
+    {
+      "epoch": 0.15899230230930722,
+      "grad_norm": 0.4496756804618233,
+      "learning_rate": 4.785703756709939e-05,
+      "loss": 0.625,
+      "step": 1136
+    },
+    {
+      "epoch": 0.15913226032190342,
+      "grad_norm": 0.44479315166484706,
+      "learning_rate": 4.78524443292594e-05,
+      "loss": 0.6219,
+      "step": 1137
+    },
+    {
+      "epoch": 0.15927221833449964,
+      "grad_norm": 0.423002589052915,
+      "learning_rate": 4.784784639501024e-05,
+      "loss": 0.6142,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15941217634709587,
+      "grad_norm": 0.45783730451738547,
+      "learning_rate": 4.7843243765296816e-05,
+      "loss": 0.6327,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1595521343596921,
+      "grad_norm": 0.41727555581040476,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 0.6314,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15969209237228832,
+      "grad_norm": 0.4433704880299786,
+      "learning_rate": 4.7834024423261715e-05,
+      "loss": 0.6297,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15983205038488454,
+      "grad_norm": 0.44605075439465414,
+      "learning_rate": 4.78294077128347e-05,
+      "loss": 0.6052,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15997200839748077,
+      "grad_norm": 0.4252564366376628,
+      "learning_rate": 4.7824786310732754e-05,
+      "loss": 0.5952,
+      "step": 1143
+    },
+    {
+      "epoch": 0.16011196641007697,
+      "grad_norm": 0.4487603412437455,
+      "learning_rate": 4.782016021790564e-05,
+      "loss": 0.5993,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1602519244226732,
+      "grad_norm": 0.4335684662398312,
+      "learning_rate": 4.781552943530405e-05,
+      "loss": 0.6119,
+      "step": 1145
+    },
+    {
+      "epoch": 0.16039188243526942,
+      "grad_norm": 0.48534201807046695,
+      "learning_rate": 4.781089396387968e-05,
+      "loss": 0.6566,
+      "step": 1146
+    },
+    {
+      "epoch": 0.16053184044786564,
+      "grad_norm": 0.4657809557886318,
+      "learning_rate": 4.780625380458513e-05,
+      "loss": 0.6188,
+      "step": 1147
+    },
+    {
+      "epoch": 0.16067179846046187,
+      "grad_norm": 0.463002737360727,
+      "learning_rate": 4.7801608958374034e-05,
+      "loss": 0.6281,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1608117564730581,
+      "grad_norm": 0.4327965386696676,
+      "learning_rate": 4.779695942620094e-05,
+      "loss": 0.5676,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1609517144856543,
+      "grad_norm": 0.43456914658498524,
+      "learning_rate": 4.779230520902138e-05,
+      "loss": 0.6018,
+      "step": 1150
+    },
+    {
+      "epoch": 0.16109167249825052,
+      "grad_norm": 0.4340562614707508,
+      "learning_rate": 4.778764630779183e-05,
+      "loss": 0.5634,
+      "step": 1151
+    },
+    {
+      "epoch": 0.16123163051084674,
+      "grad_norm": 0.5384914426514081,
+      "learning_rate": 4.778298272346976e-05,
+      "loss": 0.6159,
+      "step": 1152
+    },
+    {
+      "epoch": 0.16137158852344297,
+      "grad_norm": 0.45988369030455006,
+      "learning_rate": 4.7778314457013565e-05,
+      "loss": 0.6306,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1615115465360392,
+      "grad_norm": 0.4650532211327251,
+      "learning_rate": 4.777364150938263e-05,
+      "loss": 0.6192,
+      "step": 1154
+    },
+    {
+      "epoch": 0.16165150454863542,
+      "grad_norm": 0.4253518192932048,
+      "learning_rate": 4.77689638815373e-05,
+      "loss": 0.6373,
+      "step": 1155
+    },
+    {
+      "epoch": 0.16179146256123164,
+      "grad_norm": 0.45004968895984576,
+      "learning_rate": 4.776428157443886e-05,
+      "loss": 0.6173,
+      "step": 1156
+    },
+    {
+      "epoch": 0.16193142057382784,
+      "grad_norm": 0.4555578881747446,
+      "learning_rate": 4.775959458904958e-05,
+      "loss": 0.6011,
+      "step": 1157
+    },
+    {
+      "epoch": 0.16207137858642406,
+      "grad_norm": 0.4034724262013599,
+      "learning_rate": 4.775490292633269e-05,
+      "loss": 0.5945,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1622113365990203,
+      "grad_norm": 0.44601195862385407,
+      "learning_rate": 4.7750206587252366e-05,
+      "loss": 0.5964,
+      "step": 1159
+    },
+    {
+      "epoch": 0.16235129461161651,
+      "grad_norm": 0.45007693700753115,
+      "learning_rate": 4.7745505572773754e-05,
+      "loss": 0.6187,
+      "step": 1160
+    },
+    {
+      "epoch": 0.16249125262421274,
+      "grad_norm": 0.45228216299402807,
+      "learning_rate": 4.774079988386296e-05,
+      "loss": 0.6235,
+      "step": 1161
+    },
+    {
+      "epoch": 0.16263121063680896,
+      "grad_norm": 0.43817158113391474,
+      "learning_rate": 4.773608952148706e-05,
+      "loss": 0.6181,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1627711686494052,
+      "grad_norm": 0.4218728284381608,
+      "learning_rate": 4.7731374486614056e-05,
+      "loss": 0.6472,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1629111266620014,
+      "grad_norm": 0.4385041088709967,
+      "learning_rate": 4.772665478021296e-05,
+      "loss": 0.6263,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1630510846745976,
+      "grad_norm": 0.4575475392945415,
+      "learning_rate": 4.7721930403253714e-05,
+      "loss": 0.6136,
+      "step": 1165
+    },
+    {
+      "epoch": 0.16319104268719384,
+      "grad_norm": 0.4457426524641399,
+      "learning_rate": 4.771720135670722e-05,
+      "loss": 0.623,
+      "step": 1166
+    },
+    {
+      "epoch": 0.16333100069979006,
+      "grad_norm": 0.4257770146896383,
+      "learning_rate": 4.7712467641545354e-05,
+      "loss": 0.6371,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1634709587123863,
+      "grad_norm": 0.5052851757471238,
+      "learning_rate": 4.770772925874093e-05,
+      "loss": 0.6249,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1636109167249825,
+      "grad_norm": 0.47522529944132186,
+      "learning_rate": 4.7702986209267745e-05,
+      "loss": 0.6378,
+      "step": 1169
+    },
+    {
+      "epoch": 0.16375087473757874,
+      "grad_norm": 0.42207340820026673,
+      "learning_rate": 4.7698238494100536e-05,
+      "loss": 0.6271,
+      "step": 1170
+    },
+    {
+      "epoch": 0.16389083275017494,
+      "grad_norm": 0.42129003632456846,
+      "learning_rate": 4.7693486114215015e-05,
+      "loss": 0.6135,
+      "step": 1171
+    },
+    {
+      "epoch": 0.16403079076277116,
+      "grad_norm": 0.4929479783722899,
+      "learning_rate": 4.768872907058783e-05,
+      "loss": 0.6043,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1641707487753674,
+      "grad_norm": 0.4810208017615139,
+      "learning_rate": 4.7683967364196624e-05,
+      "loss": 0.65,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1643107067879636,
+      "grad_norm": 0.47080599343079316,
+      "learning_rate": 4.767920099601996e-05,
+      "loss": 0.6215,
+      "step": 1174
+    },
+    {
+      "epoch": 0.16445066480055984,
+      "grad_norm": 0.4354882117508445,
+      "learning_rate": 4.767442996703737e-05,
+      "loss": 0.6066,
+      "step": 1175
+    },
+    {
+      "epoch": 0.16459062281315606,
+      "grad_norm": 0.4284983370361917,
+      "learning_rate": 4.766965427822936e-05,
+      "loss": 0.5834,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1647305808257523,
+      "grad_norm": 0.45013293062633963,
+      "learning_rate": 4.7664873930577383e-05,
+      "loss": 0.6747,
+      "step": 1177
+    },
+    {
+      "epoch": 0.16487053883834848,
+      "grad_norm": 0.4616970172832583,
+      "learning_rate": 4.766008892506384e-05,
+      "loss": 0.6331,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1650104968509447,
+      "grad_norm": 0.43031916863208525,
+      "learning_rate": 4.765529926267211e-05,
+      "loss": 0.6218,
+      "step": 1179
+    },
+    {
+      "epoch": 0.16515045486354094,
+      "grad_norm": 0.4663955970462797,
+      "learning_rate": 4.765050494438651e-05,
+      "loss": 0.6508,
+      "step": 1180
+    },
+    {
+      "epoch": 0.16529041287613716,
+      "grad_norm": 0.439471188683419,
+      "learning_rate": 4.7645705971192315e-05,
+      "loss": 0.6199,
+      "step": 1181
+    },
+    {
+      "epoch": 0.16543037088873339,
+      "grad_norm": 0.45371569435156506,
+      "learning_rate": 4.764090234407577e-05,
+      "loss": 0.6266,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1655703289013296,
+      "grad_norm": 0.43706679884940447,
+      "learning_rate": 4.7636094064024076e-05,
+      "loss": 0.6358,
+      "step": 1183
+    },
+    {
+      "epoch": 0.16571028691392584,
+      "grad_norm": 0.4383648085023705,
+      "learning_rate": 4.763128113202537e-05,
+      "loss": 0.6363,
+      "step": 1184
+    },
+    {
+      "epoch": 0.16585024492652203,
+      "grad_norm": 0.4332543518196199,
+      "learning_rate": 4.762646354906877e-05,
+      "loss": 0.6284,
+      "step": 1185
+    },
+    {
+      "epoch": 0.16599020293911826,
+      "grad_norm": 0.42760152759621617,
+      "learning_rate": 4.7621641316144325e-05,
+      "loss": 0.6212,
+      "step": 1186
+    },
+    {
+      "epoch": 0.16613016095171448,
+      "grad_norm": 0.4530135732031799,
+      "learning_rate": 4.761681443424306e-05,
+      "loss": 0.664,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1662701189643107,
+      "grad_norm": 0.4459844470321281,
+      "learning_rate": 4.7611982904356954e-05,
+      "loss": 0.6384,
+      "step": 1188
+    },
+    {
+      "epoch": 0.16641007697690693,
+      "grad_norm": 0.41924312442399453,
+      "learning_rate": 4.7607146727478935e-05,
+      "loss": 0.5901,
+      "step": 1189
+    },
+    {
+      "epoch": 0.16655003498950316,
+      "grad_norm": 0.43860632090683305,
+      "learning_rate": 4.760230590460287e-05,
+      "loss": 0.6021,
+      "step": 1190
+    },
+    {
+      "epoch": 0.16668999300209936,
+      "grad_norm": 0.45006765458458414,
+      "learning_rate": 4.759746043672362e-05,
+      "loss": 0.6365,
+      "step": 1191
+    },
+    {
+      "epoch": 0.16682995101469558,
+      "grad_norm": 0.4440700448858348,
+      "learning_rate": 4.7592610324836955e-05,
+      "loss": 0.6349,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1669699090272918,
+      "grad_norm": 0.42964762641344484,
+      "learning_rate": 4.758775556993964e-05,
+      "loss": 0.573,
+      "step": 1193
+    },
+    {
+      "epoch": 0.16710986703988803,
+      "grad_norm": 0.4348088254948326,
+      "learning_rate": 4.758289617302937e-05,
+      "loss": 0.6094,
+      "step": 1194
+    },
+    {
+      "epoch": 0.16724982505248426,
+      "grad_norm": 0.42297107816490986,
+      "learning_rate": 4.7578032135104796e-05,
+      "loss": 0.6102,
+      "step": 1195
+    },
+    {
+      "epoch": 0.16738978306508048,
+      "grad_norm": 0.4494883284181831,
+      "learning_rate": 4.7573163457165534e-05,
+      "loss": 0.6071,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1675297410776767,
+      "grad_norm": 0.4339048883611629,
+      "learning_rate": 4.7568290140212145e-05,
+      "loss": 0.6213,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1676696990902729,
+      "grad_norm": 0.48344849855571825,
+      "learning_rate": 4.7563412185246145e-05,
+      "loss": 0.6433,
+      "step": 1198
+    },
+    {
+      "epoch": 0.16780965710286913,
+      "grad_norm": 0.4488129613842796,
+      "learning_rate": 4.7558529593269996e-05,
+      "loss": 0.654,
+      "step": 1199
+    },
+    {
+      "epoch": 0.16794961511546536,
+      "grad_norm": 0.43481250974353425,
+      "learning_rate": 4.755364236528713e-05,
+      "loss": 0.6316,
+      "step": 1200
+    },
+    {
+      "epoch": 0.16808957312806158,
+      "grad_norm": 0.41793689356043273,
+      "learning_rate": 4.754875050230192e-05,
+      "loss": 0.5783,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1682295311406578,
+      "grad_norm": 0.43753416598422584,
+      "learning_rate": 4.754385400531969e-05,
+      "loss": 0.5949,
+      "step": 1202
+    },
+    {
+      "epoch": 0.16836948915325403,
+      "grad_norm": 0.44668014693071806,
+      "learning_rate": 4.753895287534673e-05,
+      "loss": 0.5863,
+      "step": 1203
+    },
+    {
+      "epoch": 0.16850944716585026,
+      "grad_norm": 0.42805637156559095,
+      "learning_rate": 4.753404711339026e-05,
+      "loss": 0.6357,
+      "step": 1204
+    },
+    {
+      "epoch": 0.16864940517844645,
+      "grad_norm": 0.41847675435581816,
+      "learning_rate": 4.752913672045846e-05,
+      "loss": 0.6138,
+      "step": 1205
+    },
+    {
+      "epoch": 0.16878936319104268,
+      "grad_norm": 0.4553821804652506,
+      "learning_rate": 4.752422169756048e-05,
+      "loss": 0.6179,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1689293212036389,
+      "grad_norm": 0.45084883922329716,
+      "learning_rate": 4.75193020457064e-05,
+      "loss": 0.6321,
+      "step": 1207
+    },
+    {
+      "epoch": 0.16906927921623513,
+      "grad_norm": 0.42851103889071873,
+      "learning_rate": 4.751437776590726e-05,
+      "loss": 0.639,
+      "step": 1208
+    },
+    {
+      "epoch": 0.16920923722883136,
+      "grad_norm": 0.4555925682379364,
+      "learning_rate": 4.7509448859175043e-05,
+      "loss": 0.6185,
+      "step": 1209
+    },
+    {
+      "epoch": 0.16934919524142758,
+      "grad_norm": 0.45021258994097924,
+      "learning_rate": 4.75045153265227e-05,
+      "loss": 0.6086,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1694891532540238,
+      "grad_norm": 0.4369626572960402,
+      "learning_rate": 4.749957716896412e-05,
+      "loss": 0.6409,
+      "step": 1211
+    },
+    {
+      "epoch": 0.16962911126662,
+      "grad_norm": 0.4320366971573474,
+      "learning_rate": 4.749463438751413e-05,
+      "loss": 0.5844,
+      "step": 1212
+    },
+    {
+      "epoch": 0.16976906927921623,
+      "grad_norm": 0.4163870413689256,
+      "learning_rate": 4.7489686983188535e-05,
+      "loss": 0.5803,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16990902729181245,
+      "grad_norm": 0.44949676404016337,
+      "learning_rate": 4.748473495700408e-05,
+      "loss": 0.6234,
+      "step": 1214
+    },
+    {
+      "epoch": 0.17004898530440868,
+      "grad_norm": 0.44171736970520764,
+      "learning_rate": 4.747977830997845e-05,
+      "loss": 0.593,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1701889433170049,
+      "grad_norm": 0.44772610657802425,
+      "learning_rate": 4.747481704313028e-05,
+      "loss": 0.5924,
+      "step": 1216
+    },
+    {
+      "epoch": 0.17032890132960113,
+      "grad_norm": 0.4549687449320509,
+      "learning_rate": 4.7469851157479177e-05,
+      "loss": 0.6332,
+      "step": 1217
+    },
+    {
+      "epoch": 0.17046885934219735,
+      "grad_norm": 0.4682088646897931,
+      "learning_rate": 4.746488065404567e-05,
+      "loss": 0.6189,
+      "step": 1218
+    },
+    {
+      "epoch": 0.17060881735479355,
+      "grad_norm": 0.44383085207451206,
+      "learning_rate": 4.7459905533851246e-05,
+      "loss": 0.6176,
+      "step": 1219
+    },
+    {
+      "epoch": 0.17074877536738978,
+      "grad_norm": 0.4282311331266807,
+      "learning_rate": 4.745492579791835e-05,
+      "loss": 0.5548,
+      "step": 1220
+    },
+    {
+      "epoch": 0.170888733379986,
+      "grad_norm": 0.4542898636825727,
+      "learning_rate": 4.744994144727036e-05,
+      "loss": 0.6161,
+      "step": 1221
+    },
+    {
+      "epoch": 0.17102869139258223,
+      "grad_norm": 0.4667149443909015,
+      "learning_rate": 4.7444952482931626e-05,
+      "loss": 0.6133,
+      "step": 1222
+    },
+    {
+      "epoch": 0.17116864940517845,
+      "grad_norm": 0.4512113910763386,
+      "learning_rate": 4.743995890592742e-05,
+      "loss": 0.629,
+      "step": 1223
+    },
+    {
+      "epoch": 0.17130860741777468,
+      "grad_norm": 0.42697256281999074,
+      "learning_rate": 4.743496071728396e-05,
+      "loss": 0.61,
+      "step": 1224
+    },
+    {
+      "epoch": 0.17144856543037088,
+      "grad_norm": 0.4510644382875255,
+      "learning_rate": 4.7429957918028444e-05,
+      "loss": 0.6247,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1715885234429671,
+      "grad_norm": 0.4432778439526579,
+      "learning_rate": 4.7424950509188995e-05,
+      "loss": 0.6192,
+      "step": 1226
+    },
+    {
+      "epoch": 0.17172848145556333,
+      "grad_norm": 0.4215901680439938,
+      "learning_rate": 4.741993849179468e-05,
+      "loss": 0.5841,
+      "step": 1227
+    },
+    {
+      "epoch": 0.17186843946815955,
+      "grad_norm": 0.43500291765639604,
+      "learning_rate": 4.7414921866875524e-05,
+      "loss": 0.6216,
+      "step": 1228
+    },
+    {
+      "epoch": 0.17200839748075578,
+      "grad_norm": 0.43425438707929986,
+      "learning_rate": 4.740990063546249e-05,
+      "loss": 0.6358,
+      "step": 1229
+    },
+    {
+      "epoch": 0.172148355493352,
+      "grad_norm": 0.45544131433595525,
+      "learning_rate": 4.7404874798587494e-05,
+      "loss": 0.636,
+      "step": 1230
+    },
+    {
+      "epoch": 0.17228831350594823,
+      "grad_norm": 0.4359451447095212,
+      "learning_rate": 4.7399844357283397e-05,
+      "loss": 0.6124,
+      "step": 1231
+    },
+    {
+      "epoch": 0.17242827151854442,
+      "grad_norm": 0.4435735452706431,
+      "learning_rate": 4.739480931258401e-05,
+      "loss": 0.5854,
+      "step": 1232
+    },
+    {
+      "epoch": 0.17256822953114065,
+      "grad_norm": 0.4550094611824221,
+      "learning_rate": 4.738976966552407e-05,
+      "loss": 0.649,
+      "step": 1233
+    },
+    {
+      "epoch": 0.17270818754373687,
+      "grad_norm": 0.4417929722405249,
+      "learning_rate": 4.738472541713931e-05,
+      "loss": 0.6156,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1728481455563331,
+      "grad_norm": 0.46054175505288897,
+      "learning_rate": 4.737967656846633e-05,
+      "loss": 0.5911,
+      "step": 1235
+    },
+    {
+      "epoch": 0.17298810356892932,
+      "grad_norm": 0.44199250556635683,
+      "learning_rate": 4.737462312054275e-05,
+      "loss": 0.6378,
+      "step": 1236
+    },
+    {
+      "epoch": 0.17312806158152555,
+      "grad_norm": 0.4365770107069145,
+      "learning_rate": 4.736956507440709e-05,
+      "loss": 0.6321,
+      "step": 1237
+    },
+    {
+      "epoch": 0.17326801959412177,
+      "grad_norm": 0.4569326418275404,
+      "learning_rate": 4.7364502431098844e-05,
+      "loss": 0.6128,
+      "step": 1238
+    },
+    {
+      "epoch": 0.17340797760671797,
+      "grad_norm": 0.4331079335375279,
+      "learning_rate": 4.7359435191658425e-05,
+      "loss": 0.6375,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1735479356193142,
+      "grad_norm": 0.45213254965556654,
+      "learning_rate": 4.7354363357127204e-05,
+      "loss": 0.6024,
+      "step": 1240
+    },
+    {
+      "epoch": 0.17368789363191042,
+      "grad_norm": 0.4301062314752303,
+      "learning_rate": 4.7349286928547494e-05,
+      "loss": 0.6164,
+      "step": 1241
+    },
+    {
+      "epoch": 0.17382785164450665,
+      "grad_norm": 0.43058324357912753,
+      "learning_rate": 4.7344205906962555e-05,
+      "loss": 0.6005,
+      "step": 1242
+    },
+    {
+      "epoch": 0.17396780965710287,
+      "grad_norm": 0.4180674272631518,
+      "learning_rate": 4.7339120293416594e-05,
+      "loss": 0.6088,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1741077676696991,
+      "grad_norm": 0.4546933788993979,
+      "learning_rate": 4.733403008895474e-05,
+      "loss": 0.6279,
+      "step": 1244
+    },
+    {
+      "epoch": 0.17424772568229532,
+      "grad_norm": 0.4477652947027751,
+      "learning_rate": 4.73289352946231e-05,
+      "loss": 0.6189,
+      "step": 1245
+    },
+    {
+      "epoch": 0.17438768369489152,
+      "grad_norm": 0.4240511875731597,
+      "learning_rate": 4.732383591146869e-05,
+      "loss": 0.6028,
+      "step": 1246
+    },
+    {
+      "epoch": 0.17452764170748775,
+      "grad_norm": 0.4292894608181786,
+      "learning_rate": 4.73187319405395e-05,
+      "loss": 0.6109,
+      "step": 1247
+    },
+    {
+      "epoch": 0.17466759972008397,
+      "grad_norm": 0.4268503231804948,
+      "learning_rate": 4.7313623382884435e-05,
+      "loss": 0.6014,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1748075577326802,
+      "grad_norm": 0.44496488332160694,
+      "learning_rate": 4.730851023955337e-05,
+      "loss": 0.611,
+      "step": 1249
+    },
+    {
+      "epoch": 0.17494751574527642,
+      "grad_norm": 0.43212438694423877,
+      "learning_rate": 4.730339251159709e-05,
+      "loss": 0.5553,
+      "step": 1250
+    },
+    {
+      "epoch": 0.17508747375787265,
+      "grad_norm": 0.45661821214759873,
+      "learning_rate": 4.729827020006735e-05,
+      "loss": 0.6071,
+      "step": 1251
+    },
+    {
+      "epoch": 0.17522743177046887,
+      "grad_norm": 0.43901674416520003,
+      "learning_rate": 4.7293143306016836e-05,
+      "loss": 0.5777,
+      "step": 1252
+    },
+    {
+      "epoch": 0.17536738978306507,
+      "grad_norm": 0.4544033001494578,
+      "learning_rate": 4.728801183049918e-05,
+      "loss": 0.639,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1755073477956613,
+      "grad_norm": 0.4520331595590865,
+      "learning_rate": 4.728287577456894e-05,
+      "loss": 0.6089,
+      "step": 1254
+    },
+    {
+      "epoch": 0.17564730580825752,
+      "grad_norm": 0.44402099958528857,
+      "learning_rate": 4.7277735139281645e-05,
+      "loss": 0.6056,
+      "step": 1255
+    },
+    {
+      "epoch": 0.17578726382085375,
+      "grad_norm": 0.4599500342986939,
+      "learning_rate": 4.7272589925693735e-05,
+      "loss": 0.6,
+      "step": 1256
+    },
+    {
+      "epoch": 0.17592722183344997,
+      "grad_norm": 0.41917281344225943,
+      "learning_rate": 4.7267440134862604e-05,
+      "loss": 0.6204,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1760671798460462,
+      "grad_norm": 0.4365380087308256,
+      "learning_rate": 4.72622857678466e-05,
+      "loss": 0.6364,
+      "step": 1258
+    },
+    {
+      "epoch": 0.17620713785864242,
+      "grad_norm": 0.4302123239208661,
+      "learning_rate": 4.725712682570498e-05,
+      "loss": 0.6077,
+      "step": 1259
+    },
+    {
+      "epoch": 0.17634709587123862,
+      "grad_norm": 0.4388230610863713,
+      "learning_rate": 4.725196330949797e-05,
+      "loss": 0.5846,
+      "step": 1260
+    },
+    {
+      "epoch": 0.17648705388383484,
+      "grad_norm": 0.43045983689708245,
+      "learning_rate": 4.724679522028672e-05,
+      "loss": 0.5945,
+      "step": 1261
+    },
+    {
+      "epoch": 0.17662701189643107,
+      "grad_norm": 0.45979393311365935,
+      "learning_rate": 4.7241622559133325e-05,
+      "loss": 0.679,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1767669699090273,
+      "grad_norm": 0.4383956739878393,
+      "learning_rate": 4.723644532710082e-05,
+      "loss": 0.6391,
+      "step": 1263
+    },
+    {
+      "epoch": 0.17690692792162352,
+      "grad_norm": 0.4173574514314508,
+      "learning_rate": 4.723126352525318e-05,
+      "loss": 0.5922,
+      "step": 1264
+    },
+    {
+      "epoch": 0.17704688593421974,
+      "grad_norm": 0.4480684757490661,
+      "learning_rate": 4.722607715465532e-05,
+      "loss": 0.607,
+      "step": 1265
+    },
+    {
+      "epoch": 0.17718684394681594,
+      "grad_norm": 0.40606023013599496,
+      "learning_rate": 4.722088621637309e-05,
+      "loss": 0.5901,
+      "step": 1266
+    },
+    {
+      "epoch": 0.17732680195941217,
+      "grad_norm": 0.5273306368158803,
+      "learning_rate": 4.7215690711473275e-05,
+      "loss": 0.6502,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1774667599720084,
+      "grad_norm": 0.9204038885794614,
+      "learning_rate": 4.7210490641023615e-05,
+      "loss": 0.6287,
+      "step": 1268
+    },
+    {
+      "epoch": 0.17760671798460462,
+      "grad_norm": 0.40746942485511245,
+      "learning_rate": 4.7205286006092764e-05,
+      "loss": 0.6017,
+      "step": 1269
+    },
+    {
+      "epoch": 0.17774667599720084,
+      "grad_norm": 0.41902531090541245,
+      "learning_rate": 4.720007680775034e-05,
+      "loss": 0.5875,
+      "step": 1270
+    },
+    {
+      "epoch": 0.17788663400979707,
+      "grad_norm": 0.4117792800186485,
+      "learning_rate": 4.719486304706687e-05,
+      "loss": 0.6485,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1780265920223933,
+      "grad_norm": 0.4470634029440558,
+      "learning_rate": 4.718964472511386e-05,
+      "loss": 0.6527,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1781665500349895,
+      "grad_norm": 0.43032660391655064,
+      "learning_rate": 4.71844218429637e-05,
+      "loss": 0.6161,
+      "step": 1273
+    },
+    {
+      "epoch": 0.17830650804758572,
+      "grad_norm": 0.44832712777708833,
+      "learning_rate": 4.7179194401689764e-05,
+      "loss": 0.6199,
+      "step": 1274
+    },
+    {
+      "epoch": 0.17844646606018194,
+      "grad_norm": 0.42568085262950894,
+      "learning_rate": 4.7173962402366334e-05,
+      "loss": 0.6131,
+      "step": 1275
+    },
+    {
+      "epoch": 0.17858642407277817,
+      "grad_norm": 0.43106804328266973,
+      "learning_rate": 4.716872584606865e-05,
+      "loss": 0.6086,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1787263820853744,
+      "grad_norm": 0.4339145485054148,
+      "learning_rate": 4.716348473387286e-05,
+      "loss": 0.581,
+      "step": 1277
+    },
+    {
+      "epoch": 0.17886634009797062,
+      "grad_norm": 0.559402125506372,
+      "learning_rate": 4.715823906685609e-05,
+      "loss": 0.5788,
+      "step": 1278
+    },
+    {
+      "epoch": 0.17900629811056684,
+      "grad_norm": 0.4466695949938704,
+      "learning_rate": 4.715298884609636e-05,
+      "loss": 0.6169,
+      "step": 1279
+    },
+    {
+      "epoch": 0.17914625612316304,
+      "grad_norm": 0.42599041437598534,
+      "learning_rate": 4.7147734072672644e-05,
+      "loss": 0.5892,
+      "step": 1280
+    },
+    {
+      "epoch": 0.17928621413575926,
+      "grad_norm": 0.4193811493005937,
+      "learning_rate": 4.7142474747664856e-05,
+      "loss": 0.5716,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1794261721483555,
+      "grad_norm": 0.4283159486252786,
+      "learning_rate": 4.7137210872153844e-05,
+      "loss": 0.5828,
+      "step": 1282
+    },
+    {
+      "epoch": 0.17956613016095171,
+      "grad_norm": 0.5396297802303109,
+      "learning_rate": 4.713194244722138e-05,
+      "loss": 0.6024,
+      "step": 1283
+    },
+    {
+      "epoch": 0.17970608817354794,
+      "grad_norm": 0.45852485695580586,
+      "learning_rate": 4.712666947395018e-05,
+      "loss": 0.5972,
+      "step": 1284
+    },
+    {
+      "epoch": 0.17984604618614417,
+      "grad_norm": 0.4385616625675766,
+      "learning_rate": 4.71213919534239e-05,
+      "loss": 0.6379,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1799860041987404,
+      "grad_norm": 0.45421354404722714,
+      "learning_rate": 4.711610988672712e-05,
+      "loss": 0.6314,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1801259622113366,
+      "grad_norm": 0.4174346560548907,
+      "learning_rate": 4.711082327494536e-05,
+      "loss": 0.582,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1802659202239328,
+      "grad_norm": 0.46268786301319254,
+      "learning_rate": 4.7105532119165066e-05,
+      "loss": 0.6187,
+      "step": 1288
+    },
+    {
+      "epoch": 0.18040587823652904,
+      "grad_norm": 0.4225894844217519,
+      "learning_rate": 4.710023642047364e-05,
+      "loss": 0.6044,
+      "step": 1289
+    },
+    {
+      "epoch": 0.18054583624912526,
+      "grad_norm": 0.4212476755455102,
+      "learning_rate": 4.709493617995938e-05,
+      "loss": 0.6139,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1806857942617215,
+      "grad_norm": 0.43172618509391675,
+      "learning_rate": 4.7089631398711556e-05,
+      "loss": 0.6195,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1808257522743177,
+      "grad_norm": 0.47068060331043937,
+      "learning_rate": 4.7084322077820345e-05,
+      "loss": 0.6475,
+      "step": 1292
+    },
+    {
+      "epoch": 0.18096571028691394,
+      "grad_norm": 0.42995345865450285,
+      "learning_rate": 4.707900821837686e-05,
+      "loss": 0.6263,
+      "step": 1293
+    },
+    {
+      "epoch": 0.18110566829951014,
+      "grad_norm": 0.4574009601171825,
+      "learning_rate": 4.707368982147318e-05,
+      "loss": 0.6676,
+      "step": 1294
+    },
+    {
+      "epoch": 0.18124562631210636,
+      "grad_norm": 0.44635706573473055,
+      "learning_rate": 4.7068366888202264e-05,
+      "loss": 0.6532,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1813855843247026,
+      "grad_norm": 0.41536175859929064,
+      "learning_rate": 4.7063039419658035e-05,
+      "loss": 0.6001,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1815255423372988,
+      "grad_norm": 0.431878399826314,
+      "learning_rate": 4.705770741693535e-05,
+      "loss": 0.5995,
+      "step": 1297
+    },
+    {
+      "epoch": 0.18166550034989504,
+      "grad_norm": 0.43247283157231065,
+      "learning_rate": 4.7052370881129976e-05,
+      "loss": 0.6269,
+      "step": 1298
+    },
+    {
+      "epoch": 0.18180545836249126,
+      "grad_norm": 0.44705916198880513,
+      "learning_rate": 4.704702981333864e-05,
+      "loss": 0.5862,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1819454163750875,
+      "grad_norm": 0.4435892818551313,
+      "learning_rate": 4.704168421465896e-05,
+      "loss": 0.6206,
+      "step": 1300
+    },
+    {
+      "epoch": 0.18208537438768368,
+      "grad_norm": 0.8694903989330446,
+      "learning_rate": 4.7036334086189555e-05,
+      "loss": 0.6592,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1822253324002799,
+      "grad_norm": 0.4457381902676728,
+      "learning_rate": 4.703097942902989e-05,
+      "loss": 0.5962,
+      "step": 1302
+    },
+    {
+      "epoch": 0.18236529041287614,
+      "grad_norm": 0.4105292740202644,
+      "learning_rate": 4.702562024428042e-05,
+      "loss": 0.6314,
+      "step": 1303
+    },
+    {
+      "epoch": 0.18250524842547236,
+      "grad_norm": 0.43538468918998957,
+      "learning_rate": 4.70202565330425e-05,
+      "loss": 0.6348,
+      "step": 1304
+    },
+    {
+      "epoch": 0.18264520643806859,
+      "grad_norm": 0.42334962592608394,
+      "learning_rate": 4.701488829641845e-05,
+      "loss": 0.6271,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1827851644506648,
+      "grad_norm": 0.44222681529868485,
+      "learning_rate": 4.700951553551148e-05,
+      "loss": 0.6139,
+      "step": 1306
+    },
+    {
+      "epoch": 0.182925122463261,
+      "grad_norm": 0.41298228492856454,
+      "learning_rate": 4.700413825142574e-05,
+      "loss": 0.6059,
+      "step": 1307
+    },
+    {
+      "epoch": 0.18306508047585723,
+      "grad_norm": 0.48053391897594716,
+      "learning_rate": 4.6998756445266336e-05,
+      "loss": 0.6252,
+      "step": 1308
+    },
+    {
+      "epoch": 0.18320503848845346,
+      "grad_norm": 0.422350382910677,
+      "learning_rate": 4.6993370118139264e-05,
+      "loss": 0.594,
+      "step": 1309
+    },
+    {
+      "epoch": 0.18334499650104968,
+      "grad_norm": 0.45501043482560677,
+      "learning_rate": 4.698797927115148e-05,
+      "loss": 0.6039,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1834849545136459,
+      "grad_norm": 0.42755574101023547,
+      "learning_rate": 4.698258390541086e-05,
+      "loss": 0.635,
+      "step": 1311
+    },
+    {
+      "epoch": 0.18362491252624213,
+      "grad_norm": 0.4212712876216914,
+      "learning_rate": 4.69771840220262e-05,
+      "loss": 0.6187,
+      "step": 1312
+    },
+    {
+      "epoch": 0.18376487053883836,
+      "grad_norm": 0.4474515144815738,
+      "learning_rate": 4.697177962210722e-05,
+      "loss": 0.6044,
+      "step": 1313
+    },
+    {
+      "epoch": 0.18390482855143456,
+      "grad_norm": 0.4574401172788082,
+      "learning_rate": 4.696637070676462e-05,
+      "loss": 0.6465,
+      "step": 1314
+    },
+    {
+      "epoch": 0.18404478656403078,
+      "grad_norm": 0.47635542166003236,
+      "learning_rate": 4.696095727710994e-05,
+      "loss": 0.5871,
+      "step": 1315
+    },
+    {
+      "epoch": 0.184184744576627,
+      "grad_norm": 0.44962202431159326,
+      "learning_rate": 4.6955539334255716e-05,
+      "loss": 0.6232,
+      "step": 1316
+    },
+    {
+      "epoch": 0.18432470258922323,
+      "grad_norm": 0.45431685769602653,
+      "learning_rate": 4.6950116879315385e-05,
+      "loss": 0.621,
+      "step": 1317
+    },
+    {
+      "epoch": 0.18446466060181946,
+      "grad_norm": 0.4388675789360797,
+      "learning_rate": 4.6944689913403326e-05,
+      "loss": 0.6183,
+      "step": 1318
+    },
+    {
+      "epoch": 0.18460461861441568,
+      "grad_norm": 0.4523470765103244,
+      "learning_rate": 4.693925843763483e-05,
+      "loss": 0.6212,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1847445766270119,
+      "grad_norm": 0.4807021100408991,
+      "learning_rate": 4.693382245312612e-05,
+      "loss": 0.6097,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1848845346396081,
+      "grad_norm": 0.4306193212134266,
+      "learning_rate": 4.6928381960994336e-05,
+      "loss": 0.6438,
+      "step": 1321
+    },
+    {
+      "epoch": 0.18502449265220433,
+      "grad_norm": 0.447094803519498,
+      "learning_rate": 4.692293696235758e-05,
+      "loss": 0.6263,
+      "step": 1322
+    },
+    {
+      "epoch": 0.18516445066480056,
+      "grad_norm": 0.4200693798341761,
+      "learning_rate": 4.6917487458334824e-05,
+      "loss": 0.5863,
+      "step": 1323
+    },
+    {
+      "epoch": 0.18530440867739678,
+      "grad_norm": 0.4459700079709483,
+      "learning_rate": 4.691203345004602e-05,
+      "loss": 0.6113,
+      "step": 1324
+    },
+    {
+      "epoch": 0.185444366689993,
+      "grad_norm": 0.42812241397789785,
+      "learning_rate": 4.6906574938612e-05,
+      "loss": 0.613,
+      "step": 1325
+    },
+    {
+      "epoch": 0.18558432470258923,
+      "grad_norm": 0.44482736985800875,
+      "learning_rate": 4.690111192515457e-05,
+      "loss": 0.6087,
+      "step": 1326
+    },
+    {
+      "epoch": 0.18572428271518546,
+      "grad_norm": 0.43543296488997046,
+      "learning_rate": 4.6895644410796416e-05,
+      "loss": 0.6055,
+      "step": 1327
+    },
+    {
+      "epoch": 0.18586424072778165,
+      "grad_norm": 0.44518530430540526,
+      "learning_rate": 4.689017239666117e-05,
+      "loss": 0.6094,
+      "step": 1328
+    },
+    {
+      "epoch": 0.18600419874037788,
+      "grad_norm": 0.41902567935143953,
+      "learning_rate": 4.688469588387339e-05,
+      "loss": 0.5978,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1861441567529741,
+      "grad_norm": 0.423883271911657,
+      "learning_rate": 4.6879214873558565e-05,
+      "loss": 0.5982,
+      "step": 1330
+    },
+    {
+      "epoch": 0.18628411476557033,
+      "grad_norm": 0.44582071516941296,
+      "learning_rate": 4.6873729366843075e-05,
+      "loss": 0.562,
+      "step": 1331
+    },
+    {
+      "epoch": 0.18642407277816656,
+      "grad_norm": 0.4660473204875247,
+      "learning_rate": 4.686823936485426e-05,
+      "loss": 0.6367,
+      "step": 1332
+    },
+    {
+      "epoch": 0.18656403079076278,
+      "grad_norm": 0.4769376469582944,
+      "learning_rate": 4.6862744868720374e-05,
+      "loss": 0.6652,
+      "step": 1333
+    },
+    {
+      "epoch": 0.186703988803359,
+      "grad_norm": 0.4618831962024523,
+      "learning_rate": 4.6857245879570585e-05,
+      "loss": 0.6317,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1868439468159552,
+      "grad_norm": 0.4550475574225298,
+      "learning_rate": 4.685174239853499e-05,
+      "loss": 0.5892,
+      "step": 1335
+    },
+    {
+      "epoch": 0.18698390482855143,
+      "grad_norm": 0.43735879865032035,
+      "learning_rate": 4.684623442674463e-05,
+      "loss": 0.5915,
+      "step": 1336
+    },
+    {
+      "epoch": 0.18712386284114765,
+      "grad_norm": 0.42856554004367964,
+      "learning_rate": 4.684072196533142e-05,
+      "loss": 0.6211,
+      "step": 1337
+    },
+    {
+      "epoch": 0.18726382085374388,
+      "grad_norm": 0.4359596291931712,
+      "learning_rate": 4.6835205015428246e-05,
+      "loss": 0.5957,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1874037788663401,
+      "grad_norm": 0.44381328220368255,
+      "learning_rate": 4.682968357816889e-05,
+      "loss": 0.634,
+      "step": 1339
+    },
+    {
+      "epoch": 0.18754373687893633,
+      "grad_norm": 0.45368741213258845,
+      "learning_rate": 4.682415765468807e-05,
+      "loss": 0.6242,
+      "step": 1340
+    },
+    {
+      "epoch": 0.18768369489153253,
+      "grad_norm": 0.433054739833902,
+      "learning_rate": 4.681862724612141e-05,
+      "loss": 0.6155,
+      "step": 1341
+    },
+    {
+      "epoch": 0.18782365290412875,
+      "grad_norm": 0.4455643661885221,
+      "learning_rate": 4.681309235360546e-05,
+      "loss": 0.6061,
+      "step": 1342
+    },
+    {
+      "epoch": 0.18796361091672498,
+      "grad_norm": 0.4348159376755847,
+      "learning_rate": 4.6807552978277725e-05,
+      "loss": 0.6016,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1881035689293212,
+      "grad_norm": 0.43841673430612965,
+      "learning_rate": 4.6802009121276566e-05,
+      "loss": 0.6139,
+      "step": 1344
+    },
+    {
+      "epoch": 0.18824352694191743,
+      "grad_norm": 0.43548395332890844,
+      "learning_rate": 4.679646078374133e-05,
+      "loss": 0.6335,
+      "step": 1345
+    },
+    {
+      "epoch": 0.18838348495451365,
+      "grad_norm": 0.4311071382664937,
+      "learning_rate": 4.679090796681225e-05,
+      "loss": 0.6493,
+      "step": 1346
+    },
+    {
+      "epoch": 0.18852344296710988,
+      "grad_norm": 0.43011170445061725,
+      "learning_rate": 4.6785350671630467e-05,
+      "loss": 0.5733,
+      "step": 1347
+    },
+    {
+      "epoch": 0.18866340097970608,
+      "grad_norm": 0.44356466777661,
+      "learning_rate": 4.6779788899338095e-05,
+      "loss": 0.5936,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1888033589923023,
+      "grad_norm": 0.4408469771886245,
+      "learning_rate": 4.6774222651078106e-05,
+      "loss": 0.6,
+      "step": 1349
+    },
+    {
+      "epoch": 0.18894331700489853,
+      "grad_norm": 0.4174867184800617,
+      "learning_rate": 4.6768651927994434e-05,
+      "loss": 0.5785,
+      "step": 1350
+    },
+    {
+      "epoch": 0.18908327501749475,
+      "grad_norm": 0.4202539862345758,
+      "learning_rate": 4.6763076731231916e-05,
+      "loss": 0.6061,
+      "step": 1351
+    },
+    {
+      "epoch": 0.18922323303009098,
+      "grad_norm": 0.4362329726223066,
+      "learning_rate": 4.675749706193631e-05,
+      "loss": 0.6105,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1893631910426872,
+      "grad_norm": 0.4463015177283665,
+      "learning_rate": 4.67519129212543e-05,
+      "loss": 0.6196,
+      "step": 1353
+    },
+    {
+      "epoch": 0.18950314905528343,
+      "grad_norm": 0.4213305303975094,
+      "learning_rate": 4.674632431033348e-05,
+      "loss": 0.5771,
+      "step": 1354
+    },
+    {
+      "epoch": 0.18964310706787962,
+      "grad_norm": 0.4523660797014365,
+      "learning_rate": 4.674073123032236e-05,
+      "loss": 0.5614,
+      "step": 1355
+    },
+    {
+      "epoch": 0.18978306508047585,
+      "grad_norm": 0.42558959589979956,
+      "learning_rate": 4.673513368237039e-05,
+      "loss": 0.5987,
+      "step": 1356
+    },
+    {
+      "epoch": 0.18992302309307207,
+      "grad_norm": 0.4566223530786513,
+      "learning_rate": 4.6729531667627905e-05,
+      "loss": 0.6081,
+      "step": 1357
+    },
+    {
+      "epoch": 0.1900629811056683,
+      "grad_norm": 0.4324304893039991,
+      "learning_rate": 4.672392518724619e-05,
+      "loss": 0.6132,
+      "step": 1358
+    },
+    {
+      "epoch": 0.19020293911826452,
+      "grad_norm": 0.43913060355532496,
+      "learning_rate": 4.671831424237743e-05,
+      "loss": 0.5989,
+      "step": 1359
+    },
+    {
+      "epoch": 0.19034289713086075,
+      "grad_norm": 0.44320622872500787,
+      "learning_rate": 4.671269883417473e-05,
+      "loss": 0.5998,
+      "step": 1360
+    },
+    {
+      "epoch": 0.19048285514345698,
+      "grad_norm": 0.4298692464735462,
+      "learning_rate": 4.670707896379211e-05,
+      "loss": 0.555,
+      "step": 1361
+    },
+    {
+      "epoch": 0.19062281315605317,
+      "grad_norm": 0.4212199124955075,
+      "learning_rate": 4.670145463238451e-05,
+      "loss": 0.6177,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1907627711686494,
+      "grad_norm": 0.4448003823408276,
+      "learning_rate": 4.669582584110779e-05,
+      "loss": 0.5932,
+      "step": 1363
+    },
+    {
+      "epoch": 0.19090272918124562,
+      "grad_norm": 0.44697385174479454,
+      "learning_rate": 4.6690192591118734e-05,
+      "loss": 0.6026,
+      "step": 1364
+    },
+    {
+      "epoch": 0.19104268719384185,
+      "grad_norm": 0.4671259243593673,
+      "learning_rate": 4.668455488357502e-05,
+      "loss": 0.6156,
+      "step": 1365
+    },
+    {
+      "epoch": 0.19118264520643807,
+      "grad_norm": 0.43396652865433255,
+      "learning_rate": 4.6678912719635246e-05,
+      "loss": 0.5764,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1913226032190343,
+      "grad_norm": 0.4375416377619642,
+      "learning_rate": 4.667326610045895e-05,
+      "loss": 0.6031,
+      "step": 1367
+    },
+    {
+      "epoch": 0.19146256123163052,
+      "grad_norm": 0.441530782470694,
+      "learning_rate": 4.6667615027206564e-05,
+      "loss": 0.5773,
+      "step": 1368
+    },
+    {
+      "epoch": 0.19160251924422672,
+      "grad_norm": 0.43470376998767046,
+      "learning_rate": 4.6661959501039446e-05,
+      "loss": 0.5991,
+      "step": 1369
+    },
+    {
+      "epoch": 0.19174247725682295,
+      "grad_norm": 0.4389060530511099,
+      "learning_rate": 4.665629952311985e-05,
+      "loss": 0.5692,
+      "step": 1370
+    },
+    {
+      "epoch": 0.19188243526941917,
+      "grad_norm": 0.4752733107460391,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.6525,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1920223932820154,
+      "grad_norm": 0.43452630074557697,
+      "learning_rate": 4.66449662166769e-05,
+      "loss": 0.5925,
+      "step": 1372
+    },
+    {
+      "epoch": 0.19216235129461162,
+      "grad_norm": 0.4472660808670533,
+      "learning_rate": 4.663929289048266e-05,
+      "loss": 0.6175,
+      "step": 1373
+    },
+    {
+      "epoch": 0.19230230930720785,
+      "grad_norm": 0.4151649360316345,
+      "learning_rate": 4.6633615117194165e-05,
+      "loss": 0.6048,
+      "step": 1374
+    },
+    {
+      "epoch": 0.19244226731980407,
+      "grad_norm": 0.41991921626461726,
+      "learning_rate": 4.6627932897978254e-05,
+      "loss": 0.6141,
+      "step": 1375
+    },
+    {
+      "epoch": 0.19258222533240027,
+      "grad_norm": 0.44865262717314497,
+      "learning_rate": 4.6622246234002686e-05,
+      "loss": 0.6108,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1927221833449965,
+      "grad_norm": 0.42390019369736476,
+      "learning_rate": 4.6616555126436134e-05,
+      "loss": 0.6178,
+      "step": 1377
+    },
+    {
+      "epoch": 0.19286214135759272,
+      "grad_norm": 0.47111888994325224,
+      "learning_rate": 4.6610859576448176e-05,
+      "loss": 0.6351,
+      "step": 1378
+    },
+    {
+      "epoch": 0.19300209937018895,
+      "grad_norm": 0.4694892453227565,
+      "learning_rate": 4.660515958520929e-05,
+      "loss": 0.6313,
+      "step": 1379
+    },
+    {
+      "epoch": 0.19314205738278517,
+      "grad_norm": 0.43419016166335284,
+      "learning_rate": 4.65994551538909e-05,
+      "loss": 0.592,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1932820153953814,
+      "grad_norm": 0.4133745853690666,
+      "learning_rate": 4.659374628366532e-05,
+      "loss": 0.572,
+      "step": 1381
+    },
+    {
+      "epoch": 0.1934219734079776,
+      "grad_norm": 0.4786575561548947,
+      "learning_rate": 4.658803297570577e-05,
+      "loss": 0.6007,
+      "step": 1382
+    },
+    {
+      "epoch": 0.19356193142057382,
+      "grad_norm": 0.4524476179381755,
+      "learning_rate": 4.658231523118641e-05,
+      "loss": 0.5966,
+      "step": 1383
+    },
+    {
+      "epoch": 0.19370188943317004,
+      "grad_norm": 0.4349438663500916,
+      "learning_rate": 4.6576593051282286e-05,
+      "loss": 0.5862,
+      "step": 1384
+    },
+    {
+      "epoch": 0.19384184744576627,
+      "grad_norm": 0.45493985969941036,
+      "learning_rate": 4.657086643716936e-05,
+      "loss": 0.6482,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1939818054583625,
+      "grad_norm": 0.42740591145801066,
+      "learning_rate": 4.6565135390024515e-05,
+      "loss": 0.5906,
+      "step": 1386
+    },
+    {
+      "epoch": 0.19412176347095872,
+      "grad_norm": 0.47535147700181984,
+      "learning_rate": 4.6559399911025545e-05,
+      "loss": 0.5951,
+      "step": 1387
+    },
+    {
+      "epoch": 0.19426172148355494,
+      "grad_norm": 0.4333924774938176,
+      "learning_rate": 4.655366000135114e-05,
+      "loss": 0.6083,
+      "step": 1388
+    },
+    {
+      "epoch": 0.19440167949615114,
+      "grad_norm": 0.4197374995007798,
+      "learning_rate": 4.6547915662180905e-05,
+      "loss": 0.5758,
+      "step": 1389
+    },
+    {
+      "epoch": 0.19454163750874737,
+      "grad_norm": 0.4209565338579636,
+      "learning_rate": 4.6542166894695366e-05,
+      "loss": 0.6219,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1946815955213436,
+      "grad_norm": 0.4533900200145802,
+      "learning_rate": 4.653641370007596e-05,
+      "loss": 0.6284,
+      "step": 1391
+    },
+    {
+      "epoch": 0.19482155353393982,
+      "grad_norm": 0.43123676441236164,
+      "learning_rate": 4.653065607950502e-05,
+      "loss": 0.6172,
+      "step": 1392
+    },
+    {
+      "epoch": 0.19496151154653604,
+      "grad_norm": 0.43240025310515345,
+      "learning_rate": 4.652489403416579e-05,
+      "loss": 0.6231,
+      "step": 1393
+    },
+    {
+      "epoch": 0.19510146955913227,
+      "grad_norm": 0.4105873841245408,
+      "learning_rate": 4.651912756524244e-05,
+      "loss": 0.6351,
+      "step": 1394
+    },
+    {
+      "epoch": 0.1952414275717285,
+      "grad_norm": 0.4342716468025768,
+      "learning_rate": 4.651335667392003e-05,
+      "loss": 0.5944,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1953813855843247,
+      "grad_norm": 0.42929266034842917,
+      "learning_rate": 4.6507581361384537e-05,
+      "loss": 0.5949,
+      "step": 1396
+    },
+    {
+      "epoch": 0.19552134359692092,
+      "grad_norm": 0.43375421894645294,
+      "learning_rate": 4.650180162882285e-05,
+      "loss": 0.5674,
+      "step": 1397
+    },
+    {
+      "epoch": 0.19566130160951714,
+      "grad_norm": 0.4605858134606264,
+      "learning_rate": 4.649601747742277e-05,
+      "loss": 0.6276,
+      "step": 1398
+    },
+    {
+      "epoch": 0.19580125962211337,
+      "grad_norm": 0.4096793441616307,
+      "learning_rate": 4.649022890837298e-05,
+      "loss": 0.6066,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1959412176347096,
+      "grad_norm": 0.43325226542495887,
+      "learning_rate": 4.6484435922863105e-05,
+      "loss": 0.578,
+      "step": 1400
+    },
+    {
+      "epoch": 0.19608117564730582,
+      "grad_norm": 0.45203303336392237,
+      "learning_rate": 4.6478638522083654e-05,
+      "loss": 0.6247,
+      "step": 1401
+    },
+    {
+      "epoch": 0.19622113365990204,
+      "grad_norm": 0.43768863421738247,
+      "learning_rate": 4.6472836707226065e-05,
+      "loss": 0.6089,
+      "step": 1402
+    },
+    {
+      "epoch": 0.19636109167249824,
+      "grad_norm": 0.4561097031131965,
+      "learning_rate": 4.646703047948264e-05,
+      "loss": 0.6228,
+      "step": 1403
+    },
+    {
+      "epoch": 0.19650104968509446,
+      "grad_norm": 0.430780639976112,
+      "learning_rate": 4.6461219840046654e-05,
+      "loss": 0.5712,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1966410076976907,
+      "grad_norm": 0.4421729743171197,
+      "learning_rate": 4.645540479011223e-05,
+      "loss": 0.608,
+      "step": 1405
+    },
+    {
+      "epoch": 0.19678096571028691,
+      "grad_norm": 0.4527004953532786,
+      "learning_rate": 4.644958533087443e-05,
+      "loss": 0.5794,
+      "step": 1406
+    },
+    {
+      "epoch": 0.19692092372288314,
+      "grad_norm": 0.43655870981476536,
+      "learning_rate": 4.64437614635292e-05,
+      "loss": 0.6072,
+      "step": 1407
+    },
+    {
+      "epoch": 0.19706088173547937,
+      "grad_norm": 0.46922192783691635,
+      "learning_rate": 4.643793318927342e-05,
+      "loss": 0.6129,
+      "step": 1408
+    },
+    {
+      "epoch": 0.1972008397480756,
+      "grad_norm": 0.4096022513008127,
+      "learning_rate": 4.6432100509304843e-05,
+      "loss": 0.6139,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1973407977606718,
+      "grad_norm": 0.44858888050025963,
+      "learning_rate": 4.642626342482215e-05,
+      "loss": 0.6494,
+      "step": 1410
+    },
+    {
+      "epoch": 0.197480755773268,
+      "grad_norm": 0.42709765753910495,
+      "learning_rate": 4.642042193702493e-05,
+      "loss": 0.6144,
+      "step": 1411
+    },
+    {
+      "epoch": 0.19762071378586424,
+      "grad_norm": 0.43179502942515346,
+      "learning_rate": 4.6414576047113655e-05,
+      "loss": 0.6533,
+      "step": 1412
+    },
+    {
+      "epoch": 0.19776067179846046,
+      "grad_norm": 0.4271262128439598,
+      "learning_rate": 4.640872575628973e-05,
+      "loss": 0.5787,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1979006298110567,
+      "grad_norm": 0.44047825759880344,
+      "learning_rate": 4.640287106575543e-05,
+      "loss": 0.5908,
+      "step": 1414
+    },
+    {
+      "epoch": 0.1980405878236529,
+      "grad_norm": 0.42112850121439344,
+      "learning_rate": 4.639701197671397e-05,
+      "loss": 0.6233,
+      "step": 1415
+    },
+    {
+      "epoch": 0.19818054583624914,
+      "grad_norm": 0.4372841966734705,
+      "learning_rate": 4.639114849036944e-05,
+      "loss": 0.6486,
+      "step": 1416
+    },
+    {
+      "epoch": 0.19832050384884534,
+      "grad_norm": 0.4296545377209088,
+      "learning_rate": 4.638528060792685e-05,
+      "loss": 0.5862,
+      "step": 1417
+    },
+    {
+      "epoch": 0.19846046186144156,
+      "grad_norm": 0.4242826054839983,
+      "learning_rate": 4.637940833059211e-05,
+      "loss": 0.6224,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1986004198740378,
+      "grad_norm": 0.4386897733085964,
+      "learning_rate": 4.637353165957203e-05,
+      "loss": 0.6062,
+      "step": 1419
+    },
+    {
+      "epoch": 0.198740377886634,
+      "grad_norm": 0.43253628217173173,
+      "learning_rate": 4.636765059607434e-05,
+      "loss": 0.5796,
+      "step": 1420
+    },
+    {
+      "epoch": 0.19888033589923024,
+      "grad_norm": 0.42456136906021713,
+      "learning_rate": 4.6361765141307645e-05,
+      "loss": 0.6257,
+      "step": 1421
+    },
+    {
+      "epoch": 0.19902029391182646,
+      "grad_norm": 0.4546276363960964,
+      "learning_rate": 4.635587529648146e-05,
+      "loss": 0.6193,
+      "step": 1422
+    },
+    {
+      "epoch": 0.19916025192442266,
+      "grad_norm": 0.4511658476734362,
+      "learning_rate": 4.634998106280622e-05,
+      "loss": 0.5936,
+      "step": 1423
+    },
+    {
+      "epoch": 0.19930020993701889,
+      "grad_norm": 0.44597648248378585,
+      "learning_rate": 4.634408244149324e-05,
+      "loss": 0.6051,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1994401679496151,
+      "grad_norm": 0.43808391202287855,
+      "learning_rate": 4.6338179433754756e-05,
+      "loss": 0.5995,
+      "step": 1425
+    },
+    {
+      "epoch": 0.19958012596221134,
+      "grad_norm": 0.48296291796144863,
+      "learning_rate": 4.6332272040803895e-05,
+      "loss": 0.5904,
+      "step": 1426
+    },
+    {
+      "epoch": 0.19972008397480756,
+      "grad_norm": 0.43753233592691976,
+      "learning_rate": 4.632636026385468e-05,
+      "loss": 0.5993,
+      "step": 1427
+    },
+    {
+      "epoch": 0.19986004198740379,
+      "grad_norm": 0.441993440207615,
+      "learning_rate": 4.632044410412204e-05,
+      "loss": 0.5981,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.44046702891678347,
+      "learning_rate": 4.631452356282182e-05,
+      "loss": 0.5711,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2001399580125962,
+      "grad_norm": 0.42955814782218116,
+      "learning_rate": 4.630859864117073e-05,
+      "loss": 0.5847,
+      "step": 1430
+    },
+    {
+      "epoch": 0.20027991602519243,
+      "grad_norm": 0.43651504606811536,
+      "learning_rate": 4.630266934038642e-05,
+      "loss": 0.6069,
+      "step": 1431
+    },
+    {
+      "epoch": 0.20041987403778866,
+      "grad_norm": 0.428395840664284,
+      "learning_rate": 4.629673566168741e-05,
+      "loss": 0.5963,
+      "step": 1432
+    },
+    {
+      "epoch": 0.20055983205038488,
+      "grad_norm": 0.4138324728369602,
+      "learning_rate": 4.629079760629313e-05,
+      "loss": 0.5647,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2006997900629811,
+      "grad_norm": 0.4454769448792182,
+      "learning_rate": 4.628485517542392e-05,
+      "loss": 0.5915,
+      "step": 1434
+    },
+    {
+      "epoch": 0.20083974807557733,
+      "grad_norm": 0.43842498265287877,
+      "learning_rate": 4.627890837030101e-05,
+      "loss": 0.6333,
+      "step": 1435
+    },
+    {
+      "epoch": 0.20097970608817356,
+      "grad_norm": 0.42848973296459963,
+      "learning_rate": 4.627295719214653e-05,
+      "loss": 0.5877,
+      "step": 1436
+    },
+    {
+      "epoch": 0.20111966410076976,
+      "grad_norm": 0.44101122405778176,
+      "learning_rate": 4.6267001642183496e-05,
+      "loss": 0.6049,
+      "step": 1437
+    },
+    {
+      "epoch": 0.20125962211336598,
+      "grad_norm": 0.552921270157986,
+      "learning_rate": 4.6261041721635834e-05,
+      "loss": 0.6243,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2013995801259622,
+      "grad_norm": 0.4518649937814017,
+      "learning_rate": 4.625507743172838e-05,
+      "loss": 0.6189,
+      "step": 1439
+    },
+    {
+      "epoch": 0.20153953813855843,
+      "grad_norm": 0.4231937860204061,
+      "learning_rate": 4.6249108773686846e-05,
+      "loss": 0.5851,
+      "step": 1440
+    },
+    {
+      "epoch": 0.20167949615115466,
+      "grad_norm": 0.42182226251472993,
+      "learning_rate": 4.6243135748737864e-05,
+      "loss": 0.6401,
+      "step": 1441
+    },
+    {
+      "epoch": 0.20181945416375088,
+      "grad_norm": 0.48057969413105356,
+      "learning_rate": 4.623715835810893e-05,
+      "loss": 0.6243,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2019594121763471,
+      "grad_norm": 0.4437078265678712,
+      "learning_rate": 4.6231176603028484e-05,
+      "loss": 0.6328,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2020993701889433,
+      "grad_norm": 0.4507520837571798,
+      "learning_rate": 4.6225190484725824e-05,
+      "loss": 0.5904,
+      "step": 1444
+    },
+    {
+      "epoch": 0.20223932820153953,
+      "grad_norm": 0.43096343316315816,
+      "learning_rate": 4.6219200004431154e-05,
+      "loss": 0.6205,
+      "step": 1445
+    },
+    {
+      "epoch": 0.20237928621413576,
+      "grad_norm": 0.44215631216303225,
+      "learning_rate": 4.6213205163375586e-05,
+      "loss": 0.6171,
+      "step": 1446
+    },
+    {
+      "epoch": 0.20251924422673198,
+      "grad_norm": 0.4601057485158553,
+      "learning_rate": 4.620720596279112e-05,
+      "loss": 0.6355,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2026592022393282,
+      "grad_norm": 0.42229237803528424,
+      "learning_rate": 4.620120240391065e-05,
+      "loss": 0.5762,
+      "step": 1448
+    },
+    {
+      "epoch": 0.20279916025192443,
+      "grad_norm": 0.44247463233153156,
+      "learning_rate": 4.619519448796797e-05,
+      "loss": 0.6178,
+      "step": 1449
+    },
+    {
+      "epoch": 0.20293911826452066,
+      "grad_norm": 0.44572237683301125,
+      "learning_rate": 4.6189182216197766e-05,
+      "loss": 0.5782,
+      "step": 1450
+    },
+    {
+      "epoch": 0.20307907627711685,
+      "grad_norm": 0.44819354829925007,
+      "learning_rate": 4.618316558983562e-05,
+      "loss": 0.5575,
+      "step": 1451
+    },
+    {
+      "epoch": 0.20321903428971308,
+      "grad_norm": 0.462621341218452,
+      "learning_rate": 4.617714461011802e-05,
+      "loss": 0.6263,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2033589923023093,
+      "grad_norm": 0.4191206562398054,
+      "learning_rate": 4.6171119278282315e-05,
+      "loss": 0.5823,
+      "step": 1453
+    },
+    {
+      "epoch": 0.20349895031490553,
+      "grad_norm": 0.4234985010775575,
+      "learning_rate": 4.6165089595566795e-05,
+      "loss": 0.5935,
+      "step": 1454
+    },
+    {
+      "epoch": 0.20363890832750176,
+      "grad_norm": 0.4138224710610466,
+      "learning_rate": 4.6159055563210604e-05,
+      "loss": 0.5753,
+      "step": 1455
+    },
+    {
+      "epoch": 0.20377886634009798,
+      "grad_norm": 0.4657076630993664,
+      "learning_rate": 4.6153017182453814e-05,
+      "loss": 0.6038,
+      "step": 1456
+    },
+    {
+      "epoch": 0.20391882435269418,
+      "grad_norm": 0.4271574918448891,
+      "learning_rate": 4.6146974454537374e-05,
+      "loss": 0.613,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2040587823652904,
+      "grad_norm": 0.43052348959054654,
+      "learning_rate": 4.61409273807031e-05,
+      "loss": 0.578,
+      "step": 1458
+    },
+    {
+      "epoch": 0.20419874037788663,
+      "grad_norm": 0.446875776248068,
+      "learning_rate": 4.613487596219376e-05,
+      "loss": 0.6203,
+      "step": 1459
+    },
+    {
+      "epoch": 0.20433869839048285,
+      "grad_norm": 0.44280885463456926,
+      "learning_rate": 4.6128820200252954e-05,
+      "loss": 0.6178,
+      "step": 1460
+    },
+    {
+      "epoch": 0.20447865640307908,
+      "grad_norm": 0.4627088307800896,
+      "learning_rate": 4.612276009612522e-05,
+      "loss": 0.6083,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2046186144156753,
+      "grad_norm": 0.40566282511231716,
+      "learning_rate": 4.611669565105596e-05,
+      "loss": 0.5897,
+      "step": 1462
+    },
+    {
+      "epoch": 0.20475857242827153,
+      "grad_norm": 0.4392817676918071,
+      "learning_rate": 4.6110626866291485e-05,
+      "loss": 0.6058,
+      "step": 1463
+    },
+    {
+      "epoch": 0.20489853044086773,
+      "grad_norm": 0.42409351023845426,
+      "learning_rate": 4.6104553743078996e-05,
+      "loss": 0.5942,
+      "step": 1464
+    },
+    {
+      "epoch": 0.20503848845346395,
+      "grad_norm": 0.43598866958142785,
+      "learning_rate": 4.609847628266657e-05,
+      "loss": 0.6271,
+      "step": 1465
+    },
+    {
+      "epoch": 0.20517844646606018,
+      "grad_norm": 0.4347982090854984,
+      "learning_rate": 4.60923944863032e-05,
+      "loss": 0.611,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2053184044786564,
+      "grad_norm": 0.44333205254509317,
+      "learning_rate": 4.608630835523875e-05,
+      "loss": 0.6337,
+      "step": 1467
+    },
+    {
+      "epoch": 0.20545836249125263,
+      "grad_norm": 0.43208298348400775,
+      "learning_rate": 4.608021789072398e-05,
+      "loss": 0.5602,
+      "step": 1468
+    },
+    {
+      "epoch": 0.20559832050384885,
+      "grad_norm": 0.450291453219847,
+      "learning_rate": 4.607412309401054e-05,
+      "loss": 0.6653,
+      "step": 1469
+    },
+    {
+      "epoch": 0.20573827851644508,
+      "grad_norm": 0.4218568966937232,
+      "learning_rate": 4.606802396635098e-05,
+      "loss": 0.6039,
+      "step": 1470
+    },
+    {
+      "epoch": 0.20587823652904128,
+      "grad_norm": 0.4372151368221289,
+      "learning_rate": 4.6061920508998735e-05,
+      "loss": 0.6251,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2060181945416375,
+      "grad_norm": 0.42362258866936114,
+      "learning_rate": 4.6055812723208114e-05,
+      "loss": 0.5752,
+      "step": 1472
+    },
+    {
+      "epoch": 0.20615815255423373,
+      "grad_norm": 0.4281637412896991,
+      "learning_rate": 4.604970061023434e-05,
+      "loss": 0.5909,
+      "step": 1473
+    },
+    {
+      "epoch": 0.20629811056682995,
+      "grad_norm": 0.4473289913366391,
+      "learning_rate": 4.604358417133351e-05,
+      "loss": 0.5825,
+      "step": 1474
+    },
+    {
+      "epoch": 0.20643806857942618,
+      "grad_norm": 0.44127895646180343,
+      "learning_rate": 4.6037463407762616e-05,
+      "loss": 0.591,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2065780265920224,
+      "grad_norm": 0.4250252005670499,
+      "learning_rate": 4.6031338320779534e-05,
+      "loss": 0.61,
+      "step": 1476
+    },
+    {
+      "epoch": 0.20671798460461863,
+      "grad_norm": 0.4195465649774795,
+      "learning_rate": 4.602520891164304e-05,
+      "loss": 0.6044,
+      "step": 1477
+    },
+    {
+      "epoch": 0.20685794261721482,
+      "grad_norm": 0.4546098361873912,
+      "learning_rate": 4.601907518161277e-05,
+      "loss": 0.6216,
+      "step": 1478
+    },
+    {
+      "epoch": 0.20699790062981105,
+      "grad_norm": 0.4475804565840221,
+      "learning_rate": 4.601293713194929e-05,
+      "loss": 0.6793,
+      "step": 1479
+    },
+    {
+      "epoch": 0.20713785864240727,
+      "grad_norm": 0.43358956969279566,
+      "learning_rate": 4.600679476391402e-05,
+      "loss": 0.5491,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2072778166550035,
+      "grad_norm": 0.42850643221967644,
+      "learning_rate": 4.600064807876929e-05,
+      "loss": 0.6122,
+      "step": 1481
+    },
+    {
+      "epoch": 0.20741777466759972,
+      "grad_norm": 0.4238910915601293,
+      "learning_rate": 4.599449707777829e-05,
+      "loss": 0.5886,
+      "step": 1482
+    },
+    {
+      "epoch": 0.20755773268019595,
+      "grad_norm": 0.44358369925735036,
+      "learning_rate": 4.5988341762205125e-05,
+      "loss": 0.6162,
+      "step": 1483
+    },
+    {
+      "epoch": 0.20769769069279218,
+      "grad_norm": 0.40620154391670654,
+      "learning_rate": 4.5982182133314765e-05,
+      "loss": 0.5666,
+      "step": 1484
+    },
+    {
+      "epoch": 0.20783764870538837,
+      "grad_norm": 0.4474470089623319,
+      "learning_rate": 4.5976018192373086e-05,
+      "loss": 0.6601,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2079776067179846,
+      "grad_norm": 0.41753387085884,
+      "learning_rate": 4.5969849940646834e-05,
+      "loss": 0.6043,
+      "step": 1486
+    },
+    {
+      "epoch": 0.20811756473058082,
+      "grad_norm": 0.42752671322476543,
+      "learning_rate": 4.596367737940366e-05,
+      "loss": 0.6443,
+      "step": 1487
+    },
+    {
+      "epoch": 0.20825752274317705,
+      "grad_norm": 0.42699115737832166,
+      "learning_rate": 4.595750050991207e-05,
+      "loss": 0.6025,
+      "step": 1488
+    },
+    {
+      "epoch": 0.20839748075577327,
+      "grad_norm": 0.45076801294244706,
+      "learning_rate": 4.595131933344148e-05,
+      "loss": 0.6386,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2085374387683695,
+      "grad_norm": 0.4322528381202846,
+      "learning_rate": 4.594513385126218e-05,
+      "loss": 0.5979,
+      "step": 1490
+    },
+    {
+      "epoch": 0.20867739678096572,
+      "grad_norm": 0.4400100496903223,
+      "learning_rate": 4.593894406464537e-05,
+      "loss": 0.6226,
+      "step": 1491
+    },
+    {
+      "epoch": 0.20881735479356192,
+      "grad_norm": 0.4424527757233975,
+      "learning_rate": 4.593274997486309e-05,
+      "loss": 0.594,
+      "step": 1492
+    },
+    {
+      "epoch": 0.20895731280615815,
+      "grad_norm": 0.42378140749175675,
+      "learning_rate": 4.592655158318829e-05,
+      "loss": 0.6122,
+      "step": 1493
+    },
+    {
+      "epoch": 0.20909727081875437,
+      "grad_norm": 0.43529413412957524,
+      "learning_rate": 4.592034889089482e-05,
+      "loss": 0.6246,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2092372288313506,
+      "grad_norm": 0.4336529824699541,
+      "learning_rate": 4.591414189925739e-05,
+      "loss": 0.5778,
+      "step": 1495
+    },
+    {
+      "epoch": 0.20937718684394682,
+      "grad_norm": 0.4397235048236114,
+      "learning_rate": 4.5907930609551584e-05,
+      "loss": 0.6468,
+      "step": 1496
+    },
+    {
+      "epoch": 0.20951714485654305,
+      "grad_norm": 0.41913091394972535,
+      "learning_rate": 4.59017150230539e-05,
+      "loss": 0.6343,
+      "step": 1497
+    },
+    {
+      "epoch": 0.20965710286913924,
+      "grad_norm": 0.4402774808830381,
+      "learning_rate": 4.58954951410417e-05,
+      "loss": 0.6032,
+      "step": 1498
+    },
+    {
+      "epoch": 0.20979706088173547,
+      "grad_norm": 0.40993950169452836,
+      "learning_rate": 4.588927096479323e-05,
+      "loss": 0.5885,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2099370188943317,
+      "grad_norm": 0.4319112358311373,
+      "learning_rate": 4.5883042495587637e-05,
+      "loss": 0.6243,
+      "step": 1500
+    },
+    {
+      "epoch": 0.21007697690692792,
+      "grad_norm": 0.4312366729366661,
+      "learning_rate": 4.587680973470491e-05,
+      "loss": 0.6154,
+      "step": 1501
+    },
+    {
+      "epoch": 0.21021693491952415,
+      "grad_norm": 0.4262355331023565,
+      "learning_rate": 4.587057268342597e-05,
+      "loss": 0.5766,
+      "step": 1502
+    },
+    {
+      "epoch": 0.21035689293212037,
+      "grad_norm": 0.4142051911565897,
+      "learning_rate": 4.586433134303257e-05,
+      "loss": 0.5607,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2104968509447166,
+      "grad_norm": 0.4286460156115378,
+      "learning_rate": 4.5858085714807384e-05,
+      "loss": 0.6047,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2106368089573128,
+      "grad_norm": 0.42524352191943315,
+      "learning_rate": 4.585183580003395e-05,
+      "loss": 0.6386,
+      "step": 1505
+    },
+    {
+      "epoch": 0.21077676696990902,
+      "grad_norm": 0.4287194392081206,
+      "learning_rate": 4.584558159999668e-05,
+      "loss": 0.5917,
+      "step": 1506
+    },
+    {
+      "epoch": 0.21091672498250524,
+      "grad_norm": 0.43676644200274517,
+      "learning_rate": 4.583932311598089e-05,
+      "loss": 0.6053,
+      "step": 1507
+    },
+    {
+      "epoch": 0.21105668299510147,
+      "grad_norm": 0.4555410229607422,
+      "learning_rate": 4.583306034927275e-05,
+      "loss": 0.5908,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2111966410076977,
+      "grad_norm": 0.45902732656697165,
+      "learning_rate": 4.582679330115933e-05,
+      "loss": 0.6039,
+      "step": 1509
+    },
+    {
+      "epoch": 0.21133659902029392,
+      "grad_norm": 0.4025365324445434,
+      "learning_rate": 4.582052197292856e-05,
+      "loss": 0.581,
+      "step": 1510
+    },
+    {
+      "epoch": 0.21147655703289014,
+      "grad_norm": 0.3996651665725778,
+      "learning_rate": 4.581424636586929e-05,
+      "loss": 0.5978,
+      "step": 1511
+    },
+    {
+      "epoch": 0.21161651504548634,
+      "grad_norm": 0.4015353636671318,
+      "learning_rate": 4.580796648127118e-05,
+      "loss": 0.5862,
+      "step": 1512
+    },
+    {
+      "epoch": 0.21175647305808257,
+      "grad_norm": 0.44031192059103713,
+      "learning_rate": 4.580168232042484e-05,
+      "loss": 0.6566,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2118964310706788,
+      "grad_norm": 0.42413167616716185,
+      "learning_rate": 4.579539388462173e-05,
+      "loss": 0.6368,
+      "step": 1514
+    },
+    {
+      "epoch": 0.21203638908327502,
+      "grad_norm": 0.4357570843764536,
+      "learning_rate": 4.578910117515416e-05,
+      "loss": 0.5574,
+      "step": 1515
+    },
+    {
+      "epoch": 0.21217634709587124,
+      "grad_norm": 0.4443836991084177,
+      "learning_rate": 4.578280419331538e-05,
+      "loss": 0.6129,
+      "step": 1516
+    },
+    {
+      "epoch": 0.21231630510846747,
+      "grad_norm": 0.413705821328803,
+      "learning_rate": 4.5776502940399454e-05,
+      "loss": 0.5594,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2124562631210637,
+      "grad_norm": 0.4167212326575854,
+      "learning_rate": 4.5770197417701365e-05,
+      "loss": 0.6189,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2125962211336599,
+      "grad_norm": 0.4143995752113595,
+      "learning_rate": 4.576388762651697e-05,
+      "loss": 0.611,
+      "step": 1519
+    },
+    {
+      "epoch": 0.21273617914625612,
+      "grad_norm": 0.42271548484517474,
+      "learning_rate": 4.575757356814299e-05,
+      "loss": 0.6259,
+      "step": 1520
+    },
+    {
+      "epoch": 0.21287613715885234,
+      "grad_norm": 0.44271943318208246,
+      "learning_rate": 4.5751255243877015e-05,
+      "loss": 0.6315,
+      "step": 1521
+    },
+    {
+      "epoch": 0.21301609517144857,
+      "grad_norm": 0.4701146597325822,
+      "learning_rate": 4.574493265501755e-05,
+      "loss": 0.5963,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2131560531840448,
+      "grad_norm": 0.4347056827455416,
+      "learning_rate": 4.573860580286392e-05,
+      "loss": 0.5977,
+      "step": 1523
+    },
+    {
+      "epoch": 0.21329601119664102,
+      "grad_norm": 0.42401546255892897,
+      "learning_rate": 4.573227468871639e-05,
+      "loss": 0.6176,
+      "step": 1524
+    },
+    {
+      "epoch": 0.21343596920923724,
+      "grad_norm": 0.4263159567462123,
+      "learning_rate": 4.572593931387604e-05,
+      "loss": 0.5695,
+      "step": 1525
+    },
+    {
+      "epoch": 0.21357592722183344,
+      "grad_norm": 0.4121278643431697,
+      "learning_rate": 4.571959967964488e-05,
+      "loss": 0.5902,
+      "step": 1526
+    },
+    {
+      "epoch": 0.21371588523442966,
+      "grad_norm": 0.4158048949697275,
+      "learning_rate": 4.571325578732575e-05,
+      "loss": 0.626,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2138558432470259,
+      "grad_norm": 0.4183332266945186,
+      "learning_rate": 4.5706907638222385e-05,
+      "loss": 0.6295,
+      "step": 1528
+    },
+    {
+      "epoch": 0.21399580125962211,
+      "grad_norm": 0.42044986223624303,
+      "learning_rate": 4.57005552336394e-05,
+      "loss": 0.5776,
+      "step": 1529
+    },
+    {
+      "epoch": 0.21413575927221834,
+      "grad_norm": 0.4237626863099945,
+      "learning_rate": 4.569419857488228e-05,
+      "loss": 0.602,
+      "step": 1530
+    },
+    {
+      "epoch": 0.21427571728481457,
+      "grad_norm": 0.4235915955493986,
+      "learning_rate": 4.568783766325738e-05,
+      "loss": 0.6088,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2144156752974108,
+      "grad_norm": 0.4429373953075251,
+      "learning_rate": 4.568147250007193e-05,
+      "loss": 0.6244,
+      "step": 1532
+    },
+    {
+      "epoch": 0.214555633310007,
+      "grad_norm": 0.4312825642404986,
+      "learning_rate": 4.567510308663404e-05,
+      "loss": 0.5844,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2146955913226032,
+      "grad_norm": 0.447768930946618,
+      "learning_rate": 4.5668729424252686e-05,
+      "loss": 0.6446,
+      "step": 1534
+    },
+    {
+      "epoch": 0.21483554933519944,
+      "grad_norm": 0.4031120444622804,
+      "learning_rate": 4.5662351514237725e-05,
+      "loss": 0.6134,
+      "step": 1535
+    },
+    {
+      "epoch": 0.21497550734779566,
+      "grad_norm": 0.4102236640467674,
+      "learning_rate": 4.5655969357899874e-05,
+      "loss": 0.5515,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2151154653603919,
+      "grad_norm": 0.4184306996067782,
+      "learning_rate": 4.564958295655074e-05,
+      "loss": 0.5886,
+      "step": 1537
+    },
+    {
+      "epoch": 0.21525542337298811,
+      "grad_norm": 0.4061735300983383,
+      "learning_rate": 4.564319231150278e-05,
+      "loss": 0.5966,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2153953813855843,
+      "grad_norm": 0.4337440887551534,
+      "learning_rate": 4.563679742406935e-05,
+      "loss": 0.5803,
+      "step": 1539
+    },
+    {
+      "epoch": 0.21553533939818054,
+      "grad_norm": 0.4509938114765241,
+      "learning_rate": 4.5630398295564656e-05,
+      "loss": 0.6328,
+      "step": 1540
+    },
+    {
+      "epoch": 0.21567529741077676,
+      "grad_norm": 0.4152270913030905,
+      "learning_rate": 4.562399492730379e-05,
+      "loss": 0.5656,
+      "step": 1541
+    },
+    {
+      "epoch": 0.215815255423373,
+      "grad_norm": 0.42027118148073633,
+      "learning_rate": 4.561758732060271e-05,
+      "loss": 0.6141,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2159552134359692,
+      "grad_norm": 0.4187821883367385,
+      "learning_rate": 4.561117547677824e-05,
+      "loss": 0.5893,
+      "step": 1543
+    },
+    {
+      "epoch": 0.21609517144856544,
+      "grad_norm": 0.45872625216573915,
+      "learning_rate": 4.5604759397148076e-05,
+      "loss": 0.6301,
+      "step": 1544
+    },
+    {
+      "epoch": 0.21623512946116166,
+      "grad_norm": 0.46161111705747715,
+      "learning_rate": 4.559833908303079e-05,
+      "loss": 0.6104,
+      "step": 1545
+    },
+    {
+      "epoch": 0.21637508747375786,
+      "grad_norm": 0.41864006503435036,
+      "learning_rate": 4.559191453574582e-05,
+      "loss": 0.5886,
+      "step": 1546
+    },
+    {
+      "epoch": 0.21651504548635409,
+      "grad_norm": 0.4526739247776462,
+      "learning_rate": 4.5585485756613486e-05,
+      "loss": 0.6058,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2166550034989503,
+      "grad_norm": 0.43967142639314527,
+      "learning_rate": 4.5579052746954955e-05,
+      "loss": 0.5684,
+      "step": 1548
+    },
+    {
+      "epoch": 0.21679496151154654,
+      "grad_norm": 0.6571015727605314,
+      "learning_rate": 4.557261550809228e-05,
+      "loss": 0.6116,
+      "step": 1549
+    },
+    {
+      "epoch": 0.21693491952414276,
+      "grad_norm": 0.42734729410443334,
+      "learning_rate": 4.5566174041348374e-05,
+      "loss": 0.5922,
+      "step": 1550
+    },
+    {
+      "epoch": 0.21707487753673899,
+      "grad_norm": 0.45611337371107136,
+      "learning_rate": 4.555972834804704e-05,
+      "loss": 0.6523,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2172148355493352,
+      "grad_norm": 0.40089163017994156,
+      "learning_rate": 4.5553278429512914e-05,
+      "loss": 0.5866,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2173547935619314,
+      "grad_norm": 0.40519806476960724,
+      "learning_rate": 4.554682428707153e-05,
+      "loss": 0.6494,
+      "step": 1553
+    },
+    {
+      "epoch": 0.21749475157452763,
+      "grad_norm": 0.41428429253275884,
+      "learning_rate": 4.5540365922049275e-05,
+      "loss": 0.6035,
+      "step": 1554
+    },
+    {
+      "epoch": 0.21763470958712386,
+      "grad_norm": 0.4216731233907631,
+      "learning_rate": 4.553390333577342e-05,
+      "loss": 0.5913,
+      "step": 1555
+    },
+    {
+      "epoch": 0.21777466759972008,
+      "grad_norm": 0.4773910626390572,
+      "learning_rate": 4.552743652957208e-05,
+      "loss": 0.5793,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2179146256123163,
+      "grad_norm": 0.4386315899864147,
+      "learning_rate": 4.5520965504774246e-05,
+      "loss": 0.598,
+      "step": 1557
+    },
+    {
+      "epoch": 0.21805458362491253,
+      "grad_norm": 0.41469985672115917,
+      "learning_rate": 4.551449026270979e-05,
+      "loss": 0.5831,
+      "step": 1558
+    },
+    {
+      "epoch": 0.21819454163750876,
+      "grad_norm": 0.41848055045754795,
+      "learning_rate": 4.5508010804709434e-05,
+      "loss": 0.5948,
+      "step": 1559
+    },
+    {
+      "epoch": 0.21833449965010496,
+      "grad_norm": 0.4093401936521381,
+      "learning_rate": 4.550152713210478e-05,
+      "loss": 0.6166,
+      "step": 1560
+    },
+    {
+      "epoch": 0.21847445766270118,
+      "grad_norm": 0.39080382196363417,
+      "learning_rate": 4.5495039246228274e-05,
+      "loss": 0.5615,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2186144156752974,
+      "grad_norm": 0.4548223020948238,
+      "learning_rate": 4.548854714841326e-05,
+      "loss": 0.6275,
+      "step": 1562
+    },
+    {
+      "epoch": 0.21875437368789363,
+      "grad_norm": 0.4423455420604275,
+      "learning_rate": 4.548205083999392e-05,
+      "loss": 0.6453,
+      "step": 1563
+    },
+    {
+      "epoch": 0.21889433170048986,
+      "grad_norm": 0.43733043542862093,
+      "learning_rate": 4.547555032230531e-05,
+      "loss": 0.624,
+      "step": 1564
+    },
+    {
+      "epoch": 0.21903428971308608,
+      "grad_norm": 0.4353839094146027,
+      "learning_rate": 4.546904559668335e-05,
+      "loss": 0.6306,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2191742477256823,
+      "grad_norm": 0.43066663949560247,
+      "learning_rate": 4.546253666446484e-05,
+      "loss": 0.5603,
+      "step": 1566
+    },
+    {
+      "epoch": 0.2193142057382785,
+      "grad_norm": 0.43310274897745943,
+      "learning_rate": 4.545602352698742e-05,
+      "loss": 0.6276,
+      "step": 1567
+    },
+    {
+      "epoch": 0.21945416375087473,
+      "grad_norm": 0.4406898711500198,
+      "learning_rate": 4.544950618558961e-05,
+      "loss": 0.6111,
+      "step": 1568
+    },
+    {
+      "epoch": 0.21959412176347096,
+      "grad_norm": 0.44678818353637145,
+      "learning_rate": 4.544298464161079e-05,
+      "loss": 0.6159,
+      "step": 1569
+    },
+    {
+      "epoch": 0.21973407977606718,
+      "grad_norm": 0.43974972225616304,
+      "learning_rate": 4.54364588963912e-05,
+      "loss": 0.6103,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2198740377886634,
+      "grad_norm": 0.4450084247496375,
+      "learning_rate": 4.542992895127195e-05,
+      "loss": 0.6351,
+      "step": 1571
+    },
+    {
+      "epoch": 0.22001399580125963,
+      "grad_norm": 0.4301181992679436,
+      "learning_rate": 4.5423394807595005e-05,
+      "loss": 0.5929,
+      "step": 1572
+    },
+    {
+      "epoch": 0.22015395381385583,
+      "grad_norm": 0.4354118703683763,
+      "learning_rate": 4.541685646670321e-05,
+      "loss": 0.6417,
+      "step": 1573
+    },
+    {
+      "epoch": 0.22029391182645205,
+      "grad_norm": 0.42020542688000095,
+      "learning_rate": 4.5410313929940244e-05,
+      "loss": 0.5609,
+      "step": 1574
+    },
+    {
+      "epoch": 0.22043386983904828,
+      "grad_norm": 0.405854594592669,
+      "learning_rate": 4.5403767198650683e-05,
+      "loss": 0.6229,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2205738278516445,
+      "grad_norm": 0.4325219516484385,
+      "learning_rate": 4.5397216274179934e-05,
+      "loss": 0.5922,
+      "step": 1576
+    },
+    {
+      "epoch": 0.22071378586424073,
+      "grad_norm": 0.42884858732377196,
+      "learning_rate": 4.539066115787427e-05,
+      "loss": 0.5937,
+      "step": 1577
+    },
+    {
+      "epoch": 0.22085374387683696,
+      "grad_norm": 0.4397303655395074,
+      "learning_rate": 4.5384101851080864e-05,
+      "loss": 0.6208,
+      "step": 1578
+    },
+    {
+      "epoch": 0.22099370188943318,
+      "grad_norm": 0.4196787008923647,
+      "learning_rate": 4.537753835514769e-05,
+      "loss": 0.5936,
+      "step": 1579
+    },
+    {
+      "epoch": 0.22113365990202938,
+      "grad_norm": 0.4450498558445989,
+      "learning_rate": 4.537097067142363e-05,
+      "loss": 0.5795,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2212736179146256,
+      "grad_norm": 0.4533941983871524,
+      "learning_rate": 4.5364398801258396e-05,
+      "loss": 0.5856,
+      "step": 1581
+    },
+    {
+      "epoch": 0.22141357592722183,
+      "grad_norm": 0.4417713842234416,
+      "learning_rate": 4.5357822746002586e-05,
+      "loss": 0.6313,
+      "step": 1582
+    },
+    {
+      "epoch": 0.22155353393981805,
+      "grad_norm": 0.4463677159437907,
+      "learning_rate": 4.535124250700764e-05,
+      "loss": 0.625,
+      "step": 1583
+    },
+    {
+      "epoch": 0.22169349195241428,
+      "grad_norm": 0.40975804674998356,
+      "learning_rate": 4.534465808562587e-05,
+      "loss": 0.5485,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2218334499650105,
+      "grad_norm": 0.4543432488989666,
+      "learning_rate": 4.533806948321044e-05,
+      "loss": 0.6165,
+      "step": 1585
+    },
+    {
+      "epoch": 0.22197340797760673,
+      "grad_norm": 0.44872454774080367,
+      "learning_rate": 4.5331476701115366e-05,
+      "loss": 0.6361,
+      "step": 1586
+    },
+    {
+      "epoch": 0.22211336599020293,
+      "grad_norm": 0.44712318439982773,
+      "learning_rate": 4.532487974069554e-05,
+      "loss": 0.5872,
+      "step": 1587
+    },
+    {
+      "epoch": 0.22225332400279915,
+      "grad_norm": 0.4396204586538607,
+      "learning_rate": 4.53182786033067e-05,
+      "loss": 0.6105,
+      "step": 1588
+    },
+    {
+      "epoch": 0.22239328201539538,
+      "grad_norm": 0.4455977172934511,
+      "learning_rate": 4.531167329030545e-05,
+      "loss": 0.624,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2225332400279916,
+      "grad_norm": 0.4556263334519999,
+      "learning_rate": 4.530506380304925e-05,
+      "loss": 0.6265,
+      "step": 1590
+    },
+    {
+      "epoch": 0.22267319804058783,
+      "grad_norm": 0.42577140585892975,
+      "learning_rate": 4.529845014289642e-05,
+      "loss": 0.6105,
+      "step": 1591
+    },
+    {
+      "epoch": 0.22281315605318405,
+      "grad_norm": 0.4180730424078042,
+      "learning_rate": 4.529183231120612e-05,
+      "loss": 0.5895,
+      "step": 1592
+    },
+    {
+      "epoch": 0.22295311406578028,
+      "grad_norm": 0.40699329023671954,
+      "learning_rate": 4.528521030933839e-05,
+      "loss": 0.6161,
+      "step": 1593
+    },
+    {
+      "epoch": 0.22309307207837648,
+      "grad_norm": 0.4222068544403408,
+      "learning_rate": 4.5278584138654116e-05,
+      "loss": 0.5605,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2232330300909727,
+      "grad_norm": 0.4095845718488545,
+      "learning_rate": 4.527195380051505e-05,
+      "loss": 0.6219,
+      "step": 1595
+    },
+    {
+      "epoch": 0.22337298810356893,
+      "grad_norm": 0.4314280489339947,
+      "learning_rate": 4.526531929628379e-05,
+      "loss": 0.5701,
+      "step": 1596
+    },
+    {
+      "epoch": 0.22351294611616515,
+      "grad_norm": 0.4074259435374443,
+      "learning_rate": 4.525868062732379e-05,
+      "loss": 0.5631,
+      "step": 1597
+    },
+    {
+      "epoch": 0.22365290412876138,
+      "grad_norm": 0.4532813167302565,
+      "learning_rate": 4.5252037794999375e-05,
+      "loss": 0.589,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2237928621413576,
+      "grad_norm": 0.4513647564836139,
+      "learning_rate": 4.52453908006757e-05,
+      "loss": 0.5948,
+      "step": 1599
+    },
+    {
+      "epoch": 0.22393282015395383,
+      "grad_norm": 0.41293390054677387,
+      "learning_rate": 4.52387396457188e-05,
+      "loss": 0.585,
+      "step": 1600
+    },
+    {
+      "epoch": 0.22407277816655002,
+      "grad_norm": 0.43716711655654145,
+      "learning_rate": 4.523208433149555e-05,
+      "loss": 0.5717,
+      "step": 1601
+    },
+    {
+      "epoch": 0.22421273617914625,
+      "grad_norm": 0.4258916507285024,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.6024,
+      "step": 1602
+    },
+    {
+      "epoch": 0.22435269419174247,
+      "grad_norm": 0.42215745803677884,
+      "learning_rate": 4.52187612307218e-05,
+      "loss": 0.564,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2244926522043387,
+      "grad_norm": 0.4429228384575338,
+      "learning_rate": 4.521209344690933e-05,
+      "loss": 0.6278,
+      "step": 1604
+    },
+    {
+      "epoch": 0.22463261021693492,
+      "grad_norm": 0.4784015488798018,
+      "learning_rate": 4.5205421509306576e-05,
+      "loss": 0.6018,
+      "step": 1605
+    },
+    {
+      "epoch": 0.22477256822953115,
+      "grad_norm": 0.43024752872126554,
+      "learning_rate": 4.519874541928469e-05,
+      "loss": 0.6077,
+      "step": 1606
+    },
+    {
+      "epoch": 0.22491252624212738,
+      "grad_norm": 0.41398612904638815,
+      "learning_rate": 4.519206517821567e-05,
+      "loss": 0.5996,
+      "step": 1607
+    },
+    {
+      "epoch": 0.22505248425472357,
+      "grad_norm": 0.4141848810116437,
+      "learning_rate": 4.5185380787472384e-05,
+      "loss": 0.6349,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2251924422673198,
+      "grad_norm": 0.43155153008345165,
+      "learning_rate": 4.5178692248428536e-05,
+      "loss": 0.5882,
+      "step": 1609
+    },
+    {
+      "epoch": 0.22533240027991602,
+      "grad_norm": 0.43080194308434416,
+      "learning_rate": 4.517199956245869e-05,
+      "loss": 0.6111,
+      "step": 1610
+    },
+    {
+      "epoch": 0.22547235829251225,
+      "grad_norm": 0.42258541276252165,
+      "learning_rate": 4.516530273093825e-05,
+      "loss": 0.5848,
+      "step": 1611
+    },
+    {
+      "epoch": 0.22561231630510847,
+      "grad_norm": 0.43019795244061954,
+      "learning_rate": 4.5158601755243505e-05,
+      "loss": 0.5885,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2257522743177047,
+      "grad_norm": 0.4034883685193691,
+      "learning_rate": 4.5151896636751556e-05,
+      "loss": 0.6009,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2258922323303009,
+      "grad_norm": 0.41958257656866,
+      "learning_rate": 4.514518737684038e-05,
+      "loss": 0.5881,
+      "step": 1614
+    },
+    {
+      "epoch": 0.22603219034289712,
+      "grad_norm": 0.4074178901403192,
+      "learning_rate": 4.513847397688879e-05,
+      "loss": 0.5606,
+      "step": 1615
+    },
+    {
+      "epoch": 0.22617214835549335,
+      "grad_norm": 0.44325729803840913,
+      "learning_rate": 4.513175643827647e-05,
+      "loss": 0.6336,
+      "step": 1616
+    },
+    {
+      "epoch": 0.22631210636808957,
+      "grad_norm": 0.43312998668445646,
+      "learning_rate": 4.5125034762383936e-05,
+      "loss": 0.607,
+      "step": 1617
+    },
+    {
+      "epoch": 0.2264520643806858,
+      "grad_norm": 0.40174739442199564,
+      "learning_rate": 4.511830895059255e-05,
+      "loss": 0.5623,
+      "step": 1618
+    },
+    {
+      "epoch": 0.22659202239328202,
+      "grad_norm": 0.43059705945767945,
+      "learning_rate": 4.511157900428456e-05,
+      "loss": 0.5922,
+      "step": 1619
+    },
+    {
+      "epoch": 0.22673198040587825,
+      "grad_norm": 0.42985175056584884,
+      "learning_rate": 4.5104844924843016e-05,
+      "loss": 0.5719,
+      "step": 1620
+    },
+    {
+      "epoch": 0.22687193841847444,
+      "grad_norm": 0.4503501491245042,
+      "learning_rate": 4.5098106713651846e-05,
+      "loss": 0.5888,
+      "step": 1621
+    },
+    {
+      "epoch": 0.22701189643107067,
+      "grad_norm": 0.4306003466493788,
+      "learning_rate": 4.509136437209582e-05,
+      "loss": 0.5624,
+      "step": 1622
+    },
+    {
+      "epoch": 0.2271518544436669,
+      "grad_norm": 0.413588915370572,
+      "learning_rate": 4.508461790156056e-05,
+      "loss": 0.5822,
+      "step": 1623
+    },
+    {
+      "epoch": 0.22729181245626312,
+      "grad_norm": 0.43438032996872095,
+      "learning_rate": 4.5077867303432546e-05,
+      "loss": 0.6452,
+      "step": 1624
+    },
+    {
+      "epoch": 0.22743177046885935,
+      "grad_norm": 0.4364680489398804,
+      "learning_rate": 4.5071112579099074e-05,
+      "loss": 0.6027,
+      "step": 1625
+    },
+    {
+      "epoch": 0.22757172848145557,
+      "grad_norm": 0.44709606835984195,
+      "learning_rate": 4.5064353729948315e-05,
+      "loss": 0.6131,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2277116864940518,
+      "grad_norm": 0.43519671900411855,
+      "learning_rate": 4.505759075736929e-05,
+      "loss": 0.591,
+      "step": 1627
+    },
+    {
+      "epoch": 0.227851644506648,
+      "grad_norm": 0.41612778582560767,
+      "learning_rate": 4.505082366275184e-05,
+      "loss": 0.5877,
+      "step": 1628
+    },
+    {
+      "epoch": 0.22799160251924422,
+      "grad_norm": 0.4218216656661162,
+      "learning_rate": 4.504405244748669e-05,
+      "loss": 0.5534,
+      "step": 1629
+    },
+    {
+      "epoch": 0.22813156053184044,
+      "grad_norm": 0.42964453812050457,
+      "learning_rate": 4.503727711296538e-05,
+      "loss": 0.6389,
+      "step": 1630
+    },
+    {
+      "epoch": 0.22827151854443667,
+      "grad_norm": 0.42585594110280106,
+      "learning_rate": 4.503049766058033e-05,
+      "loss": 0.6084,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2284114765570329,
+      "grad_norm": 0.423186467723224,
+      "learning_rate": 4.5023714091724756e-05,
+      "loss": 0.6029,
+      "step": 1632
+    },
+    {
+      "epoch": 0.22855143456962912,
+      "grad_norm": 0.43009182540969443,
+      "learning_rate": 4.5016926407792774e-05,
+      "loss": 0.6404,
+      "step": 1633
+    },
+    {
+      "epoch": 0.22869139258222534,
+      "grad_norm": 0.4291224518654508,
+      "learning_rate": 4.501013461017931e-05,
+      "loss": 0.6167,
+      "step": 1634
+    },
+    {
+      "epoch": 0.22883135059482154,
+      "grad_norm": 0.41098068932209736,
+      "learning_rate": 4.500333870028016e-05,
+      "loss": 0.581,
+      "step": 1635
+    },
+    {
+      "epoch": 0.22897130860741777,
+      "grad_norm": 0.4297903813744973,
+      "learning_rate": 4.499653867949194e-05,
+      "loss": 0.5923,
+      "step": 1636
+    },
+    {
+      "epoch": 0.229111266620014,
+      "grad_norm": 0.4194573186329626,
+      "learning_rate": 4.4989734549212125e-05,
+      "loss": 0.5893,
+      "step": 1637
+    },
+    {
+      "epoch": 0.22925122463261022,
+      "grad_norm": 0.44155487858058556,
+      "learning_rate": 4.498292631083904e-05,
+      "loss": 0.5734,
+      "step": 1638
+    },
+    {
+      "epoch": 0.22939118264520644,
+      "grad_norm": 0.4463968226022601,
+      "learning_rate": 4.4976113965771835e-05,
+      "loss": 0.6054,
+      "step": 1639
+    },
+    {
+      "epoch": 0.22953114065780267,
+      "grad_norm": 0.43124302519709157,
+      "learning_rate": 4.496929751541054e-05,
+      "loss": 0.5788,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2296710986703989,
+      "grad_norm": 0.4206213171563187,
+      "learning_rate": 4.4962476961155976e-05,
+      "loss": 0.5675,
+      "step": 1641
+    },
+    {
+      "epoch": 0.2298110566829951,
+      "grad_norm": 0.4696499710097156,
+      "learning_rate": 4.495565230440985e-05,
+      "loss": 0.6295,
+      "step": 1642
+    },
+    {
+      "epoch": 0.22995101469559132,
+      "grad_norm": 0.3979651141981672,
+      "learning_rate": 4.494882354657469e-05,
+      "loss": 0.584,
+      "step": 1643
+    },
+    {
+      "epoch": 0.23009097270818754,
+      "grad_norm": 0.5843917120508454,
+      "learning_rate": 4.4941990689053886e-05,
+      "loss": 0.5725,
+      "step": 1644
+    },
+    {
+      "epoch": 0.23023093072078377,
+      "grad_norm": 0.4077407210872767,
+      "learning_rate": 4.4935153733251656e-05,
+      "loss": 0.5744,
+      "step": 1645
+    },
+    {
+      "epoch": 0.23037088873338,
+      "grad_norm": 0.4392056225941937,
+      "learning_rate": 4.4928312680573064e-05,
+      "loss": 0.6168,
+      "step": 1646
+    },
+    {
+      "epoch": 0.23051084674597622,
+      "grad_norm": 0.4356793893175191,
+      "learning_rate": 4.492146753242401e-05,
+      "loss": 0.5972,
+      "step": 1647
+    },
+    {
+      "epoch": 0.23065080475857244,
+      "grad_norm": 0.42512045974302,
+      "learning_rate": 4.491461829021125e-05,
+      "loss": 0.6265,
+      "step": 1648
+    },
+    {
+      "epoch": 0.23079076277116864,
+      "grad_norm": 0.4378392920345575,
+      "learning_rate": 4.490776495534237e-05,
+      "loss": 0.593,
+      "step": 1649
+    },
+    {
+      "epoch": 0.23093072078376486,
+      "grad_norm": 11.092127253150068,
+      "learning_rate": 4.49009075292258e-05,
+      "loss": 0.5982,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2310706787963611,
+      "grad_norm": 0.4211855457116042,
+      "learning_rate": 4.489404601327081e-05,
+      "loss": 0.5885,
+      "step": 1651
+    },
+    {
+      "epoch": 0.23121063680895731,
+      "grad_norm": 0.44452065502823107,
+      "learning_rate": 4.48871804088875e-05,
+      "loss": 0.6286,
+      "step": 1652
+    },
+    {
+      "epoch": 0.23135059482155354,
+      "grad_norm": 0.44745394315873227,
+      "learning_rate": 4.488031071748684e-05,
+      "loss": 0.6036,
+      "step": 1653
+    },
+    {
+      "epoch": 0.23149055283414977,
+      "grad_norm": 0.41971430239918867,
+      "learning_rate": 4.487343694048061e-05,
+      "loss": 0.5691,
+      "step": 1654
+    },
+    {
+      "epoch": 0.23163051084674596,
+      "grad_norm": 0.43793140979246065,
+      "learning_rate": 4.4866559079281447e-05,
+      "loss": 0.6025,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2317704688593422,
+      "grad_norm": 0.45584544502664887,
+      "learning_rate": 4.485967713530281e-05,
+      "loss": 0.5985,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2319104268719384,
+      "grad_norm": 0.4393986821966947,
+      "learning_rate": 4.485279110995903e-05,
+      "loss": 0.6181,
+      "step": 1657
+    },
+    {
+      "epoch": 0.23205038488453464,
+      "grad_norm": 0.41398671250331304,
+      "learning_rate": 4.4845901004665234e-05,
+      "loss": 0.5659,
+      "step": 1658
+    },
+    {
+      "epoch": 0.23219034289713086,
+      "grad_norm": 0.40845641035568697,
+      "learning_rate": 4.483900682083742e-05,
+      "loss": 0.6004,
+      "step": 1659
+    },
+    {
+      "epoch": 0.2323303009097271,
+      "grad_norm": 0.41163784471475046,
+      "learning_rate": 4.4832108559892406e-05,
+      "loss": 0.5676,
+      "step": 1660
+    },
+    {
+      "epoch": 0.23247025892232331,
+      "grad_norm": 0.40520538312381466,
+      "learning_rate": 4.4825206223247855e-05,
+      "loss": 0.5709,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2326102169349195,
+      "grad_norm": 0.4052889024628849,
+      "learning_rate": 4.481829981232227e-05,
+      "loss": 0.6205,
+      "step": 1662
+    },
+    {
+      "epoch": 0.23275017494751574,
+      "grad_norm": 0.44570039573163067,
+      "learning_rate": 4.481138932853499e-05,
+      "loss": 0.6339,
+      "step": 1663
+    },
+    {
+      "epoch": 0.23289013296011196,
+      "grad_norm": 0.42218605234809253,
+      "learning_rate": 4.480447477330619e-05,
+      "loss": 0.582,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2330300909727082,
+      "grad_norm": 0.4139552246868811,
+      "learning_rate": 4.479755614805688e-05,
+      "loss": 0.6185,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2331700489853044,
+      "grad_norm": 0.438909444526491,
+      "learning_rate": 4.4790633454208904e-05,
+      "loss": 0.582,
+      "step": 1666
+    },
+    {
+      "epoch": 0.23331000699790064,
+      "grad_norm": 0.41733390361062983,
+      "learning_rate": 4.478370669318494e-05,
+      "loss": 0.5954,
+      "step": 1667
+    },
+    {
+      "epoch": 0.23344996501049686,
+      "grad_norm": 0.43169892360959883,
+      "learning_rate": 4.477677586640854e-05,
+      "loss": 0.6044,
+      "step": 1668
+    },
+    {
+      "epoch": 0.23358992302309306,
+      "grad_norm": 0.41695287451508345,
+      "learning_rate": 4.4769840975304014e-05,
+      "loss": 0.5781,
+      "step": 1669
+    },
+    {
+      "epoch": 0.23372988103568929,
+      "grad_norm": 0.4287731524988866,
+      "learning_rate": 4.476290202129658e-05,
+      "loss": 0.6271,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2338698390482855,
+      "grad_norm": 0.41617003120095575,
+      "learning_rate": 4.4755959005812256e-05,
+      "loss": 0.5917,
+      "step": 1671
+    },
+    {
+      "epoch": 0.23400979706088174,
+      "grad_norm": 0.4317600967426944,
+      "learning_rate": 4.474901193027791e-05,
+      "loss": 0.6371,
+      "step": 1672
+    },
+    {
+      "epoch": 0.23414975507347796,
+      "grad_norm": 0.4245879897664653,
+      "learning_rate": 4.474206079612122e-05,
+      "loss": 0.6104,
+      "step": 1673
+    },
+    {
+      "epoch": 0.23428971308607419,
+      "grad_norm": 0.4493803376375899,
+      "learning_rate": 4.4735105604770735e-05,
+      "loss": 0.5892,
+      "step": 1674
+    },
+    {
+      "epoch": 0.2344296710986704,
+      "grad_norm": 0.4250874727480562,
+      "learning_rate": 4.4728146357655795e-05,
+      "loss": 0.5675,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2345696291112666,
+      "grad_norm": 0.4374362501965964,
+      "learning_rate": 4.4721183056206614e-05,
+      "loss": 0.6075,
+      "step": 1676
+    },
+    {
+      "epoch": 0.23470958712386283,
+      "grad_norm": 0.41760478685320906,
+      "learning_rate": 4.4714215701854225e-05,
+      "loss": 0.5986,
+      "step": 1677
+    },
+    {
+      "epoch": 0.23484954513645906,
+      "grad_norm": 0.42350699224694727,
+      "learning_rate": 4.4707244296030464e-05,
+      "loss": 0.6486,
+      "step": 1678
+    },
+    {
+      "epoch": 0.23498950314905528,
+      "grad_norm": 0.4256916273457718,
+      "learning_rate": 4.4700268840168045e-05,
+      "loss": 0.6174,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2351294611616515,
+      "grad_norm": 0.42531800370790246,
+      "learning_rate": 4.469328933570051e-05,
+      "loss": 0.551,
+      "step": 1680
+    },
+    {
+      "epoch": 0.23526941917424773,
+      "grad_norm": 0.4181940913192066,
+      "learning_rate": 4.468630578406218e-05,
+      "loss": 0.5856,
+      "step": 1681
+    },
+    {
+      "epoch": 0.23540937718684396,
+      "grad_norm": 0.4105263397513692,
+      "learning_rate": 4.467931818668827e-05,
+      "loss": 0.5633,
+      "step": 1682
+    },
+    {
+      "epoch": 0.23554933519944016,
+      "grad_norm": 0.4081654944687357,
+      "learning_rate": 4.46723265450148e-05,
+      "loss": 0.5808,
+      "step": 1683
+    },
+    {
+      "epoch": 0.23568929321203638,
+      "grad_norm": 0.3907211987137891,
+      "learning_rate": 4.466533086047861e-05,
+      "loss": 0.5571,
+      "step": 1684
+    },
+    {
+      "epoch": 0.2358292512246326,
+      "grad_norm": 0.4250060420027131,
+      "learning_rate": 4.465833113451741e-05,
+      "loss": 0.6208,
+      "step": 1685
+    },
+    {
+      "epoch": 0.23596920923722883,
+      "grad_norm": 0.3937263336098636,
+      "learning_rate": 4.465132736856969e-05,
+      "loss": 0.5621,
+      "step": 1686
+    },
+    {
+      "epoch": 0.23610916724982506,
+      "grad_norm": 0.41965579506262274,
+      "learning_rate": 4.464431956407481e-05,
+      "loss": 0.6031,
+      "step": 1687
+    },
+    {
+      "epoch": 0.23624912526242128,
+      "grad_norm": 0.426646222987036,
+      "learning_rate": 4.463730772247293e-05,
+      "loss": 0.6033,
+      "step": 1688
+    },
+    {
+      "epoch": 0.2363890832750175,
+      "grad_norm": 0.4439770950702544,
+      "learning_rate": 4.463029184520507e-05,
+      "loss": 0.604,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2365290412876137,
+      "grad_norm": 0.4761499260018663,
+      "learning_rate": 4.4623271933713065e-05,
+      "loss": 0.6307,
+      "step": 1690
+    },
+    {
+      "epoch": 0.23666899930020993,
+      "grad_norm": 0.4335225372190029,
+      "learning_rate": 4.4616247989439565e-05,
+      "loss": 0.576,
+      "step": 1691
+    },
+    {
+      "epoch": 0.23680895731280616,
+      "grad_norm": 0.43316503301129894,
+      "learning_rate": 4.4609220013828065e-05,
+      "loss": 0.6275,
+      "step": 1692
+    },
+    {
+      "epoch": 0.23694891532540238,
+      "grad_norm": 0.4313464290288283,
+      "learning_rate": 4.46021880083229e-05,
+      "loss": 0.6067,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2370888733379986,
+      "grad_norm": 0.41535036276436027,
+      "learning_rate": 4.45951519743692e-05,
+      "loss": 0.6192,
+      "step": 1694
+    },
+    {
+      "epoch": 0.23722883135059483,
+      "grad_norm": 0.4310228331540852,
+      "learning_rate": 4.4588111913412945e-05,
+      "loss": 0.5915,
+      "step": 1695
+    },
+    {
+      "epoch": 0.23736878936319103,
+      "grad_norm": 0.42874110588166797,
+      "learning_rate": 4.458106782690094e-05,
+      "loss": 0.5912,
+      "step": 1696
+    },
+    {
+      "epoch": 0.23750874737578725,
+      "grad_norm": 0.43089863664872624,
+      "learning_rate": 4.4574019716280824e-05,
+      "loss": 0.613,
+      "step": 1697
+    },
+    {
+      "epoch": 0.23764870538838348,
+      "grad_norm": 0.4181283245016506,
+      "learning_rate": 4.4566967583001046e-05,
+      "loss": 0.5918,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2377886634009797,
+      "grad_norm": 0.4521121054953592,
+      "learning_rate": 4.4559911428510895e-05,
+      "loss": 0.6326,
+      "step": 1699
+    },
+    {
+      "epoch": 0.23792862141357593,
+      "grad_norm": 0.41538878241716964,
+      "learning_rate": 4.4552851254260484e-05,
+      "loss": 0.5994,
+      "step": 1700
+    },
+    {
+      "epoch": 0.23806857942617216,
+      "grad_norm": 0.4693907751456843,
+      "learning_rate": 4.454578706170075e-05,
+      "loss": 0.6444,
+      "step": 1701
+    },
+    {
+      "epoch": 0.23820853743876838,
+      "grad_norm": 0.47587489105004577,
+      "learning_rate": 4.453871885228345e-05,
+      "loss": 0.6053,
+      "step": 1702
+    },
+    {
+      "epoch": 0.23834849545136458,
+      "grad_norm": 0.44284115011605213,
+      "learning_rate": 4.4531646627461175e-05,
+      "loss": 0.6107,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2384884534639608,
+      "grad_norm": 0.42954706530716213,
+      "learning_rate": 4.452457038868735e-05,
+      "loss": 0.6082,
+      "step": 1704
+    },
+    {
+      "epoch": 0.23862841147655703,
+      "grad_norm": 0.4558514723051049,
+      "learning_rate": 4.4517490137416196e-05,
+      "loss": 0.5926,
+      "step": 1705
+    },
+    {
+      "epoch": 0.23876836948915325,
+      "grad_norm": 0.41271838706018094,
+      "learning_rate": 4.451040587510279e-05,
+      "loss": 0.5552,
+      "step": 1706
+    },
+    {
+      "epoch": 0.23890832750174948,
+      "grad_norm": 0.4234721479869862,
+      "learning_rate": 4.450331760320302e-05,
+      "loss": 0.5907,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2390482855143457,
+      "grad_norm": 0.4301944314412581,
+      "learning_rate": 4.449622532317359e-05,
+      "loss": 0.5637,
+      "step": 1708
+    },
+    {
+      "epoch": 0.23918824352694193,
+      "grad_norm": 0.3992534678208053,
+      "learning_rate": 4.448912903647203e-05,
+      "loss": 0.5746,
+      "step": 1709
+    },
+    {
+      "epoch": 0.23932820153953813,
+      "grad_norm": 0.4151340108937535,
+      "learning_rate": 4.448202874455673e-05,
+      "loss": 0.5953,
+      "step": 1710
+    },
+    {
+      "epoch": 0.23946815955213435,
+      "grad_norm": 0.42307801303730386,
+      "learning_rate": 4.447492444888682e-05,
+      "loss": 0.5985,
+      "step": 1711
+    },
+    {
+      "epoch": 0.23960811756473058,
+      "grad_norm": 0.4254157003906138,
+      "learning_rate": 4.446781615092235e-05,
+      "loss": 0.5706,
+      "step": 1712
+    },
+    {
+      "epoch": 0.2397480755773268,
+      "grad_norm": 0.4221420053361986,
+      "learning_rate": 4.446070385212414e-05,
+      "loss": 0.5865,
+      "step": 1713
+    },
+    {
+      "epoch": 0.23988803358992303,
+      "grad_norm": 0.44169866453949225,
+      "learning_rate": 4.445358755395382e-05,
+      "loss": 0.6131,
+      "step": 1714
+    },
+    {
+      "epoch": 0.24002799160251925,
+      "grad_norm": 0.41028257323034817,
+      "learning_rate": 4.444646725787387e-05,
+      "loss": 0.6391,
+      "step": 1715
+    },
+    {
+      "epoch": 0.24016794961511548,
+      "grad_norm": 0.42430000545950824,
+      "learning_rate": 4.4439342965347595e-05,
+      "loss": 0.5713,
+      "step": 1716
+    },
+    {
+      "epoch": 0.24030790762771168,
+      "grad_norm": 0.40781271006014086,
+      "learning_rate": 4.4432214677839095e-05,
+      "loss": 0.5508,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2404478656403079,
+      "grad_norm": 0.4231812063581877,
+      "learning_rate": 4.442508239681331e-05,
+      "loss": 0.59,
+      "step": 1718
+    },
+    {
+      "epoch": 0.24058782365290413,
+      "grad_norm": 0.4175498825745329,
+      "learning_rate": 4.4417946123736e-05,
+      "loss": 0.6073,
+      "step": 1719
+    },
+    {
+      "epoch": 0.24072778166550035,
+      "grad_norm": 0.43499143607770346,
+      "learning_rate": 4.4410805860073736e-05,
+      "loss": 0.562,
+      "step": 1720
+    },
+    {
+      "epoch": 0.24086773967809658,
+      "grad_norm": 0.4405506003884184,
+      "learning_rate": 4.440366160729392e-05,
+      "loss": 0.5963,
+      "step": 1721
+    },
+    {
+      "epoch": 0.2410076976906928,
+      "grad_norm": 0.4048128591600598,
+      "learning_rate": 4.4396513366864765e-05,
+      "loss": 0.5533,
+      "step": 1722
+    },
+    {
+      "epoch": 0.24114765570328903,
+      "grad_norm": 0.41424145471376667,
+      "learning_rate": 4.4389361140255306e-05,
+      "loss": 0.5814,
+      "step": 1723
+    },
+    {
+      "epoch": 0.24128761371588522,
+      "grad_norm": 0.41036719244303405,
+      "learning_rate": 4.43822049289354e-05,
+      "loss": 0.57,
+      "step": 1724
+    },
+    {
+      "epoch": 0.24142757172848145,
+      "grad_norm": 0.41201535335271977,
+      "learning_rate": 4.4375044734375724e-05,
+      "loss": 0.5484,
+      "step": 1725
+    },
+    {
+      "epoch": 0.24156752974107767,
+      "grad_norm": 0.4111942850121887,
+      "learning_rate": 4.436788055804777e-05,
+      "loss": 0.5706,
+      "step": 1726
+    },
+    {
+      "epoch": 0.2417074877536739,
+      "grad_norm": 0.4519161404524503,
+      "learning_rate": 4.436071240142383e-05,
+      "loss": 0.6018,
+      "step": 1727
+    },
+    {
+      "epoch": 0.24184744576627012,
+      "grad_norm": 0.4403101160025735,
+      "learning_rate": 4.4353540265977064e-05,
+      "loss": 0.6072,
+      "step": 1728
+    },
+    {
+      "epoch": 0.24198740377886635,
+      "grad_norm": 0.49703818437945607,
+      "learning_rate": 4.4346364153181397e-05,
+      "loss": 0.6081,
+      "step": 1729
+    },
+    {
+      "epoch": 0.24212736179146255,
+      "grad_norm": 0.41243187874993914,
+      "learning_rate": 4.433918406451161e-05,
+      "loss": 0.5901,
+      "step": 1730
+    },
+    {
+      "epoch": 0.24226731980405877,
+      "grad_norm": 0.40745642609215477,
+      "learning_rate": 4.433200000144326e-05,
+      "loss": 0.5789,
+      "step": 1731
+    },
+    {
+      "epoch": 0.242407277816655,
+      "grad_norm": 0.42520024238957543,
+      "learning_rate": 4.4324811965452764e-05,
+      "loss": 0.5929,
+      "step": 1732
+    },
+    {
+      "epoch": 0.24254723582925122,
+      "grad_norm": 0.4483762787233949,
+      "learning_rate": 4.431761995801733e-05,
+      "loss": 0.6412,
+      "step": 1733
+    },
+    {
+      "epoch": 0.24268719384184745,
+      "grad_norm": 0.42388738271936166,
+      "learning_rate": 4.431042398061499e-05,
+      "loss": 0.6215,
+      "step": 1734
+    },
+    {
+      "epoch": 0.24282715185444367,
+      "grad_norm": 0.42006944785214295,
+      "learning_rate": 4.430322403472459e-05,
+      "loss": 0.6243,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2429671098670399,
+      "grad_norm": 0.41526878760328473,
+      "learning_rate": 4.429602012182579e-05,
+      "loss": 0.5868,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2431070678796361,
+      "grad_norm": 0.40029320601405294,
+      "learning_rate": 4.428881224339907e-05,
+      "loss": 0.6238,
+      "step": 1737
+    },
+    {
+      "epoch": 0.24324702589223232,
+      "grad_norm": 0.424773544460618,
+      "learning_rate": 4.4281600400925725e-05,
+      "loss": 0.6287,
+      "step": 1738
+    },
+    {
+      "epoch": 0.24338698390482855,
+      "grad_norm": 0.43089065054686443,
+      "learning_rate": 4.4274384595887855e-05,
+      "loss": 0.579,
+      "step": 1739
+    },
+    {
+      "epoch": 0.24352694191742477,
+      "grad_norm": 0.39335625658540807,
+      "learning_rate": 4.426716482976838e-05,
+      "loss": 0.5945,
+      "step": 1740
+    },
+    {
+      "epoch": 0.243666899930021,
+      "grad_norm": 0.4321714579592816,
+      "learning_rate": 4.425994110405105e-05,
+      "loss": 0.6165,
+      "step": 1741
+    },
+    {
+      "epoch": 0.24380685794261722,
+      "grad_norm": 0.3868262436570827,
+      "learning_rate": 4.425271342022039e-05,
+      "loss": 0.6322,
+      "step": 1742
+    },
+    {
+      "epoch": 0.24394681595521345,
+      "grad_norm": 0.4194791382722549,
+      "learning_rate": 4.424548177976179e-05,
+      "loss": 0.6155,
+      "step": 1743
+    },
+    {
+      "epoch": 0.24408677396780964,
+      "grad_norm": 0.4118315161780323,
+      "learning_rate": 4.42382461841614e-05,
+      "loss": 0.6144,
+      "step": 1744
+    },
+    {
+      "epoch": 0.24422673198040587,
+      "grad_norm": 0.40215952380747805,
+      "learning_rate": 4.4231006634906224e-05,
+      "loss": 0.5487,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2443666899930021,
+      "grad_norm": 0.42510463260619374,
+      "learning_rate": 4.422376313348405e-05,
+      "loss": 0.5674,
+      "step": 1746
+    },
+    {
+      "epoch": 0.24450664800559832,
+      "grad_norm": 0.4220078299124082,
+      "learning_rate": 4.4216515681383505e-05,
+      "loss": 0.6159,
+      "step": 1747
+    },
+    {
+      "epoch": 0.24464660601819455,
+      "grad_norm": 0.42050182431963373,
+      "learning_rate": 4.420926428009401e-05,
+      "loss": 0.6066,
+      "step": 1748
+    },
+    {
+      "epoch": 0.24478656403079077,
+      "grad_norm": 0.43689368971554526,
+      "learning_rate": 4.4202008931105795e-05,
+      "loss": 0.5568,
+      "step": 1749
+    },
+    {
+      "epoch": 0.244926522043387,
+      "grad_norm": 0.43739145581183353,
+      "learning_rate": 4.4194749635909924e-05,
+      "loss": 0.5984,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2450664800559832,
+      "grad_norm": 0.4595283770036913,
+      "learning_rate": 4.4187486395998235e-05,
+      "loss": 0.6204,
+      "step": 1751
+    },
+    {
+      "epoch": 0.24520643806857942,
+      "grad_norm": 0.4332223020604194,
+      "learning_rate": 4.4180219212863404e-05,
+      "loss": 0.6114,
+      "step": 1752
+    },
+    {
+      "epoch": 0.24534639608117564,
+      "grad_norm": 0.44490455320321953,
+      "learning_rate": 4.417294808799892e-05,
+      "loss": 0.5771,
+      "step": 1753
+    },
+    {
+      "epoch": 0.24548635409377187,
+      "grad_norm": 0.42354862515455083,
+      "learning_rate": 4.416567302289907e-05,
+      "loss": 0.6092,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2456263121063681,
+      "grad_norm": 0.4252852477883581,
+      "learning_rate": 4.4158394019058944e-05,
+      "loss": 0.6034,
+      "step": 1755
+    },
+    {
+      "epoch": 0.24576627011896432,
+      "grad_norm": 0.4376952545699014,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.623,
+      "step": 1756
+    },
+    {
+      "epoch": 0.24590622813156054,
+      "grad_norm": 0.41667144416831586,
+      "learning_rate": 4.414382420114233e-05,
+      "loss": 0.6042,
+      "step": 1757
+    },
+    {
+      "epoch": 0.24604618614415674,
+      "grad_norm": 0.42933664014516487,
+      "learning_rate": 4.413653339006008e-05,
+      "loss": 0.6022,
+      "step": 1758
+    },
+    {
+      "epoch": 0.24618614415675297,
+      "grad_norm": 0.4162921306132023,
+      "learning_rate": 4.4129238646226055e-05,
+      "loss": 0.5887,
+      "step": 1759
+    },
+    {
+      "epoch": 0.2463261021693492,
+      "grad_norm": 0.4306173347777482,
+      "learning_rate": 4.4121939971139403e-05,
+      "loss": 0.5574,
+      "step": 1760
+    },
+    {
+      "epoch": 0.24646606018194542,
+      "grad_norm": 0.43621303487831553,
+      "learning_rate": 4.411463736630006e-05,
+      "loss": 0.5876,
+      "step": 1761
+    },
+    {
+      "epoch": 0.24660601819454164,
+      "grad_norm": 0.419540410185526,
+      "learning_rate": 4.410733083320879e-05,
+      "loss": 0.599,
+      "step": 1762
+    },
+    {
+      "epoch": 0.24674597620713787,
+      "grad_norm": 0.4510539099176357,
+      "learning_rate": 4.4100020373367166e-05,
+      "loss": 0.6325,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2468859342197341,
+      "grad_norm": 0.42372431177296327,
+      "learning_rate": 4.409270598827756e-05,
+      "loss": 0.5969,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2470258922323303,
+      "grad_norm": 0.45611399870187513,
+      "learning_rate": 4.408538767944315e-05,
+      "loss": 0.6076,
+      "step": 1765
+    },
+    {
+      "epoch": 0.24716585024492652,
+      "grad_norm": 0.4401540888444283,
+      "learning_rate": 4.407806544836792e-05,
+      "loss": 0.6134,
+      "step": 1766
+    },
+    {
+      "epoch": 0.24730580825752274,
+      "grad_norm": 0.43386652619598914,
+      "learning_rate": 4.407073929655666e-05,
+      "loss": 0.6015,
+      "step": 1767
+    },
+    {
+      "epoch": 0.24744576627011897,
+      "grad_norm": 0.43265086343988496,
+      "learning_rate": 4.406340922551499e-05,
+      "loss": 0.6059,
+      "step": 1768
+    },
+    {
+      "epoch": 0.2475857242827152,
+      "grad_norm": 0.4200705255951236,
+      "learning_rate": 4.4056075236749285e-05,
+      "loss": 0.5596,
+      "step": 1769
+    },
+    {
+      "epoch": 0.24772568229531142,
+      "grad_norm": 0.4327675200614844,
+      "learning_rate": 4.404873733176678e-05,
+      "loss": 0.5673,
+      "step": 1770
+    },
+    {
+      "epoch": 0.24786564030790761,
+      "grad_norm": 0.41060345330483095,
+      "learning_rate": 4.4041395512075464e-05,
+      "loss": 0.5658,
+      "step": 1771
+    },
+    {
+      "epoch": 0.24800559832050384,
+      "grad_norm": 0.4177460191821631,
+      "learning_rate": 4.403404977918417e-05,
+      "loss": 0.5871,
+      "step": 1772
+    },
+    {
+      "epoch": 0.24814555633310006,
+      "grad_norm": 0.3995644804936815,
+      "learning_rate": 4.402670013460252e-05,
+      "loss": 0.5767,
+      "step": 1773
+    },
+    {
+      "epoch": 0.2482855143456963,
+      "grad_norm": 0.40455718790823125,
+      "learning_rate": 4.401934657984094e-05,
+      "loss": 0.6232,
+      "step": 1774
+    },
+    {
+      "epoch": 0.24842547235829252,
+      "grad_norm": 0.40147880141712233,
+      "learning_rate": 4.401198911641066e-05,
+      "loss": 0.587,
+      "step": 1775
+    },
+    {
+      "epoch": 0.24856543037088874,
+      "grad_norm": 0.40405116225388327,
+      "learning_rate": 4.400462774582371e-05,
+      "loss": 0.5708,
+      "step": 1776
+    },
+    {
+      "epoch": 0.24870538838348497,
+      "grad_norm": 0.4227092160628934,
+      "learning_rate": 4.399726246959293e-05,
+      "loss": 0.6278,
+      "step": 1777
+    },
+    {
+      "epoch": 0.24884534639608116,
+      "grad_norm": 0.40251616958969094,
+      "learning_rate": 4.3989893289231954e-05,
+      "loss": 0.6146,
+      "step": 1778
+    },
+    {
+      "epoch": 0.2489853044086774,
+      "grad_norm": 0.4113564198991073,
+      "learning_rate": 4.398252020625523e-05,
+      "loss": 0.5772,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2491252624212736,
+      "grad_norm": 0.4218976966069186,
+      "learning_rate": 4.3975143222178e-05,
+      "loss": 0.5956,
+      "step": 1780
+    },
+    {
+      "epoch": 0.24926522043386984,
+      "grad_norm": 0.4153141525664469,
+      "learning_rate": 4.39677623385163e-05,
+      "loss": 0.5925,
+      "step": 1781
+    },
+    {
+      "epoch": 0.24940517844646606,
+      "grad_norm": 0.408421965231002,
+      "learning_rate": 4.3960377556787e-05,
+      "loss": 0.597,
+      "step": 1782
+    },
+    {
+      "epoch": 0.2495451364590623,
+      "grad_norm": 0.41009081087351246,
+      "learning_rate": 4.3952988878507714e-05,
+      "loss": 0.5997,
+      "step": 1783
+    },
+    {
+      "epoch": 0.24968509447165851,
+      "grad_norm": 0.4126702795308356,
+      "learning_rate": 4.3945596305196925e-05,
+      "loss": 0.5847,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2498250524842547,
+      "grad_norm": 0.40686925420136566,
+      "learning_rate": 4.393819983837385e-05,
+      "loss": 0.5748,
+      "step": 1785
+    },
+    {
+      "epoch": 0.24996501049685094,
+      "grad_norm": 0.42764000046837597,
+      "learning_rate": 4.393079947955856e-05,
+      "loss": 0.5689,
+      "step": 1786
+    },
+    {
+      "epoch": 0.2501049685094472,
+      "grad_norm": 0.43927412707903785,
+      "learning_rate": 4.39233952302719e-05,
+      "loss": 0.6304,
+      "step": 1787
+    },
+    {
+      "epoch": 0.2502449265220434,
+      "grad_norm": 0.41444855744545284,
+      "learning_rate": 4.3915987092035505e-05,
+      "loss": 0.5849,
+      "step": 1788
+    },
+    {
+      "epoch": 0.2503848845346396,
+      "grad_norm": 0.42682475396958397,
+      "learning_rate": 4.3908575066371835e-05,
+      "loss": 0.5862,
+      "step": 1789
+    },
+    {
+      "epoch": 0.25052484254723584,
+      "grad_norm": 0.4771191183081023,
+      "learning_rate": 4.390115915480414e-05,
+      "loss": 0.6329,
+      "step": 1790
+    },
+    {
+      "epoch": 0.25066480055983204,
+      "grad_norm": 0.4308111583911404,
+      "learning_rate": 4.389373935885646e-05,
+      "loss": 0.6113,
+      "step": 1791
+    },
+    {
+      "epoch": 0.2508047585724283,
+      "grad_norm": 0.42487713253039067,
+      "learning_rate": 4.388631568005364e-05,
+      "loss": 0.6097,
+      "step": 1792
+    },
+    {
+      "epoch": 0.2509447165850245,
+      "grad_norm": 0.44680223218467024,
+      "learning_rate": 4.387888811992131e-05,
+      "loss": 0.5852,
+      "step": 1793
+    },
+    {
+      "epoch": 0.25108467459762074,
+      "grad_norm": 0.43930277215742447,
+      "learning_rate": 4.387145667998591e-05,
+      "loss": 0.6279,
+      "step": 1794
+    },
+    {
+      "epoch": 0.25122463261021694,
+      "grad_norm": 0.45405868972289387,
+      "learning_rate": 4.38640213617747e-05,
+      "loss": 0.5879,
+      "step": 1795
+    },
+    {
+      "epoch": 0.25136459062281313,
+      "grad_norm": 0.41454960378404326,
+      "learning_rate": 4.385658216681569e-05,
+      "loss": 0.5724,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2515045486354094,
+      "grad_norm": 0.4061553712342834,
+      "learning_rate": 4.384913909663772e-05,
+      "loss": 0.5411,
+      "step": 1797
+    },
+    {
+      "epoch": 0.2516445066480056,
+      "grad_norm": 0.4276501719660472,
+      "learning_rate": 4.384169215277041e-05,
+      "loss": 0.557,
+      "step": 1798
+    },
+    {
+      "epoch": 0.25178446466060184,
+      "grad_norm": 0.386857810391419,
+      "learning_rate": 4.383424133674419e-05,
+      "loss": 0.5897,
+      "step": 1799
+    },
+    {
+      "epoch": 0.25192442267319803,
+      "grad_norm": 0.4310189731603924,
+      "learning_rate": 4.382678665009028e-05,
+      "loss": 0.6095,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2520643806857943,
+      "grad_norm": 0.39537072584058847,
+      "learning_rate": 4.381932809434068e-05,
+      "loss": 0.5557,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2522043386983905,
+      "grad_norm": 0.3973617854406729,
+      "learning_rate": 4.3811865671028206e-05,
+      "loss": 0.5778,
+      "step": 1802
+    },
+    {
+      "epoch": 0.2523442967109867,
+      "grad_norm": 0.4326957921508078,
+      "learning_rate": 4.380439938168647e-05,
+      "loss": 0.5888,
+      "step": 1803
+    },
+    {
+      "epoch": 0.25248425472358293,
+      "grad_norm": 0.40126247732516246,
+      "learning_rate": 4.379692922784986e-05,
+      "loss": 0.5733,
+      "step": 1804
+    },
+    {
+      "epoch": 0.25262421273617913,
+      "grad_norm": 0.3962832983075247,
+      "learning_rate": 4.378945521105357e-05,
+      "loss": 0.5701,
+      "step": 1805
+    },
+    {
+      "epoch": 0.2527641707487754,
+      "grad_norm": 0.4292056775606421,
+      "learning_rate": 4.378197733283359e-05,
+      "loss": 0.6177,
+      "step": 1806
+    },
+    {
+      "epoch": 0.2529041287613716,
+      "grad_norm": 0.4598576468560811,
+      "learning_rate": 4.377449559472669e-05,
+      "loss": 0.5838,
+      "step": 1807
+    },
+    {
+      "epoch": 0.25304408677396784,
+      "grad_norm": 0.4943287880849202,
+      "learning_rate": 4.3767009998270464e-05,
+      "loss": 0.6519,
+      "step": 1808
+    },
+    {
+      "epoch": 0.25318404478656403,
+      "grad_norm": 0.41197329102085845,
+      "learning_rate": 4.375952054500326e-05,
+      "loss": 0.5685,
+      "step": 1809
+    },
+    {
+      "epoch": 0.25332400279916023,
+      "grad_norm": 0.4473419746095371,
+      "learning_rate": 4.375202723646424e-05,
+      "loss": 0.5987,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2534639608117565,
+      "grad_norm": 0.42470540560047887,
+      "learning_rate": 4.374453007419336e-05,
+      "loss": 0.5944,
+      "step": 1811
+    },
+    {
+      "epoch": 0.2536039188243527,
+      "grad_norm": 0.4319391359458571,
+      "learning_rate": 4.373702905973135e-05,
+      "loss": 0.6269,
+      "step": 1812
+    },
+    {
+      "epoch": 0.25374387683694893,
+      "grad_norm": 0.4025733974145466,
+      "learning_rate": 4.3729524194619766e-05,
+      "loss": 0.5944,
+      "step": 1813
+    },
+    {
+      "epoch": 0.25388383484954513,
+      "grad_norm": 0.4420608064693045,
+      "learning_rate": 4.3722015480400916e-05,
+      "loss": 0.5987,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2540237928621414,
+      "grad_norm": 0.3993911334683989,
+      "learning_rate": 4.371450291861792e-05,
+      "loss": 0.5578,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2541637508747376,
+      "grad_norm": 0.43011254515928227,
+      "learning_rate": 4.370698651081469e-05,
+      "loss": 0.6051,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2543037088873338,
+      "grad_norm": 0.4109991007436434,
+      "learning_rate": 4.369946625853593e-05,
+      "loss": 0.5478,
+      "step": 1817
+    },
+    {
+      "epoch": 0.25444366689993003,
+      "grad_norm": 0.42138731559636133,
+      "learning_rate": 4.369194216332712e-05,
+      "loss": 0.5464,
+      "step": 1818
+    },
+    {
+      "epoch": 0.25458362491252623,
+      "grad_norm": 0.43137639859587146,
+      "learning_rate": 4.368441422673453e-05,
+      "loss": 0.6625,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2547235829251225,
+      "grad_norm": 0.42368932399104675,
+      "learning_rate": 4.367688245030523e-05,
+      "loss": 0.6049,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2548635409377187,
+      "grad_norm": 0.40789716476560073,
+      "learning_rate": 4.36693468355871e-05,
+      "loss": 0.5915,
+      "step": 1821
+    },
+    {
+      "epoch": 0.25500349895031493,
+      "grad_norm": 0.4209915793540894,
+      "learning_rate": 4.366180738412876e-05,
+      "loss": 0.5932,
+      "step": 1822
+    },
+    {
+      "epoch": 0.25514345696291113,
+      "grad_norm": 0.4046845978271354,
+      "learning_rate": 4.365426409747965e-05,
+      "loss": 0.5933,
+      "step": 1823
+    },
+    {
+      "epoch": 0.25528341497550733,
+      "grad_norm": 0.3975587329369015,
+      "learning_rate": 4.3646716977189996e-05,
+      "loss": 0.5731,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2554233729881036,
+      "grad_norm": 0.42825231948537235,
+      "learning_rate": 4.36391660248108e-05,
+      "loss": 0.575,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2555633310006998,
+      "grad_norm": 0.4099650326475359,
+      "learning_rate": 4.3631611241893874e-05,
+      "loss": 0.612,
+      "step": 1826
+    },
+    {
+      "epoch": 0.25570328901329603,
+      "grad_norm": 0.4163478892519659,
+      "learning_rate": 4.362405262999178e-05,
+      "loss": 0.5941,
+      "step": 1827
+    },
+    {
+      "epoch": 0.25584324702589223,
+      "grad_norm": 0.4192607935058611,
+      "learning_rate": 4.361649019065791e-05,
+      "loss": 0.5861,
+      "step": 1828
+    },
+    {
+      "epoch": 0.2559832050384884,
+      "grad_norm": 0.4056602965223505,
+      "learning_rate": 4.3608923925446424e-05,
+      "loss": 0.5517,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2561231630510847,
+      "grad_norm": 0.4291339795549124,
+      "learning_rate": 4.360135383591224e-05,
+      "loss": 0.6211,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2562631210636809,
+      "grad_norm": 0.4214799959182231,
+      "learning_rate": 4.3593779923611114e-05,
+      "loss": 0.6101,
+      "step": 1831
+    },
+    {
+      "epoch": 0.25640307907627713,
+      "grad_norm": 0.45878889417856267,
+      "learning_rate": 4.3586202190099555e-05,
+      "loss": 0.6222,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2565430370888733,
+      "grad_norm": 0.401016178107381,
+      "learning_rate": 4.357862063693486e-05,
+      "loss": 0.5759,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2566829951014696,
+      "grad_norm": 0.4096702585349008,
+      "learning_rate": 4.357103526567511e-05,
+      "loss": 0.6322,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2568229531140658,
+      "grad_norm": 0.4133870618187997,
+      "learning_rate": 4.3563446077879194e-05,
+      "loss": 0.5974,
+      "step": 1835
+    },
+    {
+      "epoch": 0.256962911126662,
+      "grad_norm": 0.41053680469933324,
+      "learning_rate": 4.355585307510675e-05,
+      "loss": 0.5636,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2571028691392582,
+      "grad_norm": 0.4286300619381813,
+      "learning_rate": 4.354825625891822e-05,
+      "loss": 0.5705,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2572428271518544,
+      "grad_norm": 0.5243484730511359,
+      "learning_rate": 4.354065563087484e-05,
+      "loss": 0.5945,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2573827851644507,
+      "grad_norm": 0.4165723735814597,
+      "learning_rate": 4.3533051192538596e-05,
+      "loss": 0.5785,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2575227431770469,
+      "grad_norm": 0.4192843833101577,
+      "learning_rate": 4.3525442945472294e-05,
+      "loss": 0.5873,
+      "step": 1840
+    },
+    {
+      "epoch": 0.25766270118964313,
+      "grad_norm": 0.42759680733088257,
+      "learning_rate": 4.3517830891239496e-05,
+      "loss": 0.6008,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2578026592022393,
+      "grad_norm": 0.41416919735625124,
+      "learning_rate": 4.351021503140456e-05,
+      "loss": 0.6047,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2579426172148355,
+      "grad_norm": 0.4183249907357826,
+      "learning_rate": 4.350259536753262e-05,
+      "loss": 0.5839,
+      "step": 1843
+    },
+    {
+      "epoch": 0.2580825752274318,
+      "grad_norm": 0.41171514577228524,
+      "learning_rate": 4.34949719011896e-05,
+      "loss": 0.5826,
+      "step": 1844
+    },
+    {
+      "epoch": 0.258222533240028,
+      "grad_norm": 0.4240796706040656,
+      "learning_rate": 4.348734463394219e-05,
+      "loss": 0.5988,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2583624912526242,
+      "grad_norm": 0.4231976779667788,
+      "learning_rate": 4.3479713567357886e-05,
+      "loss": 0.5686,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2585024492652204,
+      "grad_norm": 0.4027448525781189,
+      "learning_rate": 4.347207870300494e-05,
+      "loss": 0.5991,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2586424072778167,
+      "grad_norm": 0.44912825496869785,
+      "learning_rate": 4.346444004245239e-05,
+      "loss": 0.5893,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2587823652904129,
+      "grad_norm": 0.4267680067654748,
+      "learning_rate": 4.3456797587270066e-05,
+      "loss": 0.5951,
+      "step": 1849
+    },
+    {
+      "epoch": 0.25892232330300907,
+      "grad_norm": 0.43586393714352484,
+      "learning_rate": 4.344915133902856e-05,
+      "loss": 0.5675,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2590622813156053,
+      "grad_norm": 0.4559790753468547,
+      "learning_rate": 4.344150129929927e-05,
+      "loss": 0.6084,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2592022393282015,
+      "grad_norm": 0.4196684079525518,
+      "learning_rate": 4.3433847469654344e-05,
+      "loss": 0.5434,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2593421973407978,
+      "grad_norm": 0.43364642603356207,
+      "learning_rate": 4.342618985166672e-05,
+      "loss": 0.602,
+      "step": 1853
+    },
+    {
+      "epoch": 0.259482155353394,
+      "grad_norm": 0.43456208436691296,
+      "learning_rate": 4.341852844691012e-05,
+      "loss": 0.5678,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2596221133659902,
+      "grad_norm": 0.4445616632907679,
+      "learning_rate": 4.341086325695905e-05,
+      "loss": 0.6011,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2597620713785864,
+      "grad_norm": 0.4148631528766163,
+      "learning_rate": 4.340319428338877e-05,
+      "loss": 0.5984,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2599020293911826,
+      "grad_norm": 0.42267402439629853,
+      "learning_rate": 4.339552152777534e-05,
+      "loss": 0.6198,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2600419874037789,
+      "grad_norm": 0.43701220197978513,
+      "learning_rate": 4.338784499169559e-05,
+      "loss": 0.5792,
+      "step": 1858
+    },
+    {
+      "epoch": 0.26018194541637507,
+      "grad_norm": 0.4498233901707297,
+      "learning_rate": 4.338016467672712e-05,
+      "loss": 0.5822,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2603219034289713,
+      "grad_norm": 0.4302218516293081,
+      "learning_rate": 4.337248058444832e-05,
+      "loss": 0.6329,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2604618614415675,
+      "grad_norm": 0.4149612252531319,
+      "learning_rate": 4.336479271643833e-05,
+      "loss": 0.5676,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2606018194541638,
+      "grad_norm": 0.4482912811055662,
+      "learning_rate": 4.335710107427711e-05,
+      "loss": 0.6405,
+      "step": 1862
+    },
+    {
+      "epoch": 0.26074177746675997,
+      "grad_norm": 0.43212615840984736,
+      "learning_rate": 4.3349405659545365e-05,
+      "loss": 0.6291,
+      "step": 1863
+    },
+    {
+      "epoch": 0.26088173547935617,
+      "grad_norm": 0.43790968743419006,
+      "learning_rate": 4.334170647382457e-05,
+      "loss": 0.6161,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2610216934919524,
+      "grad_norm": 0.4057107337252783,
+      "learning_rate": 4.333400351869699e-05,
+      "loss": 0.5897,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2611616515045486,
+      "grad_norm": 0.4169972715195528,
+      "learning_rate": 4.332629679574566e-05,
+      "loss": 0.5807,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2613016095171449,
+      "grad_norm": 0.4148009483085235,
+      "learning_rate": 4.3318586306554394e-05,
+      "loss": 0.5784,
+      "step": 1867
+    },
+    {
+      "epoch": 0.26144156752974107,
+      "grad_norm": 0.3916134068027622,
+      "learning_rate": 4.331087205270777e-05,
+      "loss": 0.6159,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2615815255423373,
+      "grad_norm": 0.413939159109484,
+      "learning_rate": 4.3303154035791164e-05,
+      "loss": 0.5965,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2617214835549335,
+      "grad_norm": 0.3965210541189402,
+      "learning_rate": 4.329543225739068e-05,
+      "loss": 0.5793,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2618614415675297,
+      "grad_norm": 0.42358611314483735,
+      "learning_rate": 4.328770671909323e-05,
+      "loss": 0.5817,
+      "step": 1871
+    },
+    {
+      "epoch": 0.26200139958012597,
+      "grad_norm": 0.4039935315082308,
+      "learning_rate": 4.32799774224865e-05,
+      "loss": 0.5825,
+      "step": 1872
+    },
+    {
+      "epoch": 0.26214135759272217,
+      "grad_norm": 0.42958959652262796,
+      "learning_rate": 4.327224436915893e-05,
+      "loss": 0.5831,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2622813156053184,
+      "grad_norm": 0.4520950932964586,
+      "learning_rate": 4.3264507560699746e-05,
+      "loss": 0.6003,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2624212736179146,
+      "grad_norm": 0.4177363042656859,
+      "learning_rate": 4.3256766998698936e-05,
+      "loss": 0.6011,
+      "step": 1875
+    },
+    {
+      "epoch": 0.26256123163051087,
+      "grad_norm": 0.4416091225807165,
+      "learning_rate": 4.324902268474727e-05,
+      "loss": 0.6244,
+      "step": 1876
+    },
+    {
+      "epoch": 0.26270118964310707,
+      "grad_norm": 0.6065946464455126,
+      "learning_rate": 4.324127462043627e-05,
+      "loss": 0.6069,
+      "step": 1877
+    },
+    {
+      "epoch": 0.26284114765570327,
+      "grad_norm": 0.41146045832841216,
+      "learning_rate": 4.323352280735826e-05,
+      "loss": 0.6005,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2629811056682995,
+      "grad_norm": 0.4003315872022978,
+      "learning_rate": 4.322576724710631e-05,
+      "loss": 0.575,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2631210636808957,
+      "grad_norm": 0.40221521105567576,
+      "learning_rate": 4.3218007941274264e-05,
+      "loss": 0.5528,
+      "step": 1880
+    },
+    {
+      "epoch": 0.26326102169349197,
+      "grad_norm": 0.4161742129214681,
+      "learning_rate": 4.321024489145673e-05,
+      "loss": 0.6148,
+      "step": 1881
+    },
+    {
+      "epoch": 0.26340097970608817,
+      "grad_norm": 0.4130366243572411,
+      "learning_rate": 4.3202478099249105e-05,
+      "loss": 0.6184,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2635409377186844,
+      "grad_norm": 0.44069492817846667,
+      "learning_rate": 4.3194707566247537e-05,
+      "loss": 0.5936,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2636808957312806,
+      "grad_norm": 0.43421883092849856,
+      "learning_rate": 4.318693329404896e-05,
+      "loss": 0.6115,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2638208537438768,
+      "grad_norm": 0.41021601219211107,
+      "learning_rate": 4.317915528425106e-05,
+      "loss": 0.5749,
+      "step": 1885
+    },
+    {
+      "epoch": 0.26396081175647307,
+      "grad_norm": 0.42123839224718174,
+      "learning_rate": 4.317137353845229e-05,
+      "loss": 0.598,
+      "step": 1886
+    },
+    {
+      "epoch": 0.26410076976906927,
+      "grad_norm": 0.431025530329678,
+      "learning_rate": 4.316358805825188e-05,
+      "loss": 0.5969,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2642407277816655,
+      "grad_norm": 0.40969971019500684,
+      "learning_rate": 4.3155798845249827e-05,
+      "loss": 0.5962,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2643806857942617,
+      "grad_norm": 0.4155454566156464,
+      "learning_rate": 4.3148005901046905e-05,
+      "loss": 0.5936,
+      "step": 1889
+    },
+    {
+      "epoch": 0.26452064380685797,
+      "grad_norm": 0.4317150467528267,
+      "learning_rate": 4.3140209227244624e-05,
+      "loss": 0.6169,
+      "step": 1890
+    },
+    {
+      "epoch": 0.26466060181945417,
+      "grad_norm": 0.3970247011419925,
+      "learning_rate": 4.313240882544529e-05,
+      "loss": 0.5761,
+      "step": 1891
+    },
+    {
+      "epoch": 0.26480055983205036,
+      "grad_norm": 0.4414850369173905,
+      "learning_rate": 4.3124604697251963e-05,
+      "loss": 0.5991,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2649405178446466,
+      "grad_norm": 0.4499059882632428,
+      "learning_rate": 4.311679684426847e-05,
+      "loss": 0.5507,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2650804758572428,
+      "grad_norm": 0.4195714682577236,
+      "learning_rate": 4.310898526809941e-05,
+      "loss": 0.5908,
+      "step": 1894
+    },
+    {
+      "epoch": 0.26522043386983907,
+      "grad_norm": 0.44322553286670324,
+      "learning_rate": 4.3101169970350125e-05,
+      "loss": 0.5807,
+      "step": 1895
+    },
+    {
+      "epoch": 0.26536039188243526,
+      "grad_norm": 0.424382131322031,
+      "learning_rate": 4.309335095262676e-05,
+      "loss": 0.5867,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2655003498950315,
+      "grad_norm": 0.43960102391817163,
+      "learning_rate": 4.308552821653618e-05,
+      "loss": 0.5984,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2656403079076277,
+      "grad_norm": 0.39684220956322386,
+      "learning_rate": 4.307770176368605e-05,
+      "loss": 0.5701,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2657802659202239,
+      "grad_norm": 0.43672589760925556,
+      "learning_rate": 4.306987159568479e-05,
+      "loss": 0.5973,
+      "step": 1899
+    },
+    {
+      "epoch": 0.26592022393282017,
+      "grad_norm": 0.40233707019737064,
+      "learning_rate": 4.3062037714141565e-05,
+      "loss": 0.613,
+      "step": 1900
+    },
+    {
+      "epoch": 0.26606018194541636,
+      "grad_norm": 0.4264494610214143,
+      "learning_rate": 4.3054200120666334e-05,
+      "loss": 0.606,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2662001399580126,
+      "grad_norm": 0.4098467135771313,
+      "learning_rate": 4.304635881686978e-05,
+      "loss": 0.5857,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2663400979706088,
+      "grad_norm": 0.4240466866788527,
+      "learning_rate": 4.303851380436339e-05,
+      "loss": 0.6153,
+      "step": 1903
+    },
+    {
+      "epoch": 0.266480055983205,
+      "grad_norm": 0.403958075146994,
+      "learning_rate": 4.303066508475939e-05,
+      "loss": 0.5663,
+      "step": 1904
+    },
+    {
+      "epoch": 0.26662001399580126,
+      "grad_norm": 0.4143420243493558,
+      "learning_rate": 4.302281265967076e-05,
+      "loss": 0.5963,
+      "step": 1905
+    },
+    {
+      "epoch": 0.26675997200839746,
+      "grad_norm": 0.4444719253392314,
+      "learning_rate": 4.301495653071126e-05,
+      "loss": 0.6231,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2668999300209937,
+      "grad_norm": 0.40995195237809584,
+      "learning_rate": 4.3007096699495406e-05,
+      "loss": 0.5821,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2670398880335899,
+      "grad_norm": 0.405726468503472,
+      "learning_rate": 4.299923316763848e-05,
+      "loss": 0.604,
+      "step": 1908
+    },
+    {
+      "epoch": 0.26717984604618616,
+      "grad_norm": 0.4078344765445211,
+      "learning_rate": 4.2991365936756505e-05,
+      "loss": 0.576,
+      "step": 1909
+    },
+    {
+      "epoch": 0.26731980405878236,
+      "grad_norm": 0.4227354700436783,
+      "learning_rate": 4.2983495008466276e-05,
+      "loss": 0.6146,
+      "step": 1910
+    },
+    {
+      "epoch": 0.26745976207137856,
+      "grad_norm": 0.4030353788382993,
+      "learning_rate": 4.2975620384385364e-05,
+      "loss": 0.5752,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2675997200839748,
+      "grad_norm": 0.4012686664661831,
+      "learning_rate": 4.296774206613207e-05,
+      "loss": 0.5892,
+      "step": 1912
+    },
+    {
+      "epoch": 0.267739678096571,
+      "grad_norm": 0.41404486946782215,
+      "learning_rate": 4.2959860055325474e-05,
+      "loss": 0.5702,
+      "step": 1913
+    },
+    {
+      "epoch": 0.26787963610916726,
+      "grad_norm": 0.41190311027197035,
+      "learning_rate": 4.295197435358541e-05,
+      "loss": 0.5803,
+      "step": 1914
+    },
+    {
+      "epoch": 0.26801959412176346,
+      "grad_norm": 0.4245596260036112,
+      "learning_rate": 4.294408496253246e-05,
+      "loss": 0.6054,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2681595521343597,
+      "grad_norm": 0.4261060945315179,
+      "learning_rate": 4.293619188378798e-05,
+      "loss": 0.5609,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2682995101469559,
+      "grad_norm": 0.42776522346359913,
+      "learning_rate": 4.292829511897409e-05,
+      "loss": 0.5822,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2684394681595521,
+      "grad_norm": 0.41648930810429086,
+      "learning_rate": 4.292039466971364e-05,
+      "loss": 0.5746,
+      "step": 1918
+    },
+    {
+      "epoch": 0.26857942617214836,
+      "grad_norm": 0.3887697670352693,
+      "learning_rate": 4.2912490537630255e-05,
+      "loss": 0.587,
+      "step": 1919
+    },
+    {
+      "epoch": 0.26871938418474456,
+      "grad_norm": 0.44349778409152874,
+      "learning_rate": 4.2904582724348316e-05,
+      "loss": 0.6146,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2688593421973408,
+      "grad_norm": 0.4179386506999582,
+      "learning_rate": 4.2896671231492966e-05,
+      "loss": 0.6399,
+      "step": 1921
+    },
+    {
+      "epoch": 0.268999300209937,
+      "grad_norm": 0.465029201453624,
+      "learning_rate": 4.288875606069008e-05,
+      "loss": 0.6134,
+      "step": 1922
+    },
+    {
+      "epoch": 0.26913925822253326,
+      "grad_norm": 0.4425974682516534,
+      "learning_rate": 4.2880837213566326e-05,
+      "loss": 0.6037,
+      "step": 1923
+    },
+    {
+      "epoch": 0.26927921623512946,
+      "grad_norm": 0.42747672129154957,
+      "learning_rate": 4.2872914691749086e-05,
+      "loss": 0.6102,
+      "step": 1924
+    },
+    {
+      "epoch": 0.26941917424772566,
+      "grad_norm": 0.4492107075459924,
+      "learning_rate": 4.286498849686654e-05,
+      "loss": 0.591,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2695591322603219,
+      "grad_norm": 0.4016863287653028,
+      "learning_rate": 4.2857058630547594e-05,
+      "loss": 0.5677,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2696990902729181,
+      "grad_norm": 0.4143898674573289,
+      "learning_rate": 4.2849125094421905e-05,
+      "loss": 0.5673,
+      "step": 1927
+    },
+    {
+      "epoch": 0.26983904828551436,
+      "grad_norm": 0.39235286773607386,
+      "learning_rate": 4.284118789011991e-05,
+      "loss": 0.5805,
+      "step": 1928
+    },
+    {
+      "epoch": 0.26997900629811056,
+      "grad_norm": 0.39515207584942524,
+      "learning_rate": 4.2833247019272775e-05,
+      "loss": 0.5898,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2701189643107068,
+      "grad_norm": 0.4201779430133444,
+      "learning_rate": 4.2825302483512444e-05,
+      "loss": 0.6402,
+      "step": 1930
+    },
+    {
+      "epoch": 0.270258922323303,
+      "grad_norm": 0.41576770047699163,
+      "learning_rate": 4.281735428447158e-05,
+      "loss": 0.5649,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2703988803358992,
+      "grad_norm": 0.41462546934983685,
+      "learning_rate": 4.2809402423783624e-05,
+      "loss": 0.5938,
+      "step": 1932
+    },
+    {
+      "epoch": 0.27053883834849546,
+      "grad_norm": 0.4096839436305104,
+      "learning_rate": 4.280144690308277e-05,
+      "loss": 0.5786,
+      "step": 1933
+    },
+    {
+      "epoch": 0.27067879636109166,
+      "grad_norm": 0.41756540359737393,
+      "learning_rate": 4.279348772400395e-05,
+      "loss": 0.5666,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2708187543736879,
+      "grad_norm": 0.39443631101683757,
+      "learning_rate": 4.278552488818288e-05,
+      "loss": 0.5919,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2709587123862841,
+      "grad_norm": 0.438020775582902,
+      "learning_rate": 4.277755839725598e-05,
+      "loss": 0.5852,
+      "step": 1936
+    },
+    {
+      "epoch": 0.27109867039888036,
+      "grad_norm": 0.4043207874267435,
+      "learning_rate": 4.2769588252860434e-05,
+      "loss": 0.5628,
+      "step": 1937
+    },
+    {
+      "epoch": 0.27123862841147656,
+      "grad_norm": 0.4455988825173519,
+      "learning_rate": 4.276161445663423e-05,
+      "loss": 0.6069,
+      "step": 1938
+    },
+    {
+      "epoch": 0.27137858642407275,
+      "grad_norm": 0.41404820018714156,
+      "learning_rate": 4.275363701021602e-05,
+      "loss": 0.5833,
+      "step": 1939
+    },
+    {
+      "epoch": 0.271518544436669,
+      "grad_norm": 0.4207829197564249,
+      "learning_rate": 4.2745655915245266e-05,
+      "loss": 0.5698,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2716585024492652,
+      "grad_norm": 0.4300264097184269,
+      "learning_rate": 4.273767117336217e-05,
+      "loss": 0.6043,
+      "step": 1941
+    },
+    {
+      "epoch": 0.27179846046186146,
+      "grad_norm": 0.4230423530894155,
+      "learning_rate": 4.272968278620768e-05,
+      "loss": 0.5954,
+      "step": 1942
+    },
+    {
+      "epoch": 0.27193841847445765,
+      "grad_norm": 0.4104705230949498,
+      "learning_rate": 4.272169075542348e-05,
+      "loss": 0.6132,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2720783764870539,
+      "grad_norm": 0.4412681694744553,
+      "learning_rate": 4.2713695082652015e-05,
+      "loss": 0.634,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2722183344996501,
+      "grad_norm": 0.41281029518549284,
+      "learning_rate": 4.270569576953648e-05,
+      "loss": 0.5803,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2723582925122463,
+      "grad_norm": 0.4326754906394364,
+      "learning_rate": 4.269769281772082e-05,
+      "loss": 0.5982,
+      "step": 1946
+    },
+    {
+      "epoch": 0.27249825052484256,
+      "grad_norm": 0.4143197407799747,
+      "learning_rate": 4.2689686228849716e-05,
+      "loss": 0.6203,
+      "step": 1947
+    },
+    {
+      "epoch": 0.27263820853743875,
+      "grad_norm": 0.44499388653333827,
+      "learning_rate": 4.26816760045686e-05,
+      "loss": 0.5921,
+      "step": 1948
+    },
+    {
+      "epoch": 0.272778166550035,
+      "grad_norm": 0.4198979954091138,
+      "learning_rate": 4.267366214652366e-05,
+      "loss": 0.5864,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2729181245626312,
+      "grad_norm": 0.43526258690475156,
+      "learning_rate": 4.266564465636183e-05,
+      "loss": 0.5948,
+      "step": 1950
+    },
+    {
+      "epoch": 0.27305808257522746,
+      "grad_norm": 0.3940644065133908,
+      "learning_rate": 4.265762353573077e-05,
+      "loss": 0.585,
+      "step": 1951
+    },
+    {
+      "epoch": 0.27319804058782365,
+      "grad_norm": 0.40794104308496737,
+      "learning_rate": 4.264959878627891e-05,
+      "loss": 0.6,
+      "step": 1952
+    },
+    {
+      "epoch": 0.27333799860041985,
+      "grad_norm": 0.3823510808207485,
+      "learning_rate": 4.264157040965543e-05,
+      "loss": 0.547,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2734779566130161,
+      "grad_norm": 0.42277154396243044,
+      "learning_rate": 4.263353840751022e-05,
+      "loss": 0.5996,
+      "step": 1954
+    },
+    {
+      "epoch": 0.2736179146256123,
+      "grad_norm": 0.39495247008059065,
+      "learning_rate": 4.2625502781493955e-05,
+      "loss": 0.5769,
+      "step": 1955
+    },
+    {
+      "epoch": 0.27375787263820855,
+      "grad_norm": 0.43172939634218205,
+      "learning_rate": 4.261746353325804e-05,
+      "loss": 0.5667,
+      "step": 1956
+    },
+    {
+      "epoch": 0.27389783065080475,
+      "grad_norm": 0.42748966931166393,
+      "learning_rate": 4.26094206644546e-05,
+      "loss": 0.619,
+      "step": 1957
+    },
+    {
+      "epoch": 0.274037788663401,
+      "grad_norm": 0.41520591372518717,
+      "learning_rate": 4.260137417673655e-05,
+      "loss": 0.5495,
+      "step": 1958
+    },
+    {
+      "epoch": 0.2741777466759972,
+      "grad_norm": 0.41675727817849284,
+      "learning_rate": 4.259332407175751e-05,
+      "loss": 0.5678,
+      "step": 1959
+    },
+    {
+      "epoch": 0.2743177046885934,
+      "grad_norm": 0.42389412857665876,
+      "learning_rate": 4.258527035117187e-05,
+      "loss": 0.5776,
+      "step": 1960
+    },
+    {
+      "epoch": 0.27445766270118965,
+      "grad_norm": 0.42650038389451433,
+      "learning_rate": 4.257721301663474e-05,
+      "loss": 0.5926,
+      "step": 1961
+    },
+    {
+      "epoch": 0.27459762071378585,
+      "grad_norm": 0.4547693632893839,
+      "learning_rate": 4.2569152069801994e-05,
+      "loss": 0.6109,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2747375787263821,
+      "grad_norm": 0.4096367527079948,
+      "learning_rate": 4.256108751233023e-05,
+      "loss": 0.5742,
+      "step": 1963
+    },
+    {
+      "epoch": 0.2748775367389783,
+      "grad_norm": 0.43852841060550257,
+      "learning_rate": 4.255301934587679e-05,
+      "loss": 0.6161,
+      "step": 1964
+    },
+    {
+      "epoch": 0.27501749475157455,
+      "grad_norm": 0.4108034529541987,
+      "learning_rate": 4.254494757209979e-05,
+      "loss": 0.5909,
+      "step": 1965
+    },
+    {
+      "epoch": 0.27515745276417075,
+      "grad_norm": 0.4179747177285123,
+      "learning_rate": 4.2536872192658036e-05,
+      "loss": 0.5902,
+      "step": 1966
+    },
+    {
+      "epoch": 0.27529741077676695,
+      "grad_norm": 0.4075584545924878,
+      "learning_rate": 4.2528793209211106e-05,
+      "loss": 0.573,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2754373687893632,
+      "grad_norm": 0.4201389140014726,
+      "learning_rate": 4.252071062341933e-05,
+      "loss": 0.5944,
+      "step": 1968
+    },
+    {
+      "epoch": 0.2755773268019594,
+      "grad_norm": 0.4253197821835051,
+      "learning_rate": 4.251262443694374e-05,
+      "loss": 0.569,
+      "step": 1969
+    },
+    {
+      "epoch": 0.27571728481455565,
+      "grad_norm": 0.41281050840051553,
+      "learning_rate": 4.2504534651446134e-05,
+      "loss": 0.5845,
+      "step": 1970
+    },
+    {
+      "epoch": 0.27585724282715185,
+      "grad_norm": 0.4053164541842358,
+      "learning_rate": 4.2496441268589046e-05,
+      "loss": 0.6001,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2759972008397481,
+      "grad_norm": 0.4381447506546874,
+      "learning_rate": 4.2488344290035755e-05,
+      "loss": 0.607,
+      "step": 1972
+    },
+    {
+      "epoch": 0.2761371588523443,
+      "grad_norm": 0.4183122839304423,
+      "learning_rate": 4.248024371745027e-05,
+      "loss": 0.5905,
+      "step": 1973
+    },
+    {
+      "epoch": 0.2762771168649405,
+      "grad_norm": 0.4284345637823913,
+      "learning_rate": 4.2472139552497334e-05,
+      "loss": 0.6224,
+      "step": 1974
+    },
+    {
+      "epoch": 0.27641707487753675,
+      "grad_norm": 0.43624222428113946,
+      "learning_rate": 4.246403179684245e-05,
+      "loss": 0.5874,
+      "step": 1975
+    },
+    {
+      "epoch": 0.27655703289013295,
+      "grad_norm": 0.4238838475394068,
+      "learning_rate": 4.245592045215182e-05,
+      "loss": 0.5939,
+      "step": 1976
+    },
+    {
+      "epoch": 0.2766969909027292,
+      "grad_norm": 0.41691347825833375,
+      "learning_rate": 4.244780552009243e-05,
+      "loss": 0.6258,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2768369489153254,
+      "grad_norm": 0.4083689917945832,
+      "learning_rate": 4.2439687002331974e-05,
+      "loss": 0.5565,
+      "step": 1978
+    },
+    {
+      "epoch": 0.27697690692792165,
+      "grad_norm": 0.4285760791242101,
+      "learning_rate": 4.2431564900538884e-05,
+      "loss": 0.5957,
+      "step": 1979
+    },
+    {
+      "epoch": 0.27711686494051785,
+      "grad_norm": 0.41197446910489155,
+      "learning_rate": 4.242343921638234e-05,
+      "loss": 0.601,
+      "step": 1980
+    },
+    {
+      "epoch": 0.27725682295311405,
+      "grad_norm": 0.38975348870242715,
+      "learning_rate": 4.241530995153225e-05,
+      "loss": 0.562,
+      "step": 1981
+    },
+    {
+      "epoch": 0.2773967809657103,
+      "grad_norm": 0.4368612257662016,
+      "learning_rate": 4.2407177107659256e-05,
+      "loss": 0.6061,
+      "step": 1982
+    },
+    {
+      "epoch": 0.2775367389783065,
+      "grad_norm": 0.40941708192964865,
+      "learning_rate": 4.2399040686434756e-05,
+      "loss": 0.5659,
+      "step": 1983
+    },
+    {
+      "epoch": 0.27767669699090275,
+      "grad_norm": 0.4233950461539689,
+      "learning_rate": 4.239090068953086e-05,
+      "loss": 0.5981,
+      "step": 1984
+    },
+    {
+      "epoch": 0.27781665500349895,
+      "grad_norm": 0.42236280001104815,
+      "learning_rate": 4.23827571186204e-05,
+      "loss": 0.5886,
+      "step": 1985
+    },
+    {
+      "epoch": 0.27795661301609514,
+      "grad_norm": 0.4144640899811079,
+      "learning_rate": 4.237460997537699e-05,
+      "loss": 0.5779,
+      "step": 1986
+    },
+    {
+      "epoch": 0.2780965710286914,
+      "grad_norm": 0.4625927000805653,
+      "learning_rate": 4.2366459261474933e-05,
+      "loss": 0.5679,
+      "step": 1987
+    },
+    {
+      "epoch": 0.2782365290412876,
+      "grad_norm": 0.4076605883733819,
+      "learning_rate": 4.235830497858929e-05,
+      "loss": 0.5655,
+      "step": 1988
+    },
+    {
+      "epoch": 0.27837648705388385,
+      "grad_norm": 0.42845289807533865,
+      "learning_rate": 4.235014712839586e-05,
+      "loss": 0.6282,
+      "step": 1989
+    },
+    {
+      "epoch": 0.27851644506648005,
+      "grad_norm": 0.3954911530156325,
+      "learning_rate": 4.2341985712571144e-05,
+      "loss": 0.5594,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2786564030790763,
+      "grad_norm": 0.42832437836826653,
+      "learning_rate": 4.233382073279241e-05,
+      "loss": 0.592,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2787963610916725,
+      "grad_norm": 0.3915035514180305,
+      "learning_rate": 4.232565219073763e-05,
+      "loss": 0.5767,
+      "step": 1992
+    },
+    {
+      "epoch": 0.2789363191042687,
+      "grad_norm": 0.4091633397512807,
+      "learning_rate": 4.231748008808554e-05,
+      "loss": 0.5995,
+      "step": 1993
+    },
+    {
+      "epoch": 0.27907627711686495,
+      "grad_norm": 0.43083489824001814,
+      "learning_rate": 4.230930442651557e-05,
+      "loss": 0.6254,
+      "step": 1994
+    },
+    {
+      "epoch": 0.27921623512946114,
+      "grad_norm": 0.4215052561977625,
+      "learning_rate": 4.230112520770792e-05,
+      "loss": 0.5946,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2793561931420574,
+      "grad_norm": 0.41624126906364184,
+      "learning_rate": 4.22929424333435e-05,
+      "loss": 0.5912,
+      "step": 1996
+    },
+    {
+      "epoch": 0.2794961511546536,
+      "grad_norm": 0.4084391778034195,
+      "learning_rate": 4.228475610510394e-05,
+      "loss": 0.5643,
+      "step": 1997
+    },
+    {
+      "epoch": 0.27963610916724985,
+      "grad_norm": 0.4151631289731341,
+      "learning_rate": 4.227656622467162e-05,
+      "loss": 0.5762,
+      "step": 1998
+    },
+    {
+      "epoch": 0.27977606717984604,
+      "grad_norm": 0.4207154793240298,
+      "learning_rate": 4.226837279372965e-05,
+      "loss": 0.5836,
+      "step": 1999
+    },
+    {
+      "epoch": 0.27991602519244224,
+      "grad_norm": 0.44259939289210365,
+      "learning_rate": 4.226017581396186e-05,
+      "loss": 0.6026,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2800559832050385,
+      "grad_norm": 0.40656254137179043,
+      "learning_rate": 4.2251975287052804e-05,
+      "loss": 0.5739,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2801959412176347,
+      "grad_norm": 0.41605805412685426,
+      "learning_rate": 4.224377121468778e-05,
+      "loss": 0.6018,
+      "step": 2002
+    },
+    {
+      "epoch": 0.28033589923023094,
+      "grad_norm": 0.4181616140874938,
+      "learning_rate": 4.223556359855282e-05,
+      "loss": 0.5429,
+      "step": 2003
+    },
+    {
+      "epoch": 0.28047585724282714,
+      "grad_norm": 0.4414984932849065,
+      "learning_rate": 4.222735244033464e-05,
+      "loss": 0.6,
+      "step": 2004
+    },
+    {
+      "epoch": 0.2806158152554234,
+      "grad_norm": 0.40209497833786334,
+      "learning_rate": 4.221913774172076e-05,
+      "loss": 0.5793,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2807557732680196,
+      "grad_norm": 0.40555504520183505,
+      "learning_rate": 4.221091950439935e-05,
+      "loss": 0.5522,
+      "step": 2006
+    },
+    {
+      "epoch": 0.2808957312806158,
+      "grad_norm": 0.4425009809651693,
+      "learning_rate": 4.220269773005935e-05,
+      "loss": 0.6562,
+      "step": 2007
+    },
+    {
+      "epoch": 0.28103568929321204,
+      "grad_norm": 0.4154023509957506,
+      "learning_rate": 4.219447242039043e-05,
+      "loss": 0.6121,
+      "step": 2008
+    },
+    {
+      "epoch": 0.28117564730580824,
+      "grad_norm": 0.4069493962165521,
+      "learning_rate": 4.2186243577082954e-05,
+      "loss": 0.582,
+      "step": 2009
+    },
+    {
+      "epoch": 0.2813156053184045,
+      "grad_norm": 0.42547539237875576,
+      "learning_rate": 4.217801120182805e-05,
+      "loss": 0.5737,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2814555633310007,
+      "grad_norm": 0.4120143925863747,
+      "learning_rate": 4.2169775296317545e-05,
+      "loss": 0.5984,
+      "step": 2011
+    },
+    {
+      "epoch": 0.28159552134359694,
+      "grad_norm": 0.4091520558950441,
+      "learning_rate": 4.216153586224401e-05,
+      "loss": 0.5533,
+      "step": 2012
+    },
+    {
+      "epoch": 0.28173547935619314,
+      "grad_norm": 0.40387339261079624,
+      "learning_rate": 4.2153292901300715e-05,
+      "loss": 0.6078,
+      "step": 2013
+    },
+    {
+      "epoch": 0.28187543736878934,
+      "grad_norm": 0.40058528148893674,
+      "learning_rate": 4.214504641518169e-05,
+      "loss": 0.552,
+      "step": 2014
+    },
+    {
+      "epoch": 0.2820153953813856,
+      "grad_norm": 0.40857595852803164,
+      "learning_rate": 4.213679640558167e-05,
+      "loss": 0.5744,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2821553533939818,
+      "grad_norm": 0.3978718427859697,
+      "learning_rate": 4.212854287419611e-05,
+      "loss": 0.5916,
+      "step": 2016
+    },
+    {
+      "epoch": 0.28229531140657804,
+      "grad_norm": 0.3821785406082015,
+      "learning_rate": 4.212028582272119e-05,
+      "loss": 0.555,
+      "step": 2017
+    },
+    {
+      "epoch": 0.28243526941917424,
+      "grad_norm": 0.4069697450232385,
+      "learning_rate": 4.2112025252853823e-05,
+      "loss": 0.564,
+      "step": 2018
+    },
+    {
+      "epoch": 0.2825752274317705,
+      "grad_norm": 0.4023231577263976,
+      "learning_rate": 4.210376116629165e-05,
+      "loss": 0.5486,
+      "step": 2019
+    },
+    {
+      "epoch": 0.2827151854443667,
+      "grad_norm": 0.4071761170849277,
+      "learning_rate": 4.2095493564733005e-05,
+      "loss": 0.5769,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2828551434569629,
+      "grad_norm": 0.4303385757744145,
+      "learning_rate": 4.208722244987698e-05,
+      "loss": 0.5891,
+      "step": 2021
+    },
+    {
+      "epoch": 0.28299510146955914,
+      "grad_norm": 0.42291929930753197,
+      "learning_rate": 4.2078947823423364e-05,
+      "loss": 0.6011,
+      "step": 2022
+    },
+    {
+      "epoch": 0.28313505948215534,
+      "grad_norm": 0.4004019282036015,
+      "learning_rate": 4.2070669687072685e-05,
+      "loss": 0.5643,
+      "step": 2023
+    },
+    {
+      "epoch": 0.2832750174947516,
+      "grad_norm": 0.4294748353114633,
+      "learning_rate": 4.206238804252617e-05,
+      "loss": 0.6098,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2834149755073478,
+      "grad_norm": 0.4207040878829443,
+      "learning_rate": 4.205410289148579e-05,
+      "loss": 0.5825,
+      "step": 2025
+    },
+    {
+      "epoch": 0.28355493351994404,
+      "grad_norm": 0.42868887539679973,
+      "learning_rate": 4.204581423565424e-05,
+      "loss": 0.5703,
+      "step": 2026
+    },
+    {
+      "epoch": 0.28369489153254024,
+      "grad_norm": 0.41961324792796834,
+      "learning_rate": 4.2037522076734895e-05,
+      "loss": 0.6024,
+      "step": 2027
+    },
+    {
+      "epoch": 0.28383484954513644,
+      "grad_norm": 0.426062343917317,
+      "learning_rate": 4.20292264164319e-05,
+      "loss": 0.5451,
+      "step": 2028
+    },
+    {
+      "epoch": 0.2839748075577327,
+      "grad_norm": 0.5356851221725408,
+      "learning_rate": 4.202092725645009e-05,
+      "loss": 0.566,
+      "step": 2029
+    },
+    {
+      "epoch": 0.2841147655703289,
+      "grad_norm": 0.41705129999473584,
+      "learning_rate": 4.2012624598495026e-05,
+      "loss": 0.5515,
+      "step": 2030
+    },
+    {
+      "epoch": 0.28425472358292514,
+      "grad_norm": 0.4305240936670277,
+      "learning_rate": 4.2004318444272985e-05,
+      "loss": 0.6063,
+      "step": 2031
+    },
+    {
+      "epoch": 0.28439468159552134,
+      "grad_norm": 0.4109507589908985,
+      "learning_rate": 4.199600879549098e-05,
+      "loss": 0.5841,
+      "step": 2032
+    },
+    {
+      "epoch": 0.2845346396081176,
+      "grad_norm": 0.4152517216058978,
+      "learning_rate": 4.198769565385671e-05,
+      "loss": 0.618,
+      "step": 2033
+    },
+    {
+      "epoch": 0.2846745976207138,
+      "grad_norm": 0.40875510353923156,
+      "learning_rate": 4.197937902107863e-05,
+      "loss": 0.6169,
+      "step": 2034
+    },
+    {
+      "epoch": 0.28481455563331,
+      "grad_norm": 0.4080934485529636,
+      "learning_rate": 4.197105889886587e-05,
+      "loss": 0.6191,
+      "step": 2035
+    },
+    {
+      "epoch": 0.28495451364590624,
+      "grad_norm": 0.395534357123435,
+      "learning_rate": 4.1962735288928305e-05,
+      "loss": 0.572,
+      "step": 2036
+    },
+    {
+      "epoch": 0.28509447165850244,
+      "grad_norm": 0.4123561699274625,
+      "learning_rate": 4.1954408192976536e-05,
+      "loss": 0.6112,
+      "step": 2037
+    },
+    {
+      "epoch": 0.2852344296710987,
+      "grad_norm": 0.43088641312757203,
+      "learning_rate": 4.1946077612721854e-05,
+      "loss": 0.6146,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2853743876836949,
+      "grad_norm": 0.4073212886136425,
+      "learning_rate": 4.193774354987629e-05,
+      "loss": 0.5876,
+      "step": 2039
+    },
+    {
+      "epoch": 0.28551434569629114,
+      "grad_norm": 0.41104021326097206,
+      "learning_rate": 4.192940600615255e-05,
+      "loss": 0.5861,
+      "step": 2040
+    },
+    {
+      "epoch": 0.28565430370888734,
+      "grad_norm": 0.4266808368141709,
+      "learning_rate": 4.192106498326411e-05,
+      "loss": 0.5982,
+      "step": 2041
+    },
+    {
+      "epoch": 0.28579426172148353,
+      "grad_norm": 0.44040096596479616,
+      "learning_rate": 4.191272048292513e-05,
+      "loss": 0.5877,
+      "step": 2042
+    },
+    {
+      "epoch": 0.2859342197340798,
+      "grad_norm": 0.4156724474946737,
+      "learning_rate": 4.1904372506850484e-05,
+      "loss": 0.5899,
+      "step": 2043
+    },
+    {
+      "epoch": 0.286074177746676,
+      "grad_norm": 0.42353874241636685,
+      "learning_rate": 4.189602105675577e-05,
+      "loss": 0.5958,
+      "step": 2044
+    },
+    {
+      "epoch": 0.28621413575927224,
+      "grad_norm": 0.40321593874563183,
+      "learning_rate": 4.18876661343573e-05,
+      "loss": 0.5898,
+      "step": 2045
+    },
+    {
+      "epoch": 0.28635409377186843,
+      "grad_norm": 0.43950894922113853,
+      "learning_rate": 4.187930774137209e-05,
+      "loss": 0.578,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2864940517844647,
+      "grad_norm": 0.4094852639707724,
+      "learning_rate": 4.187094587951786e-05,
+      "loss": 0.5808,
+      "step": 2047
+    },
+    {
+      "epoch": 0.2866340097970609,
+      "grad_norm": 0.4290190354250858,
+      "learning_rate": 4.1862580550513086e-05,
+      "loss": 0.5754,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2867739678096571,
+      "grad_norm": 0.43386920098037757,
+      "learning_rate": 4.18542117560769e-05,
+      "loss": 0.6008,
+      "step": 2049
+    },
+    {
+      "epoch": 0.28691392582225334,
+      "grad_norm": 0.4112010436019356,
+      "learning_rate": 4.1845839497929204e-05,
+      "loss": 0.5878,
+      "step": 2050
+    },
+    {
+      "epoch": 0.28705388383484953,
+      "grad_norm": 0.4052022574202855,
+      "learning_rate": 4.183746377779055e-05,
+      "loss": 0.5733,
+      "step": 2051
+    },
+    {
+      "epoch": 0.2871938418474458,
+      "grad_norm": 0.4325624910156896,
+      "learning_rate": 4.182908459738226e-05,
+      "loss": 0.5954,
+      "step": 2052
+    },
+    {
+      "epoch": 0.287333799860042,
+      "grad_norm": 0.41379739337134963,
+      "learning_rate": 4.1820701958426325e-05,
+      "loss": 0.5487,
+      "step": 2053
+    },
+    {
+      "epoch": 0.28747375787263824,
+      "grad_norm": 0.41206409125142784,
+      "learning_rate": 4.1812315862645466e-05,
+      "loss": 0.5317,
+      "step": 2054
+    },
+    {
+      "epoch": 0.28761371588523443,
+      "grad_norm": 0.4410834744075679,
+      "learning_rate": 4.180392631176312e-05,
+      "loss": 0.6188,
+      "step": 2055
+    },
+    {
+      "epoch": 0.28775367389783063,
+      "grad_norm": 0.41721301374555403,
+      "learning_rate": 4.179553330750341e-05,
+      "loss": 0.5873,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2878936319104269,
+      "grad_norm": 0.4140910091320955,
+      "learning_rate": 4.1787136851591194e-05,
+      "loss": 0.5913,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2880335899230231,
+      "grad_norm": 0.4136362029589039,
+      "learning_rate": 4.177873694575202e-05,
+      "loss": 0.6285,
+      "step": 2058
+    },
+    {
+      "epoch": 0.28817354793561933,
+      "grad_norm": 0.42507432331536726,
+      "learning_rate": 4.1770333591712164e-05,
+      "loss": 0.5854,
+      "step": 2059
+    },
+    {
+      "epoch": 0.28831350594821553,
+      "grad_norm": 0.43737552322389167,
+      "learning_rate": 4.176192679119859e-05,
+      "loss": 0.5922,
+      "step": 2060
+    },
+    {
+      "epoch": 0.28845346396081173,
+      "grad_norm": 0.3891108181335366,
+      "learning_rate": 4.175351654593899e-05,
+      "loss": 0.5284,
+      "step": 2061
+    },
+    {
+      "epoch": 0.288593421973408,
+      "grad_norm": 0.4969916385805558,
+      "learning_rate": 4.174510285766175e-05,
+      "loss": 0.5824,
+      "step": 2062
+    },
+    {
+      "epoch": 0.2887333799860042,
+      "grad_norm": 0.42340411002126843,
+      "learning_rate": 4.173668572809597e-05,
+      "loss": 0.5689,
+      "step": 2063
+    },
+    {
+      "epoch": 0.28887333799860043,
+      "grad_norm": 0.41505954996479894,
+      "learning_rate": 4.172826515897146e-05,
+      "loss": 0.5734,
+      "step": 2064
+    },
+    {
+      "epoch": 0.28901329601119663,
+      "grad_norm": 0.3799206030872319,
+      "learning_rate": 4.1719841152018716e-05,
+      "loss": 0.5669,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2891532540237929,
+      "grad_norm": 0.43628757271419605,
+      "learning_rate": 4.171141370896898e-05,
+      "loss": 0.6063,
+      "step": 2066
+    },
+    {
+      "epoch": 0.2892932120363891,
+      "grad_norm": 0.42076886960500287,
+      "learning_rate": 4.170298283155416e-05,
+      "loss": 0.5923,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2894331700489853,
+      "grad_norm": 0.4074347275662083,
+      "learning_rate": 4.169454852150689e-05,
+      "loss": 0.5872,
+      "step": 2068
+    },
+    {
+      "epoch": 0.28957312806158153,
+      "grad_norm": 0.4048049815445387,
+      "learning_rate": 4.168611078056051e-05,
+      "loss": 0.5549,
+      "step": 2069
+    },
+    {
+      "epoch": 0.28971308607417773,
+      "grad_norm": 0.411926260409551,
+      "learning_rate": 4.167766961044907e-05,
+      "loss": 0.5879,
+      "step": 2070
+    },
+    {
+      "epoch": 0.289853044086774,
+      "grad_norm": 0.4144949406829952,
+      "learning_rate": 4.166922501290729e-05,
+      "loss": 0.5718,
+      "step": 2071
+    },
+    {
+      "epoch": 0.2899930020993702,
+      "grad_norm": 0.43675024412528857,
+      "learning_rate": 4.1660776989670646e-05,
+      "loss": 0.5933,
+      "step": 2072
+    },
+    {
+      "epoch": 0.29013296011196643,
+      "grad_norm": 0.394293953998829,
+      "learning_rate": 4.165232554247528e-05,
+      "loss": 0.5345,
+      "step": 2073
+    },
+    {
+      "epoch": 0.29027291812456263,
+      "grad_norm": 0.4193571632306237,
+      "learning_rate": 4.164387067305805e-05,
+      "loss": 0.6182,
+      "step": 2074
+    },
+    {
+      "epoch": 0.2904128761371588,
+      "grad_norm": 0.6437864579317593,
+      "learning_rate": 4.163541238315653e-05,
+      "loss": 0.6224,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2905528341497551,
+      "grad_norm": 0.4015710103834747,
+      "learning_rate": 4.162695067450897e-05,
+      "loss": 0.6,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2906927921623513,
+      "grad_norm": 0.42416882893328844,
+      "learning_rate": 4.161848554885434e-05,
+      "loss": 0.5964,
+      "step": 2077
+    },
+    {
+      "epoch": 0.29083275017494753,
+      "grad_norm": 0.39865445877876937,
+      "learning_rate": 4.161001700793231e-05,
+      "loss": 0.5379,
+      "step": 2078
+    },
+    {
+      "epoch": 0.2909727081875437,
+      "grad_norm": 0.428296814729791,
+      "learning_rate": 4.160154505348326e-05,
+      "loss": 0.565,
+      "step": 2079
+    },
+    {
+      "epoch": 0.29111266620014,
+      "grad_norm": 0.4500265708630457,
+      "learning_rate": 4.159306968724824e-05,
+      "loss": 0.6454,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2912526242127362,
+      "grad_norm": 0.4113564994850615,
+      "learning_rate": 4.1584590910969055e-05,
+      "loss": 0.5792,
+      "step": 2081
+    },
+    {
+      "epoch": 0.2913925822253324,
+      "grad_norm": 0.4181766831204532,
+      "learning_rate": 4.157610872638815e-05,
+      "loss": 0.6005,
+      "step": 2082
+    },
+    {
+      "epoch": 0.29153254023792863,
+      "grad_norm": 0.42000429084184243,
+      "learning_rate": 4.156762313524873e-05,
+      "loss": 0.5565,
+      "step": 2083
+    },
+    {
+      "epoch": 0.2916724982505248,
+      "grad_norm": 0.4106324972814455,
+      "learning_rate": 4.155913413929463e-05,
+      "loss": 0.5678,
+      "step": 2084
+    },
+    {
+      "epoch": 0.2918124562631211,
+      "grad_norm": 0.3971576285170608,
+      "learning_rate": 4.155064174027047e-05,
+      "loss": 0.5972,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2919524142757173,
+      "grad_norm": 0.4438093270564298,
+      "learning_rate": 4.154214593992149e-05,
+      "loss": 0.5909,
+      "step": 2086
+    },
+    {
+      "epoch": 0.29209237228831353,
+      "grad_norm": 0.4056320133949262,
+      "learning_rate": 4.1533646739993676e-05,
+      "loss": 0.5832,
+      "step": 2087
+    },
+    {
+      "epoch": 0.2922323303009097,
+      "grad_norm": 0.4129515425696569,
+      "learning_rate": 4.1525144142233705e-05,
+      "loss": 0.5995,
+      "step": 2088
+    },
+    {
+      "epoch": 0.2923722883135059,
+      "grad_norm": 0.4513903259561082,
+      "learning_rate": 4.151663814838893e-05,
+      "loss": 0.5527,
+      "step": 2089
+    },
+    {
+      "epoch": 0.2925122463261022,
+      "grad_norm": 0.3975679752001749,
+      "learning_rate": 4.150812876020744e-05,
+      "loss": 0.5579,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2926522043386984,
+      "grad_norm": 0.4079824226246549,
+      "learning_rate": 4.1499615979437986e-05,
+      "loss": 0.5583,
+      "step": 2091
+    },
+    {
+      "epoch": 0.2927921623512946,
+      "grad_norm": 0.4452632659983039,
+      "learning_rate": 4.1491099807830044e-05,
+      "loss": 0.5871,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2929321203638908,
+      "grad_norm": 0.416969695477689,
+      "learning_rate": 4.148258024713376e-05,
+      "loss": 0.5686,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2930720783764871,
+      "grad_norm": 0.39010342813066295,
+      "learning_rate": 4.147405729909999e-05,
+      "loss": 0.5805,
+      "step": 2094
+    },
+    {
+      "epoch": 0.2932120363890833,
+      "grad_norm": 0.4146408500392661,
+      "learning_rate": 4.146553096548031e-05,
+      "loss": 0.5939,
+      "step": 2095
+    },
+    {
+      "epoch": 0.2933519944016795,
+      "grad_norm": 0.4438792319022797,
+      "learning_rate": 4.145700124802693e-05,
+      "loss": 0.6048,
+      "step": 2096
+    },
+    {
+      "epoch": 0.2934919524142757,
+      "grad_norm": 0.652029593861103,
+      "learning_rate": 4.144846814849282e-05,
+      "loss": 0.6083,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2936319104268719,
+      "grad_norm": 0.4075857320797679,
+      "learning_rate": 4.143993166863161e-05,
+      "loss": 0.5861,
+      "step": 2098
+    },
+    {
+      "epoch": 0.2937718684394682,
+      "grad_norm": 0.3986323846100174,
+      "learning_rate": 4.143139181019764e-05,
+      "loss": 0.5729,
+      "step": 2099
+    },
+    {
+      "epoch": 0.2939118264520644,
+      "grad_norm": 0.42024806604839915,
+      "learning_rate": 4.1422848574945924e-05,
+      "loss": 0.5662,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2940517844646606,
+      "grad_norm": 0.4174607888005454,
+      "learning_rate": 4.141430196463221e-05,
+      "loss": 0.542,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2941917424772568,
+      "grad_norm": 0.4330991872091227,
+      "learning_rate": 4.140575198101288e-05,
+      "loss": 0.6237,
+      "step": 2102
+    },
+    {
+      "epoch": 0.294331700489853,
+      "grad_norm": 0.41437624963664066,
+      "learning_rate": 4.139719862584506e-05,
+      "loss": 0.5758,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2944716585024493,
+      "grad_norm": 0.41455194934526307,
+      "learning_rate": 4.138864190088655e-05,
+      "loss": 0.5494,
+      "step": 2104
+    },
+    {
+      "epoch": 0.29461161651504547,
+      "grad_norm": 0.3928031972149667,
+      "learning_rate": 4.1380081807895846e-05,
+      "loss": 0.615,
+      "step": 2105
+    },
+    {
+      "epoch": 0.2947515745276417,
+      "grad_norm": 0.4097807146687554,
+      "learning_rate": 4.137151834863213e-05,
+      "loss": 0.5563,
+      "step": 2106
+    },
+    {
+      "epoch": 0.2948915325402379,
+      "grad_norm": 0.4201887917529205,
+      "learning_rate": 4.1362951524855275e-05,
+      "loss": 0.5934,
+      "step": 2107
+    },
+    {
+      "epoch": 0.2950314905528342,
+      "grad_norm": 0.408291117857594,
+      "learning_rate": 4.1354381338325864e-05,
+      "loss": 0.6003,
+      "step": 2108
+    },
+    {
+      "epoch": 0.29517144856543037,
+      "grad_norm": 0.43085354542414667,
+      "learning_rate": 4.1345807790805144e-05,
+      "loss": 0.5948,
+      "step": 2109
+    },
+    {
+      "epoch": 0.29531140657802657,
+      "grad_norm": 0.46223900630617937,
+      "learning_rate": 4.1337230884055073e-05,
+      "loss": 0.5575,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2954513645906228,
+      "grad_norm": 0.42857422520776206,
+      "learning_rate": 4.132865061983829e-05,
+      "loss": 0.6218,
+      "step": 2111
+    },
+    {
+      "epoch": 0.295591322603219,
+      "grad_norm": 0.4328798834584895,
+      "learning_rate": 4.132006699991813e-05,
+      "loss": 0.572,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2957312806158153,
+      "grad_norm": 0.4029310476167804,
+      "learning_rate": 4.131148002605861e-05,
+      "loss": 0.5762,
+      "step": 2113
+    },
+    {
+      "epoch": 0.29587123862841147,
+      "grad_norm": 0.4227536522427694,
+      "learning_rate": 4.130288970002444e-05,
+      "loss": 0.5716,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2960111966410077,
+      "grad_norm": 0.3992813234221805,
+      "learning_rate": 4.1294296023581015e-05,
+      "loss": 0.5803,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2961511546536039,
+      "grad_norm": 0.40775499827762557,
+      "learning_rate": 4.128569899849443e-05,
+      "loss": 0.5813,
+      "step": 2116
+    },
+    {
+      "epoch": 0.2962911126662001,
+      "grad_norm": 0.4187348195942707,
+      "learning_rate": 4.127709862653146e-05,
+      "loss": 0.5915,
+      "step": 2117
+    },
+    {
+      "epoch": 0.29643107067879637,
+      "grad_norm": 0.43072765488595965,
+      "learning_rate": 4.126849490945958e-05,
+      "loss": 0.5869,
+      "step": 2118
+    },
+    {
+      "epoch": 0.29657102869139257,
+      "grad_norm": 0.40631929550004103,
+      "learning_rate": 4.1259887849046906e-05,
+      "loss": 0.5783,
+      "step": 2119
+    },
+    {
+      "epoch": 0.2967109867039888,
+      "grad_norm": 0.4236924640290844,
+      "learning_rate": 4.1251277447062315e-05,
+      "loss": 0.64,
+      "step": 2120
+    },
+    {
+      "epoch": 0.296850944716585,
+      "grad_norm": 0.42013017681077297,
+      "learning_rate": 4.124266370527531e-05,
+      "loss": 0.6384,
+      "step": 2121
+    },
+    {
+      "epoch": 0.29699090272918127,
+      "grad_norm": 0.4065769677758428,
+      "learning_rate": 4.123404662545611e-05,
+      "loss": 0.6086,
+      "step": 2122
+    },
+    {
+      "epoch": 0.29713086074177747,
+      "grad_norm": 0.3886256061886866,
+      "learning_rate": 4.1225426209375605e-05,
+      "loss": 0.5219,
+      "step": 2123
+    },
+    {
+      "epoch": 0.29727081875437367,
+      "grad_norm": 0.40830396890274406,
+      "learning_rate": 4.121680245880539e-05,
+      "loss": 0.5487,
+      "step": 2124
+    },
+    {
+      "epoch": 0.2974107767669699,
+      "grad_norm": 0.40457862036093745,
+      "learning_rate": 4.120817537551773e-05,
+      "loss": 0.5797,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2975507347795661,
+      "grad_norm": 0.414048107015411,
+      "learning_rate": 4.1199544961285574e-05,
+      "loss": 0.5827,
+      "step": 2126
+    },
+    {
+      "epoch": 0.29769069279216237,
+      "grad_norm": 0.4211166151095332,
+      "learning_rate": 4.119091121788256e-05,
+      "loss": 0.5869,
+      "step": 2127
+    },
+    {
+      "epoch": 0.29783065080475857,
+      "grad_norm": 0.3999830413220613,
+      "learning_rate": 4.1182274147083e-05,
+      "loss": 0.5538,
+      "step": 2128
+    },
+    {
+      "epoch": 0.2979706088173548,
+      "grad_norm": 0.43614503719146447,
+      "learning_rate": 4.1173633750661915e-05,
+      "loss": 0.6079,
+      "step": 2129
+    },
+    {
+      "epoch": 0.298110566829951,
+      "grad_norm": 0.4284744155784332,
+      "learning_rate": 4.116499003039499e-05,
+      "loss": 0.5669,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2982505248425472,
+      "grad_norm": 0.4178472294126034,
+      "learning_rate": 4.1156342988058606e-05,
+      "loss": 0.5466,
+      "step": 2131
+    },
+    {
+      "epoch": 0.29839048285514347,
+      "grad_norm": 0.4104945666216928,
+      "learning_rate": 4.114769262542979e-05,
+      "loss": 0.5566,
+      "step": 2132
+    },
+    {
+      "epoch": 0.29853044086773967,
+      "grad_norm": 0.5142114902119391,
+      "learning_rate": 4.113903894428632e-05,
+      "loss": 0.5706,
+      "step": 2133
+    },
+    {
+      "epoch": 0.2986703988803359,
+      "grad_norm": 0.42568394409367394,
+      "learning_rate": 4.113038194640658e-05,
+      "loss": 0.5772,
+      "step": 2134
+    },
+    {
+      "epoch": 0.2988103568929321,
+      "grad_norm": 0.4072689251705309,
+      "learning_rate": 4.112172163356969e-05,
+      "loss": 0.5839,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2989503149055283,
+      "grad_norm": 0.43230154146342653,
+      "learning_rate": 4.1113058007555415e-05,
+      "loss": 0.5762,
+      "step": 2136
+    },
+    {
+      "epoch": 0.29909027291812457,
+      "grad_norm": 0.42221307246185413,
+      "learning_rate": 4.110439107014423e-05,
+      "loss": 0.5854,
+      "step": 2137
+    },
+    {
+      "epoch": 0.29923023093072076,
+      "grad_norm": 0.4115485110413119,
+      "learning_rate": 4.109572082311729e-05,
+      "loss": 0.6144,
+      "step": 2138
+    },
+    {
+      "epoch": 0.299370188943317,
+      "grad_norm": 0.4381710550273501,
+      "learning_rate": 4.10870472682564e-05,
+      "loss": 0.5764,
+      "step": 2139
+    },
+    {
+      "epoch": 0.2995101469559132,
+      "grad_norm": 0.40672792105491173,
+      "learning_rate": 4.107837040734407e-05,
+      "loss": 0.5623,
+      "step": 2140
+    },
+    {
+      "epoch": 0.29965010496850947,
+      "grad_norm": 0.4230148958590366,
+      "learning_rate": 4.1069690242163484e-05,
+      "loss": 0.6007,
+      "step": 2141
+    },
+    {
+      "epoch": 0.29979006298110566,
+      "grad_norm": 0.42487403889075154,
+      "learning_rate": 4.106100677449851e-05,
+      "loss": 0.5952,
+      "step": 2142
+    },
+    {
+      "epoch": 0.29993002099370186,
+      "grad_norm": 0.41800613830620925,
+      "learning_rate": 4.105232000613367e-05,
+      "loss": 0.5814,
+      "step": 2143
+    },
+    {
+      "epoch": 0.3000699790062981,
+      "grad_norm": 0.42812390915287574,
+      "learning_rate": 4.10436299388542e-05,
+      "loss": 0.6044,
+      "step": 2144
+    },
+    {
+      "epoch": 0.3002099370188943,
+      "grad_norm": 0.42630144567604744,
+      "learning_rate": 4.103493657444599e-05,
+      "loss": 0.5458,
+      "step": 2145
+    },
+    {
+      "epoch": 0.30034989503149057,
+      "grad_norm": 0.40026869359194683,
+      "learning_rate": 4.1026239914695617e-05,
+      "loss": 0.567,
+      "step": 2146
+    },
+    {
+      "epoch": 0.30048985304408676,
+      "grad_norm": 0.4045104989308898,
+      "learning_rate": 4.101753996139033e-05,
+      "loss": 0.5573,
+      "step": 2147
+    },
+    {
+      "epoch": 0.300629811056683,
+      "grad_norm": 0.39480816764383886,
+      "learning_rate": 4.100883671631806e-05,
+      "loss": 0.5554,
+      "step": 2148
+    },
+    {
+      "epoch": 0.3007697690692792,
+      "grad_norm": 0.4359485340891318,
+      "learning_rate": 4.100013018126742e-05,
+      "loss": 0.5306,
+      "step": 2149
+    },
+    {
+      "epoch": 0.3009097270818754,
+      "grad_norm": 0.44274809745131777,
+      "learning_rate": 4.099142035802767e-05,
+      "loss": 0.6069,
+      "step": 2150
+    },
+    {
+      "epoch": 0.30104968509447166,
+      "grad_norm": 0.4010134732813477,
+      "learning_rate": 4.098270724838879e-05,
+      "loss": 0.5737,
+      "step": 2151
+    },
+    {
+      "epoch": 0.30118964310706786,
+      "grad_norm": 0.42583763567956523,
+      "learning_rate": 4.09739908541414e-05,
+      "loss": 0.5646,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3013296011196641,
+      "grad_norm": 0.43076744664262784,
+      "learning_rate": 4.096527117707681e-05,
+      "loss": 0.6089,
+      "step": 2153
+    },
+    {
+      "epoch": 0.3014695591322603,
+      "grad_norm": 0.41900293437499037,
+      "learning_rate": 4.0956548218987004e-05,
+      "loss": 0.5818,
+      "step": 2154
+    },
+    {
+      "epoch": 0.30160951714485656,
+      "grad_norm": 0.441326688768721,
+      "learning_rate": 4.094782198166464e-05,
+      "loss": 0.601,
+      "step": 2155
+    },
+    {
+      "epoch": 0.30174947515745276,
+      "grad_norm": 0.41905876085866123,
+      "learning_rate": 4.093909246690303e-05,
+      "loss": 0.5694,
+      "step": 2156
+    },
+    {
+      "epoch": 0.30188943317004896,
+      "grad_norm": 0.40710066709424864,
+      "learning_rate": 4.093035967649621e-05,
+      "loss": 0.5571,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3020293911826452,
+      "grad_norm": 0.41211015638134546,
+      "learning_rate": 4.092162361223884e-05,
+      "loss": 0.6152,
+      "step": 2158
+    },
+    {
+      "epoch": 0.3021693491952414,
+      "grad_norm": 0.42106082661170496,
+      "learning_rate": 4.091288427592626e-05,
+      "loss": 0.5927,
+      "step": 2159
+    },
+    {
+      "epoch": 0.30230930720783766,
+      "grad_norm": 0.4141524195584368,
+      "learning_rate": 4.0904141669354504e-05,
+      "loss": 0.6316,
+      "step": 2160
+    },
+    {
+      "epoch": 0.30244926522043386,
+      "grad_norm": 0.44132339838745016,
+      "learning_rate": 4.0895395794320265e-05,
+      "loss": 0.5985,
+      "step": 2161
+    },
+    {
+      "epoch": 0.3025892232330301,
+      "grad_norm": 0.4276646446648299,
+      "learning_rate": 4.088664665262091e-05,
+      "loss": 0.6097,
+      "step": 2162
+    },
+    {
+      "epoch": 0.3027291812456263,
+      "grad_norm": 0.4150132159301798,
+      "learning_rate": 4.087789424605447e-05,
+      "loss": 0.5909,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3028691392582225,
+      "grad_norm": 0.39307522840638937,
+      "learning_rate": 4.0869138576419665e-05,
+      "loss": 0.5914,
+      "step": 2164
+    },
+    {
+      "epoch": 0.30300909727081876,
+      "grad_norm": 0.38594672832612004,
+      "learning_rate": 4.086037964551587e-05,
+      "loss": 0.5607,
+      "step": 2165
+    },
+    {
+      "epoch": 0.30314905528341496,
+      "grad_norm": 0.4044083531551891,
+      "learning_rate": 4.085161745514312e-05,
+      "loss": 0.5844,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3032890132960112,
+      "grad_norm": 0.42910899298854893,
+      "learning_rate": 4.084285200710215e-05,
+      "loss": 0.5998,
+      "step": 2167
+    },
+    {
+      "epoch": 0.3034289713086074,
+      "grad_norm": 0.4246336566951562,
+      "learning_rate": 4.083408330319435e-05,
+      "loss": 0.5741,
+      "step": 2168
+    },
+    {
+      "epoch": 0.30356892932120366,
+      "grad_norm": 0.4257821385233403,
+      "learning_rate": 4.082531134522176e-05,
+      "loss": 0.6026,
+      "step": 2169
+    },
+    {
+      "epoch": 0.30370888733379986,
+      "grad_norm": 0.4191758702628161,
+      "learning_rate": 4.0816536134987124e-05,
+      "loss": 0.6057,
+      "step": 2170
+    },
+    {
+      "epoch": 0.30384884534639606,
+      "grad_norm": 0.4038665214466415,
+      "learning_rate": 4.0807757674293834e-05,
+      "loss": 0.5794,
+      "step": 2171
+    },
+    {
+      "epoch": 0.3039888033589923,
+      "grad_norm": 0.38253864598092674,
+      "learning_rate": 4.079897596494594e-05,
+      "loss": 0.5572,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3041287613715885,
+      "grad_norm": 0.40057781747525356,
+      "learning_rate": 4.0790191008748193e-05,
+      "loss": 0.5757,
+      "step": 2173
+    },
+    {
+      "epoch": 0.30426871938418476,
+      "grad_norm": 0.4425784348877834,
+      "learning_rate": 4.078140280750597e-05,
+      "loss": 0.5909,
+      "step": 2174
+    },
+    {
+      "epoch": 0.30440867739678096,
+      "grad_norm": 0.43807792095543074,
+      "learning_rate": 4.077261136302536e-05,
+      "loss": 0.5713,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3045486354093772,
+      "grad_norm": 0.42494113081632956,
+      "learning_rate": 4.0763816677113064e-05,
+      "loss": 0.5599,
+      "step": 2176
+    },
+    {
+      "epoch": 0.3046885934219734,
+      "grad_norm": 0.4147870888377848,
+      "learning_rate": 4.07550187515765e-05,
+      "loss": 0.5958,
+      "step": 2177
+    },
+    {
+      "epoch": 0.3048285514345696,
+      "grad_norm": 0.44106090618207494,
+      "learning_rate": 4.0746217588223724e-05,
+      "loss": 0.5719,
+      "step": 2178
+    },
+    {
+      "epoch": 0.30496850944716586,
+      "grad_norm": 0.43022335538791223,
+      "learning_rate": 4.073741318886347e-05,
+      "loss": 0.5666,
+      "step": 2179
+    },
+    {
+      "epoch": 0.30510846745976206,
+      "grad_norm": 0.422655488810057,
+      "learning_rate": 4.072860555530512e-05,
+      "loss": 0.5725,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3052484254723583,
+      "grad_norm": 0.41560143611944184,
+      "learning_rate": 4.071979468935874e-05,
+      "loss": 0.5707,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3053883834849545,
+      "grad_norm": 0.40831478451563363,
+      "learning_rate": 4.071098059283505e-05,
+      "loss": 0.5724,
+      "step": 2182
+    },
+    {
+      "epoch": 0.30552834149755076,
+      "grad_norm": 0.45412422141258324,
+      "learning_rate": 4.070216326754544e-05,
+      "loss": 0.5914,
+      "step": 2183
+    },
+    {
+      "epoch": 0.30566829951014696,
+      "grad_norm": 0.4238846330415573,
+      "learning_rate": 4.069334271530196e-05,
+      "loss": 0.5852,
+      "step": 2184
+    },
+    {
+      "epoch": 0.30580825752274315,
+      "grad_norm": 0.4168947592800387,
+      "learning_rate": 4.0684518937917315e-05,
+      "loss": 0.5668,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3059482155353394,
+      "grad_norm": 0.3974079353132702,
+      "learning_rate": 4.0675691937204886e-05,
+      "loss": 0.5796,
+      "step": 2186
+    },
+    {
+      "epoch": 0.3060881735479356,
+      "grad_norm": 0.447707780183876,
+      "learning_rate": 4.0666861714978724e-05,
+      "loss": 0.6053,
+      "step": 2187
+    },
+    {
+      "epoch": 0.30622813156053186,
+      "grad_norm": 0.4241731286712751,
+      "learning_rate": 4.06580282730535e-05,
+      "loss": 0.5433,
+      "step": 2188
+    },
+    {
+      "epoch": 0.30636808957312806,
+      "grad_norm": 0.40634600846513996,
+      "learning_rate": 4.0649191613244596e-05,
+      "loss": 0.5778,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3065080475857243,
+      "grad_norm": 0.40317717922477264,
+      "learning_rate": 4.064035173736804e-05,
+      "loss": 0.5654,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3066480055983205,
+      "grad_norm": 0.40439388741840737,
+      "learning_rate": 4.06315086472405e-05,
+      "loss": 0.554,
+      "step": 2191
+    },
+    {
+      "epoch": 0.3067879636109167,
+      "grad_norm": 0.40052999103675635,
+      "learning_rate": 4.062266234467933e-05,
+      "loss": 0.5633,
+      "step": 2192
+    },
+    {
+      "epoch": 0.30692792162351296,
+      "grad_norm": 0.39039005393866494,
+      "learning_rate": 4.061381283150254e-05,
+      "loss": 0.5712,
+      "step": 2193
+    },
+    {
+      "epoch": 0.30706787963610915,
+      "grad_norm": 0.4452309708879627,
+      "learning_rate": 4.0604960109528786e-05,
+      "loss": 0.5951,
+      "step": 2194
+    },
+    {
+      "epoch": 0.3072078376487054,
+      "grad_norm": 0.4316897315393044,
+      "learning_rate": 4.059610418057739e-05,
+      "loss": 0.5643,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3073477956613016,
+      "grad_norm": 0.43497901635487507,
+      "learning_rate": 4.058724504646834e-05,
+      "loss": 0.6104,
+      "step": 2196
+    },
+    {
+      "epoch": 0.30748775367389786,
+      "grad_norm": 0.3971370299202904,
+      "learning_rate": 4.057838270902228e-05,
+      "loss": 0.5678,
+      "step": 2197
+    },
+    {
+      "epoch": 0.30762771168649405,
+      "grad_norm": 0.4316780175497349,
+      "learning_rate": 4.056951717006051e-05,
+      "loss": 0.6105,
+      "step": 2198
+    },
+    {
+      "epoch": 0.30776766969909025,
+      "grad_norm": 0.39695048154674517,
+      "learning_rate": 4.056064843140498e-05,
+      "loss": 0.5726,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3079076277116865,
+      "grad_norm": 0.41439706383436253,
+      "learning_rate": 4.0551776494878316e-05,
+      "loss": 0.568,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3080475857242827,
+      "grad_norm": 0.4191940957424624,
+      "learning_rate": 4.054290136230379e-05,
+      "loss": 0.6245,
+      "step": 2201
+    },
+    {
+      "epoch": 0.30818754373687895,
+      "grad_norm": 0.4143832100728591,
+      "learning_rate": 4.053402303550533e-05,
+      "loss": 0.5795,
+      "step": 2202
+    },
+    {
+      "epoch": 0.30832750174947515,
+      "grad_norm": 0.3915343927508335,
+      "learning_rate": 4.052514151630752e-05,
+      "loss": 0.5543,
+      "step": 2203
+    },
+    {
+      "epoch": 0.3084674597620714,
+      "grad_norm": 0.41805518170359557,
+      "learning_rate": 4.0516256806535594e-05,
+      "loss": 0.5891,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3086074177746676,
+      "grad_norm": 0.41287829384767516,
+      "learning_rate": 4.050736890801547e-05,
+      "loss": 0.569,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3087473757872638,
+      "grad_norm": 0.4029096483520546,
+      "learning_rate": 4.049847782257369e-05,
+      "loss": 0.5859,
+      "step": 2206
+    },
+    {
+      "epoch": 0.30888733379986005,
+      "grad_norm": 0.39417183897543207,
+      "learning_rate": 4.048958355203746e-05,
+      "loss": 0.5282,
+      "step": 2207
+    },
+    {
+      "epoch": 0.30902729181245625,
+      "grad_norm": 0.3952415992081539,
+      "learning_rate": 4.048068609823464e-05,
+      "loss": 0.5668,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3091672498250525,
+      "grad_norm": 0.4072747277650614,
+      "learning_rate": 4.047178546299376e-05,
+      "loss": 0.5977,
+      "step": 2209
+    },
+    {
+      "epoch": 0.3093072078376487,
+      "grad_norm": 0.4390729532532709,
+      "learning_rate": 4.0462881648143977e-05,
+      "loss": 0.6085,
+      "step": 2210
+    },
+    {
+      "epoch": 0.30944716585024495,
+      "grad_norm": 0.41550618366447567,
+      "learning_rate": 4.045397465551513e-05,
+      "loss": 0.5501,
+      "step": 2211
+    },
+    {
+      "epoch": 0.30958712386284115,
+      "grad_norm": 0.42949098915048817,
+      "learning_rate": 4.044506448693769e-05,
+      "loss": 0.5786,
+      "step": 2212
+    },
+    {
+      "epoch": 0.30972708187543735,
+      "grad_norm": 0.4180486476641866,
+      "learning_rate": 4.0436151144242776e-05,
+      "loss": 0.5759,
+      "step": 2213
+    },
+    {
+      "epoch": 0.3098670398880336,
+      "grad_norm": 0.4151740836794609,
+      "learning_rate": 4.042723462926219e-05,
+      "loss": 0.5858,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3100069979006298,
+      "grad_norm": 0.41105102208793937,
+      "learning_rate": 4.041831494382835e-05,
+      "loss": 0.5285,
+      "step": 2215
+    },
+    {
+      "epoch": 0.31014695591322605,
+      "grad_norm": 0.40516018829246475,
+      "learning_rate": 4.040939208977435e-05,
+      "loss": 0.55,
+      "step": 2216
+    },
+    {
+      "epoch": 0.31028691392582225,
+      "grad_norm": 0.42729080222511606,
+      "learning_rate": 4.040046606893392e-05,
+      "loss": 0.5872,
+      "step": 2217
+    },
+    {
+      "epoch": 0.31042687193841845,
+      "grad_norm": 0.4226725585719995,
+      "learning_rate": 4.039153688314145e-05,
+      "loss": 0.5882,
+      "step": 2218
+    },
+    {
+      "epoch": 0.3105668299510147,
+      "grad_norm": 0.40702190282971207,
+      "learning_rate": 4.0382604534232e-05,
+      "loss": 0.5968,
+      "step": 2219
+    },
+    {
+      "epoch": 0.3107067879636109,
+      "grad_norm": 0.4266606910520138,
+      "learning_rate": 4.0373669024041226e-05,
+      "loss": 0.5822,
+      "step": 2220
+    },
+    {
+      "epoch": 0.31084674597620715,
+      "grad_norm": 0.4163690162383508,
+      "learning_rate": 4.0364730354405475e-05,
+      "loss": 0.5829,
+      "step": 2221
+    },
+    {
+      "epoch": 0.31098670398880335,
+      "grad_norm": 0.40577915115104807,
+      "learning_rate": 4.035578852716175e-05,
+      "loss": 0.5733,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3111266620013996,
+      "grad_norm": 0.4011948223027502,
+      "learning_rate": 4.034684354414767e-05,
+      "loss": 0.6159,
+      "step": 2223
+    },
+    {
+      "epoch": 0.3112666200139958,
+      "grad_norm": 0.40735551230540823,
+      "learning_rate": 4.0337895407201527e-05,
+      "loss": 0.5897,
+      "step": 2224
+    },
+    {
+      "epoch": 0.311406578026592,
+      "grad_norm": 0.4007832693673057,
+      "learning_rate": 4.0328944118162255e-05,
+      "loss": 0.5823,
+      "step": 2225
+    },
+    {
+      "epoch": 0.31154653603918825,
+      "grad_norm": 0.4220826406609178,
+      "learning_rate": 4.0319989678869426e-05,
+      "loss": 0.603,
+      "step": 2226
+    },
+    {
+      "epoch": 0.31168649405178445,
+      "grad_norm": 0.42993833632780304,
+      "learning_rate": 4.031103209116328e-05,
+      "loss": 0.5959,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3118264520643807,
+      "grad_norm": 0.405630507124848,
+      "learning_rate": 4.030207135688468e-05,
+      "loss": 0.5846,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3119664100769769,
+      "grad_norm": 0.40554860006169224,
+      "learning_rate": 4.029310747787516e-05,
+      "loss": 0.5809,
+      "step": 2229
+    },
+    {
+      "epoch": 0.31210636808957315,
+      "grad_norm": 0.38272639115415097,
+      "learning_rate": 4.028414045597688e-05,
+      "loss": 0.5554,
+      "step": 2230
+    },
+    {
+      "epoch": 0.31224632610216935,
+      "grad_norm": 0.4133830894332048,
+      "learning_rate": 4.027517029303266e-05,
+      "loss": 0.5844,
+      "step": 2231
+    },
+    {
+      "epoch": 0.31238628411476554,
+      "grad_norm": 0.40469368564073166,
+      "learning_rate": 4.0266196990885955e-05,
+      "loss": 0.5692,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3125262421273618,
+      "grad_norm": 0.43419090564140506,
+      "learning_rate": 4.025722055138087e-05,
+      "loss": 0.5766,
+      "step": 2233
+    },
+    {
+      "epoch": 0.312666200139958,
+      "grad_norm": 0.4192206941532246,
+      "learning_rate": 4.0248240976362154e-05,
+      "loss": 0.5916,
+      "step": 2234
+    },
+    {
+      "epoch": 0.31280615815255425,
+      "grad_norm": 0.40494171533150775,
+      "learning_rate": 4.02392582676752e-05,
+      "loss": 0.5405,
+      "step": 2235
+    },
+    {
+      "epoch": 0.31294611616515045,
+      "grad_norm": 0.41642501232810825,
+      "learning_rate": 4.023027242716606e-05,
+      "loss": 0.5658,
+      "step": 2236
+    },
+    {
+      "epoch": 0.3130860741777467,
+      "grad_norm": 0.4179541672168489,
+      "learning_rate": 4.022128345668139e-05,
+      "loss": 0.5747,
+      "step": 2237
+    },
+    {
+      "epoch": 0.3132260321903429,
+      "grad_norm": 0.42437584151566893,
+      "learning_rate": 4.021229135806853e-05,
+      "loss": 0.5839,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3133659902029391,
+      "grad_norm": 0.39708478174893375,
+      "learning_rate": 4.020329613317545e-05,
+      "loss": 0.5491,
+      "step": 2239
+    },
+    {
+      "epoch": 0.31350594821553535,
+      "grad_norm": 0.40589223534794583,
+      "learning_rate": 4.0194297783850755e-05,
+      "loss": 0.5767,
+      "step": 2240
+    },
+    {
+      "epoch": 0.31364590622813154,
+      "grad_norm": 0.4186147684647929,
+      "learning_rate": 4.018529631194369e-05,
+      "loss": 0.5754,
+      "step": 2241
+    },
+    {
+      "epoch": 0.3137858642407278,
+      "grad_norm": 0.4066652101804317,
+      "learning_rate": 4.017629171930416e-05,
+      "loss": 0.5655,
+      "step": 2242
+    },
+    {
+      "epoch": 0.313925822253324,
+      "grad_norm": 0.4383552206858084,
+      "learning_rate": 4.01672840077827e-05,
+      "loss": 0.5829,
+      "step": 2243
+    },
+    {
+      "epoch": 0.31406578026592025,
+      "grad_norm": 0.422916994665018,
+      "learning_rate": 4.0158273179230475e-05,
+      "loss": 0.6183,
+      "step": 2244
+    },
+    {
+      "epoch": 0.31420573827851644,
+      "grad_norm": 0.4255816299563645,
+      "learning_rate": 4.0149259235499317e-05,
+      "loss": 0.6063,
+      "step": 2245
+    },
+    {
+      "epoch": 0.31434569629111264,
+      "grad_norm": 0.43561902805107455,
+      "learning_rate": 4.014024217844167e-05,
+      "loss": 0.572,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3144856543037089,
+      "grad_norm": 0.4257236548174317,
+      "learning_rate": 4.013122200991064e-05,
+      "loss": 0.5752,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3146256123163051,
+      "grad_norm": 0.38763550151633236,
+      "learning_rate": 4.012219873175995e-05,
+      "loss": 0.5677,
+      "step": 2248
+    },
+    {
+      "epoch": 0.31476557032890135,
+      "grad_norm": 0.4175429066296847,
+      "learning_rate": 4.0113172345843983e-05,
+      "loss": 0.5595,
+      "step": 2249
+    },
+    {
+      "epoch": 0.31490552834149754,
+      "grad_norm": 0.4480069652429578,
+      "learning_rate": 4.010414285401777e-05,
+      "loss": 0.5681,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3150454863540938,
+      "grad_norm": 0.40107954134135065,
+      "learning_rate": 4.009511025813694e-05,
+      "loss": 0.5706,
+      "step": 2251
+    },
+    {
+      "epoch": 0.31518544436669,
+      "grad_norm": 0.40774375327862394,
+      "learning_rate": 4.008607456005778e-05,
+      "loss": 0.5684,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3153254023792862,
+      "grad_norm": 0.4374841181743415,
+      "learning_rate": 4.007703576163724e-05,
+      "loss": 0.5685,
+      "step": 2253
+    },
+    {
+      "epoch": 0.31546536039188244,
+      "grad_norm": 0.41257242312452364,
+      "learning_rate": 4.006799386473287e-05,
+      "loss": 0.591,
+      "step": 2254
+    },
+    {
+      "epoch": 0.31560531840447864,
+      "grad_norm": 0.4114995646091396,
+      "learning_rate": 4.005894887120287e-05,
+      "loss": 0.5519,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3157452764170749,
+      "grad_norm": 0.4076070427270775,
+      "learning_rate": 4.0049900782906086e-05,
+      "loss": 0.5844,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3158852344296711,
+      "grad_norm": 0.4132478127299326,
+      "learning_rate": 4.004084960170199e-05,
+      "loss": 0.5655,
+      "step": 2257
+    },
+    {
+      "epoch": 0.31602519244226734,
+      "grad_norm": 0.4076919229201329,
+      "learning_rate": 4.0031795329450685e-05,
+      "loss": 0.5885,
+      "step": 2258
+    },
+    {
+      "epoch": 0.31616515045486354,
+      "grad_norm": 0.421676106935922,
+      "learning_rate": 4.002273796801292e-05,
+      "loss": 0.6067,
+      "step": 2259
+    },
+    {
+      "epoch": 0.31630510846745974,
+      "grad_norm": 0.3902189908536509,
+      "learning_rate": 4.001367751925008e-05,
+      "loss": 0.5666,
+      "step": 2260
+    },
+    {
+      "epoch": 0.316445066480056,
+      "grad_norm": 0.42674732993318637,
+      "learning_rate": 4.000461398502418e-05,
+      "loss": 0.5944,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3165850244926522,
+      "grad_norm": 0.40075738490963975,
+      "learning_rate": 3.9995547367197845e-05,
+      "loss": 0.565,
+      "step": 2262
+    },
+    {
+      "epoch": 0.31672498250524844,
+      "grad_norm": 0.42762891854787655,
+      "learning_rate": 3.998647766763438e-05,
+      "loss": 0.6214,
+      "step": 2263
+    },
+    {
+      "epoch": 0.31686494051784464,
+      "grad_norm": 0.3961782540593822,
+      "learning_rate": 3.9977404888197704e-05,
+      "loss": 0.5454,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3170048985304409,
+      "grad_norm": 0.3923939088232202,
+      "learning_rate": 3.996832903075235e-05,
+      "loss": 0.5883,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3171448565430371,
+      "grad_norm": 0.42198649975518954,
+      "learning_rate": 3.99592500971635e-05,
+      "loss": 0.5809,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3172848145556333,
+      "grad_norm": 0.4083907533890683,
+      "learning_rate": 3.995016808929698e-05,
+      "loss": 0.5423,
+      "step": 2267
+    },
+    {
+      "epoch": 0.31742477256822954,
+      "grad_norm": 0.39942363696045713,
+      "learning_rate": 3.9941083009019223e-05,
+      "loss": 0.561,
+      "step": 2268
+    },
+    {
+      "epoch": 0.31756473058082574,
+      "grad_norm": 0.4262274605313951,
+      "learning_rate": 3.993199485819731e-05,
+      "loss": 0.5875,
+      "step": 2269
+    },
+    {
+      "epoch": 0.317704688593422,
+      "grad_norm": 0.38915227058740126,
+      "learning_rate": 3.992290363869895e-05,
+      "loss": 0.5531,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3178446466060182,
+      "grad_norm": 0.43560718886498834,
+      "learning_rate": 3.9913809352392474e-05,
+      "loss": 0.5923,
+      "step": 2271
+    },
+    {
+      "epoch": 0.31798460461861444,
+      "grad_norm": 0.44620165771992476,
+      "learning_rate": 3.990471200114685e-05,
+      "loss": 0.608,
+      "step": 2272
+    },
+    {
+      "epoch": 0.31812456263121064,
+      "grad_norm": 0.4359041969176294,
+      "learning_rate": 3.9895611586831685e-05,
+      "loss": 0.5733,
+      "step": 2273
+    },
+    {
+      "epoch": 0.31826452064380684,
+      "grad_norm": 0.42976996875212653,
+      "learning_rate": 3.98865081113172e-05,
+      "loss": 0.5701,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3184044786564031,
+      "grad_norm": 0.40393224396318145,
+      "learning_rate": 3.987740157647426e-05,
+      "loss": 0.5764,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3185444366689993,
+      "grad_norm": 0.42936298439237425,
+      "learning_rate": 3.986829198417433e-05,
+      "loss": 0.5922,
+      "step": 2276
+    },
+    {
+      "epoch": 0.31868439468159554,
+      "grad_norm": 0.40635602349462724,
+      "learning_rate": 3.985917933628955e-05,
+      "loss": 0.5492,
+      "step": 2277
+    },
+    {
+      "epoch": 0.31882435269419174,
+      "grad_norm": 0.38817161800821404,
+      "learning_rate": 3.9850063634692635e-05,
+      "loss": 0.6025,
+      "step": 2278
+    },
+    {
+      "epoch": 0.318964310706788,
+      "grad_norm": 0.40934279574178284,
+      "learning_rate": 3.984094488125698e-05,
+      "loss": 0.5779,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3191042687193842,
+      "grad_norm": 0.4265395827777272,
+      "learning_rate": 3.983182307785657e-05,
+      "loss": 0.5947,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3192442267319804,
+      "grad_norm": 0.4131217561851261,
+      "learning_rate": 3.982269822636602e-05,
+      "loss": 0.5659,
+      "step": 2281
+    },
+    {
+      "epoch": 0.31938418474457664,
+      "grad_norm": 0.40502407520544415,
+      "learning_rate": 3.981357032866058e-05,
+      "loss": 0.618,
+      "step": 2282
+    },
+    {
+      "epoch": 0.31952414275717284,
+      "grad_norm": 0.4162474667828162,
+      "learning_rate": 3.980443938661614e-05,
+      "loss": 0.5778,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3196641007697691,
+      "grad_norm": 0.41057477420812005,
+      "learning_rate": 3.9795305402109195e-05,
+      "loss": 0.5432,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3198040587823653,
+      "grad_norm": 0.4199988291247849,
+      "learning_rate": 3.9786168377016866e-05,
+      "loss": 0.5615,
+      "step": 2285
+    },
+    {
+      "epoch": 0.31994401679496154,
+      "grad_norm": 0.4154620895424027,
+      "learning_rate": 3.977702831321692e-05,
+      "loss": 0.5774,
+      "step": 2286
+    },
+    {
+      "epoch": 0.32008397480755774,
+      "grad_norm": 0.42692144319966024,
+      "learning_rate": 3.976788521258771e-05,
+      "loss": 0.6076,
+      "step": 2287
+    },
+    {
+      "epoch": 0.32022393282015393,
+      "grad_norm": 0.41668448669509417,
+      "learning_rate": 3.975873907700825e-05,
+      "loss": 0.5973,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3203638908327502,
+      "grad_norm": 0.4268759954887578,
+      "learning_rate": 3.974958990835816e-05,
+      "loss": 0.581,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3205038488453464,
+      "grad_norm": 0.3952528234342034,
+      "learning_rate": 3.974043770851769e-05,
+      "loss": 0.5613,
+      "step": 2290
+    },
+    {
+      "epoch": 0.32064380685794264,
+      "grad_norm": 0.44836986260917555,
+      "learning_rate": 3.9731282479367706e-05,
+      "loss": 0.6248,
+      "step": 2291
+    },
+    {
+      "epoch": 0.32078376487053883,
+      "grad_norm": 0.4256932127612938,
+      "learning_rate": 3.9722124222789705e-05,
+      "loss": 0.6071,
+      "step": 2292
+    },
+    {
+      "epoch": 0.32092372288313503,
+      "grad_norm": 0.44510004429106764,
+      "learning_rate": 3.97129629406658e-05,
+      "loss": 0.5797,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3210636808957313,
+      "grad_norm": 0.4302627905158645,
+      "learning_rate": 3.970379863487872e-05,
+      "loss": 0.5665,
+      "step": 2294
+    },
+    {
+      "epoch": 0.3212036389083275,
+      "grad_norm": 0.41235485841130853,
+      "learning_rate": 3.969463130731183e-05,
+      "loss": 0.5634,
+      "step": 2295
+    },
+    {
+      "epoch": 0.32134359692092374,
+      "grad_norm": 0.41894260330344535,
+      "learning_rate": 3.9685460959849105e-05,
+      "loss": 0.5867,
+      "step": 2296
+    },
+    {
+      "epoch": 0.32148355493351993,
+      "grad_norm": 0.41203034907703107,
+      "learning_rate": 3.967628759437516e-05,
+      "loss": 0.59,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3216235129461162,
+      "grad_norm": 0.4281005271912537,
+      "learning_rate": 3.96671112127752e-05,
+      "loss": 0.6006,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3217634709587124,
+      "grad_norm": 0.4075127794042514,
+      "learning_rate": 3.965793181693506e-05,
+      "loss": 0.565,
+      "step": 2299
+    },
+    {
+      "epoch": 0.3219034289713086,
+      "grad_norm": 0.39763128617859633,
+      "learning_rate": 3.96487494087412e-05,
+      "loss": 0.5437,
+      "step": 2300
+    },
+    {
+      "epoch": 0.32204338698390483,
+      "grad_norm": 0.4552074769557467,
+      "learning_rate": 3.9639563990080716e-05,
+      "loss": 0.6201,
+      "step": 2301
+    },
+    {
+      "epoch": 0.32218334499650103,
+      "grad_norm": 0.40413041195679866,
+      "learning_rate": 3.9630375562841295e-05,
+      "loss": 0.5526,
+      "step": 2302
+    },
+    {
+      "epoch": 0.3223233030090973,
+      "grad_norm": 0.40283666703685167,
+      "learning_rate": 3.9621184128911234e-05,
+      "loss": 0.579,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3224632610216935,
+      "grad_norm": 0.3980665357469519,
+      "learning_rate": 3.96119896901795e-05,
+      "loss": 0.5626,
+      "step": 2304
+    },
+    {
+      "epoch": 0.32260321903428973,
+      "grad_norm": 0.42655702692861863,
+      "learning_rate": 3.9602792248535606e-05,
+      "loss": 0.6075,
+      "step": 2305
+    },
+    {
+      "epoch": 0.32274317704688593,
+      "grad_norm": 0.4826045217103083,
+      "learning_rate": 3.959359180586975e-05,
+      "loss": 0.5799,
+      "step": 2306
+    },
+    {
+      "epoch": 0.32288313505948213,
+      "grad_norm": 0.43008580043844,
+      "learning_rate": 3.95843883640727e-05,
+      "loss": 0.5787,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3230230930720784,
+      "grad_norm": 0.42351832550150303,
+      "learning_rate": 3.957518192503587e-05,
+      "loss": 0.5565,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3231630510846746,
+      "grad_norm": 0.40482672190396396,
+      "learning_rate": 3.956597249065126e-05,
+      "loss": 0.5533,
+      "step": 2309
+    },
+    {
+      "epoch": 0.32330300909727083,
+      "grad_norm": 0.4107029056895161,
+      "learning_rate": 3.95567600628115e-05,
+      "loss": 0.5756,
+      "step": 2310
+    },
+    {
+      "epoch": 0.32344296710986703,
+      "grad_norm": 0.43221123666830596,
+      "learning_rate": 3.954754464340987e-05,
+      "loss": 0.6074,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3235829251224633,
+      "grad_norm": 0.42051407755166026,
+      "learning_rate": 3.9538326234340194e-05,
+      "loss": 0.5948,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3237228831350595,
+      "grad_norm": 0.38578825068740574,
+      "learning_rate": 3.9529104837496974e-05,
+      "loss": 0.5692,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3238628411476557,
+      "grad_norm": 0.42139866225485473,
+      "learning_rate": 3.951988045477529e-05,
+      "loss": 0.5892,
+      "step": 2314
+    },
+    {
+      "epoch": 0.32400279916025193,
+      "grad_norm": 0.43110899316398094,
+      "learning_rate": 3.9510653088070846e-05,
+      "loss": 0.5698,
+      "step": 2315
+    },
+    {
+      "epoch": 0.32414275717284813,
+      "grad_norm": 0.4426212316446372,
+      "learning_rate": 3.9501422739279956e-05,
+      "loss": 0.5727,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3242827151854444,
+      "grad_norm": 0.42108819792246993,
+      "learning_rate": 3.9492189410299566e-05,
+      "loss": 0.554,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3244226731980406,
+      "grad_norm": 0.46665094478938685,
+      "learning_rate": 3.94829531030272e-05,
+      "loss": 0.6323,
+      "step": 2318
+    },
+    {
+      "epoch": 0.32456263121063683,
+      "grad_norm": 0.4138605211792902,
+      "learning_rate": 3.9473713819361015e-05,
+      "loss": 0.5446,
+      "step": 2319
+    },
+    {
+      "epoch": 0.32470258922323303,
+      "grad_norm": 0.43505576166032045,
+      "learning_rate": 3.946447156119979e-05,
+      "loss": 0.5664,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3248425472358292,
+      "grad_norm": 0.41878878141867854,
+      "learning_rate": 3.945522633044289e-05,
+      "loss": 0.608,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3249825052484255,
+      "grad_norm": 0.40004594511350894,
+      "learning_rate": 3.9445978128990326e-05,
+      "loss": 0.5454,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3251224632610217,
+      "grad_norm": 0.42281186251192865,
+      "learning_rate": 3.943672695874267e-05,
+      "loss": 0.567,
+      "step": 2323
+    },
+    {
+      "epoch": 0.32526242127361793,
+      "grad_norm": 0.4068793191045584,
+      "learning_rate": 3.942747282160114e-05,
+      "loss": 0.5424,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3254023792862141,
+      "grad_norm": 0.42411112127273465,
+      "learning_rate": 3.9418215719467565e-05,
+      "loss": 0.567,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3255423372988104,
+      "grad_norm": 0.4111794082883435,
+      "learning_rate": 3.9408955654244364e-05,
+      "loss": 0.5682,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3256822953114066,
+      "grad_norm": 0.39707493084776485,
+      "learning_rate": 3.939969262783457e-05,
+      "loss": 0.5541,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3258222533240028,
+      "grad_norm": 0.3941781607373903,
+      "learning_rate": 3.939042664214184e-05,
+      "loss": 0.5511,
+      "step": 2328
+    },
+    {
+      "epoch": 0.32596221133659903,
+      "grad_norm": 0.4150783846079092,
+      "learning_rate": 3.9381157699070424e-05,
+      "loss": 0.6085,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3261021693491952,
+      "grad_norm": 0.392787000567577,
+      "learning_rate": 3.937188580052518e-05,
+      "loss": 0.5834,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3262421273617915,
+      "grad_norm": 0.41765704626894373,
+      "learning_rate": 3.9362610948411585e-05,
+      "loss": 0.5497,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3263820853743877,
+      "grad_norm": 0.4175224418697374,
+      "learning_rate": 3.9353333144635706e-05,
+      "loss": 0.5597,
+      "step": 2332
+    },
+    {
+      "epoch": 0.32652204338698393,
+      "grad_norm": 0.4243451821154263,
+      "learning_rate": 3.934405239110423e-05,
+      "loss": 0.5884,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3266620013995801,
+      "grad_norm": 0.4267515430222131,
+      "learning_rate": 3.9334768689724456e-05,
+      "loss": 0.606,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3268019594121763,
+      "grad_norm": 0.4332693190274577,
+      "learning_rate": 3.932548204240426e-05,
+      "loss": 0.5995,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3269419174247726,
+      "grad_norm": 0.4121506429858514,
+      "learning_rate": 3.931619245105216e-05,
+      "loss": 0.5787,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3270818754373688,
+      "grad_norm": 0.4180451871893415,
+      "learning_rate": 3.9306899917577245e-05,
+      "loss": 0.5546,
+      "step": 2337
+    },
+    {
+      "epoch": 0.327221833449965,
+      "grad_norm": 0.40942456121846177,
+      "learning_rate": 3.9297604443889234e-05,
+      "loss": 0.6098,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3273617914625612,
+      "grad_norm": 0.4067754001916415,
+      "learning_rate": 3.928830603189844e-05,
+      "loss": 0.5756,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3275017494751575,
+      "grad_norm": 0.4140918426830124,
+      "learning_rate": 3.9279004683515783e-05,
+      "loss": 0.5641,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3276417074877537,
+      "grad_norm": 0.4212820521655152,
+      "learning_rate": 3.926970040065278e-05,
+      "loss": 0.604,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3277816655003499,
+      "grad_norm": 0.40003067107763557,
+      "learning_rate": 3.9260393185221564e-05,
+      "loss": 0.565,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3279216235129461,
+      "grad_norm": 0.39827204127063837,
+      "learning_rate": 3.925108303913485e-05,
+      "loss": 0.5799,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3280615815255423,
+      "grad_norm": 0.4233494400434675,
+      "learning_rate": 3.9241769964305976e-05,
+      "loss": 0.6037,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3282015395381386,
+      "grad_norm": 0.4118719992278687,
+      "learning_rate": 3.9232453962648864e-05,
+      "loss": 0.6098,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3283414975507348,
+      "grad_norm": 0.4016000541614996,
+      "learning_rate": 3.9223135036078064e-05,
+      "loss": 0.5734,
+      "step": 2346
+    },
+    {
+      "epoch": 0.328481455563331,
+      "grad_norm": 0.4237468615429023,
+      "learning_rate": 3.921381318650869e-05,
+      "loss": 0.6088,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3286214135759272,
+      "grad_norm": 0.46652462106861353,
+      "learning_rate": 3.920448841585649e-05,
+      "loss": 0.5814,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3287613715885234,
+      "grad_norm": 0.39685764815363,
+      "learning_rate": 3.9195160726037805e-05,
+      "loss": 0.5646,
+      "step": 2349
+    },
+    {
+      "epoch": 0.3289013296011197,
+      "grad_norm": 0.4229891713627349,
+      "learning_rate": 3.918583011896955e-05,
+      "loss": 0.5773,
+      "step": 2350
+    },
+    {
+      "epoch": 0.32904128761371587,
+      "grad_norm": 0.3879674012353026,
+      "learning_rate": 3.9176496596569265e-05,
+      "loss": 0.5397,
+      "step": 2351
+    },
+    {
+      "epoch": 0.3291812456263121,
+      "grad_norm": 0.3901223408311448,
+      "learning_rate": 3.91671601607551e-05,
+      "loss": 0.5736,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3293212036389083,
+      "grad_norm": 0.4002239555751096,
+      "learning_rate": 3.915782081344578e-05,
+      "loss": 0.5967,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3294611616515046,
+      "grad_norm": 0.41288801391640517,
+      "learning_rate": 3.914847855656062e-05,
+      "loss": 0.5715,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3296011196641008,
+      "grad_norm": 0.4184700560414015,
+      "learning_rate": 3.913913339201956e-05,
+      "loss": 0.5507,
+      "step": 2355
+    },
+    {
+      "epoch": 0.32974107767669697,
+      "grad_norm": 0.3894927148762295,
+      "learning_rate": 3.9129785321743125e-05,
+      "loss": 0.5705,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3298810356892932,
+      "grad_norm": 0.3847000287394412,
+      "learning_rate": 3.912043434765245e-05,
+      "loss": 0.5484,
+      "step": 2357
+    },
+    {
+      "epoch": 0.3300209937018894,
+      "grad_norm": 0.4403932420001484,
+      "learning_rate": 3.911108047166924e-05,
+      "loss": 0.6139,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3301609517144857,
+      "grad_norm": 0.4352977938001518,
+      "learning_rate": 3.9101723695715805e-05,
+      "loss": 0.6006,
+      "step": 2359
+    },
+    {
+      "epoch": 0.33030090972708187,
+      "grad_norm": 0.40492957509367133,
+      "learning_rate": 3.909236402171508e-05,
+      "loss": 0.6026,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3304408677396781,
+      "grad_norm": 0.3867521703349589,
+      "learning_rate": 3.908300145159055e-05,
+      "loss": 0.5449,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3305808257522743,
+      "grad_norm": 0.4404641851361547,
+      "learning_rate": 3.907363598726635e-05,
+      "loss": 0.5905,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3307207837648705,
+      "grad_norm": 0.41840972294550066,
+      "learning_rate": 3.906426763066714e-05,
+      "loss": 0.5632,
+      "step": 2363
+    },
+    {
+      "epoch": 0.33086074177746677,
+      "grad_norm": 0.4162282322770041,
+      "learning_rate": 3.905489638371823e-05,
+      "loss": 0.5913,
+      "step": 2364
+    },
+    {
+      "epoch": 0.33100069979006297,
+      "grad_norm": 0.3955450449429818,
+      "learning_rate": 3.9045522248345504e-05,
+      "loss": 0.5324,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3311406578026592,
+      "grad_norm": 0.49091811650049194,
+      "learning_rate": 3.903614522647545e-05,
+      "loss": 0.5769,
+      "step": 2366
+    },
+    {
+      "epoch": 0.3312806158152554,
+      "grad_norm": 0.4173793487196645,
+      "learning_rate": 3.902676532003514e-05,
+      "loss": 0.5886,
+      "step": 2367
+    },
+    {
+      "epoch": 0.33142057382785167,
+      "grad_norm": 0.3932967546945319,
+      "learning_rate": 3.901738253095222e-05,
+      "loss": 0.5621,
+      "step": 2368
+    },
+    {
+      "epoch": 0.33156053184044787,
+      "grad_norm": 0.43569710224836344,
+      "learning_rate": 3.900799686115498e-05,
+      "loss": 0.5741,
+      "step": 2369
+    },
+    {
+      "epoch": 0.33170048985304407,
+      "grad_norm": 0.4115145852184928,
+      "learning_rate": 3.899860831257224e-05,
+      "loss": 0.5259,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3318404478656403,
+      "grad_norm": 0.4381698932866813,
+      "learning_rate": 3.898921688713346e-05,
+      "loss": 0.6024,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3319804058782365,
+      "grad_norm": 0.4208901188098798,
+      "learning_rate": 3.897982258676867e-05,
+      "loss": 0.5527,
+      "step": 2372
+    },
+    {
+      "epoch": 0.33212036389083277,
+      "grad_norm": 0.41912261407751483,
+      "learning_rate": 3.8970425413408495e-05,
+      "loss": 0.5822,
+      "step": 2373
+    },
+    {
+      "epoch": 0.33226032190342897,
+      "grad_norm": 0.4040927772770925,
+      "learning_rate": 3.896102536898415e-05,
+      "loss": 0.5491,
+      "step": 2374
+    },
+    {
+      "epoch": 0.33240027991602517,
+      "grad_norm": 0.4173965549079611,
+      "learning_rate": 3.8951622455427425e-05,
+      "loss": 0.6025,
+      "step": 2375
+    },
+    {
+      "epoch": 0.3325402379286214,
+      "grad_norm": 0.404116401312251,
+      "learning_rate": 3.894221667467074e-05,
+      "loss": 0.553,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3326801959412176,
+      "grad_norm": 0.39814070442365174,
+      "learning_rate": 3.893280802864706e-05,
+      "loss": 0.5538,
+      "step": 2377
+    },
+    {
+      "epoch": 0.33282015395381387,
+      "grad_norm": 0.4097199303736863,
+      "learning_rate": 3.892339651928996e-05,
+      "loss": 0.5562,
+      "step": 2378
+    },
+    {
+      "epoch": 0.33296011196641007,
+      "grad_norm": 0.4251433986281386,
+      "learning_rate": 3.89139821485336e-05,
+      "loss": 0.5886,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3331000699790063,
+      "grad_norm": 0.3965548232917148,
+      "learning_rate": 3.8904564918312737e-05,
+      "loss": 0.5536,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3332400279916025,
+      "grad_norm": 0.4012944065952801,
+      "learning_rate": 3.889514483056269e-05,
+      "loss": 0.5742,
+      "step": 2381
+    },
+    {
+      "epoch": 0.3333799860041987,
+      "grad_norm": 0.4201060112427097,
+      "learning_rate": 3.8885721887219406e-05,
+      "loss": 0.5691,
+      "step": 2382
+    },
+    {
+      "epoch": 0.33351994401679497,
+      "grad_norm": 0.42817185933177176,
+      "learning_rate": 3.887629609021938e-05,
+      "loss": 0.5678,
+      "step": 2383
+    },
+    {
+      "epoch": 0.33365990202939116,
+      "grad_norm": 0.43405350104277596,
+      "learning_rate": 3.88668674414997e-05,
+      "loss": 0.6216,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3337998600419874,
+      "grad_norm": 0.42315727231869327,
+      "learning_rate": 3.8857435942998074e-05,
+      "loss": 0.575,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3339398180545836,
+      "grad_norm": 0.40648841814651454,
+      "learning_rate": 3.884800159665276e-05,
+      "loss": 0.5715,
+      "step": 2386
+    },
+    {
+      "epoch": 0.33407977606717987,
+      "grad_norm": 0.426901177821104,
+      "learning_rate": 3.883856440440261e-05,
+      "loss": 0.5788,
+      "step": 2387
+    },
+    {
+      "epoch": 0.33421973407977607,
+      "grad_norm": 0.4264982996567364,
+      "learning_rate": 3.882912436818705e-05,
+      "loss": 0.5849,
+      "step": 2388
+    },
+    {
+      "epoch": 0.33435969209237226,
+      "grad_norm": 0.41759550845666904,
+      "learning_rate": 3.881968148994613e-05,
+      "loss": 0.5524,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3344996501049685,
+      "grad_norm": 0.3922959061793526,
+      "learning_rate": 3.881023577162044e-05,
+      "loss": 0.556,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3346396081175647,
+      "grad_norm": 0.3988789451222829,
+      "learning_rate": 3.880078721515117e-05,
+      "loss": 0.5743,
+      "step": 2391
+    },
+    {
+      "epoch": 0.33477956613016097,
+      "grad_norm": 0.4206132245414296,
+      "learning_rate": 3.87913358224801e-05,
+      "loss": 0.5688,
+      "step": 2392
+    },
+    {
+      "epoch": 0.33491952414275716,
+      "grad_norm": 0.400794205539379,
+      "learning_rate": 3.878188159554959e-05,
+      "loss": 0.6014,
+      "step": 2393
+    },
+    {
+      "epoch": 0.3350594821553534,
+      "grad_norm": 0.4098374598496262,
+      "learning_rate": 3.8772424536302564e-05,
+      "loss": 0.6093,
+      "step": 2394
+    },
+    {
+      "epoch": 0.3351994401679496,
+      "grad_norm": 0.4250977633595455,
+      "learning_rate": 3.876296464668257e-05,
+      "loss": 0.606,
+      "step": 2395
+    },
+    {
+      "epoch": 0.3353393981805458,
+      "grad_norm": 0.4220441415552252,
+      "learning_rate": 3.875350192863368e-05,
+      "loss": 0.5897,
+      "step": 2396
+    },
+    {
+      "epoch": 0.33547935619314206,
+      "grad_norm": 0.41667093576332226,
+      "learning_rate": 3.87440363841006e-05,
+      "loss": 0.633,
+      "step": 2397
+    },
+    {
+      "epoch": 0.33561931420573826,
+      "grad_norm": 0.4157738741324718,
+      "learning_rate": 3.873456801502859e-05,
+      "loss": 0.5703,
+      "step": 2398
+    },
+    {
+      "epoch": 0.3357592722183345,
+      "grad_norm": 0.4258564142322055,
+      "learning_rate": 3.872509682336351e-05,
+      "loss": 0.5919,
+      "step": 2399
+    },
+    {
+      "epoch": 0.3358992302309307,
+      "grad_norm": 0.42470627438795794,
+      "learning_rate": 3.871562281105175e-05,
+      "loss": 0.5662,
+      "step": 2400
+    },
+    {
+      "epoch": 0.33603918824352697,
+      "grad_norm": 0.4186178764102384,
+      "learning_rate": 3.8706145980040344e-05,
+      "loss": 0.581,
+      "step": 2401
+    },
+    {
+      "epoch": 0.33617914625612316,
+      "grad_norm": 0.3987870833859228,
+      "learning_rate": 3.8696666332276875e-05,
+      "loss": 0.585,
+      "step": 2402
+    },
+    {
+      "epoch": 0.33631910426871936,
+      "grad_norm": 0.40669896712395964,
+      "learning_rate": 3.86871838697095e-05,
+      "loss": 0.5793,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3364590622813156,
+      "grad_norm": 0.4324505891171469,
+      "learning_rate": 3.867769859428697e-05,
+      "loss": 0.5895,
+      "step": 2404
+    },
+    {
+      "epoch": 0.3365990202939118,
+      "grad_norm": 0.4089670782754292,
+      "learning_rate": 3.866821050795859e-05,
+      "loss": 0.614,
+      "step": 2405
+    },
+    {
+      "epoch": 0.33673897830650806,
+      "grad_norm": 0.5537632601572994,
+      "learning_rate": 3.865871961267427e-05,
+      "loss": 0.5969,
+      "step": 2406
+    },
+    {
+      "epoch": 0.33687893631910426,
+      "grad_norm": 0.4149474055553234,
+      "learning_rate": 3.864922591038448e-05,
+      "loss": 0.5521,
+      "step": 2407
+    },
+    {
+      "epoch": 0.3370188943317005,
+      "grad_norm": 0.426552813649065,
+      "learning_rate": 3.863972940304028e-05,
+      "loss": 0.6032,
+      "step": 2408
+    },
+    {
+      "epoch": 0.3371588523442967,
+      "grad_norm": 0.41223595995661894,
+      "learning_rate": 3.863023009259329e-05,
+      "loss": 0.5785,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3372988103568929,
+      "grad_norm": 0.4024674567365589,
+      "learning_rate": 3.8620727980995716e-05,
+      "loss": 0.5589,
+      "step": 2410
+    },
+    {
+      "epoch": 0.33743876836948916,
+      "grad_norm": 0.41476951241547977,
+      "learning_rate": 3.861122307020034e-05,
+      "loss": 0.5648,
+      "step": 2411
+    },
+    {
+      "epoch": 0.33757872638208536,
+      "grad_norm": 0.4024211020152443,
+      "learning_rate": 3.860171536216052e-05,
+      "loss": 0.5569,
+      "step": 2412
+    },
+    {
+      "epoch": 0.3377186843946816,
+      "grad_norm": 0.41785686679485745,
+      "learning_rate": 3.8592204858830175e-05,
+      "loss": 0.6093,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3378586424072778,
+      "grad_norm": 0.41460218899468515,
+      "learning_rate": 3.858269156216383e-05,
+      "loss": 0.5795,
+      "step": 2414
+    },
+    {
+      "epoch": 0.33799860041987406,
+      "grad_norm": 0.40446354182657146,
+      "learning_rate": 3.8573175474116545e-05,
+      "loss": 0.549,
+      "step": 2415
+    },
+    {
+      "epoch": 0.33813855843247026,
+      "grad_norm": 0.4037463725904796,
+      "learning_rate": 3.856365659664399e-05,
+      "loss": 0.5356,
+      "step": 2416
+    },
+    {
+      "epoch": 0.33827851644506646,
+      "grad_norm": 0.4182076163889934,
+      "learning_rate": 3.855413493170237e-05,
+      "loss": 0.6045,
+      "step": 2417
+    },
+    {
+      "epoch": 0.3384184744576627,
+      "grad_norm": 0.42186938322484663,
+      "learning_rate": 3.85446104812485e-05,
+      "loss": 0.6155,
+      "step": 2418
+    },
+    {
+      "epoch": 0.3385584324702589,
+      "grad_norm": 0.41470003229003616,
+      "learning_rate": 3.853508324723976e-05,
+      "loss": 0.5701,
+      "step": 2419
+    },
+    {
+      "epoch": 0.33869839048285516,
+      "grad_norm": 0.41867300233582005,
+      "learning_rate": 3.852555323163406e-05,
+      "loss": 0.5762,
+      "step": 2420
+    },
+    {
+      "epoch": 0.33883834849545136,
+      "grad_norm": 0.43083623709965385,
+      "learning_rate": 3.851602043638994e-05,
+      "loss": 0.5928,
+      "step": 2421
+    },
+    {
+      "epoch": 0.3389783065080476,
+      "grad_norm": 0.4112941630964681,
+      "learning_rate": 3.850648486346649e-05,
+      "loss": 0.5671,
+      "step": 2422
+    },
+    {
+      "epoch": 0.3391182645206438,
+      "grad_norm": 0.4081030227054361,
+      "learning_rate": 3.849694651482335e-05,
+      "loss": 0.5662,
+      "step": 2423
+    },
+    {
+      "epoch": 0.33925822253324,
+      "grad_norm": 0.41990746427748654,
+      "learning_rate": 3.848740539242075e-05,
+      "loss": 0.6186,
+      "step": 2424
+    },
+    {
+      "epoch": 0.33939818054583626,
+      "grad_norm": 0.41509855881507357,
+      "learning_rate": 3.84778614982195e-05,
+      "loss": 0.5777,
+      "step": 2425
+    },
+    {
+      "epoch": 0.33953813855843246,
+      "grad_norm": 0.41529612645730785,
+      "learning_rate": 3.8468314834180954e-05,
+      "loss": 0.5834,
+      "step": 2426
+    },
+    {
+      "epoch": 0.3396780965710287,
+      "grad_norm": 0.4612886169509395,
+      "learning_rate": 3.845876540226706e-05,
+      "loss": 0.5898,
+      "step": 2427
+    },
+    {
+      "epoch": 0.3398180545836249,
+      "grad_norm": 0.40004450008966375,
+      "learning_rate": 3.844921320444031e-05,
+      "loss": 0.5677,
+      "step": 2428
+    },
+    {
+      "epoch": 0.33995801259622116,
+      "grad_norm": 0.42665920823328024,
+      "learning_rate": 3.843965824266379e-05,
+      "loss": 0.5643,
+      "step": 2429
+    },
+    {
+      "epoch": 0.34009797060881736,
+      "grad_norm": 0.4124069827888389,
+      "learning_rate": 3.843010051890114e-05,
+      "loss": 0.6075,
+      "step": 2430
+    },
+    {
+      "epoch": 0.34023792862141355,
+      "grad_norm": 0.41303867930204274,
+      "learning_rate": 3.842054003511656e-05,
+      "loss": 0.5811,
+      "step": 2431
+    },
+    {
+      "epoch": 0.3403778866340098,
+      "grad_norm": 0.4196697445658453,
+      "learning_rate": 3.841097679327483e-05,
+      "loss": 0.5525,
+      "step": 2432
+    },
+    {
+      "epoch": 0.340517844646606,
+      "grad_norm": 0.4360442351288732,
+      "learning_rate": 3.840141079534131e-05,
+      "loss": 0.6244,
+      "step": 2433
+    },
+    {
+      "epoch": 0.34065780265920226,
+      "grad_norm": 0.39392921636528555,
+      "learning_rate": 3.839184204328188e-05,
+      "loss": 0.5494,
+      "step": 2434
+    },
+    {
+      "epoch": 0.34079776067179846,
+      "grad_norm": 0.39649367381567424,
+      "learning_rate": 3.8382270539063035e-05,
+      "loss": 0.5741,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3409377186843947,
+      "grad_norm": 0.4170815607651056,
+      "learning_rate": 3.8372696284651814e-05,
+      "loss": 0.5759,
+      "step": 2436
+    },
+    {
+      "epoch": 0.3410776766969909,
+      "grad_norm": 0.3997155811593266,
+      "learning_rate": 3.8363119282015816e-05,
+      "loss": 0.5702,
+      "step": 2437
+    },
+    {
+      "epoch": 0.3412176347095871,
+      "grad_norm": 0.4282009363533291,
+      "learning_rate": 3.835353953312322e-05,
+      "loss": 0.5891,
+      "step": 2438
+    },
+    {
+      "epoch": 0.34135759272218336,
+      "grad_norm": 0.41921744383670845,
+      "learning_rate": 3.834395703994276e-05,
+      "loss": 0.5803,
+      "step": 2439
+    },
+    {
+      "epoch": 0.34149755073477955,
+      "grad_norm": 0.40474108379603385,
+      "learning_rate": 3.833437180444373e-05,
+      "loss": 0.6013,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3416375087473758,
+      "grad_norm": 0.42511536225444024,
+      "learning_rate": 3.8324783828596e-05,
+      "loss": 0.5705,
+      "step": 2441
+    },
+    {
+      "epoch": 0.341777466759972,
+      "grad_norm": 0.3949585768953575,
+      "learning_rate": 3.8315193114369996e-05,
+      "loss": 0.5617,
+      "step": 2442
+    },
+    {
+      "epoch": 0.34191742477256826,
+      "grad_norm": 0.39765016272736614,
+      "learning_rate": 3.830559966373671e-05,
+      "loss": 0.5164,
+      "step": 2443
+    },
+    {
+      "epoch": 0.34205738278516445,
+      "grad_norm": 0.3878551915452198,
+      "learning_rate": 3.829600347866768e-05,
+      "loss": 0.5661,
+      "step": 2444
+    },
+    {
+      "epoch": 0.34219734079776065,
+      "grad_norm": 0.4346559633685042,
+      "learning_rate": 3.828640456113504e-05,
+      "loss": 0.6013,
+      "step": 2445
+    },
+    {
+      "epoch": 0.3423372988103569,
+      "grad_norm": 0.4218764947636453,
+      "learning_rate": 3.827680291311143e-05,
+      "loss": 0.5684,
+      "step": 2446
+    },
+    {
+      "epoch": 0.3424772568229531,
+      "grad_norm": 0.40065712577833346,
+      "learning_rate": 3.8267198536570123e-05,
+      "loss": 0.5313,
+      "step": 2447
+    },
+    {
+      "epoch": 0.34261721483554936,
+      "grad_norm": 0.4050494830959936,
+      "learning_rate": 3.8257591433484906e-05,
+      "loss": 0.5632,
+      "step": 2448
+    },
+    {
+      "epoch": 0.34275717284814555,
+      "grad_norm": 0.41579240423995417,
+      "learning_rate": 3.824798160583012e-05,
+      "loss": 0.5719,
+      "step": 2449
+    },
+    {
+      "epoch": 0.34289713086074175,
+      "grad_norm": 0.41342697420725416,
+      "learning_rate": 3.823836905558071e-05,
+      "loss": 0.5825,
+      "step": 2450
+    },
+    {
+      "epoch": 0.343037088873338,
+      "grad_norm": 0.4242841636188481,
+      "learning_rate": 3.8228753784712126e-05,
+      "loss": 0.5917,
+      "step": 2451
+    },
+    {
+      "epoch": 0.3431770468859342,
+      "grad_norm": 0.3909799623795335,
+      "learning_rate": 3.8219135795200417e-05,
+      "loss": 0.5294,
+      "step": 2452
+    },
+    {
+      "epoch": 0.34331700489853045,
+      "grad_norm": 0.4191631316056565,
+      "learning_rate": 3.8209515089022175e-05,
+      "loss": 0.6145,
+      "step": 2453
+    },
+    {
+      "epoch": 0.34345696291112665,
+      "grad_norm": 0.41444745056939064,
+      "learning_rate": 3.819989166815455e-05,
+      "loss": 0.5448,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3435969209237229,
+      "grad_norm": 0.41319072983975424,
+      "learning_rate": 3.8190265534575256e-05,
+      "loss": 0.5493,
+      "step": 2455
+    },
+    {
+      "epoch": 0.3437368789363191,
+      "grad_norm": 0.4144210912873605,
+      "learning_rate": 3.818063669026256e-05,
+      "loss": 0.5699,
+      "step": 2456
+    },
+    {
+      "epoch": 0.3438768369489153,
+      "grad_norm": 0.43371658028513654,
+      "learning_rate": 3.817100513719529e-05,
+      "loss": 0.5994,
+      "step": 2457
+    },
+    {
+      "epoch": 0.34401679496151155,
+      "grad_norm": 0.39358381227775163,
+      "learning_rate": 3.8161370877352825e-05,
+      "loss": 0.5772,
+      "step": 2458
+    },
+    {
+      "epoch": 0.34415675297410775,
+      "grad_norm": 0.4524806967713819,
+      "learning_rate": 3.815173391271511e-05,
+      "loss": 0.5801,
+      "step": 2459
+    },
+    {
+      "epoch": 0.344296710986704,
+      "grad_norm": 0.42248629141745564,
+      "learning_rate": 3.814209424526262e-05,
+      "loss": 0.5344,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3444366689993002,
+      "grad_norm": 0.4222729879954454,
+      "learning_rate": 3.813245187697643e-05,
+      "loss": 0.6305,
+      "step": 2461
+    },
+    {
+      "epoch": 0.34457662701189645,
+      "grad_norm": 0.40274756440667137,
+      "learning_rate": 3.812280680983812e-05,
+      "loss": 0.5587,
+      "step": 2462
+    },
+    {
+      "epoch": 0.34471658502449265,
+      "grad_norm": 0.4041971062889942,
+      "learning_rate": 3.8113159045829864e-05,
+      "loss": 0.5837,
+      "step": 2463
+    },
+    {
+      "epoch": 0.34485654303708885,
+      "grad_norm": 0.437567215932465,
+      "learning_rate": 3.8103508586934365e-05,
+      "loss": 0.6047,
+      "step": 2464
+    },
+    {
+      "epoch": 0.3449965010496851,
+      "grad_norm": 0.4015334832237547,
+      "learning_rate": 3.8093855435134914e-05,
+      "loss": 0.5504,
+      "step": 2465
+    },
+    {
+      "epoch": 0.3451364590622813,
+      "grad_norm": 0.42379890127600006,
+      "learning_rate": 3.8084199592415305e-05,
+      "loss": 0.5723,
+      "step": 2466
+    },
+    {
+      "epoch": 0.34527641707487755,
+      "grad_norm": 0.41715521655535315,
+      "learning_rate": 3.8074541060759925e-05,
+      "loss": 0.5857,
+      "step": 2467
+    },
+    {
+      "epoch": 0.34541637508747375,
+      "grad_norm": 0.43326783152483606,
+      "learning_rate": 3.806487984215369e-05,
+      "loss": 0.5603,
+      "step": 2468
+    },
+    {
+      "epoch": 0.34555633310007,
+      "grad_norm": 0.403241790805183,
+      "learning_rate": 3.8055215938582086e-05,
+      "loss": 0.5727,
+      "step": 2469
+    },
+    {
+      "epoch": 0.3456962911126662,
+      "grad_norm": 0.4061061612551589,
+      "learning_rate": 3.804554935203115e-05,
+      "loss": 0.5846,
+      "step": 2470
+    },
+    {
+      "epoch": 0.3458362491252624,
+      "grad_norm": 0.40214409369497894,
+      "learning_rate": 3.803588008448745e-05,
+      "loss": 0.5655,
+      "step": 2471
+    },
+    {
+      "epoch": 0.34597620713785865,
+      "grad_norm": 0.4213923914769129,
+      "learning_rate": 3.802620813793814e-05,
+      "loss": 0.5632,
+      "step": 2472
+    },
+    {
+      "epoch": 0.34611616515045485,
+      "grad_norm": 0.41515367784081497,
+      "learning_rate": 3.801653351437087e-05,
+      "loss": 0.562,
+      "step": 2473
+    },
+    {
+      "epoch": 0.3462561231630511,
+      "grad_norm": 0.424690045157795,
+      "learning_rate": 3.8006856215773895e-05,
+      "loss": 0.5889,
+      "step": 2474
+    },
+    {
+      "epoch": 0.3463960811756473,
+      "grad_norm": 0.43537881056181427,
+      "learning_rate": 3.799717624413599e-05,
+      "loss": 0.6304,
+      "step": 2475
+    },
+    {
+      "epoch": 0.34653603918824355,
+      "grad_norm": 0.40770798716655837,
+      "learning_rate": 3.798749360144651e-05,
+      "loss": 0.5743,
+      "step": 2476
+    },
+    {
+      "epoch": 0.34667599720083975,
+      "grad_norm": 0.42860038780017473,
+      "learning_rate": 3.7977808289695306e-05,
+      "loss": 0.6029,
+      "step": 2477
+    },
+    {
+      "epoch": 0.34681595521343594,
+      "grad_norm": 0.41741937935524637,
+      "learning_rate": 3.796812031087281e-05,
+      "loss": 0.5745,
+      "step": 2478
+    },
+    {
+      "epoch": 0.3469559132260322,
+      "grad_norm": 0.4169469655781413,
+      "learning_rate": 3.7958429666970024e-05,
+      "loss": 0.5932,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3470958712386284,
+      "grad_norm": 0.41794717504089257,
+      "learning_rate": 3.7948736359978455e-05,
+      "loss": 0.5957,
+      "step": 2480
+    },
+    {
+      "epoch": 0.34723582925122465,
+      "grad_norm": 0.42237578550439164,
+      "learning_rate": 3.793904039189018e-05,
+      "loss": 0.5809,
+      "step": 2481
+    },
+    {
+      "epoch": 0.34737578726382085,
+      "grad_norm": 0.42457946983643247,
+      "learning_rate": 3.7929341764697816e-05,
+      "loss": 0.5741,
+      "step": 2482
+    },
+    {
+      "epoch": 0.3475157452764171,
+      "grad_norm": 0.40686808588079204,
+      "learning_rate": 3.791964048039454e-05,
+      "loss": 0.577,
+      "step": 2483
+    },
+    {
+      "epoch": 0.3476557032890133,
+      "grad_norm": 0.3917392399092806,
+      "learning_rate": 3.790993654097405e-05,
+      "loss": 0.5389,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3477956613016095,
+      "grad_norm": 0.41842414036596237,
+      "learning_rate": 3.790022994843061e-05,
+      "loss": 0.6111,
+      "step": 2485
+    },
+    {
+      "epoch": 0.34793561931420575,
+      "grad_norm": 0.40264506155704255,
+      "learning_rate": 3.789052070475903e-05,
+      "loss": 0.5691,
+      "step": 2486
+    },
+    {
+      "epoch": 0.34807557732680194,
+      "grad_norm": 0.41841559375670284,
+      "learning_rate": 3.788080881195465e-05,
+      "loss": 0.5796,
+      "step": 2487
+    },
+    {
+      "epoch": 0.3482155353393982,
+      "grad_norm": 0.42238316240978235,
+      "learning_rate": 3.787109427201337e-05,
+      "loss": 0.5721,
+      "step": 2488
+    },
+    {
+      "epoch": 0.3483554933519944,
+      "grad_norm": 0.4261343625599289,
+      "learning_rate": 3.7861377086931615e-05,
+      "loss": 0.5762,
+      "step": 2489
+    },
+    {
+      "epoch": 0.34849545136459065,
+      "grad_norm": 0.40050526138599146,
+      "learning_rate": 3.785165725870638e-05,
+      "loss": 0.5771,
+      "step": 2490
+    },
+    {
+      "epoch": 0.34863540937718684,
+      "grad_norm": 0.4011183639320898,
+      "learning_rate": 3.7841934789335164e-05,
+      "loss": 0.5722,
+      "step": 2491
+    },
+    {
+      "epoch": 0.34877536738978304,
+      "grad_norm": 0.41515525438995227,
+      "learning_rate": 3.783220968081606e-05,
+      "loss": 0.5863,
+      "step": 2492
+    },
+    {
+      "epoch": 0.3489153254023793,
+      "grad_norm": 0.3930742764120805,
+      "learning_rate": 3.782248193514766e-05,
+      "loss": 0.531,
+      "step": 2493
+    },
+    {
+      "epoch": 0.3490552834149755,
+      "grad_norm": 0.41871895607441606,
+      "learning_rate": 3.7812751554329116e-05,
+      "loss": 0.5925,
+      "step": 2494
+    },
+    {
+      "epoch": 0.34919524142757175,
+      "grad_norm": 0.39386200492237877,
+      "learning_rate": 3.780301854036013e-05,
+      "loss": 0.6283,
+      "step": 2495
+    },
+    {
+      "epoch": 0.34933519944016794,
+      "grad_norm": 0.4562215128945137,
+      "learning_rate": 3.7793282895240926e-05,
+      "loss": 0.6435,
+      "step": 2496
+    },
+    {
+      "epoch": 0.3494751574527642,
+      "grad_norm": 0.40380935634400195,
+      "learning_rate": 3.778354462097229e-05,
+      "loss": 0.5874,
+      "step": 2497
+    },
+    {
+      "epoch": 0.3496151154653604,
+      "grad_norm": 0.37718467161463654,
+      "learning_rate": 3.7773803719555514e-05,
+      "loss": 0.5451,
+      "step": 2498
+    },
+    {
+      "epoch": 0.3497550734779566,
+      "grad_norm": 0.49862732423279105,
+      "learning_rate": 3.776406019299247e-05,
+      "loss": 0.566,
+      "step": 2499
+    },
+    {
+      "epoch": 0.34989503149055284,
+      "grad_norm": 0.3890315425799742,
+      "learning_rate": 3.7754314043285556e-05,
+      "loss": 0.541,
+      "step": 2500
+    },
+    {
+      "epoch": 0.35003498950314904,
+      "grad_norm": 0.4020611603166375,
+      "learning_rate": 3.774456527243768e-05,
+      "loss": 0.5703,
+      "step": 2501
+    },
+    {
+      "epoch": 0.3501749475157453,
+      "grad_norm": 0.4024001534789896,
+      "learning_rate": 3.7734813882452334e-05,
+      "loss": 0.5596,
+      "step": 2502
+    },
+    {
+      "epoch": 0.3503149055283415,
+      "grad_norm": 0.41123367006288974,
+      "learning_rate": 3.772505987533352e-05,
+      "loss": 0.5635,
+      "step": 2503
+    },
+    {
+      "epoch": 0.35045486354093774,
+      "grad_norm": 0.42428174810444297,
+      "learning_rate": 3.771530325308579e-05,
+      "loss": 0.5814,
+      "step": 2504
+    },
+    {
+      "epoch": 0.35059482155353394,
+      "grad_norm": 0.4038080676069653,
+      "learning_rate": 3.770554401771423e-05,
+      "loss": 0.5727,
+      "step": 2505
+    },
+    {
+      "epoch": 0.35073477956613014,
+      "grad_norm": 0.3839599878758453,
+      "learning_rate": 3.769578217122446e-05,
+      "loss": 0.5834,
+      "step": 2506
+    },
+    {
+      "epoch": 0.3508747375787264,
+      "grad_norm": 0.4158023938010911,
+      "learning_rate": 3.768601771562262e-05,
+      "loss": 0.5712,
+      "step": 2507
+    },
+    {
+      "epoch": 0.3510146955913226,
+      "grad_norm": 0.3943326004175278,
+      "learning_rate": 3.767625065291544e-05,
+      "loss": 0.568,
+      "step": 2508
+    },
+    {
+      "epoch": 0.35115465360391884,
+      "grad_norm": 0.3964702492091683,
+      "learning_rate": 3.766648098511012e-05,
+      "loss": 0.576,
+      "step": 2509
+    },
+    {
+      "epoch": 0.35129461161651504,
+      "grad_norm": 0.3921536448421572,
+      "learning_rate": 3.765670871421445e-05,
+      "loss": 0.5209,
+      "step": 2510
+    },
+    {
+      "epoch": 0.3514345696291113,
+      "grad_norm": 0.4351764473354882,
+      "learning_rate": 3.764693384223671e-05,
+      "loss": 0.5926,
+      "step": 2511
+    },
+    {
+      "epoch": 0.3515745276417075,
+      "grad_norm": 0.40588260989157027,
+      "learning_rate": 3.763715637118575e-05,
+      "loss": 0.548,
+      "step": 2512
+    },
+    {
+      "epoch": 0.3517144856543037,
+      "grad_norm": 0.4170639037160172,
+      "learning_rate": 3.762737630307093e-05,
+      "loss": 0.5875,
+      "step": 2513
+    },
+    {
+      "epoch": 0.35185444366689994,
+      "grad_norm": 0.4115129851905205,
+      "learning_rate": 3.761759363990215e-05,
+      "loss": 0.5462,
+      "step": 2514
+    },
+    {
+      "epoch": 0.35199440167949614,
+      "grad_norm": 0.4147105457172198,
+      "learning_rate": 3.7607808383689856e-05,
+      "loss": 0.5463,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3521343596920924,
+      "grad_norm": 0.4071302253511469,
+      "learning_rate": 3.7598020536445017e-05,
+      "loss": 0.5445,
+      "step": 2516
+    },
+    {
+      "epoch": 0.3522743177046886,
+      "grad_norm": 0.39539443956776693,
+      "learning_rate": 3.758823010017913e-05,
+      "loss": 0.5455,
+      "step": 2517
+    },
+    {
+      "epoch": 0.35241427571728484,
+      "grad_norm": 0.39791097766830325,
+      "learning_rate": 3.7578437076904236e-05,
+      "loss": 0.5258,
+      "step": 2518
+    },
+    {
+      "epoch": 0.35255423372988104,
+      "grad_norm": 0.42035164161239397,
+      "learning_rate": 3.75686414686329e-05,
+      "loss": 0.5639,
+      "step": 2519
+    },
+    {
+      "epoch": 0.35269419174247724,
+      "grad_norm": 0.3830433028113876,
+      "learning_rate": 3.7558843277378206e-05,
+      "loss": 0.536,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3528341497550735,
+      "grad_norm": 0.3933038656978647,
+      "learning_rate": 3.75490425051538e-05,
+      "loss": 0.5452,
+      "step": 2521
+    },
+    {
+      "epoch": 0.3529741077676697,
+      "grad_norm": 0.41406652639293745,
+      "learning_rate": 3.753923915397383e-05,
+      "loss": 0.5667,
+      "step": 2522
+    },
+    {
+      "epoch": 0.35311406578026594,
+      "grad_norm": 0.4201955355989494,
+      "learning_rate": 3.752943322585297e-05,
+      "loss": 0.5417,
+      "step": 2523
+    },
+    {
+      "epoch": 0.35325402379286214,
+      "grad_norm": 0.41614348716949034,
+      "learning_rate": 3.751962472280647e-05,
+      "loss": 0.5817,
+      "step": 2524
+    },
+    {
+      "epoch": 0.35339398180545833,
+      "grad_norm": 0.40621639691384537,
+      "learning_rate": 3.750981364685005e-05,
+      "loss": 0.5815,
+      "step": 2525
+    },
+    {
+      "epoch": 0.3535339398180546,
+      "grad_norm": 0.4007740627071781,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.5635,
+      "step": 2526
+    },
+    {
+      "epoch": 0.3536738978306508,
+      "grad_norm": 0.4088088272074455,
+      "learning_rate": 3.749018378427312e-05,
+      "loss": 0.599,
+      "step": 2527
+    },
+    {
+      "epoch": 0.35381385584324704,
+      "grad_norm": 0.4035751382941894,
+      "learning_rate": 3.7480365001686746e-05,
+      "loss": 0.601,
+      "step": 2528
+    },
+    {
+      "epoch": 0.35395381385584324,
+      "grad_norm": 0.41808007576821743,
+      "learning_rate": 3.747054365425872e-05,
+      "loss": 0.5625,
+      "step": 2529
+    },
+    {
+      "epoch": 0.3540937718684395,
+      "grad_norm": 0.4142769655137648,
+      "learning_rate": 3.7460719744007446e-05,
+      "loss": 0.5321,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3542337298810357,
+      "grad_norm": 0.4258627449069856,
+      "learning_rate": 3.745089327295184e-05,
+      "loss": 0.5915,
+      "step": 2531
+    },
+    {
+      "epoch": 0.3543736878936319,
+      "grad_norm": 0.3979599134535474,
+      "learning_rate": 3.744106424311133e-05,
+      "loss": 0.528,
+      "step": 2532
+    },
+    {
+      "epoch": 0.35451364590622814,
+      "grad_norm": 0.40828188091566914,
+      "learning_rate": 3.7431232656505885e-05,
+      "loss": 0.584,
+      "step": 2533
+    },
+    {
+      "epoch": 0.35465360391882433,
+      "grad_norm": 0.41378663443999314,
+      "learning_rate": 3.7421398515155994e-05,
+      "loss": 0.5735,
+      "step": 2534
+    },
+    {
+      "epoch": 0.3547935619314206,
+      "grad_norm": 0.40350752325849487,
+      "learning_rate": 3.741156182108268e-05,
+      "loss": 0.5856,
+      "step": 2535
+    },
+    {
+      "epoch": 0.3549335199440168,
+      "grad_norm": 0.42065994177021454,
+      "learning_rate": 3.740172257630747e-05,
+      "loss": 0.5996,
+      "step": 2536
+    },
+    {
+      "epoch": 0.35507347795661304,
+      "grad_norm": 0.40860097335167056,
+      "learning_rate": 3.739188078285244e-05,
+      "loss": 0.6183,
+      "step": 2537
+    },
+    {
+      "epoch": 0.35521343596920923,
+      "grad_norm": 0.4122420177259316,
+      "learning_rate": 3.738203644274018e-05,
+      "loss": 0.5549,
+      "step": 2538
+    },
+    {
+      "epoch": 0.35535339398180543,
+      "grad_norm": 0.4181580931065648,
+      "learning_rate": 3.7372189557993794e-05,
+      "loss": 0.5945,
+      "step": 2539
+    },
+    {
+      "epoch": 0.3554933519944017,
+      "grad_norm": 0.41214229514921485,
+      "learning_rate": 3.7362340130636925e-05,
+      "loss": 0.5964,
+      "step": 2540
+    },
+    {
+      "epoch": 0.3556333100069979,
+      "grad_norm": 0.5379104247408596,
+      "learning_rate": 3.735248816269372e-05,
+      "loss": 0.5855,
+      "step": 2541
+    },
+    {
+      "epoch": 0.35577326801959414,
+      "grad_norm": 0.4221656282811352,
+      "learning_rate": 3.734263365618886e-05,
+      "loss": 0.611,
+      "step": 2542
+    },
+    {
+      "epoch": 0.35591322603219033,
+      "grad_norm": 0.4014247264225883,
+      "learning_rate": 3.7332776613147564e-05,
+      "loss": 0.5706,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3560531840447866,
+      "grad_norm": 0.42213996299337353,
+      "learning_rate": 3.732291703559553e-05,
+      "loss": 0.5995,
+      "step": 2544
+    },
+    {
+      "epoch": 0.3561931420573828,
+      "grad_norm": 0.41540133704593746,
+      "learning_rate": 3.731305492555901e-05,
+      "loss": 0.5574,
+      "step": 2545
+    },
+    {
+      "epoch": 0.356333100069979,
+      "grad_norm": 0.4104571121982501,
+      "learning_rate": 3.7303190285064776e-05,
+      "loss": 0.5627,
+      "step": 2546
+    },
+    {
+      "epoch": 0.35647305808257523,
+      "grad_norm": 0.3942675491175812,
+      "learning_rate": 3.72933231161401e-05,
+      "loss": 0.558,
+      "step": 2547
+    },
+    {
+      "epoch": 0.35661301609517143,
+      "grad_norm": 0.39670035734631537,
+      "learning_rate": 3.7283453420812786e-05,
+      "loss": 0.5475,
+      "step": 2548
+    },
+    {
+      "epoch": 0.3567529741077677,
+      "grad_norm": 0.42150223375815987,
+      "learning_rate": 3.727358120111117e-05,
+      "loss": 0.5721,
+      "step": 2549
+    },
+    {
+      "epoch": 0.3568929321203639,
+      "grad_norm": 0.4008262706510209,
+      "learning_rate": 3.726370645906407e-05,
+      "loss": 0.5561,
+      "step": 2550
+    },
+    {
+      "epoch": 0.35703289013296013,
+      "grad_norm": 0.44040584835645996,
+      "learning_rate": 3.7253829196700876e-05,
+      "loss": 0.5846,
+      "step": 2551
+    },
+    {
+      "epoch": 0.35717284814555633,
+      "grad_norm": 0.4126914987733967,
+      "learning_rate": 3.7243949416051435e-05,
+      "loss": 0.5469,
+      "step": 2552
+    },
+    {
+      "epoch": 0.35731280615815253,
+      "grad_norm": 0.39603138357670026,
+      "learning_rate": 3.723406711914617e-05,
+      "loss": 0.5942,
+      "step": 2553
+    },
+    {
+      "epoch": 0.3574527641707488,
+      "grad_norm": 0.3852892928635738,
+      "learning_rate": 3.7224182308015975e-05,
+      "loss": 0.5928,
+      "step": 2554
+    },
+    {
+      "epoch": 0.357592722183345,
+      "grad_norm": 0.41121367387986824,
+      "learning_rate": 3.7214294984692285e-05,
+      "loss": 0.5853,
+      "step": 2555
+    },
+    {
+      "epoch": 0.35773268019594123,
+      "grad_norm": 0.4165573669555882,
+      "learning_rate": 3.7204405151207036e-05,
+      "loss": 0.5897,
+      "step": 2556
+    },
+    {
+      "epoch": 0.35787263820853743,
+      "grad_norm": 0.45676174939260117,
+      "learning_rate": 3.71945128095927e-05,
+      "loss": 0.5853,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3580125962211337,
+      "grad_norm": 0.39957833133810244,
+      "learning_rate": 3.718461796188225e-05,
+      "loss": 0.5287,
+      "step": 2558
+    },
+    {
+      "epoch": 0.3581525542337299,
+      "grad_norm": 0.4145657616919604,
+      "learning_rate": 3.717472061010918e-05,
+      "loss": 0.593,
+      "step": 2559
+    },
+    {
+      "epoch": 0.3582925122463261,
+      "grad_norm": 0.4080507721332056,
+      "learning_rate": 3.71648207563075e-05,
+      "loss": 0.5634,
+      "step": 2560
+    },
+    {
+      "epoch": 0.35843247025892233,
+      "grad_norm": 0.3912915057782903,
+      "learning_rate": 3.715491840251172e-05,
+      "loss": 0.5949,
+      "step": 2561
+    },
+    {
+      "epoch": 0.35857242827151853,
+      "grad_norm": 0.41837049449187164,
+      "learning_rate": 3.714501355075688e-05,
+      "loss": 0.572,
+      "step": 2562
+    },
+    {
+      "epoch": 0.3587123862841148,
+      "grad_norm": 0.40328972090381965,
+      "learning_rate": 3.713510620307852e-05,
+      "loss": 0.5856,
+      "step": 2563
+    },
+    {
+      "epoch": 0.358852344296711,
+      "grad_norm": 0.431299030659745,
+      "learning_rate": 3.712519636151272e-05,
+      "loss": 0.5952,
+      "step": 2564
+    },
+    {
+      "epoch": 0.35899230230930723,
+      "grad_norm": 0.3985200361199881,
+      "learning_rate": 3.711528402809603e-05,
+      "loss": 0.5533,
+      "step": 2565
+    },
+    {
+      "epoch": 0.35913226032190343,
+      "grad_norm": 0.4192753590584814,
+      "learning_rate": 3.710536920486555e-05,
+      "loss": 0.5667,
+      "step": 2566
+    },
+    {
+      "epoch": 0.3592722183344996,
+      "grad_norm": 0.42240052564081954,
+      "learning_rate": 3.709545189385887e-05,
+      "loss": 0.5708,
+      "step": 2567
+    },
+    {
+      "epoch": 0.3594121763470959,
+      "grad_norm": 0.4236767334123938,
+      "learning_rate": 3.708553209711409e-05,
+      "loss": 0.5485,
+      "step": 2568
+    },
+    {
+      "epoch": 0.3595521343596921,
+      "grad_norm": 0.40080357675210276,
+      "learning_rate": 3.707560981666986e-05,
+      "loss": 0.557,
+      "step": 2569
+    },
+    {
+      "epoch": 0.35969209237228833,
+      "grad_norm": 0.410487921988623,
+      "learning_rate": 3.706568505456527e-05,
+      "loss": 0.5547,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3598320503848845,
+      "grad_norm": 0.41368891130760344,
+      "learning_rate": 3.705575781283999e-05,
+      "loss": 0.5673,
+      "step": 2571
+    },
+    {
+      "epoch": 0.3599720083974808,
+      "grad_norm": 0.40193894214258574,
+      "learning_rate": 3.704582809353415e-05,
+      "loss": 0.5925,
+      "step": 2572
+    },
+    {
+      "epoch": 0.360111966410077,
+      "grad_norm": 0.407102772883237,
+      "learning_rate": 3.703589589868841e-05,
+      "loss": 0.5658,
+      "step": 2573
+    },
+    {
+      "epoch": 0.3602519244226732,
+      "grad_norm": 0.4100475297715309,
+      "learning_rate": 3.702596123034395e-05,
+      "loss": 0.5946,
+      "step": 2574
+    },
+    {
+      "epoch": 0.36039188243526943,
+      "grad_norm": 0.4005884577498317,
+      "learning_rate": 3.701602409054243e-05,
+      "loss": 0.5293,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3605318404478656,
+      "grad_norm": 0.37159137413391713,
+      "learning_rate": 3.700608448132604e-05,
+      "loss": 0.5973,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3606717984604619,
+      "grad_norm": 0.41775069760426675,
+      "learning_rate": 3.699614240473748e-05,
+      "loss": 0.5731,
+      "step": 2577
+    },
+    {
+      "epoch": 0.3608117564730581,
+      "grad_norm": 0.4536333746914714,
+      "learning_rate": 3.6986197862819934e-05,
+      "loss": 0.5814,
+      "step": 2578
+    },
+    {
+      "epoch": 0.36095171448565433,
+      "grad_norm": 0.415665084054717,
+      "learning_rate": 3.69762508576171e-05,
+      "loss": 0.5813,
+      "step": 2579
+    },
+    {
+      "epoch": 0.3610916724982505,
+      "grad_norm": 0.4111649076881108,
+      "learning_rate": 3.6966301391173206e-05,
+      "loss": 0.5687,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3612316305108467,
+      "grad_norm": 0.45348364261588203,
+      "learning_rate": 3.695634946553296e-05,
+      "loss": 0.5789,
+      "step": 2581
+    },
+    {
+      "epoch": 0.361371588523443,
+      "grad_norm": 0.46524048585079214,
+      "learning_rate": 3.694639508274158e-05,
+      "loss": 0.605,
+      "step": 2582
+    },
+    {
+      "epoch": 0.3615115465360392,
+      "grad_norm": 0.43254745661664457,
+      "learning_rate": 3.69364382448448e-05,
+      "loss": 0.5723,
+      "step": 2583
+    },
+    {
+      "epoch": 0.3616515045486354,
+      "grad_norm": 0.5926681683204299,
+      "learning_rate": 3.692647895388884e-05,
+      "loss": 0.5657,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3617914625612316,
+      "grad_norm": 0.3984247442681412,
+      "learning_rate": 3.691651721192046e-05,
+      "loss": 0.5375,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3619314205738279,
+      "grad_norm": 0.4256089941995705,
+      "learning_rate": 3.6906553020986876e-05,
+      "loss": 0.5663,
+      "step": 2586
+    },
+    {
+      "epoch": 0.3620713785864241,
+      "grad_norm": 0.3955060934018287,
+      "learning_rate": 3.6896586383135835e-05,
+      "loss": 0.5367,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3622113365990203,
+      "grad_norm": 0.3983798844884697,
+      "learning_rate": 3.688661730041559e-05,
+      "loss": 0.5757,
+      "step": 2588
+    },
+    {
+      "epoch": 0.3623512946116165,
+      "grad_norm": 0.40920669971645085,
+      "learning_rate": 3.6876645774874876e-05,
+      "loss": 0.5445,
+      "step": 2589
+    },
+    {
+      "epoch": 0.3624912526242127,
+      "grad_norm": 0.43900374382517116,
+      "learning_rate": 3.6866671808562945e-05,
+      "loss": 0.5702,
+      "step": 2590
+    },
+    {
+      "epoch": 0.362631210636809,
+      "grad_norm": 0.4009385370546209,
+      "learning_rate": 3.685669540352957e-05,
+      "loss": 0.5873,
+      "step": 2591
+    },
+    {
+      "epoch": 0.3627711686494052,
+      "grad_norm": 0.4010104073772178,
+      "learning_rate": 3.6846716561824965e-05,
+      "loss": 0.5693,
+      "step": 2592
+    },
+    {
+      "epoch": 0.3629111266620014,
+      "grad_norm": 0.3942578614306398,
+      "learning_rate": 3.6836735285499924e-05,
+      "loss": 0.5459,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3630510846745976,
+      "grad_norm": 0.4052354355699724,
+      "learning_rate": 3.682675157660567e-05,
+      "loss": 0.5753,
+      "step": 2594
+    },
+    {
+      "epoch": 0.3631910426871938,
+      "grad_norm": 0.4111971028442228,
+      "learning_rate": 3.681676543719396e-05,
+      "loss": 0.6152,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3633310006997901,
+      "grad_norm": 0.38664588990396315,
+      "learning_rate": 3.680677686931707e-05,
+      "loss": 0.5354,
+      "step": 2596
+    },
+    {
+      "epoch": 0.36347095871238627,
+      "grad_norm": 0.4207284242435013,
+      "learning_rate": 3.679678587502773e-05,
+      "loss": 0.576,
+      "step": 2597
+    },
+    {
+      "epoch": 0.3636109167249825,
+      "grad_norm": 0.40595710076845204,
+      "learning_rate": 3.67867924563792e-05,
+      "loss": 0.5885,
+      "step": 2598
+    },
+    {
+      "epoch": 0.3637508747375787,
+      "grad_norm": 0.41445075167878537,
+      "learning_rate": 3.6776796615425224e-05,
+      "loss": 0.5732,
+      "step": 2599
+    },
+    {
+      "epoch": 0.363890832750175,
+      "grad_norm": 0.38888704179883676,
+      "learning_rate": 3.6766798354220054e-05,
+      "loss": 0.5658,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3640307907627712,
+      "grad_norm": 0.40093975960461736,
+      "learning_rate": 3.675679767481842e-05,
+      "loss": 0.5597,
+      "step": 2601
+    },
+    {
+      "epoch": 0.36417074877536737,
+      "grad_norm": 0.4321442568384402,
+      "learning_rate": 3.674679457927559e-05,
+      "loss": 0.5451,
+      "step": 2602
+    },
+    {
+      "epoch": 0.3643107067879636,
+      "grad_norm": 0.41193216461597326,
+      "learning_rate": 3.673678906964727e-05,
+      "loss": 0.5562,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3644506648005598,
+      "grad_norm": 0.3982633594790959,
+      "learning_rate": 3.672678114798972e-05,
+      "loss": 0.5463,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3645906228131561,
+      "grad_norm": 0.42438474880866467,
+      "learning_rate": 3.671677081635966e-05,
+      "loss": 0.5809,
+      "step": 2605
+    },
+    {
+      "epoch": 0.36473058082575227,
+      "grad_norm": 0.4217628754836848,
+      "learning_rate": 3.67067580768143e-05,
+      "loss": 0.553,
+      "step": 2606
+    },
+    {
+      "epoch": 0.36487053883834847,
+      "grad_norm": 0.42309673305357437,
+      "learning_rate": 3.669674293141139e-05,
+      "loss": 0.6209,
+      "step": 2607
+    },
+    {
+      "epoch": 0.3650104968509447,
+      "grad_norm": 0.3920738753776142,
+      "learning_rate": 3.6686725382209114e-05,
+      "loss": 0.5617,
+      "step": 2608
+    },
+    {
+      "epoch": 0.3651504548635409,
+      "grad_norm": 0.39484584407664886,
+      "learning_rate": 3.66767054312662e-05,
+      "loss": 0.5167,
+      "step": 2609
+    },
+    {
+      "epoch": 0.36529041287613717,
+      "grad_norm": 0.4255886790117686,
+      "learning_rate": 3.6666683080641846e-05,
+      "loss": 0.6193,
+      "step": 2610
+    },
+    {
+      "epoch": 0.36543037088873337,
+      "grad_norm": 0.40579442495363904,
+      "learning_rate": 3.665665833239574e-05,
+      "loss": 0.5342,
+      "step": 2611
+    },
+    {
+      "epoch": 0.3655703289013296,
+      "grad_norm": 0.4233421685048626,
+      "learning_rate": 3.664663118858808e-05,
+      "loss": 0.5465,
+      "step": 2612
+    },
+    {
+      "epoch": 0.3657102869139258,
+      "grad_norm": 0.4174294680767461,
+      "learning_rate": 3.6636601651279524e-05,
+      "loss": 0.5849,
+      "step": 2613
+    },
+    {
+      "epoch": 0.365850244926522,
+      "grad_norm": 0.44476352179830164,
+      "learning_rate": 3.662656972253127e-05,
+      "loss": 0.5717,
+      "step": 2614
+    },
+    {
+      "epoch": 0.36599020293911827,
+      "grad_norm": 0.3948209367210007,
+      "learning_rate": 3.6616535404404964e-05,
+      "loss": 0.5529,
+      "step": 2615
+    },
+    {
+      "epoch": 0.36613016095171447,
+      "grad_norm": 0.40738682961697953,
+      "learning_rate": 3.660649869896276e-05,
+      "loss": 0.5268,
+      "step": 2616
+    },
+    {
+      "epoch": 0.3662701189643107,
+      "grad_norm": 0.40020906641099585,
+      "learning_rate": 3.659645960826732e-05,
+      "loss": 0.5567,
+      "step": 2617
+    },
+    {
+      "epoch": 0.3664100769769069,
+      "grad_norm": 0.4237523890631533,
+      "learning_rate": 3.658641813438176e-05,
+      "loss": 0.573,
+      "step": 2618
+    },
+    {
+      "epoch": 0.36655003498950317,
+      "grad_norm": 0.4325798007082038,
+      "learning_rate": 3.657637427936972e-05,
+      "loss": 0.5791,
+      "step": 2619
+    },
+    {
+      "epoch": 0.36668999300209937,
+      "grad_norm": 0.43907042323555606,
+      "learning_rate": 3.65663280452953e-05,
+      "loss": 0.5786,
+      "step": 2620
+    },
+    {
+      "epoch": 0.36682995101469557,
+      "grad_norm": 0.4323098724184291,
+      "learning_rate": 3.6556279434223116e-05,
+      "loss": 0.6016,
+      "step": 2621
+    },
+    {
+      "epoch": 0.3669699090272918,
+      "grad_norm": 0.4265584212865507,
+      "learning_rate": 3.654622844821825e-05,
+      "loss": 0.623,
+      "step": 2622
+    },
+    {
+      "epoch": 0.367109867039888,
+      "grad_norm": 0.4179019694136672,
+      "learning_rate": 3.6536175089346285e-05,
+      "loss": 0.5468,
+      "step": 2623
+    },
+    {
+      "epoch": 0.36724982505248427,
+      "grad_norm": 0.41889421083309525,
+      "learning_rate": 3.6526119359673284e-05,
+      "loss": 0.5574,
+      "step": 2624
+    },
+    {
+      "epoch": 0.36738978306508047,
+      "grad_norm": 0.41710292728469345,
+      "learning_rate": 3.651606126126581e-05,
+      "loss": 0.5781,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3675297410776767,
+      "grad_norm": 0.3897312626800565,
+      "learning_rate": 3.65060007961909e-05,
+      "loss": 0.5193,
+      "step": 2626
+    },
+    {
+      "epoch": 0.3676696990902729,
+      "grad_norm": 0.42386091892910743,
+      "learning_rate": 3.649593796651608e-05,
+      "loss": 0.5917,
+      "step": 2627
+    },
+    {
+      "epoch": 0.3678096571028691,
+      "grad_norm": 0.414524194950392,
+      "learning_rate": 3.648587277430936e-05,
+      "loss": 0.5758,
+      "step": 2628
+    },
+    {
+      "epoch": 0.36794961511546537,
+      "grad_norm": 0.4038428167881955,
+      "learning_rate": 3.647580522163925e-05,
+      "loss": 0.5621,
+      "step": 2629
+    },
+    {
+      "epoch": 0.36808957312806156,
+      "grad_norm": 0.4006767303991816,
+      "learning_rate": 3.646573531057473e-05,
+      "loss": 0.5586,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3682295311406578,
+      "grad_norm": 0.40655169612049585,
+      "learning_rate": 3.645566304318526e-05,
+      "loss": 0.6016,
+      "step": 2631
+    },
+    {
+      "epoch": 0.368369489153254,
+      "grad_norm": 0.434790153457983,
+      "learning_rate": 3.644558842154081e-05,
+      "loss": 0.5481,
+      "step": 2632
+    },
+    {
+      "epoch": 0.36850944716585027,
+      "grad_norm": 0.4099645764830532,
+      "learning_rate": 3.6435511447711804e-05,
+      "loss": 0.5738,
+      "step": 2633
+    },
+    {
+      "epoch": 0.36864940517844647,
+      "grad_norm": 0.41274450825271575,
+      "learning_rate": 3.642543212376916e-05,
+      "loss": 0.5991,
+      "step": 2634
+    },
+    {
+      "epoch": 0.36878936319104266,
+      "grad_norm": 0.4032135476124307,
+      "learning_rate": 3.6415350451784294e-05,
+      "loss": 0.5499,
+      "step": 2635
+    },
+    {
+      "epoch": 0.3689293212036389,
+      "grad_norm": 0.4188775277615029,
+      "learning_rate": 3.6405266433829075e-05,
+      "loss": 0.5654,
+      "step": 2636
+    },
+    {
+      "epoch": 0.3690692792162351,
+      "grad_norm": 0.5178966570845596,
+      "learning_rate": 3.6395180071975885e-05,
+      "loss": 0.5396,
+      "step": 2637
+    },
+    {
+      "epoch": 0.36920923722883137,
+      "grad_norm": 0.40182828296097955,
+      "learning_rate": 3.638509136829758e-05,
+      "loss": 0.5652,
+      "step": 2638
+    },
+    {
+      "epoch": 0.36934919524142756,
+      "grad_norm": 0.390078684434463,
+      "learning_rate": 3.637500032486747e-05,
+      "loss": 0.557,
+      "step": 2639
+    },
+    {
+      "epoch": 0.3694891532540238,
+      "grad_norm": 0.3931686164476163,
+      "learning_rate": 3.636490694375938e-05,
+      "loss": 0.5841,
+      "step": 2640
+    },
+    {
+      "epoch": 0.36962911126662,
+      "grad_norm": 0.39382269721567603,
+      "learning_rate": 3.63548112270476e-05,
+      "loss": 0.5575,
+      "step": 2641
+    },
+    {
+      "epoch": 0.3697690692792162,
+      "grad_norm": 0.42173997864130025,
+      "learning_rate": 3.63447131768069e-05,
+      "loss": 0.6018,
+      "step": 2642
+    },
+    {
+      "epoch": 0.36990902729181246,
+      "grad_norm": 0.43303379767097616,
+      "learning_rate": 3.6334612795112534e-05,
+      "loss": 0.5894,
+      "step": 2643
+    },
+    {
+      "epoch": 0.37004898530440866,
+      "grad_norm": 0.40551708557859245,
+      "learning_rate": 3.632451008404024e-05,
+      "loss": 0.5822,
+      "step": 2644
+    },
+    {
+      "epoch": 0.3701889433170049,
+      "grad_norm": 0.4199307056582574,
+      "learning_rate": 3.631440504566621e-05,
+      "loss": 0.5359,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3703289013296011,
+      "grad_norm": 0.3856025757451338,
+      "learning_rate": 3.6304297682067144e-05,
+      "loss": 0.5279,
+      "step": 2646
+    },
+    {
+      "epoch": 0.37046885934219737,
+      "grad_norm": 0.39790868545989994,
+      "learning_rate": 3.6294187995320214e-05,
+      "loss": 0.5685,
+      "step": 2647
+    },
+    {
+      "epoch": 0.37060881735479356,
+      "grad_norm": 0.41011662651634145,
+      "learning_rate": 3.628407598750305e-05,
+      "loss": 0.5732,
+      "step": 2648
+    },
+    {
+      "epoch": 0.37074877536738976,
+      "grad_norm": 0.41740394404930137,
+      "learning_rate": 3.627396166069377e-05,
+      "loss": 0.5922,
+      "step": 2649
+    },
+    {
+      "epoch": 0.370888733379986,
+      "grad_norm": 0.41661946277247025,
+      "learning_rate": 3.626384501697099e-05,
+      "loss": 0.5501,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3710286913925822,
+      "grad_norm": 0.42202366075884673,
+      "learning_rate": 3.625372605841376e-05,
+      "loss": 0.5659,
+      "step": 2651
+    },
+    {
+      "epoch": 0.37116864940517846,
+      "grad_norm": 0.39038696179315746,
+      "learning_rate": 3.624360478710165e-05,
+      "loss": 0.5729,
+      "step": 2652
+    },
+    {
+      "epoch": 0.37130860741777466,
+      "grad_norm": 0.41764113003828224,
+      "learning_rate": 3.623348120511466e-05,
+      "loss": 0.5663,
+      "step": 2653
+    },
+    {
+      "epoch": 0.3714485654303709,
+      "grad_norm": 0.39913226890826564,
+      "learning_rate": 3.622335531453331e-05,
+      "loss": 0.5645,
+      "step": 2654
+    },
+    {
+      "epoch": 0.3715885234429671,
+      "grad_norm": 0.3939385454277571,
+      "learning_rate": 3.621322711743858e-05,
+      "loss": 0.5631,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3717284814555633,
+      "grad_norm": 0.41308594557194117,
+      "learning_rate": 3.6203096615911884e-05,
+      "loss": 0.6052,
+      "step": 2656
+    },
+    {
+      "epoch": 0.37186843946815956,
+      "grad_norm": 0.4298728275791088,
+      "learning_rate": 3.619296381203517e-05,
+      "loss": 0.5396,
+      "step": 2657
+    },
+    {
+      "epoch": 0.37200839748075576,
+      "grad_norm": 0.4544709711994261,
+      "learning_rate": 3.6182828707890816e-05,
+      "loss": 0.5973,
+      "step": 2658
+    },
+    {
+      "epoch": 0.372148355493352,
+      "grad_norm": 0.417274676645833,
+      "learning_rate": 3.61726913055617e-05,
+      "loss": 0.58,
+      "step": 2659
+    },
+    {
+      "epoch": 0.3722883135059482,
+      "grad_norm": 0.4027341706183872,
+      "learning_rate": 3.6162551607131164e-05,
+      "loss": 0.5793,
+      "step": 2660
+    },
+    {
+      "epoch": 0.37242827151854446,
+      "grad_norm": 0.41668849630230104,
+      "learning_rate": 3.615240961468301e-05,
+      "loss": 0.5573,
+      "step": 2661
+    },
+    {
+      "epoch": 0.37256822953114066,
+      "grad_norm": 0.40880781367892177,
+      "learning_rate": 3.614226533030152e-05,
+      "loss": 0.5821,
+      "step": 2662
+    },
+    {
+      "epoch": 0.37270818754373686,
+      "grad_norm": 0.4154234374761176,
+      "learning_rate": 3.6132118756071456e-05,
+      "loss": 0.5932,
+      "step": 2663
+    },
+    {
+      "epoch": 0.3728481455563331,
+      "grad_norm": 0.3839166628649522,
+      "learning_rate": 3.6121969894078024e-05,
+      "loss": 0.5498,
+      "step": 2664
+    },
+    {
+      "epoch": 0.3729881035689293,
+      "grad_norm": 0.42314063405669433,
+      "learning_rate": 3.611181874640694e-05,
+      "loss": 0.571,
+      "step": 2665
+    },
+    {
+      "epoch": 0.37312806158152556,
+      "grad_norm": 0.3944581050181605,
+      "learning_rate": 3.610166531514436e-05,
+      "loss": 0.5382,
+      "step": 2666
+    },
+    {
+      "epoch": 0.37326801959412176,
+      "grad_norm": 0.39463547519248415,
+      "learning_rate": 3.60915096023769e-05,
+      "loss": 0.5544,
+      "step": 2667
+    },
+    {
+      "epoch": 0.373407977606718,
+      "grad_norm": 0.41268049741997787,
+      "learning_rate": 3.6081351610191696e-05,
+      "loss": 0.5427,
+      "step": 2668
+    },
+    {
+      "epoch": 0.3735479356193142,
+      "grad_norm": 0.4095949213776676,
+      "learning_rate": 3.607119134067629e-05,
+      "loss": 0.5564,
+      "step": 2669
+    },
+    {
+      "epoch": 0.3736878936319104,
+      "grad_norm": 0.39819288148024584,
+      "learning_rate": 3.606102879591874e-05,
+      "loss": 0.5584,
+      "step": 2670
+    },
+    {
+      "epoch": 0.37382785164450666,
+      "grad_norm": 0.4133173629042419,
+      "learning_rate": 3.605086397800753e-05,
+      "loss": 0.5563,
+      "step": 2671
+    },
+    {
+      "epoch": 0.37396780965710286,
+      "grad_norm": 0.39980433778091934,
+      "learning_rate": 3.604069688903165e-05,
+      "loss": 0.5536,
+      "step": 2672
+    },
+    {
+      "epoch": 0.3741077676696991,
+      "grad_norm": 0.4143096692418758,
+      "learning_rate": 3.603052753108053e-05,
+      "loss": 0.6135,
+      "step": 2673
+    },
+    {
+      "epoch": 0.3742477256822953,
+      "grad_norm": 0.4220107842655034,
+      "learning_rate": 3.602035590624409e-05,
+      "loss": 0.5875,
+      "step": 2674
+    },
+    {
+      "epoch": 0.37438768369489156,
+      "grad_norm": 0.41274829656377915,
+      "learning_rate": 3.6010182016612695e-05,
+      "loss": 0.553,
+      "step": 2675
+    },
+    {
+      "epoch": 0.37452764170748776,
+      "grad_norm": 0.4077340731853976,
+      "learning_rate": 3.600000586427718e-05,
+      "loss": 0.5626,
+      "step": 2676
+    },
+    {
+      "epoch": 0.37466759972008395,
+      "grad_norm": 0.432520643568383,
+      "learning_rate": 3.598982745132885e-05,
+      "loss": 0.5764,
+      "step": 2677
+    },
+    {
+      "epoch": 0.3748075577326802,
+      "grad_norm": 0.3972351037179022,
+      "learning_rate": 3.597964677985946e-05,
+      "loss": 0.5439,
+      "step": 2678
+    },
+    {
+      "epoch": 0.3749475157452764,
+      "grad_norm": 0.421002231272715,
+      "learning_rate": 3.596946385196126e-05,
+      "loss": 0.5546,
+      "step": 2679
+    },
+    {
+      "epoch": 0.37508747375787266,
+      "grad_norm": 0.41810855014069603,
+      "learning_rate": 3.5959278669726935e-05,
+      "loss": 0.5862,
+      "step": 2680
+    },
+    {
+      "epoch": 0.37522743177046886,
+      "grad_norm": 0.3799508514194255,
+      "learning_rate": 3.594909123524965e-05,
+      "loss": 0.5434,
+      "step": 2681
+    },
+    {
+      "epoch": 0.37536738978306505,
+      "grad_norm": 0.4035591025821781,
+      "learning_rate": 3.593890155062302e-05,
+      "loss": 0.5798,
+      "step": 2682
+    },
+    {
+      "epoch": 0.3755073477956613,
+      "grad_norm": 0.4114244071207664,
+      "learning_rate": 3.592870961794113e-05,
+      "loss": 0.6205,
+      "step": 2683
+    },
+    {
+      "epoch": 0.3756473058082575,
+      "grad_norm": 0.4076569672449696,
+      "learning_rate": 3.5918515439298526e-05,
+      "loss": 0.5798,
+      "step": 2684
+    },
+    {
+      "epoch": 0.37578726382085376,
+      "grad_norm": 0.4051977237739006,
+      "learning_rate": 3.5908319016790214e-05,
+      "loss": 0.5707,
+      "step": 2685
+    },
+    {
+      "epoch": 0.37592722183344995,
+      "grad_norm": 0.41477449458013194,
+      "learning_rate": 3.589812035251167e-05,
+      "loss": 0.58,
+      "step": 2686
+    },
+    {
+      "epoch": 0.3760671798460462,
+      "grad_norm": 0.4265813533603491,
+      "learning_rate": 3.588791944855881e-05,
+      "loss": 0.5891,
+      "step": 2687
+    },
+    {
+      "epoch": 0.3762071378586424,
+      "grad_norm": 0.4009173028671792,
+      "learning_rate": 3.587771630702803e-05,
+      "loss": 0.5559,
+      "step": 2688
+    },
+    {
+      "epoch": 0.3763470958712386,
+      "grad_norm": 0.41516714213673056,
+      "learning_rate": 3.586751093001618e-05,
+      "loss": 0.5518,
+      "step": 2689
+    },
+    {
+      "epoch": 0.37648705388383485,
+      "grad_norm": 0.4246576996594534,
+      "learning_rate": 3.5857303319620566e-05,
+      "loss": 0.5804,
+      "step": 2690
+    },
+    {
+      "epoch": 0.37662701189643105,
+      "grad_norm": 0.3882381660222785,
+      "learning_rate": 3.5847093477938956e-05,
+      "loss": 0.5588,
+      "step": 2691
+    },
+    {
+      "epoch": 0.3767669699090273,
+      "grad_norm": 0.41694303450281095,
+      "learning_rate": 3.583688140706958e-05,
+      "loss": 0.5694,
+      "step": 2692
+    },
+    {
+      "epoch": 0.3769069279216235,
+      "grad_norm": 0.39378160604472995,
+      "learning_rate": 3.5826667109111115e-05,
+      "loss": 0.5595,
+      "step": 2693
+    },
+    {
+      "epoch": 0.37704688593421976,
+      "grad_norm": 0.38592197692564867,
+      "learning_rate": 3.581645058616271e-05,
+      "loss": 0.5418,
+      "step": 2694
+    },
+    {
+      "epoch": 0.37718684394681595,
+      "grad_norm": 0.3930395842122637,
+      "learning_rate": 3.580623184032396e-05,
+      "loss": 0.5424,
+      "step": 2695
+    },
+    {
+      "epoch": 0.37732680195941215,
+      "grad_norm": 0.4176452454515421,
+      "learning_rate": 3.579601087369492e-05,
+      "loss": 0.5433,
+      "step": 2696
+    },
+    {
+      "epoch": 0.3774667599720084,
+      "grad_norm": 0.4189236803610346,
+      "learning_rate": 3.5785787688376104e-05,
+      "loss": 0.5822,
+      "step": 2697
+    },
+    {
+      "epoch": 0.3776067179846046,
+      "grad_norm": 0.38962623396875146,
+      "learning_rate": 3.577556228646849e-05,
+      "loss": 0.5357,
+      "step": 2698
+    },
+    {
+      "epoch": 0.37774667599720085,
+      "grad_norm": 0.4020721519778404,
+      "learning_rate": 3.576533467007349e-05,
+      "loss": 0.5658,
+      "step": 2699
+    },
+    {
+      "epoch": 0.37788663400979705,
+      "grad_norm": 0.43467388992700556,
+      "learning_rate": 3.5755104841292974e-05,
+      "loss": 0.6269,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3780265920223933,
+      "grad_norm": 0.41515218431176615,
+      "learning_rate": 3.5744872802229296e-05,
+      "loss": 0.5777,
+      "step": 2701
+    },
+    {
+      "epoch": 0.3781665500349895,
+      "grad_norm": 0.4299136904928923,
+      "learning_rate": 3.5734638554985236e-05,
+      "loss": 0.6001,
+      "step": 2702
+    },
+    {
+      "epoch": 0.3783065080475857,
+      "grad_norm": 0.4051757493692295,
+      "learning_rate": 3.5724402101664023e-05,
+      "loss": 0.5337,
+      "step": 2703
+    },
+    {
+      "epoch": 0.37844646606018195,
+      "grad_norm": 0.41621036812349377,
+      "learning_rate": 3.571416344436938e-05,
+      "loss": 0.5754,
+      "step": 2704
+    },
+    {
+      "epoch": 0.37858642407277815,
+      "grad_norm": 0.4041482546847394,
+      "learning_rate": 3.5703922585205416e-05,
+      "loss": 0.5556,
+      "step": 2705
+    },
+    {
+      "epoch": 0.3787263820853744,
+      "grad_norm": 0.4120614594717955,
+      "learning_rate": 3.569367952627677e-05,
+      "loss": 0.5881,
+      "step": 2706
+    },
+    {
+      "epoch": 0.3788663400979706,
+      "grad_norm": 0.4225642230493687,
+      "learning_rate": 3.5683434269688485e-05,
+      "loss": 0.5639,
+      "step": 2707
+    },
+    {
+      "epoch": 0.37900629811056685,
+      "grad_norm": 0.42743372284510195,
+      "learning_rate": 3.567318681754605e-05,
+      "loss": 0.5866,
+      "step": 2708
+    },
+    {
+      "epoch": 0.37914625612316305,
+      "grad_norm": 0.39443077799329296,
+      "learning_rate": 3.566293717195543e-05,
+      "loss": 0.5282,
+      "step": 2709
+    },
+    {
+      "epoch": 0.37928621413575925,
+      "grad_norm": 0.3985257910214282,
+      "learning_rate": 3.565268533502303e-05,
+      "loss": 0.5773,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3794261721483555,
+      "grad_norm": 0.4048944059124916,
+      "learning_rate": 3.5642431308855705e-05,
+      "loss": 0.5366,
+      "step": 2711
+    },
+    {
+      "epoch": 0.3795661301609517,
+      "grad_norm": 0.39796926870963256,
+      "learning_rate": 3.563217509556076e-05,
+      "loss": 0.5375,
+      "step": 2712
+    },
+    {
+      "epoch": 0.37970608817354795,
+      "grad_norm": 0.42405605140429314,
+      "learning_rate": 3.562191669724597e-05,
+      "loss": 0.5735,
+      "step": 2713
+    },
+    {
+      "epoch": 0.37984604618614415,
+      "grad_norm": 0.415293963968241,
+      "learning_rate": 3.56116561160195e-05,
+      "loss": 0.5616,
+      "step": 2714
+    },
+    {
+      "epoch": 0.3799860041987404,
+      "grad_norm": 0.40013621152671924,
+      "learning_rate": 3.5601393353990046e-05,
+      "loss": 0.5454,
+      "step": 2715
+    },
+    {
+      "epoch": 0.3801259622113366,
+      "grad_norm": 0.4248537718600079,
+      "learning_rate": 3.5591128413266686e-05,
+      "loss": 0.5732,
+      "step": 2716
+    },
+    {
+      "epoch": 0.3802659202239328,
+      "grad_norm": 0.4520252264272862,
+      "learning_rate": 3.558086129595898e-05,
+      "loss": 0.53,
+      "step": 2717
+    },
+    {
+      "epoch": 0.38040587823652905,
+      "grad_norm": 0.43210620967696467,
+      "learning_rate": 3.557059200417691e-05,
+      "loss": 0.572,
+      "step": 2718
+    },
+    {
+      "epoch": 0.38054583624912525,
+      "grad_norm": 0.4125830527654036,
+      "learning_rate": 3.556032054003093e-05,
+      "loss": 0.5677,
+      "step": 2719
+    },
+    {
+      "epoch": 0.3806857942617215,
+      "grad_norm": 0.4121722034558122,
+      "learning_rate": 3.555004690563193e-05,
+      "loss": 0.557,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3808257522743177,
+      "grad_norm": 0.4349037334371718,
+      "learning_rate": 3.553977110309125e-05,
+      "loss": 0.6072,
+      "step": 2721
+    },
+    {
+      "epoch": 0.38096571028691395,
+      "grad_norm": 0.4145990063820183,
+      "learning_rate": 3.552949313452067e-05,
+      "loss": 0.5716,
+      "step": 2722
+    },
+    {
+      "epoch": 0.38110566829951015,
+      "grad_norm": 0.4199300937861851,
+      "learning_rate": 3.5519213002032404e-05,
+      "loss": 0.55,
+      "step": 2723
+    },
+    {
+      "epoch": 0.38124562631210634,
+      "grad_norm": 0.4125752251581557,
+      "learning_rate": 3.550893070773914e-05,
+      "loss": 0.5758,
+      "step": 2724
+    },
+    {
+      "epoch": 0.3813855843247026,
+      "grad_norm": 0.42548401480941495,
+      "learning_rate": 3.5498646253753986e-05,
+      "loss": 0.5751,
+      "step": 2725
+    },
+    {
+      "epoch": 0.3815255423372988,
+      "grad_norm": 0.42522116248828645,
+      "learning_rate": 3.54883596421905e-05,
+      "loss": 0.5816,
+      "step": 2726
+    },
+    {
+      "epoch": 0.38166550034989505,
+      "grad_norm": 0.41791254635783553,
+      "learning_rate": 3.5478070875162694e-05,
+      "loss": 0.5298,
+      "step": 2727
+    },
+    {
+      "epoch": 0.38180545836249125,
+      "grad_norm": 0.4058287630837772,
+      "learning_rate": 3.5467779954785e-05,
+      "loss": 0.5567,
+      "step": 2728
+    },
+    {
+      "epoch": 0.3819454163750875,
+      "grad_norm": 0.42582677147713893,
+      "learning_rate": 3.545748688317232e-05,
+      "loss": 0.6058,
+      "step": 2729
+    },
+    {
+      "epoch": 0.3820853743876837,
+      "grad_norm": 0.4103609286165152,
+      "learning_rate": 3.544719166243998e-05,
+      "loss": 0.5774,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3822253324002799,
+      "grad_norm": 0.401572700005355,
+      "learning_rate": 3.543689429470375e-05,
+      "loss": 0.5411,
+      "step": 2731
+    },
+    {
+      "epoch": 0.38236529041287615,
+      "grad_norm": 0.3929209916840099,
+      "learning_rate": 3.5426594782079846e-05,
+      "loss": 0.5373,
+      "step": 2732
+    },
+    {
+      "epoch": 0.38250524842547234,
+      "grad_norm": 0.40599385466099075,
+      "learning_rate": 3.541629312668492e-05,
+      "loss": 0.6015,
+      "step": 2733
+    },
+    {
+      "epoch": 0.3826452064380686,
+      "grad_norm": 0.41006832838571494,
+      "learning_rate": 3.540598933063607e-05,
+      "loss": 0.5509,
+      "step": 2734
+    },
+    {
+      "epoch": 0.3827851644506648,
+      "grad_norm": 0.4107953458690507,
+      "learning_rate": 3.5395683396050825e-05,
+      "loss": 0.5818,
+      "step": 2735
+    },
+    {
+      "epoch": 0.38292512246326105,
+      "grad_norm": 0.4164444405879751,
+      "learning_rate": 3.5385375325047166e-05,
+      "loss": 0.5664,
+      "step": 2736
+    },
+    {
+      "epoch": 0.38306508047585724,
+      "grad_norm": 0.39131447269722003,
+      "learning_rate": 3.53750651197435e-05,
+      "loss": 0.5853,
+      "step": 2737
+    },
+    {
+      "epoch": 0.38320503848845344,
+      "grad_norm": 0.4251859387890631,
+      "learning_rate": 3.5364752782258684e-05,
+      "loss": 0.6136,
+      "step": 2738
+    },
+    {
+      "epoch": 0.3833449965010497,
+      "grad_norm": 0.4074457295926625,
+      "learning_rate": 3.535443831471201e-05,
+      "loss": 0.5696,
+      "step": 2739
+    },
+    {
+      "epoch": 0.3834849545136459,
+      "grad_norm": 0.4275456468988683,
+      "learning_rate": 3.534412171922319e-05,
+      "loss": 0.5782,
+      "step": 2740
+    },
+    {
+      "epoch": 0.38362491252624215,
+      "grad_norm": 0.3958129275901648,
+      "learning_rate": 3.533380299791241e-05,
+      "loss": 0.6016,
+      "step": 2741
+    },
+    {
+      "epoch": 0.38376487053883834,
+      "grad_norm": 0.41339333649591314,
+      "learning_rate": 3.5323482152900254e-05,
+      "loss": 0.5741,
+      "step": 2742
+    },
+    {
+      "epoch": 0.3839048285514346,
+      "grad_norm": 0.4018942596879924,
+      "learning_rate": 3.5313159186307784e-05,
+      "loss": 0.5567,
+      "step": 2743
+    },
+    {
+      "epoch": 0.3840447865640308,
+      "grad_norm": 0.40959426159798856,
+      "learning_rate": 3.530283410025645e-05,
+      "loss": 0.5545,
+      "step": 2744
+    },
+    {
+      "epoch": 0.384184744576627,
+      "grad_norm": 0.39955019012640813,
+      "learning_rate": 3.529250689686817e-05,
+      "loss": 0.5491,
+      "step": 2745
+    },
+    {
+      "epoch": 0.38432470258922324,
+      "grad_norm": 0.41725417707805595,
+      "learning_rate": 3.5282177578265296e-05,
+      "loss": 0.5569,
+      "step": 2746
+    },
+    {
+      "epoch": 0.38446466060181944,
+      "grad_norm": 0.4045943366257942,
+      "learning_rate": 3.527184614657059e-05,
+      "loss": 0.5303,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3846046186144157,
+      "grad_norm": 0.39867086864097645,
+      "learning_rate": 3.526151260390729e-05,
+      "loss": 0.5567,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3847445766270119,
+      "grad_norm": 0.3971092164877937,
+      "learning_rate": 3.525117695239903e-05,
+      "loss": 0.5517,
+      "step": 2749
+    },
+    {
+      "epoch": 0.38488453463960814,
+      "grad_norm": 0.40368782181512425,
+      "learning_rate": 3.5240839194169885e-05,
+      "loss": 0.554,
+      "step": 2750
+    },
+    {
+      "epoch": 0.38502449265220434,
+      "grad_norm": 0.4012824600908186,
+      "learning_rate": 3.523049933134439e-05,
+      "loss": 0.5622,
+      "step": 2751
+    },
+    {
+      "epoch": 0.38516445066480054,
+      "grad_norm": 0.40807955297557885,
+      "learning_rate": 3.522015736604747e-05,
+      "loss": 0.5757,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3853044086773968,
+      "grad_norm": 0.401439172166039,
+      "learning_rate": 3.5209813300404516e-05,
+      "loss": 0.5523,
+      "step": 2753
+    },
+    {
+      "epoch": 0.385444366689993,
+      "grad_norm": 0.4188699833926191,
+      "learning_rate": 3.519946713654134e-05,
+      "loss": 0.585,
+      "step": 2754
+    },
+    {
+      "epoch": 0.38558432470258924,
+      "grad_norm": 0.4123520096515134,
+      "learning_rate": 3.518911887658418e-05,
+      "loss": 0.5681,
+      "step": 2755
+    },
+    {
+      "epoch": 0.38572428271518544,
+      "grad_norm": 0.41652405577544405,
+      "learning_rate": 3.5178768522659697e-05,
+      "loss": 0.571,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3858642407277817,
+      "grad_norm": 0.39453566372212684,
+      "learning_rate": 3.516841607689501e-05,
+      "loss": 0.5368,
+      "step": 2757
+    },
+    {
+      "epoch": 0.3860041987403779,
+      "grad_norm": 0.41756153998856244,
+      "learning_rate": 3.5158061541417644e-05,
+      "loss": 0.5799,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3861441567529741,
+      "grad_norm": 0.41467425735936986,
+      "learning_rate": 3.514770491835556e-05,
+      "loss": 0.5743,
+      "step": 2759
+    },
+    {
+      "epoch": 0.38628411476557034,
+      "grad_norm": 0.3947895347030933,
+      "learning_rate": 3.5137346209837165e-05,
+      "loss": 0.5059,
+      "step": 2760
+    },
+    {
+      "epoch": 0.38642407277816654,
+      "grad_norm": 0.4139447535493024,
+      "learning_rate": 3.5126985417991254e-05,
+      "loss": 0.5196,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3865640307907628,
+      "grad_norm": 0.4002485124686976,
+      "learning_rate": 3.5116622544947085e-05,
+      "loss": 0.5704,
+      "step": 2762
+    },
+    {
+      "epoch": 0.386703988803359,
+      "grad_norm": 0.41066912276683826,
+      "learning_rate": 3.5106257592834343e-05,
+      "loss": 0.5808,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3868439468159552,
+      "grad_norm": 0.40281746677069036,
+      "learning_rate": 3.509589056378312e-05,
+      "loss": 0.5779,
+      "step": 2764
+    },
+    {
+      "epoch": 0.38698390482855144,
+      "grad_norm": 0.4010030729597531,
+      "learning_rate": 3.5085521459923954e-05,
+      "loss": 0.5476,
+      "step": 2765
+    },
+    {
+      "epoch": 0.38712386284114764,
+      "grad_norm": 0.4116926245763738,
+      "learning_rate": 3.507515028338779e-05,
+      "loss": 0.5448,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3872638208537439,
+      "grad_norm": 0.4048451003049423,
+      "learning_rate": 3.5064777036306015e-05,
+      "loss": 0.5119,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3874037788663401,
+      "grad_norm": 0.41709617305694274,
+      "learning_rate": 3.505440172081044e-05,
+      "loss": 0.5769,
+      "step": 2768
+    },
+    {
+      "epoch": 0.38754373687893634,
+      "grad_norm": 0.41584372992034574,
+      "learning_rate": 3.5044024339033297e-05,
+      "loss": 0.5302,
+      "step": 2769
+    },
+    {
+      "epoch": 0.38768369489153254,
+      "grad_norm": 0.4065168099113291,
+      "learning_rate": 3.503364489310723e-05,
+      "loss": 0.602,
+      "step": 2770
+    },
+    {
+      "epoch": 0.38782365290412874,
+      "grad_norm": 0.4163553923214921,
+      "learning_rate": 3.502326338516534e-05,
+      "loss": 0.5694,
+      "step": 2771
+    },
+    {
+      "epoch": 0.387963610916725,
+      "grad_norm": 0.40029772682589904,
+      "learning_rate": 3.501287981734113e-05,
+      "loss": 0.5593,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3881035689293212,
+      "grad_norm": 0.4078208586874547,
+      "learning_rate": 3.5002494191768516e-05,
+      "loss": 0.5501,
+      "step": 2773
+    },
+    {
+      "epoch": 0.38824352694191744,
+      "grad_norm": 0.392227014886232,
+      "learning_rate": 3.499210651058185e-05,
+      "loss": 0.5752,
+      "step": 2774
+    },
+    {
+      "epoch": 0.38838348495451364,
+      "grad_norm": 0.41190554114378175,
+      "learning_rate": 3.498171677591593e-05,
+      "loss": 0.5359,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3885234429671099,
+      "grad_norm": 0.4137674619685686,
+      "learning_rate": 3.497132498990592e-05,
+      "loss": 0.5648,
+      "step": 2776
+    },
+    {
+      "epoch": 0.3886634009797061,
+      "grad_norm": 0.4063767289121146,
+      "learning_rate": 3.496093115468745e-05,
+      "loss": 0.5903,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3888033589923023,
+      "grad_norm": 0.40730098042246266,
+      "learning_rate": 3.495053527239656e-05,
+      "loss": 0.5907,
+      "step": 2778
+    },
+    {
+      "epoch": 0.38894331700489854,
+      "grad_norm": 0.3939529427168923,
+      "learning_rate": 3.494013734516971e-05,
+      "loss": 0.5748,
+      "step": 2779
+    },
+    {
+      "epoch": 0.38908327501749473,
+      "grad_norm": 0.40700996952853286,
+      "learning_rate": 3.492973737514378e-05,
+      "loss": 0.57,
+      "step": 2780
+    },
+    {
+      "epoch": 0.389223233030091,
+      "grad_norm": 0.4127250869631396,
+      "learning_rate": 3.491933536445606e-05,
+      "loss": 0.576,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3893631910426872,
+      "grad_norm": 0.42198868253989263,
+      "learning_rate": 3.490893131524429e-05,
+      "loss": 0.5982,
+      "step": 2782
+    },
+    {
+      "epoch": 0.38950314905528344,
+      "grad_norm": 0.3655454661119806,
+      "learning_rate": 3.489852522964658e-05,
+      "loss": 0.565,
+      "step": 2783
+    },
+    {
+      "epoch": 0.38964310706787963,
+      "grad_norm": 0.4144948208886313,
+      "learning_rate": 3.488811710980151e-05,
+      "loss": 0.5674,
+      "step": 2784
+    },
+    {
+      "epoch": 0.38978306508047583,
+      "grad_norm": 0.4074829627997144,
+      "learning_rate": 3.487770695784805e-05,
+      "loss": 0.5748,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3899230230930721,
+      "grad_norm": 0.41858532067834603,
+      "learning_rate": 3.486729477592558e-05,
+      "loss": 0.5814,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3900629811056683,
+      "grad_norm": 0.41786381819649815,
+      "learning_rate": 3.485688056617391e-05,
+      "loss": 0.5818,
+      "step": 2787
+    },
+    {
+      "epoch": 0.39020293911826454,
+      "grad_norm": 0.43164286265234475,
+      "learning_rate": 3.484646433073328e-05,
+      "loss": 0.6056,
+      "step": 2788
+    },
+    {
+      "epoch": 0.39034289713086073,
+      "grad_norm": 0.41022430212941835,
+      "learning_rate": 3.483604607174432e-05,
+      "loss": 0.586,
+      "step": 2789
+    },
+    {
+      "epoch": 0.390482855143457,
+      "grad_norm": 0.4232828423199565,
+      "learning_rate": 3.4825625791348096e-05,
+      "loss": 0.5569,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3906228131560532,
+      "grad_norm": 0.400556356726097,
+      "learning_rate": 3.481520349168607e-05,
+      "loss": 0.5853,
+      "step": 2791
+    },
+    {
+      "epoch": 0.3907627711686494,
+      "grad_norm": 0.4163389824429049,
+      "learning_rate": 3.480477917490014e-05,
+      "loss": 0.5423,
+      "step": 2792
+    },
+    {
+      "epoch": 0.39090272918124563,
+      "grad_norm": 0.40119034319759783,
+      "learning_rate": 3.479435284313261e-05,
+      "loss": 0.5618,
+      "step": 2793
+    },
+    {
+      "epoch": 0.39104268719384183,
+      "grad_norm": 0.41230633715814125,
+      "learning_rate": 3.4783924498526184e-05,
+      "loss": 0.566,
+      "step": 2794
+    },
+    {
+      "epoch": 0.3911826452064381,
+      "grad_norm": 0.38406879859249193,
+      "learning_rate": 3.4773494143224e-05,
+      "loss": 0.5728,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3913226032190343,
+      "grad_norm": 0.42250079129339874,
+      "learning_rate": 3.476306177936961e-05,
+      "loss": 0.5937,
+      "step": 2796
+    },
+    {
+      "epoch": 0.39146256123163053,
+      "grad_norm": 0.39662665146228837,
+      "learning_rate": 3.475262740910696e-05,
+      "loss": 0.5644,
+      "step": 2797
+    },
+    {
+      "epoch": 0.39160251924422673,
+      "grad_norm": 0.42760877117863655,
+      "learning_rate": 3.474219103458043e-05,
+      "loss": 0.5762,
+      "step": 2798
+    },
+    {
+      "epoch": 0.39174247725682293,
+      "grad_norm": 0.3990049510043462,
+      "learning_rate": 3.4731752657934794e-05,
+      "loss": 0.5234,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3918824352694192,
+      "grad_norm": 0.3870544184519613,
+      "learning_rate": 3.4721312281315236e-05,
+      "loss": 0.5921,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3920223932820154,
+      "grad_norm": 0.3907379222640409,
+      "learning_rate": 3.471086990686737e-05,
+      "loss": 0.5402,
+      "step": 2801
+    },
+    {
+      "epoch": 0.39216235129461163,
+      "grad_norm": 0.42863016339615406,
+      "learning_rate": 3.470042553673721e-05,
+      "loss": 0.6075,
+      "step": 2802
+    },
+    {
+      "epoch": 0.39230230930720783,
+      "grad_norm": 0.40157662221583645,
+      "learning_rate": 3.468997917307118e-05,
+      "loss": 0.5723,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3924422673198041,
+      "grad_norm": 0.42138157521984687,
+      "learning_rate": 3.46795308180161e-05,
+      "loss": 0.5605,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3925822253324003,
+      "grad_norm": 0.39033795656224785,
+      "learning_rate": 3.466908047371923e-05,
+      "loss": 0.5814,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3927221833449965,
+      "grad_norm": 0.3904474581637845,
+      "learning_rate": 3.465862814232822e-05,
+      "loss": 0.5662,
+      "step": 2806
+    },
+    {
+      "epoch": 0.39286214135759273,
+      "grad_norm": 0.4019162642970993,
+      "learning_rate": 3.464817382599112e-05,
+      "loss": 0.5509,
+      "step": 2807
+    },
+    {
+      "epoch": 0.39300209937018893,
+      "grad_norm": 0.4060877755013805,
+      "learning_rate": 3.4637717526856406e-05,
+      "loss": 0.5642,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3931420573827852,
+      "grad_norm": 0.41153563140737226,
+      "learning_rate": 3.462725924707295e-05,
+      "loss": 0.574,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3932820153953814,
+      "grad_norm": 0.4177210107228342,
+      "learning_rate": 3.461679898879004e-05,
+      "loss": 0.6133,
+      "step": 2810
+    },
+    {
+      "epoch": 0.39342197340797763,
+      "grad_norm": 0.3996252120493831,
+      "learning_rate": 3.460633675415736e-05,
+      "loss": 0.5635,
+      "step": 2811
+    },
+    {
+      "epoch": 0.39356193142057383,
+      "grad_norm": 0.417067202150554,
+      "learning_rate": 3.459587254532502e-05,
+      "loss": 0.548,
+      "step": 2812
+    },
+    {
+      "epoch": 0.39370188943317,
+      "grad_norm": 0.4855752987156052,
+      "learning_rate": 3.458540636444349e-05,
+      "loss": 0.5833,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3938418474457663,
+      "grad_norm": 0.40124325073974404,
+      "learning_rate": 3.457493821366369e-05,
+      "loss": 0.5403,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3939818054583625,
+      "grad_norm": 0.41052139871938925,
+      "learning_rate": 3.456446809513695e-05,
+      "loss": 0.5639,
+      "step": 2815
+    },
+    {
+      "epoch": 0.39412176347095873,
+      "grad_norm": 0.429168200503203,
+      "learning_rate": 3.455399601101497e-05,
+      "loss": 0.6069,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3942617214835549,
+      "grad_norm": 0.4130675301409534,
+      "learning_rate": 3.4543521963449857e-05,
+      "loss": 0.5628,
+      "step": 2817
+    },
+    {
+      "epoch": 0.3944016794961512,
+      "grad_norm": 0.4028085056325733,
+      "learning_rate": 3.4533045954594164e-05,
+      "loss": 0.5806,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3945416375087474,
+      "grad_norm": 0.40759739919737564,
+      "learning_rate": 3.452256798660079e-05,
+      "loss": 0.6007,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3946815955213436,
+      "grad_norm": 0.41359320045974896,
+      "learning_rate": 3.4512088061623075e-05,
+      "loss": 0.5912,
+      "step": 2820
+    },
+    {
+      "epoch": 0.39482155353393983,
+      "grad_norm": 0.3896223345474862,
+      "learning_rate": 3.450160618181476e-05,
+      "loss": 0.5583,
+      "step": 2821
+    },
+    {
+      "epoch": 0.394961511546536,
+      "grad_norm": 0.41092585101647605,
+      "learning_rate": 3.449112234932996e-05,
+      "loss": 0.5571,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3951014695591323,
+      "grad_norm": 0.42614487405443774,
+      "learning_rate": 3.4480636566323215e-05,
+      "loss": 0.6322,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3952414275717285,
+      "grad_norm": 0.41152848618890253,
+      "learning_rate": 3.447014883494946e-05,
+      "loss": 0.5636,
+      "step": 2824
+    },
+    {
+      "epoch": 0.39538138558432473,
+      "grad_norm": 0.4043580579983999,
+      "learning_rate": 3.445965915736403e-05,
+      "loss": 0.5338,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3955213435969209,
+      "grad_norm": 0.3944340553638311,
+      "learning_rate": 3.444916753572266e-05,
+      "loss": 0.5345,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3956613016095171,
+      "grad_norm": 0.4032299392463221,
+      "learning_rate": 3.44386739721815e-05,
+      "loss": 0.5645,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3958012596221134,
+      "grad_norm": 0.43128038862284984,
+      "learning_rate": 3.442817846889705e-05,
+      "loss": 0.5793,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3959412176347096,
+      "grad_norm": 0.3904629770048097,
+      "learning_rate": 3.4417681028026276e-05,
+      "loss": 0.5594,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3960811756473058,
+      "grad_norm": 0.4032671372019056,
+      "learning_rate": 3.4407181651726495e-05,
+      "loss": 0.6305,
+      "step": 2830
+    },
+    {
+      "epoch": 0.396221133659902,
+      "grad_norm": 0.41716969232369355,
+      "learning_rate": 3.439668034215543e-05,
+      "loss": 0.5661,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3963610916724983,
+      "grad_norm": 0.3785072143764436,
+      "learning_rate": 3.4386177101471216e-05,
+      "loss": 0.5378,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3965010496850945,
+      "grad_norm": 0.4062464075615544,
+      "learning_rate": 3.437567193183237e-05,
+      "loss": 0.5613,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3966410076976907,
+      "grad_norm": 0.41547338740317635,
+      "learning_rate": 3.436516483539781e-05,
+      "loss": 0.5652,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3967809657102869,
+      "grad_norm": 0.4212387411967819,
+      "learning_rate": 3.4354655814326845e-05,
+      "loss": 0.5759,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3969209237228831,
+      "grad_norm": 0.39412673101104495,
+      "learning_rate": 3.434414487077921e-05,
+      "loss": 0.5455,
+      "step": 2836
+    },
+    {
+      "epoch": 0.3970608817354794,
+      "grad_norm": 0.40728215607499646,
+      "learning_rate": 3.433363200691498e-05,
+      "loss": 0.5677,
+      "step": 2837
+    },
+    {
+      "epoch": 0.3972008397480756,
+      "grad_norm": 0.4180649072512069,
+      "learning_rate": 3.432311722489467e-05,
+      "loss": 0.5845,
+      "step": 2838
+    },
+    {
+      "epoch": 0.39734079776067177,
+      "grad_norm": 0.4237453415621929,
+      "learning_rate": 3.431260052687919e-05,
+      "loss": 0.5543,
+      "step": 2839
+    },
+    {
+      "epoch": 0.397480755773268,
+      "grad_norm": 0.47750072518183245,
+      "learning_rate": 3.430208191502979e-05,
+      "loss": 0.5799,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3976207137858642,
+      "grad_norm": 0.40826704545334747,
+      "learning_rate": 3.4291561391508185e-05,
+      "loss": 0.5524,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3977606717984605,
+      "grad_norm": 0.4353573374399257,
+      "learning_rate": 3.428103895847644e-05,
+      "loss": 0.5489,
+      "step": 2842
+    },
+    {
+      "epoch": 0.39790062981105667,
+      "grad_norm": 0.4164859565724006,
+      "learning_rate": 3.427051461809702e-05,
+      "loss": 0.5694,
+      "step": 2843
+    },
+    {
+      "epoch": 0.3980405878236529,
+      "grad_norm": 0.4285455810589633,
+      "learning_rate": 3.425998837253278e-05,
+      "loss": 0.5879,
+      "step": 2844
+    },
+    {
+      "epoch": 0.3981805458362491,
+      "grad_norm": 0.3797736872864129,
+      "learning_rate": 3.4249460223946975e-05,
+      "loss": 0.555,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3983205038488453,
+      "grad_norm": 0.39764697478634364,
+      "learning_rate": 3.4238930174503245e-05,
+      "loss": 0.5683,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3984604618614416,
+      "grad_norm": 0.40768853681588346,
+      "learning_rate": 3.422839822636562e-05,
+      "loss": 0.5822,
+      "step": 2847
+    },
+    {
+      "epoch": 0.39860041987403777,
+      "grad_norm": 0.40077863607268926,
+      "learning_rate": 3.421786438169852e-05,
+      "loss": 0.5218,
+      "step": 2848
+    },
+    {
+      "epoch": 0.398740377886634,
+      "grad_norm": 0.43178105768051894,
+      "learning_rate": 3.420732864266677e-05,
+      "loss": 0.5917,
+      "step": 2849
+    },
+    {
+      "epoch": 0.3988803358992302,
+      "grad_norm": 0.4012531476036567,
+      "learning_rate": 3.4196791011435554e-05,
+      "loss": 0.5575,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3990202939118265,
+      "grad_norm": 0.42052995566229917,
+      "learning_rate": 3.418625149017047e-05,
+      "loss": 0.5898,
+      "step": 2851
+    },
+    {
+      "epoch": 0.39916025192442267,
+      "grad_norm": 0.48024947250916206,
+      "learning_rate": 3.4175710081037505e-05,
+      "loss": 0.5438,
+      "step": 2852
+    },
+    {
+      "epoch": 0.39930020993701887,
+      "grad_norm": 0.3857911120539825,
+      "learning_rate": 3.4165166786203015e-05,
+      "loss": 0.5641,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3994401679496151,
+      "grad_norm": 0.3948093390858907,
+      "learning_rate": 3.415462160783375e-05,
+      "loss": 0.5655,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3995801259622113,
+      "grad_norm": 0.4003575934592238,
+      "learning_rate": 3.4144074548096866e-05,
+      "loss": 0.6106,
+      "step": 2855
+    },
+    {
+      "epoch": 0.39972008397480757,
+      "grad_norm": 0.4321737583237654,
+      "learning_rate": 3.413352560915988e-05,
+      "loss": 0.5514,
+      "step": 2856
+    },
+    {
+      "epoch": 0.39986004198740377,
+      "grad_norm": 0.45894659025440354,
+      "learning_rate": 3.412297479319071e-05,
+      "loss": 0.5669,
+      "step": 2857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42987083801881937,
+      "learning_rate": 3.411242210235765e-05,
+      "loss": 0.5974,
+      "step": 2858
+    },
+    {
+      "epoch": 0.4001399580125962,
+      "grad_norm": 0.40207997706206416,
+      "learning_rate": 3.4101867538829394e-05,
+      "loss": 0.5286,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4002799160251924,
+      "grad_norm": 0.40235651191545074,
+      "learning_rate": 3.4091311104775e-05,
+      "loss": 0.5427,
+      "step": 2860
+    },
+    {
+      "epoch": 0.40041987403778867,
+      "grad_norm": 0.40231984058011133,
+      "learning_rate": 3.4080752802363936e-05,
+      "loss": 0.5341,
+      "step": 2861
+    },
+    {
+      "epoch": 0.40055983205038487,
+      "grad_norm": 0.4240007925517186,
+      "learning_rate": 3.4070192633766025e-05,
+      "loss": 0.5571,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4006997900629811,
+      "grad_norm": 0.3961994599738629,
+      "learning_rate": 3.4059630601151504e-05,
+      "loss": 0.5585,
+      "step": 2863
+    },
+    {
+      "epoch": 0.4008397480755773,
+      "grad_norm": 0.3908273444014209,
+      "learning_rate": 3.404906670669097e-05,
+      "loss": 0.5444,
+      "step": 2864
+    },
+    {
+      "epoch": 0.40097970608817357,
+      "grad_norm": 0.4011478963975767,
+      "learning_rate": 3.403850095255542e-05,
+      "loss": 0.601,
+      "step": 2865
+    },
+    {
+      "epoch": 0.40111966410076977,
+      "grad_norm": 0.4492421654680456,
+      "learning_rate": 3.402793334091621e-05,
+      "loss": 0.5953,
+      "step": 2866
+    },
+    {
+      "epoch": 0.40125962211336597,
+      "grad_norm": 0.4207454261749868,
+      "learning_rate": 3.40173638739451e-05,
+      "loss": 0.5931,
+      "step": 2867
+    },
+    {
+      "epoch": 0.4013995801259622,
+      "grad_norm": 0.40623369999208175,
+      "learning_rate": 3.400679255381421e-05,
+      "loss": 0.5528,
+      "step": 2868
+    },
+    {
+      "epoch": 0.4015395381385584,
+      "grad_norm": 0.4169914185414076,
+      "learning_rate": 3.399621938269606e-05,
+      "loss": 0.5823,
+      "step": 2869
+    },
+    {
+      "epoch": 0.40167949615115467,
+      "grad_norm": 0.42169370060384803,
+      "learning_rate": 3.3985644362763566e-05,
+      "loss": 0.5565,
+      "step": 2870
+    },
+    {
+      "epoch": 0.40181945416375087,
+      "grad_norm": 0.38157604657299227,
+      "learning_rate": 3.3975067496189965e-05,
+      "loss": 0.525,
+      "step": 2871
+    },
+    {
+      "epoch": 0.4019594121763471,
+      "grad_norm": 0.42241067538345,
+      "learning_rate": 3.396448878514894e-05,
+      "loss": 0.577,
+      "step": 2872
+    },
+    {
+      "epoch": 0.4020993701889433,
+      "grad_norm": 0.4164649305478365,
+      "learning_rate": 3.395390823181451e-05,
+      "loss": 0.5334,
+      "step": 2873
+    },
+    {
+      "epoch": 0.4022393282015395,
+      "grad_norm": 0.4224179515123039,
+      "learning_rate": 3.3943325838361084e-05,
+      "loss": 0.6014,
+      "step": 2874
+    },
+    {
+      "epoch": 0.40237928621413577,
+      "grad_norm": 0.4243031111453756,
+      "learning_rate": 3.393274160696346e-05,
+      "loss": 0.5759,
+      "step": 2875
+    },
+    {
+      "epoch": 0.40251924422673196,
+      "grad_norm": 0.4214607029417847,
+      "learning_rate": 3.392215553979679e-05,
+      "loss": 0.6041,
+      "step": 2876
+    },
+    {
+      "epoch": 0.4026592022393282,
+      "grad_norm": 0.39897243417459854,
+      "learning_rate": 3.391156763903665e-05,
+      "loss": 0.5425,
+      "step": 2877
+    },
+    {
+      "epoch": 0.4027991602519244,
+      "grad_norm": 0.39128320703090635,
+      "learning_rate": 3.390097790685892e-05,
+      "loss": 0.579,
+      "step": 2878
+    },
+    {
+      "epoch": 0.40293911826452067,
+      "grad_norm": 0.40615873009003534,
+      "learning_rate": 3.389038634543993e-05,
+      "loss": 0.5828,
+      "step": 2879
+    },
+    {
+      "epoch": 0.40307907627711687,
+      "grad_norm": 0.41878420947323897,
+      "learning_rate": 3.387979295695632e-05,
+      "loss": 0.5828,
+      "step": 2880
+    },
+    {
+      "epoch": 0.40321903428971306,
+      "grad_norm": 0.383331359104639,
+      "learning_rate": 3.386919774358518e-05,
+      "loss": 0.5384,
+      "step": 2881
+    },
+    {
+      "epoch": 0.4033589923023093,
+      "grad_norm": 0.41913779228819376,
+      "learning_rate": 3.38586007075039e-05,
+      "loss": 0.5591,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4034989503149055,
+      "grad_norm": 0.38348801468846716,
+      "learning_rate": 3.3848001850890296e-05,
+      "loss": 0.5393,
+      "step": 2883
+    },
+    {
+      "epoch": 0.40363890832750177,
+      "grad_norm": 0.4113992124017339,
+      "learning_rate": 3.383740117592254e-05,
+      "loss": 0.5567,
+      "step": 2884
+    },
+    {
+      "epoch": 0.40377886634009796,
+      "grad_norm": 0.40383694106615,
+      "learning_rate": 3.382679868477917e-05,
+      "loss": 0.5686,
+      "step": 2885
+    },
+    {
+      "epoch": 0.4039188243526942,
+      "grad_norm": 0.4275501335912839,
+      "learning_rate": 3.381619437963911e-05,
+      "loss": 0.5709,
+      "step": 2886
+    },
+    {
+      "epoch": 0.4040587823652904,
+      "grad_norm": 0.410136210120104,
+      "learning_rate": 3.3805588262681655e-05,
+      "loss": 0.5578,
+      "step": 2887
+    },
+    {
+      "epoch": 0.4041987403778866,
+      "grad_norm": 0.39945476818253844,
+      "learning_rate": 3.379498033608646e-05,
+      "loss": 0.5576,
+      "step": 2888
+    },
+    {
+      "epoch": 0.40433869839048286,
+      "grad_norm": 0.41503488018923496,
+      "learning_rate": 3.378437060203357e-05,
+      "loss": 0.5621,
+      "step": 2889
+    },
+    {
+      "epoch": 0.40447865640307906,
+      "grad_norm": 0.4205244822133629,
+      "learning_rate": 3.3773759062703396e-05,
+      "loss": 0.571,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4046186144156753,
+      "grad_norm": 0.40721392794208344,
+      "learning_rate": 3.376314572027671e-05,
+      "loss": 0.5539,
+      "step": 2891
+    },
+    {
+      "epoch": 0.4047585724282715,
+      "grad_norm": 0.4061261621378902,
+      "learning_rate": 3.375253057693466e-05,
+      "loss": 0.5502,
+      "step": 2892
+    },
+    {
+      "epoch": 0.40489853044086777,
+      "grad_norm": 0.4040934767631151,
+      "learning_rate": 3.374191363485877e-05,
+      "loss": 0.5502,
+      "step": 2893
+    },
+    {
+      "epoch": 0.40503848845346396,
+      "grad_norm": 0.4140086272358693,
+      "learning_rate": 3.373129489623093e-05,
+      "loss": 0.5752,
+      "step": 2894
+    },
+    {
+      "epoch": 0.40517844646606016,
+      "grad_norm": 0.4285410515445642,
+      "learning_rate": 3.3720674363233396e-05,
+      "loss": 0.5641,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4053184044786564,
+      "grad_norm": 0.3886401656745267,
+      "learning_rate": 3.37100520380488e-05,
+      "loss": 0.5721,
+      "step": 2896
+    },
+    {
+      "epoch": 0.4054583624912526,
+      "grad_norm": 0.4252355420954435,
+      "learning_rate": 3.369942792286013e-05,
+      "loss": 0.5494,
+      "step": 2897
+    },
+    {
+      "epoch": 0.40559832050384886,
+      "grad_norm": 0.44644708491027574,
+      "learning_rate": 3.368880201985076e-05,
+      "loss": 0.5722,
+      "step": 2898
+    },
+    {
+      "epoch": 0.40573827851644506,
+      "grad_norm": 0.42435542539228505,
+      "learning_rate": 3.367817433120441e-05,
+      "loss": 0.5744,
+      "step": 2899
+    },
+    {
+      "epoch": 0.4058782365290413,
+      "grad_norm": 0.41641484378908183,
+      "learning_rate": 3.366754485910518e-05,
+      "loss": 0.5874,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4060181945416375,
+      "grad_norm": 0.40783494684210103,
+      "learning_rate": 3.365691360573754e-05,
+      "loss": 0.5561,
+      "step": 2901
+    },
+    {
+      "epoch": 0.4061581525542337,
+      "grad_norm": 0.3940864417785125,
+      "learning_rate": 3.3646280573286314e-05,
+      "loss": 0.5437,
+      "step": 2902
+    },
+    {
+      "epoch": 0.40629811056682996,
+      "grad_norm": 0.41032563545142275,
+      "learning_rate": 3.363564576393671e-05,
+      "loss": 0.5679,
+      "step": 2903
+    },
+    {
+      "epoch": 0.40643806857942616,
+      "grad_norm": 0.43293587153461693,
+      "learning_rate": 3.3625009179874265e-05,
+      "loss": 0.6005,
+      "step": 2904
+    },
+    {
+      "epoch": 0.4065780265920224,
+      "grad_norm": 0.41913595404532833,
+      "learning_rate": 3.361437082328493e-05,
+      "loss": 0.5841,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4067179846046186,
+      "grad_norm": 0.39955188014787174,
+      "learning_rate": 3.360373069635498e-05,
+      "loss": 0.5762,
+      "step": 2906
+    },
+    {
+      "epoch": 0.40685794261721486,
+      "grad_norm": 0.40532054476042145,
+      "learning_rate": 3.359308880127108e-05,
+      "loss": 0.5811,
+      "step": 2907
+    },
+    {
+      "epoch": 0.40699790062981106,
+      "grad_norm": 0.3950198242019259,
+      "learning_rate": 3.358244514022025e-05,
+      "loss": 0.5795,
+      "step": 2908
+    },
+    {
+      "epoch": 0.40713785864240726,
+      "grad_norm": 0.4211880514353567,
+      "learning_rate": 3.357179971538985e-05,
+      "loss": 0.5606,
+      "step": 2909
+    },
+    {
+      "epoch": 0.4072778166550035,
+      "grad_norm": 0.4142064942386179,
+      "learning_rate": 3.3561152528967646e-05,
+      "loss": 0.5802,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4074177746675997,
+      "grad_norm": 0.41844753492725145,
+      "learning_rate": 3.355050358314172e-05,
+      "loss": 0.5274,
+      "step": 2911
+    },
+    {
+      "epoch": 0.40755773268019596,
+      "grad_norm": 0.41589536675167615,
+      "learning_rate": 3.353985288010056e-05,
+      "loss": 0.5938,
+      "step": 2912
+    },
+    {
+      "epoch": 0.40769769069279216,
+      "grad_norm": 0.39984810028580087,
+      "learning_rate": 3.352920042203298e-05,
+      "loss": 0.58,
+      "step": 2913
+    },
+    {
+      "epoch": 0.40783764870538836,
+      "grad_norm": 0.4140807749556365,
+      "learning_rate": 3.3518546211128166e-05,
+      "loss": 0.5491,
+      "step": 2914
+    },
+    {
+      "epoch": 0.4079776067179846,
+      "grad_norm": 0.4328655244745967,
+      "learning_rate": 3.350789024957568e-05,
+      "loss": 0.5972,
+      "step": 2915
+    },
+    {
+      "epoch": 0.4081175647305808,
+      "grad_norm": 0.3996263118365605,
+      "learning_rate": 3.349723253956542e-05,
+      "loss": 0.5629,
+      "step": 2916
+    },
+    {
+      "epoch": 0.40825752274317706,
+      "grad_norm": 0.4138825890216495,
+      "learning_rate": 3.348657308328766e-05,
+      "loss": 0.5461,
+      "step": 2917
+    },
+    {
+      "epoch": 0.40839748075577326,
+      "grad_norm": 0.45384712853313164,
+      "learning_rate": 3.3475911882933015e-05,
+      "loss": 0.5456,
+      "step": 2918
+    },
+    {
+      "epoch": 0.4085374387683695,
+      "grad_norm": 0.4085437808720945,
+      "learning_rate": 3.346524894069248e-05,
+      "loss": 0.5453,
+      "step": 2919
+    },
+    {
+      "epoch": 0.4086773967809657,
+      "grad_norm": 0.4038494007288198,
+      "learning_rate": 3.3454584258757404e-05,
+      "loss": 0.5623,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4088173547935619,
+      "grad_norm": 0.4566167974871496,
+      "learning_rate": 3.344391783931947e-05,
+      "loss": 0.5907,
+      "step": 2921
+    },
+    {
+      "epoch": 0.40895731280615816,
+      "grad_norm": 0.3969364122046131,
+      "learning_rate": 3.343324968457076e-05,
+      "loss": 0.5131,
+      "step": 2922
+    },
+    {
+      "epoch": 0.40909727081875435,
+      "grad_norm": 0.4203451679837855,
+      "learning_rate": 3.342257979670365e-05,
+      "loss": 0.5644,
+      "step": 2923
+    },
+    {
+      "epoch": 0.4092372288313506,
+      "grad_norm": 0.40925859811350057,
+      "learning_rate": 3.341190817791094e-05,
+      "loss": 0.5498,
+      "step": 2924
+    },
+    {
+      "epoch": 0.4093771868439468,
+      "grad_norm": 0.4083989313014574,
+      "learning_rate": 3.3401234830385756e-05,
+      "loss": 0.5854,
+      "step": 2925
+    },
+    {
+      "epoch": 0.40951714485654306,
+      "grad_norm": 0.39780928336254934,
+      "learning_rate": 3.3390559756321566e-05,
+      "loss": 0.5659,
+      "step": 2926
+    },
+    {
+      "epoch": 0.40965710286913926,
+      "grad_norm": 0.4109873033834379,
+      "learning_rate": 3.337988295791221e-05,
+      "loss": 0.5422,
+      "step": 2927
+    },
+    {
+      "epoch": 0.40979706088173545,
+      "grad_norm": 0.3952927193405086,
+      "learning_rate": 3.3369204437351886e-05,
+      "loss": 0.5592,
+      "step": 2928
+    },
+    {
+      "epoch": 0.4099370188943317,
+      "grad_norm": 0.40670959006044405,
+      "learning_rate": 3.335852419683513e-05,
+      "loss": 0.5711,
+      "step": 2929
+    },
+    {
+      "epoch": 0.4100769769069279,
+      "grad_norm": 0.3974994882518793,
+      "learning_rate": 3.3347842238556836e-05,
+      "loss": 0.5315,
+      "step": 2930
+    },
+    {
+      "epoch": 0.41021693491952416,
+      "grad_norm": 0.44234521625168705,
+      "learning_rate": 3.3337158564712267e-05,
+      "loss": 0.5552,
+      "step": 2931
+    },
+    {
+      "epoch": 0.41035689293212035,
+      "grad_norm": 0.40022379185542395,
+      "learning_rate": 3.332647317749702e-05,
+      "loss": 0.5745,
+      "step": 2932
+    },
+    {
+      "epoch": 0.4104968509447166,
+      "grad_norm": 0.418577329208144,
+      "learning_rate": 3.3315786079107055e-05,
+      "loss": 0.5947,
+      "step": 2933
+    },
+    {
+      "epoch": 0.4106368089573128,
+      "grad_norm": 0.40576821316747064,
+      "learning_rate": 3.3305097271738665e-05,
+      "loss": 0.6025,
+      "step": 2934
+    },
+    {
+      "epoch": 0.410776766969909,
+      "grad_norm": 0.4066641994376824,
+      "learning_rate": 3.329440675758853e-05,
+      "loss": 0.5475,
+      "step": 2935
+    },
+    {
+      "epoch": 0.41091672498250525,
+      "grad_norm": 0.4150792589177556,
+      "learning_rate": 3.3283714538853636e-05,
+      "loss": 0.5866,
+      "step": 2936
+    },
+    {
+      "epoch": 0.41105668299510145,
+      "grad_norm": 0.397074447582423,
+      "learning_rate": 3.327302061773136e-05,
+      "loss": 0.5799,
+      "step": 2937
+    },
+    {
+      "epoch": 0.4111966410076977,
+      "grad_norm": 0.40754727507905664,
+      "learning_rate": 3.3262324996419405e-05,
+      "loss": 0.5469,
+      "step": 2938
+    },
+    {
+      "epoch": 0.4113365990202939,
+      "grad_norm": 0.40164275820649376,
+      "learning_rate": 3.325162767711583e-05,
+      "loss": 0.5232,
+      "step": 2939
+    },
+    {
+      "epoch": 0.41147655703289016,
+      "grad_norm": 0.4080075519925706,
+      "learning_rate": 3.324092866201904e-05,
+      "loss": 0.5661,
+      "step": 2940
+    },
+    {
+      "epoch": 0.41161651504548635,
+      "grad_norm": 0.40149245258055244,
+      "learning_rate": 3.3230227953327796e-05,
+      "loss": 0.5474,
+      "step": 2941
+    },
+    {
+      "epoch": 0.41175647305808255,
+      "grad_norm": 0.4158994899194992,
+      "learning_rate": 3.321952555324121e-05,
+      "loss": 0.5258,
+      "step": 2942
+    },
+    {
+      "epoch": 0.4118964310706788,
+      "grad_norm": 0.4019593770294339,
+      "learning_rate": 3.320882146395871e-05,
+      "loss": 0.5762,
+      "step": 2943
+    },
+    {
+      "epoch": 0.412036389083275,
+      "grad_norm": 0.4249496622474027,
+      "learning_rate": 3.3198115687680115e-05,
+      "loss": 0.5462,
+      "step": 2944
+    },
+    {
+      "epoch": 0.41217634709587125,
+      "grad_norm": 0.43225282993592656,
+      "learning_rate": 3.318740822660556e-05,
+      "loss": 0.5446,
+      "step": 2945
+    },
+    {
+      "epoch": 0.41231630510846745,
+      "grad_norm": 0.41726887193732093,
+      "learning_rate": 3.3176699082935545e-05,
+      "loss": 0.5493,
+      "step": 2946
+    },
+    {
+      "epoch": 0.4124562631210637,
+      "grad_norm": 0.39903806845546885,
+      "learning_rate": 3.31659882588709e-05,
+      "loss": 0.5888,
+      "step": 2947
+    },
+    {
+      "epoch": 0.4125962211336599,
+      "grad_norm": 0.4023782994195669,
+      "learning_rate": 3.315527575661282e-05,
+      "loss": 0.5684,
+      "step": 2948
+    },
+    {
+      "epoch": 0.4127361791462561,
+      "grad_norm": 0.3936024186034419,
+      "learning_rate": 3.314456157836281e-05,
+      "loss": 0.5117,
+      "step": 2949
+    },
+    {
+      "epoch": 0.41287613715885235,
+      "grad_norm": 0.3952468427687398,
+      "learning_rate": 3.313384572632277e-05,
+      "loss": 0.5547,
+      "step": 2950
+    },
+    {
+      "epoch": 0.41301609517144855,
+      "grad_norm": 0.4190384583627527,
+      "learning_rate": 3.31231282026949e-05,
+      "loss": 0.5805,
+      "step": 2951
+    },
+    {
+      "epoch": 0.4131560531840448,
+      "grad_norm": 0.39658407316177025,
+      "learning_rate": 3.3112409009681766e-05,
+      "loss": 0.5406,
+      "step": 2952
+    },
+    {
+      "epoch": 0.413296011196641,
+      "grad_norm": 0.4076292942631321,
+      "learning_rate": 3.310168814948627e-05,
+      "loss": 0.588,
+      "step": 2953
+    },
+    {
+      "epoch": 0.41343596920923725,
+      "grad_norm": 0.41540765697070914,
+      "learning_rate": 3.3090965624311654e-05,
+      "loss": 0.5594,
+      "step": 2954
+    },
+    {
+      "epoch": 0.41357592722183345,
+      "grad_norm": 0.4035372057640521,
+      "learning_rate": 3.3080241436361506e-05,
+      "loss": 0.5636,
+      "step": 2955
+    },
+    {
+      "epoch": 0.41371588523442965,
+      "grad_norm": 0.41459880243511804,
+      "learning_rate": 3.3069515587839754e-05,
+      "loss": 0.5532,
+      "step": 2956
+    },
+    {
+      "epoch": 0.4138558432470259,
+      "grad_norm": 0.39031658646005546,
+      "learning_rate": 3.305878808095068e-05,
+      "loss": 0.5939,
+      "step": 2957
+    },
+    {
+      "epoch": 0.4139958012596221,
+      "grad_norm": 0.4182734311842286,
+      "learning_rate": 3.304805891789888e-05,
+      "loss": 0.5872,
+      "step": 2958
+    },
+    {
+      "epoch": 0.41413575927221835,
+      "grad_norm": 0.4222389070033302,
+      "learning_rate": 3.303732810088931e-05,
+      "loss": 0.6106,
+      "step": 2959
+    },
+    {
+      "epoch": 0.41427571728481455,
+      "grad_norm": 0.39648114038614457,
+      "learning_rate": 3.302659563212727e-05,
+      "loss": 0.5457,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4144156752974108,
+      "grad_norm": 0.42978337071116823,
+      "learning_rate": 3.301586151381839e-05,
+      "loss": 0.5644,
+      "step": 2961
+    },
+    {
+      "epoch": 0.414555633310007,
+      "grad_norm": 0.4005463311390214,
+      "learning_rate": 3.300512574816863e-05,
+      "loss": 0.5676,
+      "step": 2962
+    },
+    {
+      "epoch": 0.4146955913226032,
+      "grad_norm": 0.39670767820561714,
+      "learning_rate": 3.2994388337384306e-05,
+      "loss": 0.5683,
+      "step": 2963
+    },
+    {
+      "epoch": 0.41483554933519945,
+      "grad_norm": 0.414169510604101,
+      "learning_rate": 3.298364928367207e-05,
+      "loss": 0.5685,
+      "step": 2964
+    },
+    {
+      "epoch": 0.41497550734779565,
+      "grad_norm": 0.39829585786394445,
+      "learning_rate": 3.2972908589238896e-05,
+      "loss": 0.5627,
+      "step": 2965
+    },
+    {
+      "epoch": 0.4151154653603919,
+      "grad_norm": 0.40155559541312513,
+      "learning_rate": 3.2962166256292113e-05,
+      "loss": 0.5805,
+      "step": 2966
+    },
+    {
+      "epoch": 0.4152554233729881,
+      "grad_norm": 0.4105341813553997,
+      "learning_rate": 3.295142228703938e-05,
+      "loss": 0.6003,
+      "step": 2967
+    },
+    {
+      "epoch": 0.41539538138558435,
+      "grad_norm": 0.4006968222058459,
+      "learning_rate": 3.2940676683688677e-05,
+      "loss": 0.546,
+      "step": 2968
+    },
+    {
+      "epoch": 0.41553533939818055,
+      "grad_norm": 0.39978495742611814,
+      "learning_rate": 3.292992944844836e-05,
+      "loss": 0.5892,
+      "step": 2969
+    },
+    {
+      "epoch": 0.41567529741077675,
+      "grad_norm": 0.4016166800083618,
+      "learning_rate": 3.291918058352706e-05,
+      "loss": 0.5708,
+      "step": 2970
+    },
+    {
+      "epoch": 0.415815255423373,
+      "grad_norm": 0.40924317132383004,
+      "learning_rate": 3.290843009113382e-05,
+      "loss": 0.5291,
+      "step": 2971
+    },
+    {
+      "epoch": 0.4159552134359692,
+      "grad_norm": 0.41496924484348124,
+      "learning_rate": 3.2897677973477936e-05,
+      "loss": 0.5638,
+      "step": 2972
+    },
+    {
+      "epoch": 0.41609517144856545,
+      "grad_norm": 0.4170073162428262,
+      "learning_rate": 3.288692423276911e-05,
+      "loss": 0.5659,
+      "step": 2973
+    },
+    {
+      "epoch": 0.41623512946116165,
+      "grad_norm": 0.4393981644887219,
+      "learning_rate": 3.2876168871217325e-05,
+      "loss": 0.5674,
+      "step": 2974
+    },
+    {
+      "epoch": 0.4163750874737579,
+      "grad_norm": 0.39928933014260387,
+      "learning_rate": 3.2865411891032916e-05,
+      "loss": 0.5589,
+      "step": 2975
+    },
+    {
+      "epoch": 0.4165150454863541,
+      "grad_norm": 0.6076938066952381,
+      "learning_rate": 3.2854653294426566e-05,
+      "loss": 0.5781,
+      "step": 2976
+    },
+    {
+      "epoch": 0.4166550034989503,
+      "grad_norm": 0.4206690009626964,
+      "learning_rate": 3.284389308360927e-05,
+      "loss": 0.5945,
+      "step": 2977
+    },
+    {
+      "epoch": 0.41679496151154655,
+      "grad_norm": 0.38925483204393885,
+      "learning_rate": 3.2833131260792345e-05,
+      "loss": 0.527,
+      "step": 2978
+    },
+    {
+      "epoch": 0.41693491952414274,
+      "grad_norm": 0.43019105225997517,
+      "learning_rate": 3.282236782818747e-05,
+      "loss": 0.5975,
+      "step": 2979
+    },
+    {
+      "epoch": 0.417074877536739,
+      "grad_norm": 0.4151005702647614,
+      "learning_rate": 3.2811602788006645e-05,
+      "loss": 0.5512,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4172148355493352,
+      "grad_norm": 0.43155499812179654,
+      "learning_rate": 3.280083614246218e-05,
+      "loss": 0.5582,
+      "step": 2981
+    },
+    {
+      "epoch": 0.41735479356193145,
+      "grad_norm": 0.4147678068443396,
+      "learning_rate": 3.279006789376674e-05,
+      "loss": 0.5355,
+      "step": 2982
+    },
+    {
+      "epoch": 0.41749475157452764,
+      "grad_norm": 0.40986730591235127,
+      "learning_rate": 3.2779298044133304e-05,
+      "loss": 0.5521,
+      "step": 2983
+    },
+    {
+      "epoch": 0.41763470958712384,
+      "grad_norm": 0.41750246578218936,
+      "learning_rate": 3.276852659577519e-05,
+      "loss": 0.5642,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4177746675997201,
+      "grad_norm": 0.4008235425339085,
+      "learning_rate": 3.275775355090603e-05,
+      "loss": 0.5385,
+      "step": 2985
+    },
+    {
+      "epoch": 0.4179146256123163,
+      "grad_norm": 0.42115213204670543,
+      "learning_rate": 3.274697891173982e-05,
+      "loss": 0.597,
+      "step": 2986
+    },
+    {
+      "epoch": 0.41805458362491255,
+      "grad_norm": 0.40847799222276204,
+      "learning_rate": 3.273620268049083e-05,
+      "loss": 0.5372,
+      "step": 2987
+    },
+    {
+      "epoch": 0.41819454163750874,
+      "grad_norm": 0.39981522543632353,
+      "learning_rate": 3.272542485937369e-05,
+      "loss": 0.5812,
+      "step": 2988
+    },
+    {
+      "epoch": 0.418334499650105,
+      "grad_norm": 0.4281898310803181,
+      "learning_rate": 3.271464545060336e-05,
+      "loss": 0.6052,
+      "step": 2989
+    },
+    {
+      "epoch": 0.4184744576627012,
+      "grad_norm": 0.3960705407504793,
+      "learning_rate": 3.2703864456395106e-05,
+      "loss": 0.5535,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4186144156752974,
+      "grad_norm": 0.40782715021367766,
+      "learning_rate": 3.2693081878964546e-05,
+      "loss": 0.5444,
+      "step": 2991
+    },
+    {
+      "epoch": 0.41875437368789364,
+      "grad_norm": 0.4108892545513627,
+      "learning_rate": 3.2682297720527596e-05,
+      "loss": 0.588,
+      "step": 2992
+    },
+    {
+      "epoch": 0.41889433170048984,
+      "grad_norm": 0.4052095905789736,
+      "learning_rate": 3.267151198330053e-05,
+      "loss": 0.6009,
+      "step": 2993
+    },
+    {
+      "epoch": 0.4190342897130861,
+      "grad_norm": 0.3957928902994827,
+      "learning_rate": 3.2660724669499906e-05,
+      "loss": 0.5503,
+      "step": 2994
+    },
+    {
+      "epoch": 0.4191742477256823,
+      "grad_norm": 0.3945291064646862,
+      "learning_rate": 3.264993578134263e-05,
+      "loss": 0.5456,
+      "step": 2995
+    },
+    {
+      "epoch": 0.4193142057382785,
+      "grad_norm": 0.38055971820367784,
+      "learning_rate": 3.263914532104593e-05,
+      "loss": 0.575,
+      "step": 2996
+    },
+    {
+      "epoch": 0.41945416375087474,
+      "grad_norm": 0.3990796286153784,
+      "learning_rate": 3.2628353290827365e-05,
+      "loss": 0.5653,
+      "step": 2997
+    },
+    {
+      "epoch": 0.41959412176347094,
+      "grad_norm": 0.39699323320776897,
+      "learning_rate": 3.2617559692904784e-05,
+      "loss": 0.5567,
+      "step": 2998
+    },
+    {
+      "epoch": 0.4197340797760672,
+      "grad_norm": 0.42070498470477863,
+      "learning_rate": 3.260676452949641e-05,
+      "loss": 0.5643,
+      "step": 2999
+    },
+    {
+      "epoch": 0.4198740377886634,
+      "grad_norm": 0.41895847299595507,
+      "learning_rate": 3.259596780282074e-05,
+      "loss": 0.5866,
+      "step": 3000
+    },
+    {
+      "epoch": 0.42001399580125964,
+      "grad_norm": 0.4054149661332881,
+      "learning_rate": 3.2585169515096615e-05,
+      "loss": 0.5221,
+      "step": 3001
+    },
+    {
+      "epoch": 0.42015395381385584,
+      "grad_norm": 0.41159580867468976,
+      "learning_rate": 3.257436966854319e-05,
+      "loss": 0.5409,
+      "step": 3002
+    },
+    {
+      "epoch": 0.42029391182645204,
+      "grad_norm": 0.41565485170065436,
+      "learning_rate": 3.256356826537994e-05,
+      "loss": 0.572,
+      "step": 3003
+    },
+    {
+      "epoch": 0.4204338698390483,
+      "grad_norm": 0.4011710239277276,
+      "learning_rate": 3.255276530782667e-05,
+      "loss": 0.5399,
+      "step": 3004
+    },
+    {
+      "epoch": 0.4205738278516445,
+      "grad_norm": 0.45007473746050475,
+      "learning_rate": 3.25419607981035e-05,
+      "loss": 0.5283,
+      "step": 3005
+    },
+    {
+      "epoch": 0.42071378586424074,
+      "grad_norm": 0.4258234246567913,
+      "learning_rate": 3.253115473843086e-05,
+      "loss": 0.6111,
+      "step": 3006
+    },
+    {
+      "epoch": 0.42085374387683694,
+      "grad_norm": 0.41531516468042334,
+      "learning_rate": 3.252034713102951e-05,
+      "loss": 0.5644,
+      "step": 3007
+    },
+    {
+      "epoch": 0.4209937018894332,
+      "grad_norm": 0.41220387089241467,
+      "learning_rate": 3.250953797812051e-05,
+      "loss": 0.5798,
+      "step": 3008
+    },
+    {
+      "epoch": 0.4211336599020294,
+      "grad_norm": 0.4165684547155104,
+      "learning_rate": 3.249872728192527e-05,
+      "loss": 0.578,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4212736179146256,
+      "grad_norm": 0.41641177346916974,
+      "learning_rate": 3.248791504466548e-05,
+      "loss": 0.5847,
+      "step": 3010
+    },
+    {
+      "epoch": 0.42141357592722184,
+      "grad_norm": 0.4230423352548566,
+      "learning_rate": 3.2477101268563184e-05,
+      "loss": 0.5578,
+      "step": 3011
+    },
+    {
+      "epoch": 0.42155353393981804,
+      "grad_norm": 0.40182190163643416,
+      "learning_rate": 3.24662859558407e-05,
+      "loss": 0.5322,
+      "step": 3012
+    },
+    {
+      "epoch": 0.4216934919524143,
+      "grad_norm": 0.43511787431355525,
+      "learning_rate": 3.245546910872071e-05,
+      "loss": 0.5473,
+      "step": 3013
+    },
+    {
+      "epoch": 0.4218334499650105,
+      "grad_norm": 0.41338327961202165,
+      "learning_rate": 3.244465072942615e-05,
+      "loss": 0.5883,
+      "step": 3014
+    },
+    {
+      "epoch": 0.42197340797760674,
+      "grad_norm": 0.39971647364371277,
+      "learning_rate": 3.2433830820180346e-05,
+      "loss": 0.5384,
+      "step": 3015
+    },
+    {
+      "epoch": 0.42211336599020294,
+      "grad_norm": 0.4417586008166198,
+      "learning_rate": 3.2423009383206876e-05,
+      "loss": 0.5662,
+      "step": 3016
+    },
+    {
+      "epoch": 0.42225332400279914,
+      "grad_norm": 0.41967998090343933,
+      "learning_rate": 3.241218642072966e-05,
+      "loss": 0.5456,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4223932820153954,
+      "grad_norm": 0.41986891673954385,
+      "learning_rate": 3.240136193497293e-05,
+      "loss": 0.5633,
+      "step": 3018
+    },
+    {
+      "epoch": 0.4225332400279916,
+      "grad_norm": 0.4038319504416983,
+      "learning_rate": 3.239053592816122e-05,
+      "loss": 0.5405,
+      "step": 3019
+    },
+    {
+      "epoch": 0.42267319804058784,
+      "grad_norm": 0.42072486493920963,
+      "learning_rate": 3.2379708402519394e-05,
+      "loss": 0.5397,
+      "step": 3020
+    },
+    {
+      "epoch": 0.42281315605318404,
+      "grad_norm": 0.3963667588002371,
+      "learning_rate": 3.2368879360272606e-05,
+      "loss": 0.5437,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4229531140657803,
+      "grad_norm": 0.43813839588807524,
+      "learning_rate": 3.235804880364635e-05,
+      "loss": 0.6218,
+      "step": 3022
+    },
+    {
+      "epoch": 0.4230930720783765,
+      "grad_norm": 0.420130728007189,
+      "learning_rate": 3.23472167348664e-05,
+      "loss": 0.5528,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4232330300909727,
+      "grad_norm": 0.3931327455825961,
+      "learning_rate": 3.233638315615887e-05,
+      "loss": 0.5493,
+      "step": 3024
+    },
+    {
+      "epoch": 0.42337298810356894,
+      "grad_norm": 0.4231921022929699,
+      "learning_rate": 3.232554806975016e-05,
+      "loss": 0.5567,
+      "step": 3025
+    },
+    {
+      "epoch": 0.42351294611616513,
+      "grad_norm": 0.4136408960013234,
+      "learning_rate": 3.231471147786699e-05,
+      "loss": 0.6057,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4236529041287614,
+      "grad_norm": 0.3990285743665062,
+      "learning_rate": 3.23038733827364e-05,
+      "loss": 0.5888,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4237928621413576,
+      "grad_norm": 0.38749988478653646,
+      "learning_rate": 3.2293033786585716e-05,
+      "loss": 0.5386,
+      "step": 3028
+    },
+    {
+      "epoch": 0.42393282015395384,
+      "grad_norm": 0.4001033239780827,
+      "learning_rate": 3.228219269164259e-05,
+      "loss": 0.5741,
+      "step": 3029
+    },
+    {
+      "epoch": 0.42407277816655004,
+      "grad_norm": 0.38245836943637973,
+      "learning_rate": 3.2271350100134975e-05,
+      "loss": 0.5658,
+      "step": 3030
+    },
+    {
+      "epoch": 0.42421273617914623,
+      "grad_norm": 0.3936968710553453,
+      "learning_rate": 3.226050601429115e-05,
+      "loss": 0.572,
+      "step": 3031
+    },
+    {
+      "epoch": 0.4243526941917425,
+      "grad_norm": 0.3902441652129766,
+      "learning_rate": 3.224966043633966e-05,
+      "loss": 0.5199,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4244926522043387,
+      "grad_norm": 0.43154246579160466,
+      "learning_rate": 3.223881336850939e-05,
+      "loss": 0.5621,
+      "step": 3033
+    },
+    {
+      "epoch": 0.42463261021693494,
+      "grad_norm": 0.41498447902123786,
+      "learning_rate": 3.222796481302953e-05,
+      "loss": 0.5886,
+      "step": 3034
+    },
+    {
+      "epoch": 0.42477256822953113,
+      "grad_norm": 0.38695499906221675,
+      "learning_rate": 3.221711477212956e-05,
+      "loss": 0.5449,
+      "step": 3035
+    },
+    {
+      "epoch": 0.4249125262421274,
+      "grad_norm": 0.3974805369088236,
+      "learning_rate": 3.2206263248039276e-05,
+      "loss": 0.5602,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4250524842547236,
+      "grad_norm": 0.40115361129111843,
+      "learning_rate": 3.2195410242988776e-05,
+      "loss": 0.5675,
+      "step": 3037
+    },
+    {
+      "epoch": 0.4251924422673198,
+      "grad_norm": 0.3992286694324687,
+      "learning_rate": 3.2184555759208465e-05,
+      "loss": 0.5365,
+      "step": 3038
+    },
+    {
+      "epoch": 0.42533240027991603,
+      "grad_norm": 0.39187949837957153,
+      "learning_rate": 3.217369979892905e-05,
+      "loss": 0.5514,
+      "step": 3039
+    },
+    {
+      "epoch": 0.42547235829251223,
+      "grad_norm": 0.42850832602682615,
+      "learning_rate": 3.216284236438154e-05,
+      "loss": 0.5432,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4256123163051085,
+      "grad_norm": 0.43436314625968364,
+      "learning_rate": 3.215198345779723e-05,
+      "loss": 0.586,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4257522743177047,
+      "grad_norm": 0.4099577503194474,
+      "learning_rate": 3.214112308140777e-05,
+      "loss": 0.5915,
+      "step": 3042
+    },
+    {
+      "epoch": 0.42589223233030093,
+      "grad_norm": 0.42233874886106193,
+      "learning_rate": 3.213026123744506e-05,
+      "loss": 0.6066,
+      "step": 3043
+    },
+    {
+      "epoch": 0.42603219034289713,
+      "grad_norm": 0.3978279257361451,
+      "learning_rate": 3.211939792814131e-05,
+      "loss": 0.5882,
+      "step": 3044
+    },
+    {
+      "epoch": 0.42617214835549333,
+      "grad_norm": 0.40970969740070096,
+      "learning_rate": 3.210853315572906e-05,
+      "loss": 0.5647,
+      "step": 3045
+    },
+    {
+      "epoch": 0.4263121063680896,
+      "grad_norm": 0.3976277940502661,
+      "learning_rate": 3.20976669224411e-05,
+      "loss": 0.5873,
+      "step": 3046
+    },
+    {
+      "epoch": 0.4264520643806858,
+      "grad_norm": 0.400367465814101,
+      "learning_rate": 3.208679923051059e-05,
+      "loss": 0.5595,
+      "step": 3047
+    },
+    {
+      "epoch": 0.42659202239328203,
+      "grad_norm": 0.40911631062240816,
+      "learning_rate": 3.207593008217092e-05,
+      "loss": 0.566,
+      "step": 3048
+    },
+    {
+      "epoch": 0.42673198040587823,
+      "grad_norm": 0.4245921605186627,
+      "learning_rate": 3.206505947965583e-05,
+      "loss": 0.5643,
+      "step": 3049
+    },
+    {
+      "epoch": 0.4268719384184745,
+      "grad_norm": 0.42320217692212986,
+      "learning_rate": 3.205418742519933e-05,
+      "loss": 0.5512,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4270118964310707,
+      "grad_norm": 0.4050992683553409,
+      "learning_rate": 3.2043313921035743e-05,
+      "loss": 0.5712,
+      "step": 3051
+    },
+    {
+      "epoch": 0.4271518544436669,
+      "grad_norm": 0.39118318821429826,
+      "learning_rate": 3.203243896939968e-05,
+      "loss": 0.5839,
+      "step": 3052
+    },
+    {
+      "epoch": 0.42729181245626313,
+      "grad_norm": 0.4116673535552461,
+      "learning_rate": 3.202156257252606e-05,
+      "loss": 0.5599,
+      "step": 3053
+    },
+    {
+      "epoch": 0.42743177046885933,
+      "grad_norm": 0.42864138348780956,
+      "learning_rate": 3.201068473265007e-05,
+      "loss": 0.5877,
+      "step": 3054
+    },
+    {
+      "epoch": 0.4275717284814556,
+      "grad_norm": 0.4084556626969551,
+      "learning_rate": 3.1999805452007245e-05,
+      "loss": 0.578,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4277116864940518,
+      "grad_norm": 0.40349943769579744,
+      "learning_rate": 3.1988924732833384e-05,
+      "loss": 0.562,
+      "step": 3056
+    },
+    {
+      "epoch": 0.42785164450664803,
+      "grad_norm": 0.41607503622672326,
+      "learning_rate": 3.197804257736456e-05,
+      "loss": 0.5766,
+      "step": 3057
+    },
+    {
+      "epoch": 0.42799160251924423,
+      "grad_norm": 0.390556372929388,
+      "learning_rate": 3.19671589878372e-05,
+      "loss": 0.5343,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4281315605318404,
+      "grad_norm": 0.40189243879585396,
+      "learning_rate": 3.195627396648796e-05,
+      "loss": 0.5923,
+      "step": 3059
+    },
+    {
+      "epoch": 0.4282715185444367,
+      "grad_norm": 0.383550649614106,
+      "learning_rate": 3.1945387515553846e-05,
+      "loss": 0.5356,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4284114765570329,
+      "grad_norm": 0.39467142307894415,
+      "learning_rate": 3.193449963727213e-05,
+      "loss": 0.5402,
+      "step": 3061
+    },
+    {
+      "epoch": 0.42855143456962913,
+      "grad_norm": 0.3980289276258243,
+      "learning_rate": 3.192361033388037e-05,
+      "loss": 0.5769,
+      "step": 3062
+    },
+    {
+      "epoch": 0.42869139258222533,
+      "grad_norm": 0.40844869461585276,
+      "learning_rate": 3.191271960761645e-05,
+      "loss": 0.5635,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4288313505948216,
+      "grad_norm": 0.4137330574117752,
+      "learning_rate": 3.19018274607185e-05,
+      "loss": 0.5866,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4289713086074178,
+      "grad_norm": 0.4089150211795348,
+      "learning_rate": 3.1890933895424976e-05,
+      "loss": 0.5783,
+      "step": 3065
+    },
+    {
+      "epoch": 0.429111266620014,
+      "grad_norm": 0.40728615720674294,
+      "learning_rate": 3.188003891397463e-05,
+      "loss": 0.6048,
+      "step": 3066
+    },
+    {
+      "epoch": 0.42925122463261023,
+      "grad_norm": 0.3988606193766422,
+      "learning_rate": 3.186914251860648e-05,
+      "loss": 0.5234,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4293911826452064,
+      "grad_norm": 0.4081704250933805,
+      "learning_rate": 3.185824471155983e-05,
+      "loss": 0.5593,
+      "step": 3068
+    },
+    {
+      "epoch": 0.4295311406578027,
+      "grad_norm": 0.42949202285429056,
+      "learning_rate": 3.184734549507431e-05,
+      "loss": 0.5385,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4296710986703989,
+      "grad_norm": 0.39800337635965866,
+      "learning_rate": 3.183644487138982e-05,
+      "loss": 0.5213,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4298110566829951,
+      "grad_norm": 0.3968235944285963,
+      "learning_rate": 3.182554284274654e-05,
+      "loss": 0.5508,
+      "step": 3071
+    },
+    {
+      "epoch": 0.4299510146955913,
+      "grad_norm": 0.41991412353150687,
+      "learning_rate": 3.181463941138495e-05,
+      "loss": 0.5829,
+      "step": 3072
+    },
+    {
+      "epoch": 0.4300909727081875,
+      "grad_norm": 0.42186424178865006,
+      "learning_rate": 3.180373457954581e-05,
+      "loss": 0.5444,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4302309307207838,
+      "grad_norm": 0.409043031506044,
+      "learning_rate": 3.17928283494702e-05,
+      "loss": 0.5473,
+      "step": 3074
+    },
+    {
+      "epoch": 0.43037088873338,
+      "grad_norm": 0.4090988482060365,
+      "learning_rate": 3.178192072339942e-05,
+      "loss": 0.5293,
+      "step": 3075
+    },
+    {
+      "epoch": 0.43051084674597623,
+      "grad_norm": 0.40615205351515493,
+      "learning_rate": 3.177101170357513e-05,
+      "loss": 0.5869,
+      "step": 3076
+    },
+    {
+      "epoch": 0.4306508047585724,
+      "grad_norm": 0.39773350024022835,
+      "learning_rate": 3.176010129223923e-05,
+      "loss": 0.5456,
+      "step": 3077
+    },
+    {
+      "epoch": 0.4307907627711686,
+      "grad_norm": 0.4199551126566616,
+      "learning_rate": 3.174918949163392e-05,
+      "loss": 0.5642,
+      "step": 3078
+    },
+    {
+      "epoch": 0.4309307207837649,
+      "grad_norm": 0.43088576609462415,
+      "learning_rate": 3.17382763040017e-05,
+      "loss": 0.5563,
+      "step": 3079
+    },
+    {
+      "epoch": 0.4310706787963611,
+      "grad_norm": 0.40278017058920423,
+      "learning_rate": 3.172736173158532e-05,
+      "loss": 0.5279,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4312106368089573,
+      "grad_norm": 0.4423703616752564,
+      "learning_rate": 3.171644577662785e-05,
+      "loss": 0.5701,
+      "step": 3081
+    },
+    {
+      "epoch": 0.4313505948215535,
+      "grad_norm": 0.4240618723379521,
+      "learning_rate": 3.1705528441372626e-05,
+      "loss": 0.5876,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4314905528341498,
+      "grad_norm": 0.39848697262264293,
+      "learning_rate": 3.169460972806327e-05,
+      "loss": 0.5549,
+      "step": 3083
+    },
+    {
+      "epoch": 0.431630510846746,
+      "grad_norm": 0.40754000265622903,
+      "learning_rate": 3.1683689638943684e-05,
+      "loss": 0.5772,
+      "step": 3084
+    },
+    {
+      "epoch": 0.43177046885934217,
+      "grad_norm": 0.4018754339809231,
+      "learning_rate": 3.167276817625806e-05,
+      "loss": 0.5783,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4319104268719384,
+      "grad_norm": 0.430233233796668,
+      "learning_rate": 3.166184534225087e-05,
+      "loss": 0.5775,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4320503848845346,
+      "grad_norm": 0.4210970260038791,
+      "learning_rate": 3.165092113916688e-05,
+      "loss": 0.5593,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4321903428971309,
+      "grad_norm": 0.4162185793360081,
+      "learning_rate": 3.163999556925111e-05,
+      "loss": 0.5587,
+      "step": 3088
+    },
+    {
+      "epoch": 0.43233030090972707,
+      "grad_norm": 0.38464238839448883,
+      "learning_rate": 3.162906863474887e-05,
+      "loss": 0.5527,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4324702589223233,
+      "grad_norm": 0.3899859914956373,
+      "learning_rate": 3.161814033790577e-05,
+      "loss": 0.5734,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4326102169349195,
+      "grad_norm": 0.40900462690982553,
+      "learning_rate": 3.160721068096768e-05,
+      "loss": 0.5399,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4327501749475157,
+      "grad_norm": 0.4220096792046019,
+      "learning_rate": 3.159627966618075e-05,
+      "loss": 0.5727,
+      "step": 3092
+    },
+    {
+      "epoch": 0.432890132960112,
+      "grad_norm": 0.4233571391991352,
+      "learning_rate": 3.158534729579142e-05,
+      "loss": 0.5949,
+      "step": 3093
+    },
+    {
+      "epoch": 0.43303009097270817,
+      "grad_norm": 0.40740179328267645,
+      "learning_rate": 3.157441357204641e-05,
+      "loss": 0.5645,
+      "step": 3094
+    },
+    {
+      "epoch": 0.4331700489853044,
+      "grad_norm": 0.38846491608664135,
+      "learning_rate": 3.15634784971927e-05,
+      "loss": 0.5352,
+      "step": 3095
+    },
+    {
+      "epoch": 0.4333100069979006,
+      "grad_norm": 0.40243615629816265,
+      "learning_rate": 3.1552542073477555e-05,
+      "loss": 0.5672,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4334499650104969,
+      "grad_norm": 0.38997133690973523,
+      "learning_rate": 3.154160430314854e-05,
+      "loss": 0.5756,
+      "step": 3097
+    },
+    {
+      "epoch": 0.43358992302309307,
+      "grad_norm": 0.3815517517819305,
+      "learning_rate": 3.1530665188453464e-05,
+      "loss": 0.5596,
+      "step": 3098
+    },
+    {
+      "epoch": 0.43372988103568927,
+      "grad_norm": 0.39845289899145564,
+      "learning_rate": 3.1519724731640424e-05,
+      "loss": 0.5914,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4338698390482855,
+      "grad_norm": 0.39018815516325334,
+      "learning_rate": 3.15087829349578e-05,
+      "loss": 0.5351,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4340097970608817,
+      "grad_norm": 0.4129017336286664,
+      "learning_rate": 3.149783980065425e-05,
+      "loss": 0.562,
+      "step": 3101
+    },
+    {
+      "epoch": 0.43414975507347797,
+      "grad_norm": 0.42547377124636576,
+      "learning_rate": 3.1486895330978685e-05,
+      "loss": 0.5924,
+      "step": 3102
+    },
+    {
+      "epoch": 0.43428971308607417,
+      "grad_norm": 0.41352999220449604,
+      "learning_rate": 3.147594952818031e-05,
+      "loss": 0.5417,
+      "step": 3103
+    },
+    {
+      "epoch": 0.4344296710986704,
+      "grad_norm": 0.42347168990756195,
+      "learning_rate": 3.14650023945086e-05,
+      "loss": 0.5479,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4345696291112666,
+      "grad_norm": 0.389428898967484,
+      "learning_rate": 3.1454053932213304e-05,
+      "loss": 0.5552,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4347095871238628,
+      "grad_norm": 0.38385485279343906,
+      "learning_rate": 3.144310414354444e-05,
+      "loss": 0.5197,
+      "step": 3106
+    },
+    {
+      "epoch": 0.43484954513645907,
+      "grad_norm": 0.41030767770871096,
+      "learning_rate": 3.1432153030752295e-05,
+      "loss": 0.5597,
+      "step": 3107
+    },
+    {
+      "epoch": 0.43498950314905527,
+      "grad_norm": 0.41349967870711807,
+      "learning_rate": 3.142120059608744e-05,
+      "loss": 0.5663,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4351294611616515,
+      "grad_norm": 0.4082370728237636,
+      "learning_rate": 3.141024684180071e-05,
+      "loss": 0.5275,
+      "step": 3109
+    },
+    {
+      "epoch": 0.4352694191742477,
+      "grad_norm": 0.4122249486824059,
+      "learning_rate": 3.139929177014322e-05,
+      "loss": 0.5451,
+      "step": 3110
+    },
+    {
+      "epoch": 0.43540937718684397,
+      "grad_norm": 0.41854855746910635,
+      "learning_rate": 3.138833538336633e-05,
+      "loss": 0.5418,
+      "step": 3111
+    },
+    {
+      "epoch": 0.43554933519944017,
+      "grad_norm": 0.4259659771695767,
+      "learning_rate": 3.137737768372171e-05,
+      "loss": 0.5931,
+      "step": 3112
+    },
+    {
+      "epoch": 0.43568929321203637,
+      "grad_norm": 0.43545087247200576,
+      "learning_rate": 3.1366418673461254e-05,
+      "loss": 0.5899,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4358292512246326,
+      "grad_norm": 0.4096432845087902,
+      "learning_rate": 3.135545835483718e-05,
+      "loss": 0.5686,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4359692092372288,
+      "grad_norm": 0.40360816791162424,
+      "learning_rate": 3.134449673010192e-05,
+      "loss": 0.5471,
+      "step": 3115
+    },
+    {
+      "epoch": 0.43610916724982507,
+      "grad_norm": 0.4293932793165949,
+      "learning_rate": 3.1333533801508204e-05,
+      "loss": 0.5721,
+      "step": 3116
+    },
+    {
+      "epoch": 0.43624912526242127,
+      "grad_norm": 0.3963350874406529,
+      "learning_rate": 3.132256957130904e-05,
+      "loss": 0.5685,
+      "step": 3117
+    },
+    {
+      "epoch": 0.4363890832750175,
+      "grad_norm": 0.3871923444763041,
+      "learning_rate": 3.131160404175767e-05,
+      "loss": 0.5544,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4365290412876137,
+      "grad_norm": 0.41730033342791306,
+      "learning_rate": 3.130063721510763e-05,
+      "loss": 0.5594,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4366689993002099,
+      "grad_norm": 0.43319861497508905,
+      "learning_rate": 3.1289669093612714e-05,
+      "loss": 0.5896,
+      "step": 3120
+    },
+    {
+      "epoch": 0.43680895731280617,
+      "grad_norm": 0.3896513313595881,
+      "learning_rate": 3.127869967952698e-05,
+      "loss": 0.5799,
+      "step": 3121
+    },
+    {
+      "epoch": 0.43694891532540237,
+      "grad_norm": 0.4391301390474321,
+      "learning_rate": 3.126772897510476e-05,
+      "loss": 0.582,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4370888733379986,
+      "grad_norm": 0.3951829444712206,
+      "learning_rate": 3.125675698260065e-05,
+      "loss": 0.5577,
+      "step": 3123
+    },
+    {
+      "epoch": 0.4372288313505948,
+      "grad_norm": 0.4088921127679525,
+      "learning_rate": 3.124578370426947e-05,
+      "loss": 0.5471,
+      "step": 3124
+    },
+    {
+      "epoch": 0.43736878936319107,
+      "grad_norm": 0.40507594886972204,
+      "learning_rate": 3.1234809142366376e-05,
+      "loss": 0.5787,
+      "step": 3125
+    },
+    {
+      "epoch": 0.43750874737578727,
+      "grad_norm": 0.48880201961534253,
+      "learning_rate": 3.1223833299146756e-05,
+      "loss": 0.5665,
+      "step": 3126
+    },
+    {
+      "epoch": 0.43764870538838346,
+      "grad_norm": 0.40104306636405085,
+      "learning_rate": 3.1212856176866226e-05,
+      "loss": 0.5403,
+      "step": 3127
+    },
+    {
+      "epoch": 0.4377886634009797,
+      "grad_norm": 0.38729222817598974,
+      "learning_rate": 3.120187777778073e-05,
+      "loss": 0.5298,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4379286214135759,
+      "grad_norm": 0.38974703851837783,
+      "learning_rate": 3.11908981041464e-05,
+      "loss": 0.5612,
+      "step": 3129
+    },
+    {
+      "epoch": 0.43806857942617217,
+      "grad_norm": 0.4100334816926811,
+      "learning_rate": 3.117991715821972e-05,
+      "loss": 0.5903,
+      "step": 3130
+    },
+    {
+      "epoch": 0.43820853743876836,
+      "grad_norm": 0.4156843103680473,
+      "learning_rate": 3.116893494225734e-05,
+      "loss": 0.5886,
+      "step": 3131
+    },
+    {
+      "epoch": 0.4383484954513646,
+      "grad_norm": 0.39607057982434785,
+      "learning_rate": 3.115795145851625e-05,
+      "loss": 0.5621,
+      "step": 3132
+    },
+    {
+      "epoch": 0.4384884534639608,
+      "grad_norm": 0.4036233265018997,
+      "learning_rate": 3.114696670925365e-05,
+      "loss": 0.5538,
+      "step": 3133
+    },
+    {
+      "epoch": 0.438628411476557,
+      "grad_norm": 0.4101277411308792,
+      "learning_rate": 3.113598069672702e-05,
+      "loss": 0.5961,
+      "step": 3134
+    },
+    {
+      "epoch": 0.43876836948915326,
+      "grad_norm": 0.399711153788506,
+      "learning_rate": 3.11249934231941e-05,
+      "loss": 0.5594,
+      "step": 3135
+    },
+    {
+      "epoch": 0.43890832750174946,
+      "grad_norm": 0.3867721840790253,
+      "learning_rate": 3.111400489091288e-05,
+      "loss": 0.5122,
+      "step": 3136
+    },
+    {
+      "epoch": 0.4390482855143457,
+      "grad_norm": 0.4029896269340335,
+      "learning_rate": 3.110301510214162e-05,
+      "loss": 0.5724,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4391882435269419,
+      "grad_norm": 0.4323192969018167,
+      "learning_rate": 3.1092024059138836e-05,
+      "loss": 0.5901,
+      "step": 3138
+    },
+    {
+      "epoch": 0.43932820153953817,
+      "grad_norm": 0.4030250752398442,
+      "learning_rate": 3.108103176416329e-05,
+      "loss": 0.5369,
+      "step": 3139
+    },
+    {
+      "epoch": 0.43946815955213436,
+      "grad_norm": 0.40699746634257905,
+      "learning_rate": 3.1070038219474026e-05,
+      "loss": 0.5236,
+      "step": 3140
+    },
+    {
+      "epoch": 0.43960811756473056,
+      "grad_norm": 0.4054193975355776,
+      "learning_rate": 3.105904342733032e-05,
+      "loss": 0.555,
+      "step": 3141
+    },
+    {
+      "epoch": 0.4397480755773268,
+      "grad_norm": 0.43325267645820853,
+      "learning_rate": 3.104804738999169e-05,
+      "loss": 0.5704,
+      "step": 3142
+    },
+    {
+      "epoch": 0.439888033589923,
+      "grad_norm": 0.41581039947645454,
+      "learning_rate": 3.103705010971797e-05,
+      "loss": 0.5733,
+      "step": 3143
+    },
+    {
+      "epoch": 0.44002799160251926,
+      "grad_norm": 0.397475655616127,
+      "learning_rate": 3.1026051588769204e-05,
+      "loss": 0.5681,
+      "step": 3144
+    },
+    {
+      "epoch": 0.44016794961511546,
+      "grad_norm": 0.4037137237750616,
+      "learning_rate": 3.101505182940568e-05,
+      "loss": 0.5641,
+      "step": 3145
+    },
+    {
+      "epoch": 0.44030790762771166,
+      "grad_norm": 0.395909015440236,
+      "learning_rate": 3.1004050833887985e-05,
+      "loss": 0.5453,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4404478656403079,
+      "grad_norm": 0.40746541718886475,
+      "learning_rate": 3.099304860447692e-05,
+      "loss": 0.5576,
+      "step": 3147
+    },
+    {
+      "epoch": 0.4405878236529041,
+      "grad_norm": 0.40891529518862035,
+      "learning_rate": 3.098204514343356e-05,
+      "loss": 0.5744,
+      "step": 3148
+    },
+    {
+      "epoch": 0.44072778166550036,
+      "grad_norm": 0.42039074958765454,
+      "learning_rate": 3.097104045301922e-05,
+      "loss": 0.5945,
+      "step": 3149
+    },
+    {
+      "epoch": 0.44086773967809656,
+      "grad_norm": 0.4233543661398848,
+      "learning_rate": 3.096003453549549e-05,
+      "loss": 0.5174,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4410076976906928,
+      "grad_norm": 0.41103834231055475,
+      "learning_rate": 3.0949027393124185e-05,
+      "loss": 0.5859,
+      "step": 3151
+    },
+    {
+      "epoch": 0.441147655703289,
+      "grad_norm": 0.4071396114989415,
+      "learning_rate": 3.093801902816739e-05,
+      "loss": 0.5513,
+      "step": 3152
+    },
+    {
+      "epoch": 0.4412876137158852,
+      "grad_norm": 0.5084306343751018,
+      "learning_rate": 3.092700944288744e-05,
+      "loss": 0.5871,
+      "step": 3153
+    },
+    {
+      "epoch": 0.44142757172848146,
+      "grad_norm": 0.41487834345004226,
+      "learning_rate": 3.091599863954691e-05,
+      "loss": 0.5789,
+      "step": 3154
+    },
+    {
+      "epoch": 0.44156752974107766,
+      "grad_norm": 0.38353879873996594,
+      "learning_rate": 3.090498662040863e-05,
+      "loss": 0.5864,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4417074877536739,
+      "grad_norm": 0.4037917238728566,
+      "learning_rate": 3.0893973387735687e-05,
+      "loss": 0.5649,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4418474457662701,
+      "grad_norm": 0.4017882825846787,
+      "learning_rate": 3.0882958943791405e-05,
+      "loss": 0.5314,
+      "step": 3157
+    },
+    {
+      "epoch": 0.44198740377886636,
+      "grad_norm": 0.413358406501944,
+      "learning_rate": 3.087194329083937e-05,
+      "loss": 0.5433,
+      "step": 3158
+    },
+    {
+      "epoch": 0.44212736179146256,
+      "grad_norm": 0.38344575892281724,
+      "learning_rate": 3.0860926431143415e-05,
+      "loss": 0.5612,
+      "step": 3159
+    },
+    {
+      "epoch": 0.44226731980405876,
+      "grad_norm": 0.4087942341185106,
+      "learning_rate": 3.0849908366967605e-05,
+      "loss": 0.5256,
+      "step": 3160
+    },
+    {
+      "epoch": 0.442407277816655,
+      "grad_norm": 0.4055705356412522,
+      "learning_rate": 3.083888910057627e-05,
+      "loss": 0.5661,
+      "step": 3161
+    },
+    {
+      "epoch": 0.4425472358292512,
+      "grad_norm": 0.3826502993612589,
+      "learning_rate": 3.082786863423399e-05,
+      "loss": 0.5183,
+      "step": 3162
+    },
+    {
+      "epoch": 0.44268719384184746,
+      "grad_norm": 0.4043315813890686,
+      "learning_rate": 3.0816846970205556e-05,
+      "loss": 0.5805,
+      "step": 3163
+    },
+    {
+      "epoch": 0.44282715185444366,
+      "grad_norm": 0.6100429100368088,
+      "learning_rate": 3.0805824110756064e-05,
+      "loss": 0.5527,
+      "step": 3164
+    },
+    {
+      "epoch": 0.4429671098670399,
+      "grad_norm": 0.39290972200856517,
+      "learning_rate": 3.079480005815081e-05,
+      "loss": 0.5529,
+      "step": 3165
+    },
+    {
+      "epoch": 0.4431070678796361,
+      "grad_norm": 0.38604414946838106,
+      "learning_rate": 3.078377481465534e-05,
+      "loss": 0.5216,
+      "step": 3166
+    },
+    {
+      "epoch": 0.4432470258922323,
+      "grad_norm": 0.4231758177347116,
+      "learning_rate": 3.0772748382535463e-05,
+      "loss": 0.5315,
+      "step": 3167
+    },
+    {
+      "epoch": 0.44338698390482856,
+      "grad_norm": 0.3991789110660417,
+      "learning_rate": 3.076172076405722e-05,
+      "loss": 0.5488,
+      "step": 3168
+    },
+    {
+      "epoch": 0.44352694191742476,
+      "grad_norm": 0.5750580981426986,
+      "learning_rate": 3.07506919614869e-05,
+      "loss": 0.5191,
+      "step": 3169
+    },
+    {
+      "epoch": 0.443666899930021,
+      "grad_norm": 0.41221569096533744,
+      "learning_rate": 3.073966197709103e-05,
+      "loss": 0.5505,
+      "step": 3170
+    },
+    {
+      "epoch": 0.4438068579426172,
+      "grad_norm": 0.4397217342239035,
+      "learning_rate": 3.072863081313639e-05,
+      "loss": 0.594,
+      "step": 3171
+    },
+    {
+      "epoch": 0.44394681595521346,
+      "grad_norm": 0.4109178364737107,
+      "learning_rate": 3.071759847188998e-05,
+      "loss": 0.5438,
+      "step": 3172
+    },
+    {
+      "epoch": 0.44408677396780966,
+      "grad_norm": 0.4037573317008806,
+      "learning_rate": 3.070656495561909e-05,
+      "loss": 0.56,
+      "step": 3173
+    },
+    {
+      "epoch": 0.44422673198040585,
+      "grad_norm": 0.4029097868342135,
+      "learning_rate": 3.069553026659119e-05,
+      "loss": 0.5552,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4443666899930021,
+      "grad_norm": 0.4100219354350298,
+      "learning_rate": 3.068449440707404e-05,
+      "loss": 0.5533,
+      "step": 3175
+    },
+    {
+      "epoch": 0.4445066480055983,
+      "grad_norm": 0.3892720780012183,
+      "learning_rate": 3.067345737933561e-05,
+      "loss": 0.5478,
+      "step": 3176
+    },
+    {
+      "epoch": 0.44464660601819456,
+      "grad_norm": 0.389146645661204,
+      "learning_rate": 3.0662419185644115e-05,
+      "loss": 0.5506,
+      "step": 3177
+    },
+    {
+      "epoch": 0.44478656403079075,
+      "grad_norm": 0.3873299876663967,
+      "learning_rate": 3.065137982826802e-05,
+      "loss": 0.5245,
+      "step": 3178
+    },
+    {
+      "epoch": 0.444926522043387,
+      "grad_norm": 0.38222672454752543,
+      "learning_rate": 3.064033930947604e-05,
+      "loss": 0.5451,
+      "step": 3179
+    },
+    {
+      "epoch": 0.4450664800559832,
+      "grad_norm": 0.39505544780509594,
+      "learning_rate": 3.06292976315371e-05,
+      "loss": 0.5617,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4452064380685794,
+      "grad_norm": 0.410933368734618,
+      "learning_rate": 3.0618254796720375e-05,
+      "loss": 0.5598,
+      "step": 3181
+    },
+    {
+      "epoch": 0.44534639608117566,
+      "grad_norm": 0.4357975405512465,
+      "learning_rate": 3.060721080729529e-05,
+      "loss": 0.5558,
+      "step": 3182
+    },
+    {
+      "epoch": 0.44548635409377185,
+      "grad_norm": 0.43046775321062414,
+      "learning_rate": 3.059616566553149e-05,
+      "loss": 0.5637,
+      "step": 3183
+    },
+    {
+      "epoch": 0.4456263121063681,
+      "grad_norm": 0.3985416977600822,
+      "learning_rate": 3.058511937369886e-05,
+      "loss": 0.525,
+      "step": 3184
+    },
+    {
+      "epoch": 0.4457662701189643,
+      "grad_norm": 0.40340048691929675,
+      "learning_rate": 3.057407193406753e-05,
+      "loss": 0.5249,
+      "step": 3185
+    },
+    {
+      "epoch": 0.44590622813156056,
+      "grad_norm": 0.4135936572888444,
+      "learning_rate": 3.056302334890786e-05,
+      "loss": 0.5549,
+      "step": 3186
+    },
+    {
+      "epoch": 0.44604618614415675,
+      "grad_norm": 0.3920888452917507,
+      "learning_rate": 3.0551973620490456e-05,
+      "loss": 0.5491,
+      "step": 3187
+    },
+    {
+      "epoch": 0.44618614415675295,
+      "grad_norm": 0.4035469272259035,
+      "learning_rate": 3.0540922751086135e-05,
+      "loss": 0.5465,
+      "step": 3188
+    },
+    {
+      "epoch": 0.4463261021693492,
+      "grad_norm": 0.4216381953710183,
+      "learning_rate": 3.052987074296596e-05,
+      "loss": 0.5535,
+      "step": 3189
+    },
+    {
+      "epoch": 0.4464660601819454,
+      "grad_norm": 0.41233431611024285,
+      "learning_rate": 3.051881759840124e-05,
+      "loss": 0.5498,
+      "step": 3190
+    },
+    {
+      "epoch": 0.44660601819454165,
+      "grad_norm": 0.4114941998250367,
+      "learning_rate": 3.0507763319663517e-05,
+      "loss": 0.5718,
+      "step": 3191
+    },
+    {
+      "epoch": 0.44674597620713785,
+      "grad_norm": 0.407680668986072,
+      "learning_rate": 3.0496707909024542e-05,
+      "loss": 0.5298,
+      "step": 3192
+    },
+    {
+      "epoch": 0.4468859342197341,
+      "grad_norm": 0.3977959198257157,
+      "learning_rate": 3.0485651368756323e-05,
+      "loss": 0.549,
+      "step": 3193
+    },
+    {
+      "epoch": 0.4470258922323303,
+      "grad_norm": 0.3998186038348927,
+      "learning_rate": 3.0474593701131084e-05,
+      "loss": 0.5743,
+      "step": 3194
+    },
+    {
+      "epoch": 0.4471658502449265,
+      "grad_norm": 0.4154710015319719,
+      "learning_rate": 3.0463534908421298e-05,
+      "loss": 0.5515,
+      "step": 3195
+    },
+    {
+      "epoch": 0.44730580825752275,
+      "grad_norm": 0.3943039960321348,
+      "learning_rate": 3.0452474992899643e-05,
+      "loss": 0.5391,
+      "step": 3196
+    },
+    {
+      "epoch": 0.44744576627011895,
+      "grad_norm": 0.4199054832560681,
+      "learning_rate": 3.044141395683906e-05,
+      "loss": 0.5741,
+      "step": 3197
+    },
+    {
+      "epoch": 0.4475857242827152,
+      "grad_norm": 0.41750106544298576,
+      "learning_rate": 3.0430351802512698e-05,
+      "loss": 0.6249,
+      "step": 3198
+    },
+    {
+      "epoch": 0.4477256822953114,
+      "grad_norm": 0.3946839624490567,
+      "learning_rate": 3.041928853219394e-05,
+      "loss": 0.5692,
+      "step": 3199
+    },
+    {
+      "epoch": 0.44786564030790765,
+      "grad_norm": 0.4224068149128873,
+      "learning_rate": 3.0408224148156407e-05,
+      "loss": 0.5772,
+      "step": 3200
+    },
+    {
+      "epoch": 0.44800559832050385,
+      "grad_norm": 0.44042046806838825,
+      "learning_rate": 3.039715865267393e-05,
+      "loss": 0.6051,
+      "step": 3201
+    },
+    {
+      "epoch": 0.44814555633310005,
+      "grad_norm": 0.7011787045122906,
+      "learning_rate": 3.0386092048020593e-05,
+      "loss": 0.6074,
+      "step": 3202
+    },
+    {
+      "epoch": 0.4482855143456963,
+      "grad_norm": 0.41427835190534584,
+      "learning_rate": 3.037502433647068e-05,
+      "loss": 0.5624,
+      "step": 3203
+    },
+    {
+      "epoch": 0.4484254723582925,
+      "grad_norm": 0.4053190934113154,
+      "learning_rate": 3.0363955520298742e-05,
+      "loss": 0.5951,
+      "step": 3204
+    },
+    {
+      "epoch": 0.44856543037088875,
+      "grad_norm": 0.43048899640751676,
+      "learning_rate": 3.0352885601779512e-05,
+      "loss": 0.5652,
+      "step": 3205
+    },
+    {
+      "epoch": 0.44870538838348495,
+      "grad_norm": 0.3964539941995037,
+      "learning_rate": 3.0341814583187978e-05,
+      "loss": 0.5595,
+      "step": 3206
+    },
+    {
+      "epoch": 0.4488453463960812,
+      "grad_norm": 0.4009111440660274,
+      "learning_rate": 3.033074246679935e-05,
+      "loss": 0.5861,
+      "step": 3207
+    },
+    {
+      "epoch": 0.4489853044086774,
+      "grad_norm": 0.38864197775544784,
+      "learning_rate": 3.0319669254889055e-05,
+      "loss": 0.598,
+      "step": 3208
+    },
+    {
+      "epoch": 0.4491252624212736,
+      "grad_norm": 0.4292785996534306,
+      "learning_rate": 3.0308594949732755e-05,
+      "loss": 0.5901,
+      "step": 3209
+    },
+    {
+      "epoch": 0.44926522043386985,
+      "grad_norm": 0.37775134402140526,
+      "learning_rate": 3.029751955360633e-05,
+      "loss": 0.5327,
+      "step": 3210
+    },
+    {
+      "epoch": 0.44940517844646605,
+      "grad_norm": 0.37253895115039465,
+      "learning_rate": 3.0286443068785885e-05,
+      "loss": 0.5649,
+      "step": 3211
+    },
+    {
+      "epoch": 0.4495451364590623,
+      "grad_norm": 0.43373031007480495,
+      "learning_rate": 3.027536549754775e-05,
+      "loss": 0.541,
+      "step": 3212
+    },
+    {
+      "epoch": 0.4496850944716585,
+      "grad_norm": 0.42039610931604904,
+      "learning_rate": 3.0264286842168477e-05,
+      "loss": 0.5549,
+      "step": 3213
+    },
+    {
+      "epoch": 0.44982505248425475,
+      "grad_norm": 0.40625318217496825,
+      "learning_rate": 3.025320710492484e-05,
+      "loss": 0.568,
+      "step": 3214
+    },
+    {
+      "epoch": 0.44996501049685095,
+      "grad_norm": 0.4025993136298079,
+      "learning_rate": 3.0242126288093846e-05,
+      "loss": 0.5824,
+      "step": 3215
+    },
+    {
+      "epoch": 0.45010496850944715,
+      "grad_norm": 0.3984904672668101,
+      "learning_rate": 3.0231044393952712e-05,
+      "loss": 0.5615,
+      "step": 3216
+    },
+    {
+      "epoch": 0.4502449265220434,
+      "grad_norm": 0.4137524109943405,
+      "learning_rate": 3.0219961424778877e-05,
+      "loss": 0.5588,
+      "step": 3217
+    },
+    {
+      "epoch": 0.4503848845346396,
+      "grad_norm": 0.39240078172637105,
+      "learning_rate": 3.020887738285001e-05,
+      "loss": 0.521,
+      "step": 3218
+    },
+    {
+      "epoch": 0.45052484254723585,
+      "grad_norm": 0.41138657372667015,
+      "learning_rate": 3.0197792270443982e-05,
+      "loss": 0.5471,
+      "step": 3219
+    },
+    {
+      "epoch": 0.45066480055983205,
+      "grad_norm": 0.4192515121968027,
+      "learning_rate": 3.0186706089838913e-05,
+      "loss": 0.527,
+      "step": 3220
+    },
+    {
+      "epoch": 0.4508047585724283,
+      "grad_norm": 0.4089670890583155,
+      "learning_rate": 3.017561884331311e-05,
+      "loss": 0.5726,
+      "step": 3221
+    },
+    {
+      "epoch": 0.4509447165850245,
+      "grad_norm": 0.40191329030749956,
+      "learning_rate": 3.0164530533145123e-05,
+      "loss": 0.5444,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4510846745976207,
+      "grad_norm": 0.41118884176015386,
+      "learning_rate": 3.0153441161613704e-05,
+      "loss": 0.5541,
+      "step": 3223
+    },
+    {
+      "epoch": 0.45122463261021695,
+      "grad_norm": 0.4309170413845665,
+      "learning_rate": 3.0142350730997837e-05,
+      "loss": 0.5794,
+      "step": 3224
+    },
+    {
+      "epoch": 0.45136459062281314,
+      "grad_norm": 0.41219057005882026,
+      "learning_rate": 3.0131259243576726e-05,
+      "loss": 0.5601,
+      "step": 3225
+    },
+    {
+      "epoch": 0.4515045486354094,
+      "grad_norm": 0.3941521672183907,
+      "learning_rate": 3.012016670162977e-05,
+      "loss": 0.521,
+      "step": 3226
+    },
+    {
+      "epoch": 0.4516445066480056,
+      "grad_norm": 0.4198018097191238,
+      "learning_rate": 3.0109073107436608e-05,
+      "loss": 0.5573,
+      "step": 3227
+    },
+    {
+      "epoch": 0.4517844646606018,
+      "grad_norm": 0.4135610840050145,
+      "learning_rate": 3.0097978463277076e-05,
+      "loss": 0.5718,
+      "step": 3228
+    },
+    {
+      "epoch": 0.45192442267319805,
+      "grad_norm": 0.405778139396707,
+      "learning_rate": 3.0086882771431258e-05,
+      "loss": 0.5468,
+      "step": 3229
+    },
+    {
+      "epoch": 0.45206438068579424,
+      "grad_norm": 0.4401596995437583,
+      "learning_rate": 3.0075786034179405e-05,
+      "loss": 0.5915,
+      "step": 3230
+    },
+    {
+      "epoch": 0.4522043386983905,
+      "grad_norm": 0.4270515847507665,
+      "learning_rate": 3.0064688253802026e-05,
+      "loss": 0.5704,
+      "step": 3231
+    },
+    {
+      "epoch": 0.4523442967109867,
+      "grad_norm": 0.3969545038717031,
+      "learning_rate": 3.0053589432579827e-05,
+      "loss": 0.539,
+      "step": 3232
+    },
+    {
+      "epoch": 0.45248425472358295,
+      "grad_norm": 0.4231180537782753,
+      "learning_rate": 3.004248957279372e-05,
+      "loss": 0.5993,
+      "step": 3233
+    },
+    {
+      "epoch": 0.45262421273617914,
+      "grad_norm": 0.3821272498454476,
+      "learning_rate": 3.0031388676724836e-05,
+      "loss": 0.5444,
+      "step": 3234
+    },
+    {
+      "epoch": 0.45276417074877534,
+      "grad_norm": 0.38147859259770744,
+      "learning_rate": 3.0020286746654525e-05,
+      "loss": 0.5893,
+      "step": 3235
+    },
+    {
+      "epoch": 0.4529041287613716,
+      "grad_norm": 0.40206232247150925,
+      "learning_rate": 3.0009183784864365e-05,
+      "loss": 0.5567,
+      "step": 3236
+    },
+    {
+      "epoch": 0.4530440867739678,
+      "grad_norm": 0.3712974814446054,
+      "learning_rate": 2.9998079793636098e-05,
+      "loss": 0.5148,
+      "step": 3237
+    },
+    {
+      "epoch": 0.45318404478656404,
+      "grad_norm": 0.4047911445642038,
+      "learning_rate": 2.998697477525173e-05,
+      "loss": 0.5543,
+      "step": 3238
+    },
+    {
+      "epoch": 0.45332400279916024,
+      "grad_norm": 0.41083909886481995,
+      "learning_rate": 2.997586873199344e-05,
+      "loss": 0.5411,
+      "step": 3239
+    },
+    {
+      "epoch": 0.4534639608117565,
+      "grad_norm": 0.3979787826364129,
+      "learning_rate": 2.996476166614364e-05,
+      "loss": 0.526,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4536039188243527,
+      "grad_norm": 0.39694745383011404,
+      "learning_rate": 2.9953653579984942e-05,
+      "loss": 0.5523,
+      "step": 3241
+    },
+    {
+      "epoch": 0.4537438768369489,
+      "grad_norm": 0.3902000936154921,
+      "learning_rate": 2.9942544475800172e-05,
+      "loss": 0.5502,
+      "step": 3242
+    },
+    {
+      "epoch": 0.45388383484954514,
+      "grad_norm": 0.42203355743945187,
+      "learning_rate": 2.9931434355872367e-05,
+      "loss": 0.5609,
+      "step": 3243
+    },
+    {
+      "epoch": 0.45402379286214134,
+      "grad_norm": 0.4454869866225551,
+      "learning_rate": 2.992032322248476e-05,
+      "loss": 0.6,
+      "step": 3244
+    },
+    {
+      "epoch": 0.4541637508747376,
+      "grad_norm": 0.405554964175632,
+      "learning_rate": 2.9909211077920805e-05,
+      "loss": 0.572,
+      "step": 3245
+    },
+    {
+      "epoch": 0.4543037088873338,
+      "grad_norm": 0.41415547431542904,
+      "learning_rate": 2.989809792446417e-05,
+      "loss": 0.5578,
+      "step": 3246
+    },
+    {
+      "epoch": 0.45444366689993004,
+      "grad_norm": 0.41647413135178146,
+      "learning_rate": 2.9886983764398707e-05,
+      "loss": 0.5691,
+      "step": 3247
+    },
+    {
+      "epoch": 0.45458362491252624,
+      "grad_norm": 0.3958733230625823,
+      "learning_rate": 2.9875868600008496e-05,
+      "loss": 0.5708,
+      "step": 3248
+    },
+    {
+      "epoch": 0.45472358292512244,
+      "grad_norm": 0.39688745966500016,
+      "learning_rate": 2.986475243357782e-05,
+      "loss": 0.5497,
+      "step": 3249
+    },
+    {
+      "epoch": 0.4548635409377187,
+      "grad_norm": 0.3859490510740407,
+      "learning_rate": 2.985363526739115e-05,
+      "loss": 0.5227,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4550034989503149,
+      "grad_norm": 0.3934619917655379,
+      "learning_rate": 2.9842517103733192e-05,
+      "loss": 0.5662,
+      "step": 3251
+    },
+    {
+      "epoch": 0.45514345696291114,
+      "grad_norm": 0.5131902290403141,
+      "learning_rate": 2.9831397944888833e-05,
+      "loss": 0.5677,
+      "step": 3252
+    },
+    {
+      "epoch": 0.45528341497550734,
+      "grad_norm": 0.42208973839755937,
+      "learning_rate": 2.9820277793143177e-05,
+      "loss": 0.5347,
+      "step": 3253
+    },
+    {
+      "epoch": 0.4554233729881036,
+      "grad_norm": 0.4047958403557277,
+      "learning_rate": 2.9809156650781528e-05,
+      "loss": 0.5533,
+      "step": 3254
+    },
+    {
+      "epoch": 0.4555633310006998,
+      "grad_norm": 0.41534042335678106,
+      "learning_rate": 2.9798034520089385e-05,
+      "loss": 0.6129,
+      "step": 3255
+    },
+    {
+      "epoch": 0.455703289013296,
+      "grad_norm": 0.405760676923056,
+      "learning_rate": 2.978691140335247e-05,
+      "loss": 0.575,
+      "step": 3256
+    },
+    {
+      "epoch": 0.45584324702589224,
+      "grad_norm": 0.4049361175064025,
+      "learning_rate": 2.9775787302856683e-05,
+      "loss": 0.5287,
+      "step": 3257
+    },
+    {
+      "epoch": 0.45598320503848844,
+      "grad_norm": 0.4212785218180886,
+      "learning_rate": 2.9764662220888157e-05,
+      "loss": 0.5649,
+      "step": 3258
+    },
+    {
+      "epoch": 0.4561231630510847,
+      "grad_norm": 0.4014033439696915,
+      "learning_rate": 2.9753536159733196e-05,
+      "loss": 0.5457,
+      "step": 3259
+    },
+    {
+      "epoch": 0.4562631210636809,
+      "grad_norm": 0.38748328739856236,
+      "learning_rate": 2.974240912167833e-05,
+      "loss": 0.5409,
+      "step": 3260
+    },
+    {
+      "epoch": 0.45640307907627714,
+      "grad_norm": 0.40938166793163544,
+      "learning_rate": 2.9731281109010256e-05,
+      "loss": 0.6002,
+      "step": 3261
+    },
+    {
+      "epoch": 0.45654303708887334,
+      "grad_norm": 0.40068504604922145,
+      "learning_rate": 2.9720152124015916e-05,
+      "loss": 0.5767,
+      "step": 3262
+    },
+    {
+      "epoch": 0.45668299510146954,
+      "grad_norm": 0.39286585333437835,
+      "learning_rate": 2.9709022168982426e-05,
+      "loss": 0.5823,
+      "step": 3263
+    },
+    {
+      "epoch": 0.4568229531140658,
+      "grad_norm": 0.39448349470052235,
+      "learning_rate": 2.9697891246197097e-05,
+      "loss": 0.5702,
+      "step": 3264
+    },
+    {
+      "epoch": 0.456962911126662,
+      "grad_norm": 0.42230524457915075,
+      "learning_rate": 2.9686759357947446e-05,
+      "loss": 0.5695,
+      "step": 3265
+    },
+    {
+      "epoch": 0.45710286913925824,
+      "grad_norm": 0.3829307935804848,
+      "learning_rate": 2.9675626506521205e-05,
+      "loss": 0.5486,
+      "step": 3266
+    },
+    {
+      "epoch": 0.45724282715185444,
+      "grad_norm": 0.4002606373918937,
+      "learning_rate": 2.966449269420627e-05,
+      "loss": 0.5953,
+      "step": 3267
+    },
+    {
+      "epoch": 0.4573827851644507,
+      "grad_norm": 0.4081380916674823,
+      "learning_rate": 2.9653357923290753e-05,
+      "loss": 0.5605,
+      "step": 3268
+    },
+    {
+      "epoch": 0.4575227431770469,
+      "grad_norm": 0.4173988076758946,
+      "learning_rate": 2.9642222196062973e-05,
+      "loss": 0.5941,
+      "step": 3269
+    },
+    {
+      "epoch": 0.4576627011896431,
+      "grad_norm": 0.41804559043749046,
+      "learning_rate": 2.9631085514811423e-05,
+      "loss": 0.5979,
+      "step": 3270
+    },
+    {
+      "epoch": 0.45780265920223934,
+      "grad_norm": 0.3906986098038151,
+      "learning_rate": 2.9619947881824818e-05,
+      "loss": 0.5567,
+      "step": 3271
+    },
+    {
+      "epoch": 0.45794261721483553,
+      "grad_norm": 0.4110120727152376,
+      "learning_rate": 2.9608809299392043e-05,
+      "loss": 0.5732,
+      "step": 3272
+    },
+    {
+      "epoch": 0.4580825752274318,
+      "grad_norm": 0.40920966877207876,
+      "learning_rate": 2.9597669769802196e-05,
+      "loss": 0.5673,
+      "step": 3273
+    },
+    {
+      "epoch": 0.458222533240028,
+      "grad_norm": 0.3959304691526394,
+      "learning_rate": 2.958652929534456e-05,
+      "loss": 0.5719,
+      "step": 3274
+    },
+    {
+      "epoch": 0.45836249125262424,
+      "grad_norm": 0.3844580981772225,
+      "learning_rate": 2.9575387878308615e-05,
+      "loss": 0.5441,
+      "step": 3275
+    },
+    {
+      "epoch": 0.45850244926522044,
+      "grad_norm": 0.4425235109241173,
+      "learning_rate": 2.9564245520984047e-05,
+      "loss": 0.5896,
+      "step": 3276
+    },
+    {
+      "epoch": 0.45864240727781663,
+      "grad_norm": 0.4206583711187626,
+      "learning_rate": 2.9553102225660706e-05,
+      "loss": 0.5718,
+      "step": 3277
+    },
+    {
+      "epoch": 0.4587823652904129,
+      "grad_norm": 0.43363119264281536,
+      "learning_rate": 2.954195799462866e-05,
+      "loss": 0.5584,
+      "step": 3278
+    },
+    {
+      "epoch": 0.4589223233030091,
+      "grad_norm": 0.40918971785473857,
+      "learning_rate": 2.9530812830178162e-05,
+      "loss": 0.5386,
+      "step": 3279
+    },
+    {
+      "epoch": 0.45906228131560534,
+      "grad_norm": 0.4195378947004359,
+      "learning_rate": 2.951966673459965e-05,
+      "loss": 0.5862,
+      "step": 3280
+    },
+    {
+      "epoch": 0.45920223932820153,
+      "grad_norm": 0.3984648051790019,
+      "learning_rate": 2.9508519710183772e-05,
+      "loss": 0.5745,
+      "step": 3281
+    },
+    {
+      "epoch": 0.4593421973407978,
+      "grad_norm": 0.42412243772475006,
+      "learning_rate": 2.9497371759221347e-05,
+      "loss": 0.5765,
+      "step": 3282
+    },
+    {
+      "epoch": 0.459482155353394,
+      "grad_norm": 0.4155137965851478,
+      "learning_rate": 2.9486222884003393e-05,
+      "loss": 0.5742,
+      "step": 3283
+    },
+    {
+      "epoch": 0.4596221133659902,
+      "grad_norm": 0.3929372634486545,
+      "learning_rate": 2.9475073086821115e-05,
+      "loss": 0.5467,
+      "step": 3284
+    },
+    {
+      "epoch": 0.45976207137858643,
+      "grad_norm": 0.4188647769421151,
+      "learning_rate": 2.9463922369965917e-05,
+      "loss": 0.6122,
+      "step": 3285
+    },
+    {
+      "epoch": 0.45990202939118263,
+      "grad_norm": 0.39936115798519434,
+      "learning_rate": 2.9452770735729374e-05,
+      "loss": 0.5592,
+      "step": 3286
+    },
+    {
+      "epoch": 0.4600419874037789,
+      "grad_norm": 0.41793644764176796,
+      "learning_rate": 2.944161818640327e-05,
+      "loss": 0.5679,
+      "step": 3287
+    },
+    {
+      "epoch": 0.4601819454163751,
+      "grad_norm": 0.4053297174029381,
+      "learning_rate": 2.943046472427956e-05,
+      "loss": 0.5731,
+      "step": 3288
+    },
+    {
+      "epoch": 0.46032190342897134,
+      "grad_norm": 0.4026483342781206,
+      "learning_rate": 2.9419310351650392e-05,
+      "loss": 0.5884,
+      "step": 3289
+    },
+    {
+      "epoch": 0.46046186144156753,
+      "grad_norm": 0.4014937430945868,
+      "learning_rate": 2.9408155070808113e-05,
+      "loss": 0.5753,
+      "step": 3290
+    },
+    {
+      "epoch": 0.46060181945416373,
+      "grad_norm": 0.4168370671409344,
+      "learning_rate": 2.9396998884045235e-05,
+      "loss": 0.563,
+      "step": 3291
+    },
+    {
+      "epoch": 0.46074177746676,
+      "grad_norm": 0.4015569692729826,
+      "learning_rate": 2.9385841793654474e-05,
+      "loss": 0.603,
+      "step": 3292
+    },
+    {
+      "epoch": 0.4608817354793562,
+      "grad_norm": 0.3918827502679464,
+      "learning_rate": 2.937468380192872e-05,
+      "loss": 0.5587,
+      "step": 3293
+    },
+    {
+      "epoch": 0.46102169349195243,
+      "grad_norm": 0.4129995494719521,
+      "learning_rate": 2.936352491116106e-05,
+      "loss": 0.5554,
+      "step": 3294
+    },
+    {
+      "epoch": 0.46116165150454863,
+      "grad_norm": 0.4281727152857462,
+      "learning_rate": 2.9352365123644755e-05,
+      "loss": 0.5632,
+      "step": 3295
+    },
+    {
+      "epoch": 0.4613016095171449,
+      "grad_norm": 0.4078100385278002,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 0.5717,
+      "step": 3296
+    },
+    {
+      "epoch": 0.4614415675297411,
+      "grad_norm": 0.4670973011247201,
+      "learning_rate": 2.93300428675402e-05,
+      "loss": 0.5701,
+      "step": 3297
+    },
+    {
+      "epoch": 0.4615815255423373,
+      "grad_norm": 0.4149076463346234,
+      "learning_rate": 2.93188804035394e-05,
+      "loss": 0.5477,
+      "step": 3298
+    },
+    {
+      "epoch": 0.46172148355493353,
+      "grad_norm": 0.41094417493173746,
+      "learning_rate": 2.9307717051964862e-05,
+      "loss": 0.5499,
+      "step": 3299
+    },
+    {
+      "epoch": 0.46186144156752973,
+      "grad_norm": 0.40933405766552283,
+      "learning_rate": 2.929655281511075e-05,
+      "loss": 0.5389,
+      "step": 3300
+    },
+    {
+      "epoch": 0.462001399580126,
+      "grad_norm": 0.4235858324194054,
+      "learning_rate": 2.9285387695271444e-05,
+      "loss": 0.5568,
+      "step": 3301
+    },
+    {
+      "epoch": 0.4621413575927222,
+      "grad_norm": 0.3997529823349059,
+      "learning_rate": 2.9274221694741484e-05,
+      "loss": 0.5384,
+      "step": 3302
+    },
+    {
+      "epoch": 0.4622813156053184,
+      "grad_norm": 0.42004597023741636,
+      "learning_rate": 2.92630548158156e-05,
+      "loss": 0.5613,
+      "step": 3303
+    },
+    {
+      "epoch": 0.46242127361791463,
+      "grad_norm": 0.4076618497301079,
+      "learning_rate": 2.925188706078869e-05,
+      "loss": 0.533,
+      "step": 3304
+    },
+    {
+      "epoch": 0.4625612316305108,
+      "grad_norm": 0.4131076996508653,
+      "learning_rate": 2.9240718431955855e-05,
+      "loss": 0.5809,
+      "step": 3305
+    },
+    {
+      "epoch": 0.4627011896431071,
+      "grad_norm": 0.4246693787392879,
+      "learning_rate": 2.9229548931612348e-05,
+      "loss": 0.553,
+      "step": 3306
+    },
+    {
+      "epoch": 0.4628411476557033,
+      "grad_norm": 0.4081853198228349,
+      "learning_rate": 2.9218378562053623e-05,
+      "loss": 0.5269,
+      "step": 3307
+    },
+    {
+      "epoch": 0.46298110566829953,
+      "grad_norm": 0.4185513766982846,
+      "learning_rate": 2.9207207325575304e-05,
+      "loss": 0.5391,
+      "step": 3308
+    },
+    {
+      "epoch": 0.46312106368089573,
+      "grad_norm": 0.40971729273974855,
+      "learning_rate": 2.9196035224473196e-05,
+      "loss": 0.5562,
+      "step": 3309
+    },
+    {
+      "epoch": 0.4632610216934919,
+      "grad_norm": 0.39899049100540745,
+      "learning_rate": 2.918486226104327e-05,
+      "loss": 0.5574,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4634009797060882,
+      "grad_norm": 0.4210144546249132,
+      "learning_rate": 2.917368843758168e-05,
+      "loss": 0.5593,
+      "step": 3311
+    },
+    {
+      "epoch": 0.4635409377186844,
+      "grad_norm": 0.43313267085828555,
+      "learning_rate": 2.916251375638478e-05,
+      "loss": 0.5827,
+      "step": 3312
+    },
+    {
+      "epoch": 0.46368089573128063,
+      "grad_norm": 0.4159179322901382,
+      "learning_rate": 2.9151338219749065e-05,
+      "loss": 0.555,
+      "step": 3313
+    },
+    {
+      "epoch": 0.4638208537438768,
+      "grad_norm": 0.4383620819549179,
+      "learning_rate": 2.9140161829971223e-05,
+      "loss": 0.5638,
+      "step": 3314
+    },
+    {
+      "epoch": 0.4639608117564731,
+      "grad_norm": 0.4176966506673776,
+      "learning_rate": 2.9128984589348114e-05,
+      "loss": 0.5802,
+      "step": 3315
+    },
+    {
+      "epoch": 0.4641007697690693,
+      "grad_norm": 0.3985420143299249,
+      "learning_rate": 2.9117806500176774e-05,
+      "loss": 0.5614,
+      "step": 3316
+    },
+    {
+      "epoch": 0.4642407277816655,
+      "grad_norm": 0.40635759034414026,
+      "learning_rate": 2.910662756475443e-05,
+      "loss": 0.5502,
+      "step": 3317
+    },
+    {
+      "epoch": 0.4643806857942617,
+      "grad_norm": 0.3920809381768575,
+      "learning_rate": 2.9095447785378443e-05,
+      "loss": 0.5187,
+      "step": 3318
+    },
+    {
+      "epoch": 0.4645206438068579,
+      "grad_norm": 0.4065824247557009,
+      "learning_rate": 2.9084267164346386e-05,
+      "loss": 0.5285,
+      "step": 3319
+    },
+    {
+      "epoch": 0.4646606018194542,
+      "grad_norm": 0.47569552376008784,
+      "learning_rate": 2.9073085703955987e-05,
+      "loss": 0.5592,
+      "step": 3320
+    },
+    {
+      "epoch": 0.4648005598320504,
+      "grad_norm": 0.41348964126371485,
+      "learning_rate": 2.9061903406505154e-05,
+      "loss": 0.573,
+      "step": 3321
+    },
+    {
+      "epoch": 0.46494051784464663,
+      "grad_norm": 0.3990216785800373,
+      "learning_rate": 2.9050720274291943e-05,
+      "loss": 0.5686,
+      "step": 3322
+    },
+    {
+      "epoch": 0.4650804758572428,
+      "grad_norm": 0.3834656155781017,
+      "learning_rate": 2.903953630961463e-05,
+      "loss": 0.5297,
+      "step": 3323
+    },
+    {
+      "epoch": 0.465220433869839,
+      "grad_norm": 0.4046993834562908,
+      "learning_rate": 2.9028351514771606e-05,
+      "loss": 0.5515,
+      "step": 3324
+    },
+    {
+      "epoch": 0.4653603918824353,
+      "grad_norm": 0.43593159937116366,
+      "learning_rate": 2.901716589206147e-05,
+      "loss": 0.5706,
+      "step": 3325
+    },
+    {
+      "epoch": 0.4655003498950315,
+      "grad_norm": 0.41907649826033294,
+      "learning_rate": 2.9005979443782993e-05,
+      "loss": 0.5738,
+      "step": 3326
+    },
+    {
+      "epoch": 0.4656403079076277,
+      "grad_norm": 0.39650396383746195,
+      "learning_rate": 2.899479217223509e-05,
+      "loss": 0.5678,
+      "step": 3327
+    },
+    {
+      "epoch": 0.4657802659202239,
+      "grad_norm": 0.41237846145466506,
+      "learning_rate": 2.898360407971687e-05,
+      "loss": 0.5743,
+      "step": 3328
+    },
+    {
+      "epoch": 0.4659202239328202,
+      "grad_norm": 0.39685046167702703,
+      "learning_rate": 2.8972415168527584e-05,
+      "loss": 0.5439,
+      "step": 3329
+    },
+    {
+      "epoch": 0.4660601819454164,
+      "grad_norm": 0.4050270937895372,
+      "learning_rate": 2.896122544096668e-05,
+      "loss": 0.5541,
+      "step": 3330
+    },
+    {
+      "epoch": 0.46620013995801257,
+      "grad_norm": 0.40567639713038367,
+      "learning_rate": 2.895003489933375e-05,
+      "loss": 0.5712,
+      "step": 3331
+    },
+    {
+      "epoch": 0.4663400979706088,
+      "grad_norm": 0.4174104058437715,
+      "learning_rate": 2.8938843545928573e-05,
+      "loss": 0.5634,
+      "step": 3332
+    },
+    {
+      "epoch": 0.466480055983205,
+      "grad_norm": 0.4174890988710686,
+      "learning_rate": 2.892765138305108e-05,
+      "loss": 0.5633,
+      "step": 3333
+    },
+    {
+      "epoch": 0.4666200139958013,
+      "grad_norm": 0.4085166170083092,
+      "learning_rate": 2.8916458413001375e-05,
+      "loss": 0.5661,
+      "step": 3334
+    },
+    {
+      "epoch": 0.4667599720083975,
+      "grad_norm": 0.416230735105726,
+      "learning_rate": 2.8905264638079732e-05,
+      "loss": 0.5736,
+      "step": 3335
+    },
+    {
+      "epoch": 0.4668999300209937,
+      "grad_norm": 0.3905749725084316,
+      "learning_rate": 2.8894070060586576e-05,
+      "loss": 0.5483,
+      "step": 3336
+    },
+    {
+      "epoch": 0.4670398880335899,
+      "grad_norm": 0.40890374703665905,
+      "learning_rate": 2.888287468282252e-05,
+      "loss": 0.5615,
+      "step": 3337
+    },
+    {
+      "epoch": 0.4671798460461861,
+      "grad_norm": 0.41635405487950694,
+      "learning_rate": 2.8871678507088312e-05,
+      "loss": 0.5467,
+      "step": 3338
+    },
+    {
+      "epoch": 0.4673198040587824,
+      "grad_norm": 0.4336846869668269,
+      "learning_rate": 2.886048153568489e-05,
+      "loss": 0.5511,
+      "step": 3339
+    },
+    {
+      "epoch": 0.46745976207137857,
+      "grad_norm": 0.38470189006529587,
+      "learning_rate": 2.8849283770913337e-05,
+      "loss": 0.5488,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4675997200839748,
+      "grad_norm": 0.43922504036128623,
+      "learning_rate": 2.8838085215074923e-05,
+      "loss": 0.592,
+      "step": 3341
+    },
+    {
+      "epoch": 0.467739678096571,
+      "grad_norm": 0.4046341572147428,
+      "learning_rate": 2.8826885870471043e-05,
+      "loss": 0.5618,
+      "step": 3342
+    },
+    {
+      "epoch": 0.4678796361091673,
+      "grad_norm": 0.38781336644907494,
+      "learning_rate": 2.8815685739403298e-05,
+      "loss": 0.556,
+      "step": 3343
+    },
+    {
+      "epoch": 0.46801959412176347,
+      "grad_norm": 0.4120764675677611,
+      "learning_rate": 2.8804484824173417e-05,
+      "loss": 0.5407,
+      "step": 3344
+    },
+    {
+      "epoch": 0.46815955213435967,
+      "grad_norm": 0.4309704301588916,
+      "learning_rate": 2.8793283127083292e-05,
+      "loss": 0.5491,
+      "step": 3345
+    },
+    {
+      "epoch": 0.4682995101469559,
+      "grad_norm": 0.3937722091804054,
+      "learning_rate": 2.8782080650435006e-05,
+      "loss": 0.5799,
+      "step": 3346
+    },
+    {
+      "epoch": 0.4684394681595521,
+      "grad_norm": 0.39196249510935055,
+      "learning_rate": 2.8770877396530766e-05,
+      "loss": 0.5386,
+      "step": 3347
+    },
+    {
+      "epoch": 0.46857942617214837,
+      "grad_norm": 0.410878891101837,
+      "learning_rate": 2.875967336767296e-05,
+      "loss": 0.603,
+      "step": 3348
+    },
+    {
+      "epoch": 0.46871938418474457,
+      "grad_norm": 0.4020037749774897,
+      "learning_rate": 2.8748468566164134e-05,
+      "loss": 0.5499,
+      "step": 3349
+    },
+    {
+      "epoch": 0.4688593421973408,
+      "grad_norm": 0.45651213620060477,
+      "learning_rate": 2.8737262994306985e-05,
+      "loss": 0.6113,
+      "step": 3350
+    },
+    {
+      "epoch": 0.468999300209937,
+      "grad_norm": 0.4187097393109256,
+      "learning_rate": 2.872605665440436e-05,
+      "loss": 0.548,
+      "step": 3351
+    },
+    {
+      "epoch": 0.4691392582225332,
+      "grad_norm": 0.4096624177728241,
+      "learning_rate": 2.8714849548759293e-05,
+      "loss": 0.5629,
+      "step": 3352
+    },
+    {
+      "epoch": 0.46927921623512947,
+      "grad_norm": 0.4106487034595635,
+      "learning_rate": 2.8703641679674954e-05,
+      "loss": 0.5743,
+      "step": 3353
+    },
+    {
+      "epoch": 0.46941917424772567,
+      "grad_norm": 0.4160938815791059,
+      "learning_rate": 2.869243304945467e-05,
+      "loss": 0.5366,
+      "step": 3354
+    },
+    {
+      "epoch": 0.4695591322603219,
+      "grad_norm": 0.401353123051922,
+      "learning_rate": 2.8681223660401935e-05,
+      "loss": 0.5601,
+      "step": 3355
+    },
+    {
+      "epoch": 0.4696990902729181,
+      "grad_norm": 0.3960216537704636,
+      "learning_rate": 2.8670013514820375e-05,
+      "loss": 0.5472,
+      "step": 3356
+    },
+    {
+      "epoch": 0.46983904828551437,
+      "grad_norm": 0.40124295326684595,
+      "learning_rate": 2.8658802615013807e-05,
+      "loss": 0.5582,
+      "step": 3357
+    },
+    {
+      "epoch": 0.46997900629811057,
+      "grad_norm": 0.4061763668246899,
+      "learning_rate": 2.8647590963286175e-05,
+      "loss": 0.5582,
+      "step": 3358
+    },
+    {
+      "epoch": 0.47011896431070677,
+      "grad_norm": 0.42540877415275535,
+      "learning_rate": 2.8636378561941592e-05,
+      "loss": 0.5555,
+      "step": 3359
+    },
+    {
+      "epoch": 0.470258922323303,
+      "grad_norm": 0.40556363836960674,
+      "learning_rate": 2.862516541328431e-05,
+      "loss": 0.5677,
+      "step": 3360
+    },
+    {
+      "epoch": 0.4703988803358992,
+      "grad_norm": 0.4163599612619921,
+      "learning_rate": 2.8613951519618765e-05,
+      "loss": 0.5648,
+      "step": 3361
+    },
+    {
+      "epoch": 0.47053883834849547,
+      "grad_norm": 0.4088068121264862,
+      "learning_rate": 2.8602736883249503e-05,
+      "loss": 0.5592,
+      "step": 3362
+    },
+    {
+      "epoch": 0.47067879636109167,
+      "grad_norm": 0.37314677325016854,
+      "learning_rate": 2.859152150648126e-05,
+      "loss": 0.5145,
+      "step": 3363
+    },
+    {
+      "epoch": 0.4708187543736879,
+      "grad_norm": 0.7930198911401396,
+      "learning_rate": 2.8580305391618912e-05,
+      "loss": 0.5536,
+      "step": 3364
+    },
+    {
+      "epoch": 0.4709587123862841,
+      "grad_norm": 0.41677939251494683,
+      "learning_rate": 2.8569088540967476e-05,
+      "loss": 0.5369,
+      "step": 3365
+    },
+    {
+      "epoch": 0.4710986703988803,
+      "grad_norm": 0.3922638579854847,
+      "learning_rate": 2.8557870956832132e-05,
+      "loss": 0.5565,
+      "step": 3366
+    },
+    {
+      "epoch": 0.47123862841147657,
+      "grad_norm": 0.4290159510473203,
+      "learning_rate": 2.8546652641518208e-05,
+      "loss": 0.5655,
+      "step": 3367
+    },
+    {
+      "epoch": 0.47137858642407277,
+      "grad_norm": 0.41113045647723195,
+      "learning_rate": 2.8535433597331173e-05,
+      "loss": 0.5651,
+      "step": 3368
+    },
+    {
+      "epoch": 0.471518544436669,
+      "grad_norm": 0.4241008004580075,
+      "learning_rate": 2.8524213826576667e-05,
+      "loss": 0.5634,
+      "step": 3369
+    },
+    {
+      "epoch": 0.4716585024492652,
+      "grad_norm": 0.4461069068987371,
+      "learning_rate": 2.851299333156046e-05,
+      "loss": 0.5641,
+      "step": 3370
+    },
+    {
+      "epoch": 0.47179846046186147,
+      "grad_norm": 0.4284897736677648,
+      "learning_rate": 2.8501772114588476e-05,
+      "loss": 0.5711,
+      "step": 3371
+    },
+    {
+      "epoch": 0.47193841847445767,
+      "grad_norm": 0.3880573082608547,
+      "learning_rate": 2.8490550177966797e-05,
+      "loss": 0.4844,
+      "step": 3372
+    },
+    {
+      "epoch": 0.47207837648705386,
+      "grad_norm": 0.4343172019973716,
+      "learning_rate": 2.8479327524001636e-05,
+      "loss": 0.5544,
+      "step": 3373
+    },
+    {
+      "epoch": 0.4722183344996501,
+      "grad_norm": 0.40211668252102023,
+      "learning_rate": 2.8468104154999366e-05,
+      "loss": 0.5577,
+      "step": 3374
+    },
+    {
+      "epoch": 0.4723582925122463,
+      "grad_norm": 0.4165983505211132,
+      "learning_rate": 2.845688007326651e-05,
+      "loss": 0.5462,
+      "step": 3375
+    },
+    {
+      "epoch": 0.47249825052484257,
+      "grad_norm": 0.41318809239332804,
+      "learning_rate": 2.8445655281109718e-05,
+      "loss": 0.5821,
+      "step": 3376
+    },
+    {
+      "epoch": 0.47263820853743876,
+      "grad_norm": 0.43834923618793703,
+      "learning_rate": 2.8434429780835807e-05,
+      "loss": 0.5682,
+      "step": 3377
+    },
+    {
+      "epoch": 0.472778166550035,
+      "grad_norm": 0.4069711220875038,
+      "learning_rate": 2.8423203574751727e-05,
+      "loss": 0.5339,
+      "step": 3378
+    },
+    {
+      "epoch": 0.4729181245626312,
+      "grad_norm": 0.39860818609724685,
+      "learning_rate": 2.8411976665164585e-05,
+      "loss": 0.5379,
+      "step": 3379
+    },
+    {
+      "epoch": 0.4730580825752274,
+      "grad_norm": 0.41823989656114025,
+      "learning_rate": 2.840074905438161e-05,
+      "loss": 0.5801,
+      "step": 3380
+    },
+    {
+      "epoch": 0.47319804058782367,
+      "grad_norm": 0.41218220409199874,
+      "learning_rate": 2.8389520744710196e-05,
+      "loss": 0.5221,
+      "step": 3381
+    },
+    {
+      "epoch": 0.47333799860041986,
+      "grad_norm": 0.42461569878650546,
+      "learning_rate": 2.8378291738457887e-05,
+      "loss": 0.5785,
+      "step": 3382
+    },
+    {
+      "epoch": 0.4734779566130161,
+      "grad_norm": 0.4076091943362167,
+      "learning_rate": 2.8367062037932342e-05,
+      "loss": 0.5415,
+      "step": 3383
+    },
+    {
+      "epoch": 0.4736179146256123,
+      "grad_norm": 0.41452613569881464,
+      "learning_rate": 2.8355831645441388e-05,
+      "loss": 0.5649,
+      "step": 3384
+    },
+    {
+      "epoch": 0.4737578726382085,
+      "grad_norm": 0.4037756702805257,
+      "learning_rate": 2.834460056329298e-05,
+      "loss": 0.5523,
+      "step": 3385
+    },
+    {
+      "epoch": 0.47389783065080476,
+      "grad_norm": 0.4011452181351811,
+      "learning_rate": 2.8333368793795224e-05,
+      "loss": 0.5246,
+      "step": 3386
+    },
+    {
+      "epoch": 0.47403778866340096,
+      "grad_norm": 0.413887567092737,
+      "learning_rate": 2.8322136339256356e-05,
+      "loss": 0.5281,
+      "step": 3387
+    },
+    {
+      "epoch": 0.4741777466759972,
+      "grad_norm": 0.42257963112995583,
+      "learning_rate": 2.8310903201984763e-05,
+      "loss": 0.5612,
+      "step": 3388
+    },
+    {
+      "epoch": 0.4743177046885934,
+      "grad_norm": 0.6050510544766738,
+      "learning_rate": 2.829966938428897e-05,
+      "loss": 0.5807,
+      "step": 3389
+    },
+    {
+      "epoch": 0.47445766270118966,
+      "grad_norm": 0.42337956289277995,
+      "learning_rate": 2.8288434888477627e-05,
+      "loss": 0.5648,
+      "step": 3390
+    },
+    {
+      "epoch": 0.47459762071378586,
+      "grad_norm": 0.39066953133834564,
+      "learning_rate": 2.827719971685956e-05,
+      "loss": 0.5896,
+      "step": 3391
+    },
+    {
+      "epoch": 0.47473757872638206,
+      "grad_norm": 0.40333322646376846,
+      "learning_rate": 2.8265963871743696e-05,
+      "loss": 0.5761,
+      "step": 3392
+    },
+    {
+      "epoch": 0.4748775367389783,
+      "grad_norm": 0.41387027644417956,
+      "learning_rate": 2.825472735543912e-05,
+      "loss": 0.5925,
+      "step": 3393
+    },
+    {
+      "epoch": 0.4750174947515745,
+      "grad_norm": 0.3928152148704509,
+      "learning_rate": 2.8243490170255043e-05,
+      "loss": 0.5828,
+      "step": 3394
+    },
+    {
+      "epoch": 0.47515745276417076,
+      "grad_norm": 0.3929042322405297,
+      "learning_rate": 2.8232252318500834e-05,
+      "loss": 0.5831,
+      "step": 3395
+    },
+    {
+      "epoch": 0.47529741077676696,
+      "grad_norm": 0.4673310747874466,
+      "learning_rate": 2.8221013802485975e-05,
+      "loss": 0.5564,
+      "step": 3396
+    },
+    {
+      "epoch": 0.4754373687893632,
+      "grad_norm": 0.3888740851313229,
+      "learning_rate": 2.8209774624520097e-05,
+      "loss": 0.5575,
+      "step": 3397
+    },
+    {
+      "epoch": 0.4755773268019594,
+      "grad_norm": 0.40361151709091403,
+      "learning_rate": 2.8198534786912965e-05,
+      "loss": 0.5496,
+      "step": 3398
+    },
+    {
+      "epoch": 0.4757172848145556,
+      "grad_norm": 0.4106448187857887,
+      "learning_rate": 2.818729429197448e-05,
+      "loss": 0.5607,
+      "step": 3399
+    },
+    {
+      "epoch": 0.47585724282715186,
+      "grad_norm": 0.42784648142356363,
+      "learning_rate": 2.8176053142014687e-05,
+      "loss": 0.6264,
+      "step": 3400
+    },
+    {
+      "epoch": 0.47599720083974806,
+      "grad_norm": 0.40344827404813044,
+      "learning_rate": 2.8164811339343732e-05,
+      "loss": 0.5812,
+      "step": 3401
+    },
+    {
+      "epoch": 0.4761371588523443,
+      "grad_norm": 0.41503031802135976,
+      "learning_rate": 2.815356888627195e-05,
+      "loss": 0.5635,
+      "step": 3402
+    },
+    {
+      "epoch": 0.4762771168649405,
+      "grad_norm": 0.3854795531096736,
+      "learning_rate": 2.814232578510975e-05,
+      "loss": 0.5635,
+      "step": 3403
+    },
+    {
+      "epoch": 0.47641707487753676,
+      "grad_norm": 0.4009076755410169,
+      "learning_rate": 2.8131082038167735e-05,
+      "loss": 0.5731,
+      "step": 3404
+    },
+    {
+      "epoch": 0.47655703289013296,
+      "grad_norm": 0.38835090703799413,
+      "learning_rate": 2.8119837647756574e-05,
+      "loss": 0.565,
+      "step": 3405
+    },
+    {
+      "epoch": 0.47669699090272916,
+      "grad_norm": 0.4286357705154638,
+      "learning_rate": 2.8108592616187133e-05,
+      "loss": 0.549,
+      "step": 3406
+    },
+    {
+      "epoch": 0.4768369489153254,
+      "grad_norm": 0.4275083145446453,
+      "learning_rate": 2.8097346945770364e-05,
+      "loss": 0.5719,
+      "step": 3407
+    },
+    {
+      "epoch": 0.4769769069279216,
+      "grad_norm": 0.3940008020310687,
+      "learning_rate": 2.808610063881737e-05,
+      "loss": 0.5208,
+      "step": 3408
+    },
+    {
+      "epoch": 0.47711686494051786,
+      "grad_norm": 0.4229335142753006,
+      "learning_rate": 2.807485369763938e-05,
+      "loss": 0.5763,
+      "step": 3409
+    },
+    {
+      "epoch": 0.47725682295311406,
+      "grad_norm": 0.4161121773460535,
+      "learning_rate": 2.8063606124547765e-05,
+      "loss": 0.5443,
+      "step": 3410
+    },
+    {
+      "epoch": 0.4773967809657103,
+      "grad_norm": 0.40765919854648125,
+      "learning_rate": 2.8052357921854e-05,
+      "loss": 0.5702,
+      "step": 3411
+    },
+    {
+      "epoch": 0.4775367389783065,
+      "grad_norm": 0.4163390293585378,
+      "learning_rate": 2.804110909186971e-05,
+      "loss": 0.5728,
+      "step": 3412
+    },
+    {
+      "epoch": 0.4776766969909027,
+      "grad_norm": 0.4377093314827132,
+      "learning_rate": 2.8029859636906654e-05,
+      "loss": 0.5785,
+      "step": 3413
+    },
+    {
+      "epoch": 0.47781665500349896,
+      "grad_norm": 0.3997496345348747,
+      "learning_rate": 2.8018609559276686e-05,
+      "loss": 0.5872,
+      "step": 3414
+    },
+    {
+      "epoch": 0.47795661301609516,
+      "grad_norm": 0.41218479804115327,
+      "learning_rate": 2.800735886129184e-05,
+      "loss": 0.5606,
+      "step": 3415
+    },
+    {
+      "epoch": 0.4780965710286914,
+      "grad_norm": 0.4227372321145177,
+      "learning_rate": 2.7996107545264223e-05,
+      "loss": 0.5708,
+      "step": 3416
+    },
+    {
+      "epoch": 0.4782365290412876,
+      "grad_norm": 0.42665670114635346,
+      "learning_rate": 2.7984855613506107e-05,
+      "loss": 0.5704,
+      "step": 3417
+    },
+    {
+      "epoch": 0.47837648705388386,
+      "grad_norm": 0.4216359313020928,
+      "learning_rate": 2.7973603068329884e-05,
+      "loss": 0.5378,
+      "step": 3418
+    },
+    {
+      "epoch": 0.47851644506648006,
+      "grad_norm": 0.4320525421405627,
+      "learning_rate": 2.796234991204805e-05,
+      "loss": 0.5626,
+      "step": 3419
+    },
+    {
+      "epoch": 0.47865640307907625,
+      "grad_norm": 0.4087593290796549,
+      "learning_rate": 2.795109614697326e-05,
+      "loss": 0.5472,
+      "step": 3420
+    },
+    {
+      "epoch": 0.4787963610916725,
+      "grad_norm": 0.4311719285838966,
+      "learning_rate": 2.793984177541827e-05,
+      "loss": 0.5585,
+      "step": 3421
+    },
+    {
+      "epoch": 0.4789363191042687,
+      "grad_norm": 0.4101348935742422,
+      "learning_rate": 2.792858679969596e-05,
+      "loss": 0.5539,
+      "step": 3422
+    },
+    {
+      "epoch": 0.47907627711686496,
+      "grad_norm": 0.4240688742729598,
+      "learning_rate": 2.7917331222119346e-05,
+      "loss": 0.5853,
+      "step": 3423
+    },
+    {
+      "epoch": 0.47921623512946115,
+      "grad_norm": 0.4192530280413205,
+      "learning_rate": 2.790607504500157e-05,
+      "loss": 0.5476,
+      "step": 3424
+    },
+    {
+      "epoch": 0.4793561931420574,
+      "grad_norm": 0.39900981412049086,
+      "learning_rate": 2.7894818270655882e-05,
+      "loss": 0.5582,
+      "step": 3425
+    },
+    {
+      "epoch": 0.4794961511546536,
+      "grad_norm": 0.3965711350604459,
+      "learning_rate": 2.7883560901395667e-05,
+      "loss": 0.5419,
+      "step": 3426
+    },
+    {
+      "epoch": 0.4796361091672498,
+      "grad_norm": 0.4312456683665289,
+      "learning_rate": 2.7872302939534433e-05,
+      "loss": 0.5536,
+      "step": 3427
+    },
+    {
+      "epoch": 0.47977606717984606,
+      "grad_norm": 0.40864635902503055,
+      "learning_rate": 2.78610443873858e-05,
+      "loss": 0.5812,
+      "step": 3428
+    },
+    {
+      "epoch": 0.47991602519244225,
+      "grad_norm": 0.40901705450143366,
+      "learning_rate": 2.7849785247263515e-05,
+      "loss": 0.5284,
+      "step": 3429
+    },
+    {
+      "epoch": 0.4800559832050385,
+      "grad_norm": 0.40430272789500593,
+      "learning_rate": 2.7838525521481444e-05,
+      "loss": 0.5587,
+      "step": 3430
+    },
+    {
+      "epoch": 0.4801959412176347,
+      "grad_norm": 0.3934946456153737,
+      "learning_rate": 2.7827265212353587e-05,
+      "loss": 0.5509,
+      "step": 3431
+    },
+    {
+      "epoch": 0.48033589923023096,
+      "grad_norm": 0.42366137833048706,
+      "learning_rate": 2.781600432219404e-05,
+      "loss": 0.5478,
+      "step": 3432
+    },
+    {
+      "epoch": 0.48047585724282715,
+      "grad_norm": 0.4067675335773988,
+      "learning_rate": 2.7804742853317027e-05,
+      "loss": 0.5348,
+      "step": 3433
+    },
+    {
+      "epoch": 0.48061581525542335,
+      "grad_norm": 0.3990940892590606,
+      "learning_rate": 2.7793480808036898e-05,
+      "loss": 0.5449,
+      "step": 3434
+    },
+    {
+      "epoch": 0.4807557732680196,
+      "grad_norm": 0.40172847447577514,
+      "learning_rate": 2.7782218188668118e-05,
+      "loss": 0.5898,
+      "step": 3435
+    },
+    {
+      "epoch": 0.4808957312806158,
+      "grad_norm": 0.3955704115554125,
+      "learning_rate": 2.7770954997525277e-05,
+      "loss": 0.5357,
+      "step": 3436
+    },
+    {
+      "epoch": 0.48103568929321205,
+      "grad_norm": 0.37493116627514156,
+      "learning_rate": 2.7759691236923064e-05,
+      "loss": 0.5197,
+      "step": 3437
+    },
+    {
+      "epoch": 0.48117564730580825,
+      "grad_norm": 0.40865836884433765,
+      "learning_rate": 2.7748426909176307e-05,
+      "loss": 0.5567,
+      "step": 3438
+    },
+    {
+      "epoch": 0.4813156053184045,
+      "grad_norm": 0.41498192510745757,
+      "learning_rate": 2.7737162016599927e-05,
+      "loss": 0.595,
+      "step": 3439
+    },
+    {
+      "epoch": 0.4814555633310007,
+      "grad_norm": 0.4299760976868325,
+      "learning_rate": 2.7725896561508983e-05,
+      "loss": 0.5465,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4815955213435969,
+      "grad_norm": 0.40031579978939574,
+      "learning_rate": 2.7714630546218635e-05,
+      "loss": 0.5965,
+      "step": 3441
+    },
+    {
+      "epoch": 0.48173547935619315,
+      "grad_norm": 0.40096718317347985,
+      "learning_rate": 2.770336397304417e-05,
+      "loss": 0.5321,
+      "step": 3442
+    },
+    {
+      "epoch": 0.48187543736878935,
+      "grad_norm": 0.3981130281220963,
+      "learning_rate": 2.769209684430098e-05,
+      "loss": 0.5558,
+      "step": 3443
+    },
+    {
+      "epoch": 0.4820153953813856,
+      "grad_norm": 0.45577292418268894,
+      "learning_rate": 2.7680829162304567e-05,
+      "loss": 0.5586,
+      "step": 3444
+    },
+    {
+      "epoch": 0.4821553533939818,
+      "grad_norm": 0.4112624187601042,
+      "learning_rate": 2.7669560929370564e-05,
+      "loss": 0.5457,
+      "step": 3445
+    },
+    {
+      "epoch": 0.48229531140657805,
+      "grad_norm": 0.39264802624439415,
+      "learning_rate": 2.76582921478147e-05,
+      "loss": 0.547,
+      "step": 3446
+    },
+    {
+      "epoch": 0.48243526941917425,
+      "grad_norm": 0.4141847607595186,
+      "learning_rate": 2.7647022819952835e-05,
+      "loss": 0.5274,
+      "step": 3447
+    },
+    {
+      "epoch": 0.48257522743177045,
+      "grad_norm": 0.40886257454682706,
+      "learning_rate": 2.763575294810091e-05,
+      "loss": 0.5305,
+      "step": 3448
+    },
+    {
+      "epoch": 0.4827151854443667,
+      "grad_norm": 0.39545619852603475,
+      "learning_rate": 2.7624482534575026e-05,
+      "loss": 0.5839,
+      "step": 3449
+    },
+    {
+      "epoch": 0.4828551434569629,
+      "grad_norm": 0.40937623131741985,
+      "learning_rate": 2.761321158169134e-05,
+      "loss": 0.5638,
+      "step": 3450
+    },
+    {
+      "epoch": 0.48299510146955915,
+      "grad_norm": 0.407186107707272,
+      "learning_rate": 2.7601940091766164e-05,
+      "loss": 0.5158,
+      "step": 3451
+    },
+    {
+      "epoch": 0.48313505948215535,
+      "grad_norm": 0.4076092863747084,
+      "learning_rate": 2.7590668067115895e-05,
+      "loss": 0.5874,
+      "step": 3452
+    },
+    {
+      "epoch": 0.4832750174947516,
+      "grad_norm": 0.4221563990293702,
+      "learning_rate": 2.757939551005706e-05,
+      "loss": 0.5484,
+      "step": 3453
+    },
+    {
+      "epoch": 0.4834149755073478,
+      "grad_norm": 0.38895523634768725,
+      "learning_rate": 2.7568122422906273e-05,
+      "loss": 0.5166,
+      "step": 3454
+    },
+    {
+      "epoch": 0.483554933519944,
+      "grad_norm": 0.4268904833505735,
+      "learning_rate": 2.755684880798026e-05,
+      "loss": 0.5862,
+      "step": 3455
+    },
+    {
+      "epoch": 0.48369489153254025,
+      "grad_norm": 0.41707456856060987,
+      "learning_rate": 2.754557466759589e-05,
+      "loss": 0.5704,
+      "step": 3456
+    },
+    {
+      "epoch": 0.48383484954513645,
+      "grad_norm": 0.41401351680807885,
+      "learning_rate": 2.7534300004070086e-05,
+      "loss": 0.5452,
+      "step": 3457
+    },
+    {
+      "epoch": 0.4839748075577327,
+      "grad_norm": 0.4000475828565154,
+      "learning_rate": 2.7523024819719922e-05,
+      "loss": 0.5862,
+      "step": 3458
+    },
+    {
+      "epoch": 0.4841147655703289,
+      "grad_norm": 0.4126946869503196,
+      "learning_rate": 2.7511749116862558e-05,
+      "loss": 0.5499,
+      "step": 3459
+    },
+    {
+      "epoch": 0.4842547235829251,
+      "grad_norm": 0.4046742301519851,
+      "learning_rate": 2.7500472897815265e-05,
+      "loss": 0.5295,
+      "step": 3460
+    },
+    {
+      "epoch": 0.48439468159552135,
+      "grad_norm": 0.42125194708107067,
+      "learning_rate": 2.748919616489542e-05,
+      "loss": 0.5356,
+      "step": 3461
+    },
+    {
+      "epoch": 0.48453463960811755,
+      "grad_norm": 0.4040649587533329,
+      "learning_rate": 2.7477918920420504e-05,
+      "loss": 0.553,
+      "step": 3462
+    },
+    {
+      "epoch": 0.4846745976207138,
+      "grad_norm": 0.3886454789211703,
+      "learning_rate": 2.7466641166708113e-05,
+      "loss": 0.5567,
+      "step": 3463
+    },
+    {
+      "epoch": 0.48481455563331,
+      "grad_norm": 0.4996582332055803,
+      "learning_rate": 2.7455362906075932e-05,
+      "loss": 0.5722,
+      "step": 3464
+    },
+    {
+      "epoch": 0.48495451364590625,
+      "grad_norm": 0.423979275094494,
+      "learning_rate": 2.7444084140841765e-05,
+      "loss": 0.5837,
+      "step": 3465
+    },
+    {
+      "epoch": 0.48509447165850245,
+      "grad_norm": 0.39551615527180584,
+      "learning_rate": 2.74328048733235e-05,
+      "loss": 0.5754,
+      "step": 3466
+    },
+    {
+      "epoch": 0.48523442967109864,
+      "grad_norm": 0.45978728132052005,
+      "learning_rate": 2.7421525105839152e-05,
+      "loss": 0.5367,
+      "step": 3467
+    },
+    {
+      "epoch": 0.4853743876836949,
+      "grad_norm": 0.4193876985062947,
+      "learning_rate": 2.741024484070682e-05,
+      "loss": 0.5576,
+      "step": 3468
+    },
+    {
+      "epoch": 0.4855143456962911,
+      "grad_norm": 0.397308979018639,
+      "learning_rate": 2.739896408024473e-05,
+      "loss": 0.5577,
+      "step": 3469
+    },
+    {
+      "epoch": 0.48565430370888735,
+      "grad_norm": 0.4127845230541042,
+      "learning_rate": 2.7387682826771173e-05,
+      "loss": 0.5876,
+      "step": 3470
+    },
+    {
+      "epoch": 0.48579426172148354,
+      "grad_norm": 0.40489978417434586,
+      "learning_rate": 2.7376401082604564e-05,
+      "loss": 0.5622,
+      "step": 3471
+    },
+    {
+      "epoch": 0.4859342197340798,
+      "grad_norm": 0.3839599583629985,
+      "learning_rate": 2.736511885006343e-05,
+      "loss": 0.5432,
+      "step": 3472
+    },
+    {
+      "epoch": 0.486074177746676,
+      "grad_norm": 0.39225392267678383,
+      "learning_rate": 2.7353836131466372e-05,
+      "loss": 0.5639,
+      "step": 3473
+    },
+    {
+      "epoch": 0.4862141357592722,
+      "grad_norm": 0.4051784046488118,
+      "learning_rate": 2.7342552929132104e-05,
+      "loss": 0.5664,
+      "step": 3474
+    },
+    {
+      "epoch": 0.48635409377186845,
+      "grad_norm": 0.40996524593878775,
+      "learning_rate": 2.7331269245379443e-05,
+      "loss": 0.5601,
+      "step": 3475
+    },
+    {
+      "epoch": 0.48649405178446464,
+      "grad_norm": 0.42173107295905043,
+      "learning_rate": 2.7319985082527292e-05,
+      "loss": 0.5497,
+      "step": 3476
+    },
+    {
+      "epoch": 0.4866340097970609,
+      "grad_norm": 0.40466464089704407,
+      "learning_rate": 2.7308700442894675e-05,
+      "loss": 0.5397,
+      "step": 3477
+    },
+    {
+      "epoch": 0.4867739678096571,
+      "grad_norm": 0.3942022841659981,
+      "learning_rate": 2.7297415328800692e-05,
+      "loss": 0.5313,
+      "step": 3478
+    },
+    {
+      "epoch": 0.48691392582225335,
+      "grad_norm": 0.40743336481779613,
+      "learning_rate": 2.728612974256454e-05,
+      "loss": 0.5313,
+      "step": 3479
+    },
+    {
+      "epoch": 0.48705388383484954,
+      "grad_norm": 0.4332085486342721,
+      "learning_rate": 2.7274843686505536e-05,
+      "loss": 0.5455,
+      "step": 3480
+    },
+    {
+      "epoch": 0.48719384184744574,
+      "grad_norm": 0.4094748981538972,
+      "learning_rate": 2.726355716294307e-05,
+      "loss": 0.5727,
+      "step": 3481
+    },
+    {
+      "epoch": 0.487333799860042,
+      "grad_norm": 0.4111671844080047,
+      "learning_rate": 2.7252270174196643e-05,
+      "loss": 0.5635,
+      "step": 3482
+    },
+    {
+      "epoch": 0.4874737578726382,
+      "grad_norm": 0.39174881532897554,
+      "learning_rate": 2.724098272258584e-05,
+      "loss": 0.5414,
+      "step": 3483
+    },
+    {
+      "epoch": 0.48761371588523444,
+      "grad_norm": 0.3907484104141741,
+      "learning_rate": 2.7229694810430344e-05,
+      "loss": 0.5766,
+      "step": 3484
+    },
+    {
+      "epoch": 0.48775367389783064,
+      "grad_norm": 0.4010543272236602,
+      "learning_rate": 2.7218406440049954e-05,
+      "loss": 0.5371,
+      "step": 3485
+    },
+    {
+      "epoch": 0.4878936319104269,
+      "grad_norm": 0.4346615757792164,
+      "learning_rate": 2.720711761376452e-05,
+      "loss": 0.5729,
+      "step": 3486
+    },
+    {
+      "epoch": 0.4880335899230231,
+      "grad_norm": 0.4341695351875495,
+      "learning_rate": 2.719582833389403e-05,
+      "loss": 0.6109,
+      "step": 3487
+    },
+    {
+      "epoch": 0.4881735479356193,
+      "grad_norm": 0.43134633918438864,
+      "learning_rate": 2.7184538602758536e-05,
+      "loss": 0.5904,
+      "step": 3488
+    },
+    {
+      "epoch": 0.48831350594821554,
+      "grad_norm": 0.40924934907989946,
+      "learning_rate": 2.71732484226782e-05,
+      "loss": 0.5533,
+      "step": 3489
+    },
+    {
+      "epoch": 0.48845346396081174,
+      "grad_norm": 0.4117613614479626,
+      "learning_rate": 2.7161957795973246e-05,
+      "loss": 0.519,
+      "step": 3490
+    },
+    {
+      "epoch": 0.488593421973408,
+      "grad_norm": 0.3876515998640682,
+      "learning_rate": 2.7150666724964035e-05,
+      "loss": 0.5543,
+      "step": 3491
+    },
+    {
+      "epoch": 0.4887333799860042,
+      "grad_norm": 0.407828217550299,
+      "learning_rate": 2.7139375211970996e-05,
+      "loss": 0.5515,
+      "step": 3492
+    },
+    {
+      "epoch": 0.48887333799860044,
+      "grad_norm": 0.40672013393542944,
+      "learning_rate": 2.712808325931464e-05,
+      "loss": 0.5445,
+      "step": 3493
+    },
+    {
+      "epoch": 0.48901329601119664,
+      "grad_norm": 0.40868169613662847,
+      "learning_rate": 2.7116790869315582e-05,
+      "loss": 0.5712,
+      "step": 3494
+    },
+    {
+      "epoch": 0.48915325402379284,
+      "grad_norm": 0.4092633353198152,
+      "learning_rate": 2.710549804429452e-05,
+      "loss": 0.5515,
+      "step": 3495
+    },
+    {
+      "epoch": 0.4892932120363891,
+      "grad_norm": 0.4366896965056272,
+      "learning_rate": 2.7094204786572254e-05,
+      "loss": 0.5883,
+      "step": 3496
+    },
+    {
+      "epoch": 0.4894331700489853,
+      "grad_norm": 0.41669884030755355,
+      "learning_rate": 2.7082911098469648e-05,
+      "loss": 0.5386,
+      "step": 3497
+    },
+    {
+      "epoch": 0.48957312806158154,
+      "grad_norm": 0.39572509135352624,
+      "learning_rate": 2.7071616982307684e-05,
+      "loss": 0.5563,
+      "step": 3498
+    },
+    {
+      "epoch": 0.48971308607417774,
+      "grad_norm": 0.41695121258573753,
+      "learning_rate": 2.706032244040741e-05,
+      "loss": 0.5796,
+      "step": 3499
+    },
+    {
+      "epoch": 0.489853044086774,
+      "grad_norm": 0.4236891067324582,
+      "learning_rate": 2.704902747508996e-05,
+      "loss": 0.5799,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4899930020993702,
+      "grad_norm": 0.3977818931519091,
+      "learning_rate": 2.7037732088676582e-05,
+      "loss": 0.6042,
+      "step": 3501
+    },
+    {
+      "epoch": 0.4901329601119664,
+      "grad_norm": 0.4233529044212103,
+      "learning_rate": 2.7026436283488583e-05,
+      "loss": 0.5673,
+      "step": 3502
+    },
+    {
+      "epoch": 0.49027291812456264,
+      "grad_norm": 0.4082847133569342,
+      "learning_rate": 2.7015140061847365e-05,
+      "loss": 0.513,
+      "step": 3503
+    },
+    {
+      "epoch": 0.49041287613715884,
+      "grad_norm": 0.4200304747170798,
+      "learning_rate": 2.7003843426074416e-05,
+      "loss": 0.5791,
+      "step": 3504
+    },
+    {
+      "epoch": 0.4905528341497551,
+      "grad_norm": 0.39533501738229987,
+      "learning_rate": 2.6992546378491318e-05,
+      "loss": 0.5704,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4906927921623513,
+      "grad_norm": 0.41764209923224177,
+      "learning_rate": 2.698124892141971e-05,
+      "loss": 0.5457,
+      "step": 3506
+    },
+    {
+      "epoch": 0.49083275017494754,
+      "grad_norm": 0.4064990179760341,
+      "learning_rate": 2.6969951057181358e-05,
+      "loss": 0.5574,
+      "step": 3507
+    },
+    {
+      "epoch": 0.49097270818754374,
+      "grad_norm": 0.420505722395169,
+      "learning_rate": 2.695865278809807e-05,
+      "loss": 0.5699,
+      "step": 3508
+    },
+    {
+      "epoch": 0.49111266620013994,
+      "grad_norm": 0.40016382436365205,
+      "learning_rate": 2.6947354116491763e-05,
+      "loss": 0.5567,
+      "step": 3509
+    },
+    {
+      "epoch": 0.4912526242127362,
+      "grad_norm": 0.4021262284998323,
+      "learning_rate": 2.693605504468443e-05,
+      "loss": 0.5676,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4913925822253324,
+      "grad_norm": 0.39530073122269704,
+      "learning_rate": 2.692475557499813e-05,
+      "loss": 0.5328,
+      "step": 3511
+    },
+    {
+      "epoch": 0.49153254023792864,
+      "grad_norm": 0.397282341609004,
+      "learning_rate": 2.6913455709755042e-05,
+      "loss": 0.5567,
+      "step": 3512
+    },
+    {
+      "epoch": 0.49167249825052484,
+      "grad_norm": 0.41008312840461486,
+      "learning_rate": 2.6902155451277377e-05,
+      "loss": 0.5393,
+      "step": 3513
+    },
+    {
+      "epoch": 0.4918124562631211,
+      "grad_norm": 0.4096259675286922,
+      "learning_rate": 2.6890854801887478e-05,
+      "loss": 0.5951,
+      "step": 3514
+    },
+    {
+      "epoch": 0.4919524142757173,
+      "grad_norm": 0.3919817250852191,
+      "learning_rate": 2.6879553763907726e-05,
+      "loss": 0.536,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4920923722883135,
+      "grad_norm": 0.39561344260712744,
+      "learning_rate": 2.686825233966061e-05,
+      "loss": 0.5536,
+      "step": 3516
+    },
+    {
+      "epoch": 0.49223233030090974,
+      "grad_norm": 0.409797831934292,
+      "learning_rate": 2.685695053146868e-05,
+      "loss": 0.5395,
+      "step": 3517
+    },
+    {
+      "epoch": 0.49237228831350593,
+      "grad_norm": 0.3998242277512286,
+      "learning_rate": 2.684564834165457e-05,
+      "loss": 0.5404,
+      "step": 3518
+    },
+    {
+      "epoch": 0.4925122463261022,
+      "grad_norm": 0.4267398317265081,
+      "learning_rate": 2.6834345772541002e-05,
+      "loss": 0.5711,
+      "step": 3519
+    },
+    {
+      "epoch": 0.4926522043386984,
+      "grad_norm": 0.42572746496869157,
+      "learning_rate": 2.6823042826450774e-05,
+      "loss": 0.5921,
+      "step": 3520
+    },
+    {
+      "epoch": 0.49279216235129464,
+      "grad_norm": 0.400101008335866,
+      "learning_rate": 2.681173950570674e-05,
+      "loss": 0.5325,
+      "step": 3521
+    },
+    {
+      "epoch": 0.49293212036389084,
+      "grad_norm": 0.4029901094408229,
+      "learning_rate": 2.6800435812631854e-05,
+      "loss": 0.5678,
+      "step": 3522
+    },
+    {
+      "epoch": 0.49307207837648703,
+      "grad_norm": 0.41541374559781874,
+      "learning_rate": 2.678913174954914e-05,
+      "loss": 0.5285,
+      "step": 3523
+    },
+    {
+      "epoch": 0.4932120363890833,
+      "grad_norm": 0.4370900148410592,
+      "learning_rate": 2.6777827318781697e-05,
+      "loss": 0.5666,
+      "step": 3524
+    },
+    {
+      "epoch": 0.4933519944016795,
+      "grad_norm": 0.42959019276244964,
+      "learning_rate": 2.6766522522652704e-05,
+      "loss": 0.5402,
+      "step": 3525
+    },
+    {
+      "epoch": 0.49349195241427574,
+      "grad_norm": 0.40725453736472333,
+      "learning_rate": 2.6755217363485406e-05,
+      "loss": 0.5171,
+      "step": 3526
+    },
+    {
+      "epoch": 0.49363191042687193,
+      "grad_norm": 0.40719701171899414,
+      "learning_rate": 2.674391184360313e-05,
+      "loss": 0.5283,
+      "step": 3527
+    },
+    {
+      "epoch": 0.4937718684394682,
+      "grad_norm": 0.40279902099076637,
+      "learning_rate": 2.6732605965329283e-05,
+      "loss": 0.5477,
+      "step": 3528
+    },
+    {
+      "epoch": 0.4939118264520644,
+      "grad_norm": 0.3802988661016701,
+      "learning_rate": 2.6721299730987324e-05,
+      "loss": 0.548,
+      "step": 3529
+    },
+    {
+      "epoch": 0.4940517844646606,
+      "grad_norm": 0.42211573921030193,
+      "learning_rate": 2.670999314290081e-05,
+      "loss": 0.5365,
+      "step": 3530
+    },
+    {
+      "epoch": 0.49419174247725683,
+      "grad_norm": 0.4001947861590466,
+      "learning_rate": 2.6698686203393354e-05,
+      "loss": 0.5495,
+      "step": 3531
+    },
+    {
+      "epoch": 0.49433170048985303,
+      "grad_norm": 0.4075041830130491,
+      "learning_rate": 2.6687378914788645e-05,
+      "loss": 0.5598,
+      "step": 3532
+    },
+    {
+      "epoch": 0.4944716585024493,
+      "grad_norm": 0.42826625295627585,
+      "learning_rate": 2.6676071279410448e-05,
+      "loss": 0.546,
+      "step": 3533
+    },
+    {
+      "epoch": 0.4946116165150455,
+      "grad_norm": 0.4325314823046203,
+      "learning_rate": 2.6664763299582602e-05,
+      "loss": 0.5549,
+      "step": 3534
+    },
+    {
+      "epoch": 0.4947515745276417,
+      "grad_norm": 0.4257334322531565,
+      "learning_rate": 2.6653454977629e-05,
+      "loss": 0.5714,
+      "step": 3535
+    },
+    {
+      "epoch": 0.49489153254023793,
+      "grad_norm": 0.41145085533484926,
+      "learning_rate": 2.6642146315873622e-05,
+      "loss": 0.5637,
+      "step": 3536
+    },
+    {
+      "epoch": 0.49503149055283413,
+      "grad_norm": 0.41951469265452185,
+      "learning_rate": 2.6630837316640523e-05,
+      "loss": 0.5531,
+      "step": 3537
+    },
+    {
+      "epoch": 0.4951714485654304,
+      "grad_norm": 0.42251053392716215,
+      "learning_rate": 2.6619527982253794e-05,
+      "loss": 0.5631,
+      "step": 3538
+    },
+    {
+      "epoch": 0.4953114065780266,
+      "grad_norm": 0.4071751122281306,
+      "learning_rate": 2.6608218315037648e-05,
+      "loss": 0.5791,
+      "step": 3539
+    },
+    {
+      "epoch": 0.49545136459062283,
+      "grad_norm": 0.41230607754128246,
+      "learning_rate": 2.659690831731631e-05,
+      "loss": 0.533,
+      "step": 3540
+    },
+    {
+      "epoch": 0.49559132260321903,
+      "grad_norm": 0.4049581094697219,
+      "learning_rate": 2.6585597991414114e-05,
+      "loss": 0.5314,
+      "step": 3541
+    },
+    {
+      "epoch": 0.49573128061581523,
+      "grad_norm": 0.38934941480909324,
+      "learning_rate": 2.6574287339655447e-05,
+      "loss": 0.5359,
+      "step": 3542
+    },
+    {
+      "epoch": 0.4958712386284115,
+      "grad_norm": 0.4183426624654952,
+      "learning_rate": 2.656297636436475e-05,
+      "loss": 0.5885,
+      "step": 3543
+    },
+    {
+      "epoch": 0.4960111966410077,
+      "grad_norm": 0.4426159573559465,
+      "learning_rate": 2.6551665067866556e-05,
+      "loss": 0.5809,
+      "step": 3544
+    },
+    {
+      "epoch": 0.49615115465360393,
+      "grad_norm": 0.5149102513551677,
+      "learning_rate": 2.6540353452485443e-05,
+      "loss": 0.5692,
+      "step": 3545
+    },
+    {
+      "epoch": 0.49629111266620013,
+      "grad_norm": 0.4109744615781001,
+      "learning_rate": 2.652904152054607e-05,
+      "loss": 0.5789,
+      "step": 3546
+    },
+    {
+      "epoch": 0.4964310706787964,
+      "grad_norm": 0.44177673319774247,
+      "learning_rate": 2.651772927437315e-05,
+      "loss": 0.5912,
+      "step": 3547
+    },
+    {
+      "epoch": 0.4965710286913926,
+      "grad_norm": 0.4052859032621841,
+      "learning_rate": 2.6506416716291465e-05,
+      "loss": 0.5616,
+      "step": 3548
+    },
+    {
+      "epoch": 0.4967109867039888,
+      "grad_norm": 0.393563528563957,
+      "learning_rate": 2.649510384862586e-05,
+      "loss": 0.5379,
+      "step": 3549
+    },
+    {
+      "epoch": 0.49685094471658503,
+      "grad_norm": 0.41866504482839206,
+      "learning_rate": 2.6483790673701242e-05,
+      "loss": 0.5486,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4969909027291812,
+      "grad_norm": 0.41154455431337716,
+      "learning_rate": 2.6472477193842583e-05,
+      "loss": 0.5647,
+      "step": 3551
+    },
+    {
+      "epoch": 0.4971308607417775,
+      "grad_norm": 0.39736456239087703,
+      "learning_rate": 2.6461163411374923e-05,
+      "loss": 0.5901,
+      "step": 3552
+    },
+    {
+      "epoch": 0.4972708187543737,
+      "grad_norm": 0.4060568204824789,
+      "learning_rate": 2.6449849328623355e-05,
+      "loss": 0.5435,
+      "step": 3553
+    },
+    {
+      "epoch": 0.49741077676696993,
+      "grad_norm": 0.4115697803723554,
+      "learning_rate": 2.6438534947913047e-05,
+      "loss": 0.5655,
+      "step": 3554
+    },
+    {
+      "epoch": 0.49755073477956613,
+      "grad_norm": 0.3956194134720466,
+      "learning_rate": 2.6427220271569203e-05,
+      "loss": 0.5394,
+      "step": 3555
+    },
+    {
+      "epoch": 0.4976906927921623,
+      "grad_norm": 0.38712507902856763,
+      "learning_rate": 2.6415905301917114e-05,
+      "loss": 0.5345,
+      "step": 3556
+    },
+    {
+      "epoch": 0.4978306508047586,
+      "grad_norm": 0.9891507537143042,
+      "learning_rate": 2.6404590041282116e-05,
+      "loss": 0.5452,
+      "step": 3557
+    },
+    {
+      "epoch": 0.4979706088173548,
+      "grad_norm": 0.40961536120801306,
+      "learning_rate": 2.6393274491989617e-05,
+      "loss": 0.5193,
+      "step": 3558
+    },
+    {
+      "epoch": 0.49811056682995103,
+      "grad_norm": 0.43943809296308844,
+      "learning_rate": 2.6381958656365073e-05,
+      "loss": 0.5351,
+      "step": 3559
+    },
+    {
+      "epoch": 0.4982505248425472,
+      "grad_norm": 0.4108272260762031,
+      "learning_rate": 2.6370642536734004e-05,
+      "loss": 0.5348,
+      "step": 3560
+    },
+    {
+      "epoch": 0.4983904828551435,
+      "grad_norm": 0.4119616541998687,
+      "learning_rate": 2.6359326135421986e-05,
+      "loss": 0.5255,
+      "step": 3561
+    },
+    {
+      "epoch": 0.4985304408677397,
+      "grad_norm": 0.4495829632664617,
+      "learning_rate": 2.6348009454754653e-05,
+      "loss": 0.5828,
+      "step": 3562
+    },
+    {
+      "epoch": 0.4986703988803359,
+      "grad_norm": 0.41137247203721966,
+      "learning_rate": 2.6336692497057696e-05,
+      "loss": 0.5541,
+      "step": 3563
+    },
+    {
+      "epoch": 0.4988103568929321,
+      "grad_norm": 0.4367447061362204,
+      "learning_rate": 2.632537526465687e-05,
+      "loss": 0.5993,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4989503149055283,
+      "grad_norm": 0.4326908998762443,
+      "learning_rate": 2.6314057759877985e-05,
+      "loss": 0.5765,
+      "step": 3565
+    },
+    {
+      "epoch": 0.4990902729181246,
+      "grad_norm": 0.4192288623903318,
+      "learning_rate": 2.6302739985046898e-05,
+      "loss": 0.5376,
+      "step": 3566
+    },
+    {
+      "epoch": 0.4992302309307208,
+      "grad_norm": 0.4248836776012947,
+      "learning_rate": 2.629142194248952e-05,
+      "loss": 0.5938,
+      "step": 3567
+    },
+    {
+      "epoch": 0.49937018894331703,
+      "grad_norm": 0.4047565365438917,
+      "learning_rate": 2.6280103634531833e-05,
+      "loss": 0.5811,
+      "step": 3568
+    },
+    {
+      "epoch": 0.4995101469559132,
+      "grad_norm": 0.4144316245441231,
+      "learning_rate": 2.626878506349986e-05,
+      "loss": 0.5902,
+      "step": 3569
+    },
+    {
+      "epoch": 0.4996501049685094,
+      "grad_norm": 0.4178965245669895,
+      "learning_rate": 2.625746623171968e-05,
+      "loss": 0.525,
+      "step": 3570
+    },
+    {
+      "epoch": 0.4997900629811057,
+      "grad_norm": 0.4159969957434166,
+      "learning_rate": 2.624614714151743e-05,
+      "loss": 0.5576,
+      "step": 3571
+    },
+    {
+      "epoch": 0.4999300209937019,
+      "grad_norm": 0.40173942508061367,
+      "learning_rate": 2.6234827795219297e-05,
+      "loss": 0.5358,
+      "step": 3572
+    },
+    {
+      "epoch": 0.5000699790062981,
+      "grad_norm": 0.41640033764604484,
+      "learning_rate": 2.622350819515153e-05,
+      "loss": 0.5571,
+      "step": 3573
+    },
+    {
+      "epoch": 0.5002099370188944,
+      "grad_norm": 0.40378307872853386,
+      "learning_rate": 2.621218834364041e-05,
+      "loss": 0.5472,
+      "step": 3574
+    },
+    {
+      "epoch": 0.5003498950314905,
+      "grad_norm": 0.398494511998799,
+      "learning_rate": 2.620086824301229e-05,
+      "loss": 0.5331,
+      "step": 3575
+    },
+    {
+      "epoch": 0.5004898530440868,
+      "grad_norm": 0.40959256585724063,
+      "learning_rate": 2.6189547895593562e-05,
+      "loss": 0.5496,
+      "step": 3576
+    },
+    {
+      "epoch": 0.500629811056683,
+      "grad_norm": 0.41352440579598176,
+      "learning_rate": 2.6178227303710673e-05,
+      "loss": 0.5697,
+      "step": 3577
+    },
+    {
+      "epoch": 0.5007697690692792,
+      "grad_norm": 0.408240609494089,
+      "learning_rate": 2.616690646969011e-05,
+      "loss": 0.56,
+      "step": 3578
+    },
+    {
+      "epoch": 0.5009097270818754,
+      "grad_norm": 0.3832170219371199,
+      "learning_rate": 2.6155585395858435e-05,
+      "loss": 0.523,
+      "step": 3579
+    },
+    {
+      "epoch": 0.5010496850944717,
+      "grad_norm": 0.405736469986696,
+      "learning_rate": 2.6144264084542224e-05,
+      "loss": 0.5743,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5011896431070679,
+      "grad_norm": 0.4036128469748438,
+      "learning_rate": 2.6132942538068146e-05,
+      "loss": 0.5807,
+      "step": 3581
+    },
+    {
+      "epoch": 0.5013296011196641,
+      "grad_norm": 0.40173202764431026,
+      "learning_rate": 2.6121620758762877e-05,
+      "loss": 0.5703,
+      "step": 3582
+    },
+    {
+      "epoch": 0.5014695591322603,
+      "grad_norm": 0.42011031965776624,
+      "learning_rate": 2.6110298748953153e-05,
+      "loss": 0.5602,
+      "step": 3583
+    },
+    {
+      "epoch": 0.5016095171448566,
+      "grad_norm": 0.39388389881683084,
+      "learning_rate": 2.6098976510965788e-05,
+      "loss": 0.5268,
+      "step": 3584
+    },
+    {
+      "epoch": 0.5017494751574527,
+      "grad_norm": 0.5531371102734028,
+      "learning_rate": 2.6087654047127587e-05,
+      "loss": 0.5644,
+      "step": 3585
+    },
+    {
+      "epoch": 0.501889433170049,
+      "grad_norm": 0.40679492045140125,
+      "learning_rate": 2.6076331359765448e-05,
+      "loss": 0.5047,
+      "step": 3586
+    },
+    {
+      "epoch": 0.5020293911826452,
+      "grad_norm": 0.4165461319341358,
+      "learning_rate": 2.6065008451206296e-05,
+      "loss": 0.5296,
+      "step": 3587
+    },
+    {
+      "epoch": 0.5021693491952415,
+      "grad_norm": 0.39190580835869326,
+      "learning_rate": 2.60536853237771e-05,
+      "loss": 0.5346,
+      "step": 3588
+    },
+    {
+      "epoch": 0.5023093072078376,
+      "grad_norm": 0.3991014002270092,
+      "learning_rate": 2.6042361979804874e-05,
+      "loss": 0.5006,
+      "step": 3589
+    },
+    {
+      "epoch": 0.5024492652204339,
+      "grad_norm": 0.4371615498576483,
+      "learning_rate": 2.6031038421616683e-05,
+      "loss": 0.5738,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5025892232330301,
+      "grad_norm": 0.42427696712386304,
+      "learning_rate": 2.6019714651539646e-05,
+      "loss": 0.5857,
+      "step": 3591
+    },
+    {
+      "epoch": 0.5027291812456263,
+      "grad_norm": 0.4155090445862282,
+      "learning_rate": 2.600839067190089e-05,
+      "loss": 0.584,
+      "step": 3592
+    },
+    {
+      "epoch": 0.5028691392582225,
+      "grad_norm": 0.4233864118860884,
+      "learning_rate": 2.5997066485027626e-05,
+      "loss": 0.5393,
+      "step": 3593
+    },
+    {
+      "epoch": 0.5030090972708188,
+      "grad_norm": 0.4286876156223743,
+      "learning_rate": 2.5985742093247078e-05,
+      "loss": 0.5787,
+      "step": 3594
+    },
+    {
+      "epoch": 0.503149055283415,
+      "grad_norm": 0.4103875069797372,
+      "learning_rate": 2.5974417498886532e-05,
+      "loss": 0.5617,
+      "step": 3595
+    },
+    {
+      "epoch": 0.5032890132960112,
+      "grad_norm": 0.43666777401439383,
+      "learning_rate": 2.59630927042733e-05,
+      "loss": 0.6215,
+      "step": 3596
+    },
+    {
+      "epoch": 0.5034289713086074,
+      "grad_norm": 0.3967739646837927,
+      "learning_rate": 2.5951767711734753e-05,
+      "loss": 0.5521,
+      "step": 3597
+    },
+    {
+      "epoch": 0.5035689293212037,
+      "grad_norm": 0.4134700097976131,
+      "learning_rate": 2.594044252359828e-05,
+      "loss": 0.5564,
+      "step": 3598
+    },
+    {
+      "epoch": 0.5037088873337998,
+      "grad_norm": 0.4193374425121163,
+      "learning_rate": 2.592911714219132e-05,
+      "loss": 0.5436,
+      "step": 3599
+    },
+    {
+      "epoch": 0.5038488453463961,
+      "grad_norm": 0.41979302117803585,
+      "learning_rate": 2.591779156984137e-05,
+      "loss": 0.5439,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5039888033589923,
+      "grad_norm": 0.40958102976844213,
+      "learning_rate": 2.590646580887593e-05,
+      "loss": 0.5436,
+      "step": 3601
+    },
+    {
+      "epoch": 0.5041287613715886,
+      "grad_norm": 0.4201739272976801,
+      "learning_rate": 2.589513986162258e-05,
+      "loss": 0.5466,
+      "step": 3602
+    },
+    {
+      "epoch": 0.5042687193841847,
+      "grad_norm": 0.4258535420468519,
+      "learning_rate": 2.5883813730408894e-05,
+      "loss": 0.58,
+      "step": 3603
+    },
+    {
+      "epoch": 0.504408677396781,
+      "grad_norm": 0.41534287545104986,
+      "learning_rate": 2.587248741756253e-05,
+      "loss": 0.576,
+      "step": 3604
+    },
+    {
+      "epoch": 0.5045486354093772,
+      "grad_norm": 0.42287342650084503,
+      "learning_rate": 2.5861160925411138e-05,
+      "loss": 0.5832,
+      "step": 3605
+    },
+    {
+      "epoch": 0.5046885934219734,
+      "grad_norm": 0.4501273841496384,
+      "learning_rate": 2.5849834256282447e-05,
+      "loss": 0.6017,
+      "step": 3606
+    },
+    {
+      "epoch": 0.5048285514345696,
+      "grad_norm": 0.4274714205267487,
+      "learning_rate": 2.5838507412504187e-05,
+      "loss": 0.5484,
+      "step": 3607
+    },
+    {
+      "epoch": 0.5049685094471659,
+      "grad_norm": 0.3888114387545627,
+      "learning_rate": 2.5827180396404156e-05,
+      "loss": 0.5101,
+      "step": 3608
+    },
+    {
+      "epoch": 0.5051084674597621,
+      "grad_norm": 0.43140565124104524,
+      "learning_rate": 2.5815853210310152e-05,
+      "loss": 0.5742,
+      "step": 3609
+    },
+    {
+      "epoch": 0.5052484254723583,
+      "grad_norm": 0.41280804310487207,
+      "learning_rate": 2.580452585655004e-05,
+      "loss": 0.5187,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5053883834849545,
+      "grad_norm": 0.4010119649911929,
+      "learning_rate": 2.5793198337451696e-05,
+      "loss": 0.5268,
+      "step": 3611
+    },
+    {
+      "epoch": 0.5055283414975508,
+      "grad_norm": 0.43498896699155226,
+      "learning_rate": 2.5781870655343045e-05,
+      "loss": 0.5975,
+      "step": 3612
+    },
+    {
+      "epoch": 0.5056682995101469,
+      "grad_norm": 0.3871046921614993,
+      "learning_rate": 2.5770542812552047e-05,
+      "loss": 0.5777,
+      "step": 3613
+    },
+    {
+      "epoch": 0.5058082575227432,
+      "grad_norm": 0.3836731432807253,
+      "learning_rate": 2.5759214811406678e-05,
+      "loss": 0.5591,
+      "step": 3614
+    },
+    {
+      "epoch": 0.5059482155353394,
+      "grad_norm": 0.4186045435948251,
+      "learning_rate": 2.5747886654234967e-05,
+      "loss": 0.5929,
+      "step": 3615
+    },
+    {
+      "epoch": 0.5060881735479357,
+      "grad_norm": 0.3989221255230432,
+      "learning_rate": 2.5736558343364953e-05,
+      "loss": 0.5484,
+      "step": 3616
+    },
+    {
+      "epoch": 0.5062281315605318,
+      "grad_norm": 0.3802059534422492,
+      "learning_rate": 2.5725229881124734e-05,
+      "loss": 0.5194,
+      "step": 3617
+    },
+    {
+      "epoch": 0.5063680895731281,
+      "grad_norm": 0.40770746004051617,
+      "learning_rate": 2.5713901269842404e-05,
+      "loss": 0.5294,
+      "step": 3618
+    },
+    {
+      "epoch": 0.5065080475857243,
+      "grad_norm": 0.4179059282812127,
+      "learning_rate": 2.5702572511846134e-05,
+      "loss": 0.56,
+      "step": 3619
+    },
+    {
+      "epoch": 0.5066480055983205,
+      "grad_norm": 0.402518977593649,
+      "learning_rate": 2.569124360946407e-05,
+      "loss": 0.5518,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5067879636109167,
+      "grad_norm": 0.40070397027204235,
+      "learning_rate": 2.5679914565024443e-05,
+      "loss": 0.5229,
+      "step": 3621
+    },
+    {
+      "epoch": 0.506927921623513,
+      "grad_norm": 0.41261781341428777,
+      "learning_rate": 2.5668585380855475e-05,
+      "loss": 0.5716,
+      "step": 3622
+    },
+    {
+      "epoch": 0.5070678796361092,
+      "grad_norm": 0.4086792886609616,
+      "learning_rate": 2.5657256059285417e-05,
+      "loss": 0.5469,
+      "step": 3623
+    },
+    {
+      "epoch": 0.5072078376487054,
+      "grad_norm": 0.41885939317076915,
+      "learning_rate": 2.564592660264258e-05,
+      "loss": 0.5896,
+      "step": 3624
+    },
+    {
+      "epoch": 0.5073477956613016,
+      "grad_norm": 0.3962896649868784,
+      "learning_rate": 2.563459701325526e-05,
+      "loss": 0.5581,
+      "step": 3625
+    },
+    {
+      "epoch": 0.5074877536738979,
+      "grad_norm": 0.39540667903653226,
+      "learning_rate": 2.5623267293451826e-05,
+      "loss": 0.5078,
+      "step": 3626
+    },
+    {
+      "epoch": 0.507627711686494,
+      "grad_norm": 0.42585105817986957,
+      "learning_rate": 2.5611937445560634e-05,
+      "loss": 0.5645,
+      "step": 3627
+    },
+    {
+      "epoch": 0.5077676696990903,
+      "grad_norm": 0.38842241786452814,
+      "learning_rate": 2.5600607471910088e-05,
+      "loss": 0.574,
+      "step": 3628
+    },
+    {
+      "epoch": 0.5079076277116865,
+      "grad_norm": 0.4605380093032065,
+      "learning_rate": 2.5589277374828613e-05,
+      "loss": 0.5743,
+      "step": 3629
+    },
+    {
+      "epoch": 0.5080475857242828,
+      "grad_norm": 0.4103790649353433,
+      "learning_rate": 2.5577947156644655e-05,
+      "loss": 0.5652,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5081875437368789,
+      "grad_norm": 0.4037333605021007,
+      "learning_rate": 2.55666168196867e-05,
+      "loss": 0.5374,
+      "step": 3631
+    },
+    {
+      "epoch": 0.5083275017494752,
+      "grad_norm": 0.40487097389779486,
+      "learning_rate": 2.5555286366283237e-05,
+      "loss": 0.5631,
+      "step": 3632
+    },
+    {
+      "epoch": 0.5084674597620714,
+      "grad_norm": 0.4082664258736793,
+      "learning_rate": 2.5543955798762798e-05,
+      "loss": 0.5598,
+      "step": 3633
+    },
+    {
+      "epoch": 0.5086074177746676,
+      "grad_norm": 0.41864775722779846,
+      "learning_rate": 2.553262511945391e-05,
+      "loss": 0.5544,
+      "step": 3634
+    },
+    {
+      "epoch": 0.5087473757872638,
+      "grad_norm": 0.41778038341235235,
+      "learning_rate": 2.5521294330685165e-05,
+      "loss": 0.545,
+      "step": 3635
+    },
+    {
+      "epoch": 0.5088873337998601,
+      "grad_norm": 0.3965960636107479,
+      "learning_rate": 2.550996343478514e-05,
+      "loss": 0.5629,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5090272918124563,
+      "grad_norm": 0.39741586166804516,
+      "learning_rate": 2.5498632434082452e-05,
+      "loss": 0.5261,
+      "step": 3637
+    },
+    {
+      "epoch": 0.5091672498250525,
+      "grad_norm": 0.4036662033874021,
+      "learning_rate": 2.5487301330905745e-05,
+      "loss": 0.5913,
+      "step": 3638
+    },
+    {
+      "epoch": 0.5093072078376487,
+      "grad_norm": 0.41863388691988607,
+      "learning_rate": 2.5475970127583666e-05,
+      "loss": 0.5442,
+      "step": 3639
+    },
+    {
+      "epoch": 0.509447165850245,
+      "grad_norm": 0.4135621919846456,
+      "learning_rate": 2.5464638826444904e-05,
+      "loss": 0.5563,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5095871238628411,
+      "grad_norm": 0.4351748088714143,
+      "learning_rate": 2.545330742981814e-05,
+      "loss": 0.5551,
+      "step": 3641
+    },
+    {
+      "epoch": 0.5097270818754374,
+      "grad_norm": 0.4234047595420709,
+      "learning_rate": 2.544197594003211e-05,
+      "loss": 0.5876,
+      "step": 3642
+    },
+    {
+      "epoch": 0.5098670398880336,
+      "grad_norm": 0.4191593238313807,
+      "learning_rate": 2.5430644359415535e-05,
+      "loss": 0.5505,
+      "step": 3643
+    },
+    {
+      "epoch": 0.5100069979006299,
+      "grad_norm": 0.4163770493254072,
+      "learning_rate": 2.5419312690297176e-05,
+      "loss": 0.5234,
+      "step": 3644
+    },
+    {
+      "epoch": 0.510146955913226,
+      "grad_norm": 0.4326672484948109,
+      "learning_rate": 2.54079809350058e-05,
+      "loss": 0.5565,
+      "step": 3645
+    },
+    {
+      "epoch": 0.5102869139258223,
+      "grad_norm": 0.4293727352297737,
+      "learning_rate": 2.5396649095870202e-05,
+      "loss": 0.5651,
+      "step": 3646
+    },
+    {
+      "epoch": 0.5104268719384185,
+      "grad_norm": 0.3928814112239087,
+      "learning_rate": 2.5385317175219193e-05,
+      "loss": 0.5223,
+      "step": 3647
+    },
+    {
+      "epoch": 0.5105668299510147,
+      "grad_norm": 0.39498197465504736,
+      "learning_rate": 2.5373985175381594e-05,
+      "loss": 0.546,
+      "step": 3648
+    },
+    {
+      "epoch": 0.5107067879636109,
+      "grad_norm": 0.3934270157464578,
+      "learning_rate": 2.5362653098686245e-05,
+      "loss": 0.532,
+      "step": 3649
+    },
+    {
+      "epoch": 0.5108467459762072,
+      "grad_norm": 0.41869564796385766,
+      "learning_rate": 2.5351320947462005e-05,
+      "loss": 0.5768,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5109867039888033,
+      "grad_norm": 0.42028182516820867,
+      "learning_rate": 2.533998872403775e-05,
+      "loss": 0.5859,
+      "step": 3651
+    },
+    {
+      "epoch": 0.5111266620013996,
+      "grad_norm": 0.39465753139334825,
+      "learning_rate": 2.532865643074236e-05,
+      "loss": 0.5363,
+      "step": 3652
+    },
+    {
+      "epoch": 0.5112666200139958,
+      "grad_norm": 0.3962177186218334,
+      "learning_rate": 2.531732406990474e-05,
+      "loss": 0.5417,
+      "step": 3653
+    },
+    {
+      "epoch": 0.5114065780265921,
+      "grad_norm": 0.4175265234416395,
+      "learning_rate": 2.5305991643853806e-05,
+      "loss": 0.542,
+      "step": 3654
+    },
+    {
+      "epoch": 0.5115465360391882,
+      "grad_norm": 0.4149054282303681,
+      "learning_rate": 2.5294659154918475e-05,
+      "loss": 0.5876,
+      "step": 3655
+    },
+    {
+      "epoch": 0.5116864940517845,
+      "grad_norm": 0.433229070178671,
+      "learning_rate": 2.528332660542771e-05,
+      "loss": 0.5516,
+      "step": 3656
+    },
+    {
+      "epoch": 0.5118264520643807,
+      "grad_norm": 0.4482742289905204,
+      "learning_rate": 2.527199399771044e-05,
+      "loss": 0.5785,
+      "step": 3657
+    },
+    {
+      "epoch": 0.5119664100769769,
+      "grad_norm": 0.41488401148674847,
+      "learning_rate": 2.526066133409566e-05,
+      "loss": 0.5107,
+      "step": 3658
+    },
+    {
+      "epoch": 0.5121063680895731,
+      "grad_norm": 0.3865352913695365,
+      "learning_rate": 2.5249328616912316e-05,
+      "loss": 0.4911,
+      "step": 3659
+    },
+    {
+      "epoch": 0.5122463261021694,
+      "grad_norm": 0.47638789508637014,
+      "learning_rate": 2.523799584848942e-05,
+      "loss": 0.5472,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5123862841147656,
+      "grad_norm": 0.4067671688157654,
+      "learning_rate": 2.5226663031155954e-05,
+      "loss": 0.532,
+      "step": 3661
+    },
+    {
+      "epoch": 0.5125262421273618,
+      "grad_norm": 0.36222693303435777,
+      "learning_rate": 2.5215330167240947e-05,
+      "loss": 0.5085,
+      "step": 3662
+    },
+    {
+      "epoch": 0.512666200139958,
+      "grad_norm": 0.403199238488262,
+      "learning_rate": 2.52039972590734e-05,
+      "loss": 0.5497,
+      "step": 3663
+    },
+    {
+      "epoch": 0.5128061581525543,
+      "grad_norm": 0.4052356626415982,
+      "learning_rate": 2.5192664308982344e-05,
+      "loss": 0.5224,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5129461161651504,
+      "grad_norm": 0.40489338303806344,
+      "learning_rate": 2.5181331319296825e-05,
+      "loss": 0.5621,
+      "step": 3665
+    },
+    {
+      "epoch": 0.5130860741777467,
+      "grad_norm": 0.40195392432260063,
+      "learning_rate": 2.516999829234587e-05,
+      "loss": 0.5221,
+      "step": 3666
+    },
+    {
+      "epoch": 0.5132260321903429,
+      "grad_norm": 0.44319445941617425,
+      "learning_rate": 2.515866523045855e-05,
+      "loss": 0.5607,
+      "step": 3667
+    },
+    {
+      "epoch": 0.5133659902029392,
+      "grad_norm": 0.42549718785093565,
+      "learning_rate": 2.514733213596391e-05,
+      "loss": 0.5678,
+      "step": 3668
+    },
+    {
+      "epoch": 0.5135059482155353,
+      "grad_norm": 0.4012452520954793,
+      "learning_rate": 2.5135999011191018e-05,
+      "loss": 0.574,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5136459062281316,
+      "grad_norm": 0.3988377623943865,
+      "learning_rate": 2.5124665858468954e-05,
+      "loss": 0.5638,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5137858642407278,
+      "grad_norm": 0.4068508250249883,
+      "learning_rate": 2.5113332680126795e-05,
+      "loss": 0.531,
+      "step": 3671
+    },
+    {
+      "epoch": 0.513925822253324,
+      "grad_norm": 0.40607709238953865,
+      "learning_rate": 2.510199947849361e-05,
+      "loss": 0.5555,
+      "step": 3672
+    },
+    {
+      "epoch": 0.5140657802659202,
+      "grad_norm": 0.40363285844425767,
+      "learning_rate": 2.50906662558985e-05,
+      "loss": 0.549,
+      "step": 3673
+    },
+    {
+      "epoch": 0.5142057382785165,
+      "grad_norm": 0.42540315547468743,
+      "learning_rate": 2.507933301467056e-05,
+      "loss": 0.562,
+      "step": 3674
+    },
+    {
+      "epoch": 0.5143456962911127,
+      "grad_norm": 0.41197452211967045,
+      "learning_rate": 2.5067999757138884e-05,
+      "loss": 0.5572,
+      "step": 3675
+    },
+    {
+      "epoch": 0.5144856543037089,
+      "grad_norm": 0.42328084101201563,
+      "learning_rate": 2.505666648563256e-05,
+      "loss": 0.5651,
+      "step": 3676
+    },
+    {
+      "epoch": 0.5146256123163051,
+      "grad_norm": 0.39617586204555455,
+      "learning_rate": 2.5045333202480698e-05,
+      "loss": 0.5505,
+      "step": 3677
+    },
+    {
+      "epoch": 0.5147655703289014,
+      "grad_norm": 0.40450493806111026,
+      "learning_rate": 2.5033999910012414e-05,
+      "loss": 0.5387,
+      "step": 3678
+    },
+    {
+      "epoch": 0.5149055283414975,
+      "grad_norm": 0.41125264241739257,
+      "learning_rate": 2.502266661055679e-05,
+      "loss": 0.5338,
+      "step": 3679
+    },
+    {
+      "epoch": 0.5150454863540938,
+      "grad_norm": 0.41606733051241485,
+      "learning_rate": 2.5011333306442953e-05,
+      "loss": 0.5472,
+      "step": 3680
+    },
+    {
+      "epoch": 0.51518544436669,
+      "grad_norm": 0.3940959299812521,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5264,
+      "step": 3681
+    },
+    {
+      "epoch": 0.5153254023792863,
+      "grad_norm": 0.4160510111032202,
+      "learning_rate": 2.498866669355706e-05,
+      "loss": 0.5365,
+      "step": 3682
+    },
+    {
+      "epoch": 0.5154653603918824,
+      "grad_norm": 0.40780249685180014,
+      "learning_rate": 2.497733338944321e-05,
+      "loss": 0.5137,
+      "step": 3683
+    },
+    {
+      "epoch": 0.5156053184044787,
+      "grad_norm": 0.41930837922293446,
+      "learning_rate": 2.4966000089987596e-05,
+      "loss": 0.549,
+      "step": 3684
+    },
+    {
+      "epoch": 0.5157452764170749,
+      "grad_norm": 0.41175787454056617,
+      "learning_rate": 2.4954666797519305e-05,
+      "loss": 0.5445,
+      "step": 3685
+    },
+    {
+      "epoch": 0.515885234429671,
+      "grad_norm": 0.40112997035094405,
+      "learning_rate": 2.4943333514367448e-05,
+      "loss": 0.5323,
+      "step": 3686
+    },
+    {
+      "epoch": 0.5160251924422673,
+      "grad_norm": 0.41601050967799824,
+      "learning_rate": 2.4932000242861122e-05,
+      "loss": 0.5705,
+      "step": 3687
+    },
+    {
+      "epoch": 0.5161651504548636,
+      "grad_norm": 0.3938594765036972,
+      "learning_rate": 2.4920666985329443e-05,
+      "loss": 0.5147,
+      "step": 3688
+    },
+    {
+      "epoch": 0.5163051084674598,
+      "grad_norm": 0.4187589160579126,
+      "learning_rate": 2.4909333744101503e-05,
+      "loss": 0.5461,
+      "step": 3689
+    },
+    {
+      "epoch": 0.516445066480056,
+      "grad_norm": 0.4447365782109597,
+      "learning_rate": 2.48980005215064e-05,
+      "loss": 0.5388,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5165850244926522,
+      "grad_norm": 0.40572901279081225,
+      "learning_rate": 2.4886667319873214e-05,
+      "loss": 0.5746,
+      "step": 3691
+    },
+    {
+      "epoch": 0.5167249825052485,
+      "grad_norm": 0.3937401384845004,
+      "learning_rate": 2.4875334141531052e-05,
+      "loss": 0.5514,
+      "step": 3692
+    },
+    {
+      "epoch": 0.5168649405178446,
+      "grad_norm": 0.4251082426934148,
+      "learning_rate": 2.486400098880899e-05,
+      "loss": 0.5531,
+      "step": 3693
+    },
+    {
+      "epoch": 0.5170048985304408,
+      "grad_norm": 0.4233026623262488,
+      "learning_rate": 2.4852667864036093e-05,
+      "loss": 0.5698,
+      "step": 3694
+    },
+    {
+      "epoch": 0.5171448565430371,
+      "grad_norm": 0.4184619207216555,
+      "learning_rate": 2.484133476954146e-05,
+      "loss": 0.5445,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5172848145556334,
+      "grad_norm": 0.38758737521657194,
+      "learning_rate": 2.4830001707654134e-05,
+      "loss": 0.5383,
+      "step": 3696
+    },
+    {
+      "epoch": 0.5174247725682295,
+      "grad_norm": 0.41121949288648046,
+      "learning_rate": 2.4818668680703187e-05,
+      "loss": 0.5548,
+      "step": 3697
+    },
+    {
+      "epoch": 0.5175647305808257,
+      "grad_norm": 0.3912926343783076,
+      "learning_rate": 2.4807335691017662e-05,
+      "loss": 0.5993,
+      "step": 3698
+    },
+    {
+      "epoch": 0.517704688593422,
+      "grad_norm": 0.4352916580631968,
+      "learning_rate": 2.479600274092661e-05,
+      "loss": 0.5947,
+      "step": 3699
+    },
+    {
+      "epoch": 0.5178446466060181,
+      "grad_norm": 0.4117712323458744,
+      "learning_rate": 2.4784669832759065e-05,
+      "loss": 0.566,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5179846046186144,
+      "grad_norm": 0.48543517712737916,
+      "learning_rate": 2.4773336968844045e-05,
+      "loss": 0.5217,
+      "step": 3701
+    },
+    {
+      "epoch": 0.5181245626312107,
+      "grad_norm": 0.4201945605915251,
+      "learning_rate": 2.4762004151510584e-05,
+      "loss": 0.5817,
+      "step": 3702
+    },
+    {
+      "epoch": 0.5182645206438069,
+      "grad_norm": 0.40028199720572183,
+      "learning_rate": 2.475067138308769e-05,
+      "loss": 0.5305,
+      "step": 3703
+    },
+    {
+      "epoch": 0.518404478656403,
+      "grad_norm": 0.45130189575312646,
+      "learning_rate": 2.4739338665904356e-05,
+      "loss": 0.5762,
+      "step": 3704
+    },
+    {
+      "epoch": 0.5185444366689993,
+      "grad_norm": 0.40520084504522474,
+      "learning_rate": 2.4728006002289555e-05,
+      "loss": 0.5618,
+      "step": 3705
+    },
+    {
+      "epoch": 0.5186843946815956,
+      "grad_norm": 0.39762527724883795,
+      "learning_rate": 2.4716673394572297e-05,
+      "loss": 0.5519,
+      "step": 3706
+    },
+    {
+      "epoch": 0.5188243526941917,
+      "grad_norm": 0.4303843474015456,
+      "learning_rate": 2.4705340845081527e-05,
+      "loss": 0.56,
+      "step": 3707
+    },
+    {
+      "epoch": 0.518964310706788,
+      "grad_norm": 0.4169723723648604,
+      "learning_rate": 2.4694008356146207e-05,
+      "loss": 0.5436,
+      "step": 3708
+    },
+    {
+      "epoch": 0.5191042687193842,
+      "grad_norm": 0.40055909799700246,
+      "learning_rate": 2.4682675930095263e-05,
+      "loss": 0.536,
+      "step": 3709
+    },
+    {
+      "epoch": 0.5192442267319805,
+      "grad_norm": 0.4109688655223356,
+      "learning_rate": 2.4671343569257647e-05,
+      "loss": 0.5324,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5193841847445766,
+      "grad_norm": 0.4149175387680195,
+      "learning_rate": 2.4660011275962258e-05,
+      "loss": 0.5388,
+      "step": 3711
+    },
+    {
+      "epoch": 0.5195241427571728,
+      "grad_norm": 0.4264139741717757,
+      "learning_rate": 2.4648679052537994e-05,
+      "loss": 0.5625,
+      "step": 3712
+    },
+    {
+      "epoch": 0.5196641007697691,
+      "grad_norm": 0.4289357178860241,
+      "learning_rate": 2.4637346901313757e-05,
+      "loss": 0.5689,
+      "step": 3713
+    },
+    {
+      "epoch": 0.5198040587823652,
+      "grad_norm": 0.42933380182165265,
+      "learning_rate": 2.4626014824618415e-05,
+      "loss": 0.5751,
+      "step": 3714
+    },
+    {
+      "epoch": 0.5199440167949615,
+      "grad_norm": 0.41237458627558815,
+      "learning_rate": 2.461468282478082e-05,
+      "loss": 0.5708,
+      "step": 3715
+    },
+    {
+      "epoch": 0.5200839748075577,
+      "grad_norm": 0.40783837110167326,
+      "learning_rate": 2.46033509041298e-05,
+      "loss": 0.5163,
+      "step": 3716
+    },
+    {
+      "epoch": 0.520223932820154,
+      "grad_norm": 0.3934394614065878,
+      "learning_rate": 2.4592019064994207e-05,
+      "loss": 0.5123,
+      "step": 3717
+    },
+    {
+      "epoch": 0.5203638908327501,
+      "grad_norm": 0.42393320677683094,
+      "learning_rate": 2.4580687309702836e-05,
+      "loss": 0.5459,
+      "step": 3718
+    },
+    {
+      "epoch": 0.5205038488453464,
+      "grad_norm": 0.4201737718371511,
+      "learning_rate": 2.456935564058447e-05,
+      "loss": 0.5484,
+      "step": 3719
+    },
+    {
+      "epoch": 0.5206438068579426,
+      "grad_norm": 0.4037513584821948,
+      "learning_rate": 2.45580240599679e-05,
+      "loss": 0.539,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5207837648705388,
+      "grad_norm": 0.40384281587615306,
+      "learning_rate": 2.4546692570181863e-05,
+      "loss": 0.5594,
+      "step": 3721
+    },
+    {
+      "epoch": 0.520923722883135,
+      "grad_norm": 0.39646345117822984,
+      "learning_rate": 2.453536117355511e-05,
+      "loss": 0.5344,
+      "step": 3722
+    },
+    {
+      "epoch": 0.5210636808957313,
+      "grad_norm": 0.402351444419845,
+      "learning_rate": 2.4524029872416333e-05,
+      "loss": 0.548,
+      "step": 3723
+    },
+    {
+      "epoch": 0.5212036389083275,
+      "grad_norm": 0.4310017935995615,
+      "learning_rate": 2.451269866909426e-05,
+      "loss": 0.5654,
+      "step": 3724
+    },
+    {
+      "epoch": 0.5213435969209237,
+      "grad_norm": 0.39852961133599507,
+      "learning_rate": 2.4501367565917554e-05,
+      "loss": 0.5248,
+      "step": 3725
+    },
+    {
+      "epoch": 0.5214835549335199,
+      "grad_norm": 0.3978940832017032,
+      "learning_rate": 2.4490036565214873e-05,
+      "loss": 0.5423,
+      "step": 3726
+    },
+    {
+      "epoch": 0.5216235129461162,
+      "grad_norm": 0.4032221188907521,
+      "learning_rate": 2.447870566931484e-05,
+      "loss": 0.5625,
+      "step": 3727
+    },
+    {
+      "epoch": 0.5217634709587123,
+      "grad_norm": 0.40124015142391184,
+      "learning_rate": 2.4467374880546095e-05,
+      "loss": 0.5173,
+      "step": 3728
+    },
+    {
+      "epoch": 0.5219034289713086,
+      "grad_norm": 0.419761697749084,
+      "learning_rate": 2.4456044201237215e-05,
+      "loss": 0.577,
+      "step": 3729
+    },
+    {
+      "epoch": 0.5220433869839048,
+      "grad_norm": 0.4068938084974087,
+      "learning_rate": 2.4444713633716765e-05,
+      "loss": 0.5089,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5221833449965011,
+      "grad_norm": 0.4372637791239717,
+      "learning_rate": 2.4433383180313303e-05,
+      "loss": 0.6125,
+      "step": 3731
+    },
+    {
+      "epoch": 0.5223233030090972,
+      "grad_norm": 0.4211367854727307,
+      "learning_rate": 2.4422052843355347e-05,
+      "loss": 0.5837,
+      "step": 3732
+    },
+    {
+      "epoch": 0.5224632610216935,
+      "grad_norm": 0.4138228399759103,
+      "learning_rate": 2.4410722625171396e-05,
+      "loss": 0.5904,
+      "step": 3733
+    },
+    {
+      "epoch": 0.5226032190342897,
+      "grad_norm": 0.43035321865290377,
+      "learning_rate": 2.4399392528089915e-05,
+      "loss": 0.5543,
+      "step": 3734
+    },
+    {
+      "epoch": 0.5227431770468859,
+      "grad_norm": 0.393376380647825,
+      "learning_rate": 2.4388062554439372e-05,
+      "loss": 0.5625,
+      "step": 3735
+    },
+    {
+      "epoch": 0.5228831350594821,
+      "grad_norm": 0.4057572933128106,
+      "learning_rate": 2.4376732706548183e-05,
+      "loss": 0.5386,
+      "step": 3736
+    },
+    {
+      "epoch": 0.5230230930720784,
+      "grad_norm": 0.4127940511172469,
+      "learning_rate": 2.4365402986744738e-05,
+      "loss": 0.5424,
+      "step": 3737
+    },
+    {
+      "epoch": 0.5231630510846746,
+      "grad_norm": 0.43512898556014096,
+      "learning_rate": 2.4354073397357427e-05,
+      "loss": 0.5886,
+      "step": 3738
+    },
+    {
+      "epoch": 0.5233030090972708,
+      "grad_norm": 0.429632898236204,
+      "learning_rate": 2.434274394071459e-05,
+      "loss": 0.5897,
+      "step": 3739
+    },
+    {
+      "epoch": 0.523442967109867,
+      "grad_norm": 0.5024683991705309,
+      "learning_rate": 2.4331414619144537e-05,
+      "loss": 0.5389,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5235829251224633,
+      "grad_norm": 0.426305756774051,
+      "learning_rate": 2.432008543497556e-05,
+      "loss": 0.5319,
+      "step": 3741
+    },
+    {
+      "epoch": 0.5237228831350594,
+      "grad_norm": 0.4067282442774484,
+      "learning_rate": 2.430875639053593e-05,
+      "loss": 0.5364,
+      "step": 3742
+    },
+    {
+      "epoch": 0.5238628411476557,
+      "grad_norm": 0.40523378137952,
+      "learning_rate": 2.4297427488153872e-05,
+      "loss": 0.5512,
+      "step": 3743
+    },
+    {
+      "epoch": 0.5240027991602519,
+      "grad_norm": 0.42309697388292367,
+      "learning_rate": 2.42860987301576e-05,
+      "loss": 0.5832,
+      "step": 3744
+    },
+    {
+      "epoch": 0.5241427571728482,
+      "grad_norm": 0.39749349792214256,
+      "learning_rate": 2.427477011887527e-05,
+      "loss": 0.5453,
+      "step": 3745
+    },
+    {
+      "epoch": 0.5242827151854443,
+      "grad_norm": 0.4128499134441993,
+      "learning_rate": 2.4263441656635053e-05,
+      "loss": 0.5614,
+      "step": 3746
+    },
+    {
+      "epoch": 0.5244226731980406,
+      "grad_norm": 0.4307471936718753,
+      "learning_rate": 2.4252113345765046e-05,
+      "loss": 0.5426,
+      "step": 3747
+    },
+    {
+      "epoch": 0.5245626312106368,
+      "grad_norm": 0.40120347811779566,
+      "learning_rate": 2.4240785188593325e-05,
+      "loss": 0.5562,
+      "step": 3748
+    },
+    {
+      "epoch": 0.524702589223233,
+      "grad_norm": 0.40709801793493633,
+      "learning_rate": 2.4229457187447956e-05,
+      "loss": 0.5326,
+      "step": 3749
+    },
+    {
+      "epoch": 0.5248425472358292,
+      "grad_norm": 0.38783041950630354,
+      "learning_rate": 2.4218129344656958e-05,
+      "loss": 0.5801,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5249825052484255,
+      "grad_norm": 0.41702390350925966,
+      "learning_rate": 2.4206801662548314e-05,
+      "loss": 0.543,
+      "step": 3751
+    },
+    {
+      "epoch": 0.5251224632610217,
+      "grad_norm": 0.40842859770281464,
+      "learning_rate": 2.419547414344997e-05,
+      "loss": 0.5719,
+      "step": 3752
+    },
+    {
+      "epoch": 0.5252624212736179,
+      "grad_norm": 0.40966706938584196,
+      "learning_rate": 2.418414678968985e-05,
+      "loss": 0.5679,
+      "step": 3753
+    },
+    {
+      "epoch": 0.5254023792862141,
+      "grad_norm": 0.4217418453013296,
+      "learning_rate": 2.4172819603595853e-05,
+      "loss": 0.5721,
+      "step": 3754
+    },
+    {
+      "epoch": 0.5255423372988104,
+      "grad_norm": 0.41200690319133665,
+      "learning_rate": 2.4161492587495812e-05,
+      "loss": 0.5681,
+      "step": 3755
+    },
+    {
+      "epoch": 0.5256822953114065,
+      "grad_norm": 0.4307870892864373,
+      "learning_rate": 2.4150165743717556e-05,
+      "loss": 0.5991,
+      "step": 3756
+    },
+    {
+      "epoch": 0.5258222533240028,
+      "grad_norm": 0.3964187463617023,
+      "learning_rate": 2.4138839074588868e-05,
+      "loss": 0.5381,
+      "step": 3757
+    },
+    {
+      "epoch": 0.525962211336599,
+      "grad_norm": 0.41691171273978006,
+      "learning_rate": 2.4127512582437485e-05,
+      "loss": 0.5336,
+      "step": 3758
+    },
+    {
+      "epoch": 0.5261021693491953,
+      "grad_norm": 0.4243161504581818,
+      "learning_rate": 2.4116186269591105e-05,
+      "loss": 0.5466,
+      "step": 3759
+    },
+    {
+      "epoch": 0.5262421273617914,
+      "grad_norm": 0.41746283237341214,
+      "learning_rate": 2.410486013837743e-05,
+      "loss": 0.5292,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5263820853743877,
+      "grad_norm": 0.4258425078942338,
+      "learning_rate": 2.4093534191124072e-05,
+      "loss": 0.597,
+      "step": 3761
+    },
+    {
+      "epoch": 0.5265220433869839,
+      "grad_norm": 0.4082388057932214,
+      "learning_rate": 2.408220843015864e-05,
+      "loss": 0.5604,
+      "step": 3762
+    },
+    {
+      "epoch": 0.5266620013995801,
+      "grad_norm": 0.40439792503437516,
+      "learning_rate": 2.4070882857808678e-05,
+      "loss": 0.5599,
+      "step": 3763
+    },
+    {
+      "epoch": 0.5268019594121763,
+      "grad_norm": 0.403671819972436,
+      "learning_rate": 2.4059557476401726e-05,
+      "loss": 0.5701,
+      "step": 3764
+    },
+    {
+      "epoch": 0.5269419174247726,
+      "grad_norm": 0.4115634599388499,
+      "learning_rate": 2.4048232288265253e-05,
+      "loss": 0.5613,
+      "step": 3765
+    },
+    {
+      "epoch": 0.5270818754373688,
+      "grad_norm": 0.4196728696799674,
+      "learning_rate": 2.40369072957267e-05,
+      "loss": 0.5456,
+      "step": 3766
+    },
+    {
+      "epoch": 0.527221833449965,
+      "grad_norm": 0.45632616155283706,
+      "learning_rate": 2.4025582501113474e-05,
+      "loss": 0.5986,
+      "step": 3767
+    },
+    {
+      "epoch": 0.5273617914625612,
+      "grad_norm": 0.41907920921841135,
+      "learning_rate": 2.4014257906752928e-05,
+      "loss": 0.5691,
+      "step": 3768
+    },
+    {
+      "epoch": 0.5275017494751575,
+      "grad_norm": 0.4593046177587895,
+      "learning_rate": 2.4002933514972383e-05,
+      "loss": 0.5491,
+      "step": 3769
+    },
+    {
+      "epoch": 0.5276417074877536,
+      "grad_norm": 0.4141383704701082,
+      "learning_rate": 2.399160932809911e-05,
+      "loss": 0.5405,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5277816655003499,
+      "grad_norm": 0.40874131996221996,
+      "learning_rate": 2.3980285348460363e-05,
+      "loss": 0.5549,
+      "step": 3771
+    },
+    {
+      "epoch": 0.5279216235129461,
+      "grad_norm": 0.42334424503896645,
+      "learning_rate": 2.3968961578383323e-05,
+      "loss": 0.5596,
+      "step": 3772
+    },
+    {
+      "epoch": 0.5280615815255424,
+      "grad_norm": 0.40408381518564374,
+      "learning_rate": 2.395763802019513e-05,
+      "loss": 0.5541,
+      "step": 3773
+    },
+    {
+      "epoch": 0.5282015395381385,
+      "grad_norm": 0.3892924948143325,
+      "learning_rate": 2.3946314676222905e-05,
+      "loss": 0.5561,
+      "step": 3774
+    },
+    {
+      "epoch": 0.5283414975507348,
+      "grad_norm": 0.38424570502320887,
+      "learning_rate": 2.393499154879371e-05,
+      "loss": 0.5513,
+      "step": 3775
+    },
+    {
+      "epoch": 0.528481455563331,
+      "grad_norm": 0.40542835424755946,
+      "learning_rate": 2.3923668640234558e-05,
+      "loss": 0.5511,
+      "step": 3776
+    },
+    {
+      "epoch": 0.5286214135759272,
+      "grad_norm": 0.43305443591622345,
+      "learning_rate": 2.3912345952872416e-05,
+      "loss": 0.5534,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5287613715885234,
+      "grad_norm": 0.39079289404660855,
+      "learning_rate": 2.3901023489034218e-05,
+      "loss": 0.5693,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5289013296011197,
+      "grad_norm": 0.4993326382660523,
+      "learning_rate": 2.388970125104685e-05,
+      "loss": 0.5665,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5290412876137159,
+      "grad_norm": 0.4013031423647386,
+      "learning_rate": 2.3878379241237136e-05,
+      "loss": 0.5059,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5291812456263121,
+      "grad_norm": 0.4122753651945142,
+      "learning_rate": 2.3867057461931857e-05,
+      "loss": 0.5643,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5293212036389083,
+      "grad_norm": 0.3958098061908628,
+      "learning_rate": 2.3855735915457778e-05,
+      "loss": 0.5783,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5294611616515046,
+      "grad_norm": 0.40430666249800307,
+      "learning_rate": 2.384441460414158e-05,
+      "loss": 0.532,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5296011196641007,
+      "grad_norm": 0.43159094392859726,
+      "learning_rate": 2.383309353030989e-05,
+      "loss": 0.5648,
+      "step": 3784
+    },
+    {
+      "epoch": 0.529741077676697,
+      "grad_norm": 0.4278430956102745,
+      "learning_rate": 2.3821772696289336e-05,
+      "loss": 0.5525,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5298810356892932,
+      "grad_norm": 0.42297623639530246,
+      "learning_rate": 2.3810452104406444e-05,
+      "loss": 0.5557,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5300209937018895,
+      "grad_norm": 0.4194124443087706,
+      "learning_rate": 2.3799131756987716e-05,
+      "loss": 0.5805,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5301609517144856,
+      "grad_norm": 0.42223996989148205,
+      "learning_rate": 2.378781165635959e-05,
+      "loss": 0.553,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5303009097270819,
+      "grad_norm": 0.4051480666509419,
+      "learning_rate": 2.3776491804848474e-05,
+      "loss": 0.5439,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5304408677396781,
+      "grad_norm": 0.6913758982209272,
+      "learning_rate": 2.3765172204780705e-05,
+      "loss": 0.5037,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5305808257522743,
+      "grad_norm": 0.39883852712442486,
+      "learning_rate": 2.375385285848257e-05,
+      "loss": 0.5373,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5307207837648705,
+      "grad_norm": 0.41473538961504164,
+      "learning_rate": 2.3742533768280322e-05,
+      "loss": 0.558,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5308607417774668,
+      "grad_norm": 0.4338240458587132,
+      "learning_rate": 2.3731214936500147e-05,
+      "loss": 0.5703,
+      "step": 3793
+    },
+    {
+      "epoch": 0.531000699790063,
+      "grad_norm": 0.44632810155747676,
+      "learning_rate": 2.3719896365468176e-05,
+      "loss": 0.5428,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5311406578026592,
+      "grad_norm": 0.4693366131147214,
+      "learning_rate": 2.370857805751048e-05,
+      "loss": 0.5601,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5312806158152554,
+      "grad_norm": 0.4328684886571571,
+      "learning_rate": 2.3697260014953108e-05,
+      "loss": 0.5984,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5314205738278517,
+      "grad_norm": 0.4150511950700261,
+      "learning_rate": 2.3685942240122017e-05,
+      "loss": 0.5578,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5315605318404478,
+      "grad_norm": 0.4140437625288529,
+      "learning_rate": 2.3674624735343133e-05,
+      "loss": 0.57,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5317004898530441,
+      "grad_norm": 0.4075001120603276,
+      "learning_rate": 2.3663307502942306e-05,
+      "loss": 0.5392,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5318404478656403,
+      "grad_norm": 0.4132932952092581,
+      "learning_rate": 2.3651990545245356e-05,
+      "loss": 0.5826,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5319804058782366,
+      "grad_norm": 0.4260946689124058,
+      "learning_rate": 2.3640673864578023e-05,
+      "loss": 0.5291,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5321203638908327,
+      "grad_norm": 0.42352605916443814,
+      "learning_rate": 2.3629357463265995e-05,
+      "loss": 0.5619,
+      "step": 3802
+    },
+    {
+      "epoch": 0.532260321903429,
+      "grad_norm": 0.4071157003238009,
+      "learning_rate": 2.361804134363493e-05,
+      "loss": 0.5693,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5324002799160252,
+      "grad_norm": 0.40527112461883336,
+      "learning_rate": 2.360672550801039e-05,
+      "loss": 0.5573,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5325402379286214,
+      "grad_norm": 0.4185227436712144,
+      "learning_rate": 2.359540995871789e-05,
+      "loss": 0.5519,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5326801959412176,
+      "grad_norm": 0.40474881222817827,
+      "learning_rate": 2.3584094698082888e-05,
+      "loss": 0.5416,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5328201539538139,
+      "grad_norm": 0.4035189283812099,
+      "learning_rate": 2.35727797284308e-05,
+      "loss": 0.5386,
+      "step": 3807
+    },
+    {
+      "epoch": 0.53296011196641,
+      "grad_norm": 0.4126483303377044,
+      "learning_rate": 2.3561465052086962e-05,
+      "loss": 0.5513,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5331000699790063,
+      "grad_norm": 0.41908332941234616,
+      "learning_rate": 2.3550150671376644e-05,
+      "loss": 0.598,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5332400279916025,
+      "grad_norm": 0.42507750506666686,
+      "learning_rate": 2.353883658862508e-05,
+      "loss": 0.5289,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5333799860041988,
+      "grad_norm": 0.3908992343447489,
+      "learning_rate": 2.3527522806157422e-05,
+      "loss": 0.5955,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5335199440167949,
+      "grad_norm": 0.4088150381618844,
+      "learning_rate": 2.351620932629877e-05,
+      "loss": 0.5434,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5336599020293912,
+      "grad_norm": 0.4313685833974606,
+      "learning_rate": 2.3504896151374144e-05,
+      "loss": 0.5751,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5337998600419874,
+      "grad_norm": 0.4274490027569291,
+      "learning_rate": 2.349358328370854e-05,
+      "loss": 0.5459,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5339398180545836,
+      "grad_norm": 0.4224342493673025,
+      "learning_rate": 2.3482270725626856e-05,
+      "loss": 0.5372,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5340797760671798,
+      "grad_norm": 0.508038847256669,
+      "learning_rate": 2.3470958479453938e-05,
+      "loss": 0.5545,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5342197340797761,
+      "grad_norm": 0.40285477548710635,
+      "learning_rate": 2.345964654751456e-05,
+      "loss": 0.5348,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5343596920923723,
+      "grad_norm": 0.4204468078785021,
+      "learning_rate": 2.3448334932133446e-05,
+      "loss": 0.5873,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5344996501049685,
+      "grad_norm": 0.3926665806611564,
+      "learning_rate": 2.3437023635635254e-05,
+      "loss": 0.5856,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5346396081175647,
+      "grad_norm": 0.4281667973965022,
+      "learning_rate": 2.342571266034456e-05,
+      "loss": 0.5829,
+      "step": 3820
+    },
+    {
+      "epoch": 0.534779566130161,
+      "grad_norm": 0.4235798408466391,
+      "learning_rate": 2.3414402008585888e-05,
+      "loss": 0.5345,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5349195241427571,
+      "grad_norm": 0.41560783094264714,
+      "learning_rate": 2.34030916826837e-05,
+      "loss": 0.5745,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5350594821553534,
+      "grad_norm": 0.4120864111860111,
+      "learning_rate": 2.3391781684962368e-05,
+      "loss": 0.5304,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5351994401679496,
+      "grad_norm": 0.3998366908738693,
+      "learning_rate": 2.3380472017746202e-05,
+      "loss": 0.5579,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5353393981805459,
+      "grad_norm": 0.4078921865892698,
+      "learning_rate": 2.3369162683359486e-05,
+      "loss": 0.5532,
+      "step": 3825
+    },
+    {
+      "epoch": 0.535479356193142,
+      "grad_norm": 0.42102589317173017,
+      "learning_rate": 2.3357853684126384e-05,
+      "loss": 0.5367,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5356193142057383,
+      "grad_norm": 0.4303168029475006,
+      "learning_rate": 2.3346545022371015e-05,
+      "loss": 0.5275,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5357592722183345,
+      "grad_norm": 0.4084962687065381,
+      "learning_rate": 2.3335236700417404e-05,
+      "loss": 0.5643,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5358992302309307,
+      "grad_norm": 0.42704269523521693,
+      "learning_rate": 2.3323928720589555e-05,
+      "loss": 0.5145,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5360391882435269,
+      "grad_norm": 0.4251581534304365,
+      "learning_rate": 2.331262108521136e-05,
+      "loss": 0.5768,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5361791462561232,
+      "grad_norm": 0.4296437560622346,
+      "learning_rate": 2.3301313796606652e-05,
+      "loss": 0.5804,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5363191042687194,
+      "grad_norm": 0.4160558480550044,
+      "learning_rate": 2.3290006857099194e-05,
+      "loss": 0.5618,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5364590622813156,
+      "grad_norm": 0.43852307849271205,
+      "learning_rate": 2.327870026901268e-05,
+      "loss": 0.582,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5365990202939118,
+      "grad_norm": 0.42358328346140833,
+      "learning_rate": 2.326739403467073e-05,
+      "loss": 0.5265,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5367389783065081,
+      "grad_norm": 0.4344046385915696,
+      "learning_rate": 2.3256088156396868e-05,
+      "loss": 0.5855,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5368789363191042,
+      "grad_norm": 0.4318461945295688,
+      "learning_rate": 2.3244782636514596e-05,
+      "loss": 0.5665,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5370188943317005,
+      "grad_norm": 0.4129881442304272,
+      "learning_rate": 2.3233477477347305e-05,
+      "loss": 0.5313,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5371588523442967,
+      "grad_norm": 0.4277048919308099,
+      "learning_rate": 2.3222172681218302e-05,
+      "loss": 0.5632,
+      "step": 3838
+    },
+    {
+      "epoch": 0.537298810356893,
+      "grad_norm": 0.4200455327121287,
+      "learning_rate": 2.3210868250450865e-05,
+      "loss": 0.5406,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5374387683694891,
+      "grad_norm": 0.4205148102378079,
+      "learning_rate": 2.3199564187368156e-05,
+      "loss": 0.5609,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5375787263820854,
+      "grad_norm": 0.41158756784707023,
+      "learning_rate": 2.3188260494293273e-05,
+      "loss": 0.5872,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5377186843946816,
+      "grad_norm": 0.43038344004576323,
+      "learning_rate": 2.3176957173549235e-05,
+      "loss": 0.5768,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5378586424072778,
+      "grad_norm": 0.4358694072281102,
+      "learning_rate": 2.3165654227459004e-05,
+      "loss": 0.5357,
+      "step": 3843
+    },
+    {
+      "epoch": 0.537998600419874,
+      "grad_norm": 0.4343491080187233,
+      "learning_rate": 2.3154351658345437e-05,
+      "loss": 0.5608,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5381385584324703,
+      "grad_norm": 0.42585193830336604,
+      "learning_rate": 2.3143049468531334e-05,
+      "loss": 0.6127,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5382785164450665,
+      "grad_norm": 0.4168594356306299,
+      "learning_rate": 2.3131747660339394e-05,
+      "loss": 0.5421,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5384184744576627,
+      "grad_norm": 0.4143482797388666,
+      "learning_rate": 2.3120446236092276e-05,
+      "loss": 0.5634,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5385584324702589,
+      "grad_norm": 0.42347488623228974,
+      "learning_rate": 2.310914519811253e-05,
+      "loss": 0.553,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5386983904828552,
+      "grad_norm": 0.4232797867273413,
+      "learning_rate": 2.309784454872262e-05,
+      "loss": 0.597,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5388383484954513,
+      "grad_norm": 0.4226580404199005,
+      "learning_rate": 2.3086544290244967e-05,
+      "loss": 0.5418,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5389783065080476,
+      "grad_norm": 0.43647901356290086,
+      "learning_rate": 2.3075244425001874e-05,
+      "loss": 0.6088,
+      "step": 3851
+    },
+    {
+      "epoch": 0.5391182645206438,
+      "grad_norm": 0.4110315073763923,
+      "learning_rate": 2.3063944955315584e-05,
+      "loss": 0.54,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5392582225332401,
+      "grad_norm": 0.370641365809909,
+      "learning_rate": 2.3052645883508242e-05,
+      "loss": 0.5216,
+      "step": 3853
+    },
+    {
+      "epoch": 0.5393981805458362,
+      "grad_norm": 0.4035453298163194,
+      "learning_rate": 2.3041347211901935e-05,
+      "loss": 0.559,
+      "step": 3854
+    },
+    {
+      "epoch": 0.5395381385584325,
+      "grad_norm": 0.426639073461407,
+      "learning_rate": 2.303004894281865e-05,
+      "loss": 0.5409,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5396780965710287,
+      "grad_norm": 0.42512057290523336,
+      "learning_rate": 2.3018751078580287e-05,
+      "loss": 0.607,
+      "step": 3856
+    },
+    {
+      "epoch": 0.5398180545836249,
+      "grad_norm": 0.42427900709447586,
+      "learning_rate": 2.300745362150869e-05,
+      "loss": 0.5736,
+      "step": 3857
+    },
+    {
+      "epoch": 0.5399580125962211,
+      "grad_norm": 0.4308088797406061,
+      "learning_rate": 2.299615657392559e-05,
+      "loss": 0.5737,
+      "step": 3858
+    },
+    {
+      "epoch": 0.5400979706088174,
+      "grad_norm": 0.39893806446075136,
+      "learning_rate": 2.2984859938152644e-05,
+      "loss": 0.5167,
+      "step": 3859
+    },
+    {
+      "epoch": 0.5402379286214136,
+      "grad_norm": 0.4227311792800813,
+      "learning_rate": 2.297356371651142e-05,
+      "loss": 0.5559,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5403778866340098,
+      "grad_norm": 0.4307602196718643,
+      "learning_rate": 2.296226791132342e-05,
+      "loss": 0.567,
+      "step": 3861
+    },
+    {
+      "epoch": 0.540517844646606,
+      "grad_norm": 0.4048870602226458,
+      "learning_rate": 2.2950972524910045e-05,
+      "loss": 0.5474,
+      "step": 3862
+    },
+    {
+      "epoch": 0.5406578026592023,
+      "grad_norm": 0.4130105755887349,
+      "learning_rate": 2.2939677559592605e-05,
+      "loss": 0.5438,
+      "step": 3863
+    },
+    {
+      "epoch": 0.5407977606717984,
+      "grad_norm": 0.4388089801883774,
+      "learning_rate": 2.2928383017692322e-05,
+      "loss": 0.562,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5409377186843947,
+      "grad_norm": 0.4591220588088252,
+      "learning_rate": 2.2917088901530358e-05,
+      "loss": 0.5396,
+      "step": 3865
+    },
+    {
+      "epoch": 0.5410776766969909,
+      "grad_norm": 0.42951530189135795,
+      "learning_rate": 2.290579521342776e-05,
+      "loss": 0.5726,
+      "step": 3866
+    },
+    {
+      "epoch": 0.5412176347095872,
+      "grad_norm": 0.40253890454723895,
+      "learning_rate": 2.2894501955705477e-05,
+      "loss": 0.5234,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5413575927221833,
+      "grad_norm": 0.42549136435980095,
+      "learning_rate": 2.288320913068442e-05,
+      "loss": 0.5505,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5414975507347796,
+      "grad_norm": 0.4060031912669076,
+      "learning_rate": 2.2871916740685366e-05,
+      "loss": 0.5863,
+      "step": 3869
+    },
+    {
+      "epoch": 0.5416375087473758,
+      "grad_norm": 0.4242326527971933,
+      "learning_rate": 2.2860624788029013e-05,
+      "loss": 0.5408,
+      "step": 3870
+    },
+    {
+      "epoch": 0.541777466759972,
+      "grad_norm": 0.43284739833626334,
+      "learning_rate": 2.2849333275035964e-05,
+      "loss": 0.5781,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5419174247725682,
+      "grad_norm": 0.38138784670752485,
+      "learning_rate": 2.283804220402676e-05,
+      "loss": 0.5374,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5420573827851645,
+      "grad_norm": 0.42596112184398915,
+      "learning_rate": 2.2826751577321813e-05,
+      "loss": 0.5407,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5421973407977607,
+      "grad_norm": 0.4016705521461287,
+      "learning_rate": 2.2815461397241466e-05,
+      "loss": 0.543,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5423372988103569,
+      "grad_norm": 0.41007081598495276,
+      "learning_rate": 2.2804171666105976e-05,
+      "loss": 0.5606,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5424772568229531,
+      "grad_norm": 0.41478676365527245,
+      "learning_rate": 2.2792882386235485e-05,
+      "loss": 0.5264,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5426172148355494,
+      "grad_norm": 0.4120034933406958,
+      "learning_rate": 2.2781593559950052e-05,
+      "loss": 0.5392,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5427571728481455,
+      "grad_norm": 0.4154088448683998,
+      "learning_rate": 2.277030518956965e-05,
+      "loss": 0.5667,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5428971308607418,
+      "grad_norm": 0.4099026076541546,
+      "learning_rate": 2.2759017277414166e-05,
+      "loss": 0.5248,
+      "step": 3879
+    },
+    {
+      "epoch": 0.543037088873338,
+      "grad_norm": 0.44769525609475164,
+      "learning_rate": 2.2747729825803366e-05,
+      "loss": 0.5772,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5431770468859343,
+      "grad_norm": 0.4131464054022151,
+      "learning_rate": 2.273644283705694e-05,
+      "loss": 0.5797,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5433170048985304,
+      "grad_norm": 0.4261856744618942,
+      "learning_rate": 2.2725156313494466e-05,
+      "loss": 0.549,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5434569629111267,
+      "grad_norm": 0.4207414434352504,
+      "learning_rate": 2.271387025743546e-05,
+      "loss": 0.5738,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5435969209237229,
+      "grad_norm": 0.4087475690432694,
+      "learning_rate": 2.2702584671199317e-05,
+      "loss": 0.5674,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5437368789363191,
+      "grad_norm": 0.40838344825260353,
+      "learning_rate": 2.2691299557105328e-05,
+      "loss": 0.5503,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5438768369489153,
+      "grad_norm": 0.42246887497450447,
+      "learning_rate": 2.268001491747271e-05,
+      "loss": 0.5276,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5440167949615116,
+      "grad_norm": 0.4325429890817475,
+      "learning_rate": 2.266873075462056e-05,
+      "loss": 0.5538,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5441567529741078,
+      "grad_norm": 0.420910542262214,
+      "learning_rate": 2.2657447070867902e-05,
+      "loss": 0.5475,
+      "step": 3888
+    },
+    {
+      "epoch": 0.544296710986704,
+      "grad_norm": 0.422876814725259,
+      "learning_rate": 2.264616386853363e-05,
+      "loss": 0.5477,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5444366689993002,
+      "grad_norm": 0.4143598457765582,
+      "learning_rate": 2.2634881149936575e-05,
+      "loss": 0.5419,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5445766270118965,
+      "grad_norm": 0.4105061536286951,
+      "learning_rate": 2.2623598917395438e-05,
+      "loss": 0.5475,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5447165850244926,
+      "grad_norm": 0.41659655042186283,
+      "learning_rate": 2.261231717322883e-05,
+      "loss": 0.5753,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5448565430370889,
+      "grad_norm": 0.4530820570399967,
+      "learning_rate": 2.2601035919755274e-05,
+      "loss": 0.5281,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5449965010496851,
+      "grad_norm": 0.39850319238936455,
+      "learning_rate": 2.258975515929318e-05,
+      "loss": 0.5278,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5451364590622814,
+      "grad_norm": 0.3978091253457419,
+      "learning_rate": 2.2578474894160857e-05,
+      "loss": 0.5453,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5452764170748775,
+      "grad_norm": 0.4231460383028299,
+      "learning_rate": 2.2567195126676507e-05,
+      "loss": 0.5963,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5454163750874738,
+      "grad_norm": 0.42063829224368665,
+      "learning_rate": 2.2555915859158244e-05,
+      "loss": 0.5614,
+      "step": 3897
+    },
+    {
+      "epoch": 0.54555633310007,
+      "grad_norm": 0.4083469272821722,
+      "learning_rate": 2.2544637093924074e-05,
+      "loss": 0.5389,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5456962911126662,
+      "grad_norm": 0.4217541606558565,
+      "learning_rate": 2.2533358833291896e-05,
+      "loss": 0.5544,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5458362491252624,
+      "grad_norm": 0.4303357717984438,
+      "learning_rate": 2.2522081079579498e-05,
+      "loss": 0.5524,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5459762071378587,
+      "grad_norm": 0.41929973760406747,
+      "learning_rate": 2.251080383510459e-05,
+      "loss": 0.5652,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5461161651504549,
+      "grad_norm": 0.4126809759165879,
+      "learning_rate": 2.2499527102184744e-05,
+      "loss": 0.5917,
+      "step": 3902
+    },
+    {
+      "epoch": 0.546256123163051,
+      "grad_norm": 0.39993684724127465,
+      "learning_rate": 2.2488250883137445e-05,
+      "loss": 0.532,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5463960811756473,
+      "grad_norm": 0.4070252666557728,
+      "learning_rate": 2.247697518028008e-05,
+      "loss": 0.5381,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5465360391882436,
+      "grad_norm": 0.4203749044451171,
+      "learning_rate": 2.2465699995929916e-05,
+      "loss": 0.5445,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5466759972008397,
+      "grad_norm": 0.4313492162077952,
+      "learning_rate": 2.2454425332404122e-05,
+      "loss": 0.55,
+      "step": 3906
+    },
+    {
+      "epoch": 0.546815955213436,
+      "grad_norm": 0.39695375283438905,
+      "learning_rate": 2.2443151192019735e-05,
+      "loss": 0.5418,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5469559132260322,
+      "grad_norm": 0.4051789619978703,
+      "learning_rate": 2.2431877577093737e-05,
+      "loss": 0.5059,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5470958712386285,
+      "grad_norm": 0.4218855232884939,
+      "learning_rate": 2.2420604489942946e-05,
+      "loss": 0.5617,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5472358292512246,
+      "grad_norm": 0.3989929337699537,
+      "learning_rate": 2.2409331932884108e-05,
+      "loss": 0.5244,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5473757872638209,
+      "grad_norm": 0.4010087179047754,
+      "learning_rate": 2.2398059908233842e-05,
+      "loss": 0.5371,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5475157452764171,
+      "grad_norm": 0.4466643997788068,
+      "learning_rate": 2.238678841830867e-05,
+      "loss": 0.5934,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5476557032890133,
+      "grad_norm": 0.39728611887637755,
+      "learning_rate": 2.237551746542499e-05,
+      "loss": 0.5587,
+      "step": 3913
+    },
+    {
+      "epoch": 0.5477956613016095,
+      "grad_norm": 0.41809901084229356,
+      "learning_rate": 2.236424705189909e-05,
+      "loss": 0.5865,
+      "step": 3914
+    },
+    {
+      "epoch": 0.5479356193142058,
+      "grad_norm": 0.40468351305300143,
+      "learning_rate": 2.2352977180047175e-05,
+      "loss": 0.5213,
+      "step": 3915
+    },
+    {
+      "epoch": 0.548075577326802,
+      "grad_norm": 0.4113028432405159,
+      "learning_rate": 2.2341707852185305e-05,
+      "loss": 0.5488,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5482155353393982,
+      "grad_norm": 0.4448481069088446,
+      "learning_rate": 2.2330439070629448e-05,
+      "loss": 0.5665,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5483554933519944,
+      "grad_norm": 0.4360732301218836,
+      "learning_rate": 2.2319170837695435e-05,
+      "loss": 0.5586,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5484954513645907,
+      "grad_norm": 0.4194014787432002,
+      "learning_rate": 2.2307903155699027e-05,
+      "loss": 0.5265,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5486354093771868,
+      "grad_norm": 0.4085819437326063,
+      "learning_rate": 2.2296636026955835e-05,
+      "loss": 0.5485,
+      "step": 3920
+    },
+    {
+      "epoch": 0.548775367389783,
+      "grad_norm": 0.40406354838300534,
+      "learning_rate": 2.2285369453781364e-05,
+      "loss": 0.5409,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5489153254023793,
+      "grad_norm": 0.41085941144789107,
+      "learning_rate": 2.2274103438491022e-05,
+      "loss": 0.5228,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5490552834149756,
+      "grad_norm": 0.44664374072567387,
+      "learning_rate": 2.2262837983400082e-05,
+      "loss": 0.5663,
+      "step": 3923
+    },
+    {
+      "epoch": 0.5491952414275717,
+      "grad_norm": 0.3848765417302251,
+      "learning_rate": 2.2251573090823706e-05,
+      "loss": 0.5645,
+      "step": 3924
+    },
+    {
+      "epoch": 0.549335199440168,
+      "grad_norm": 0.40234337327172925,
+      "learning_rate": 2.2240308763076935e-05,
+      "loss": 0.573,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5494751574527642,
+      "grad_norm": 0.4124065219528931,
+      "learning_rate": 2.222904500247473e-05,
+      "loss": 0.5579,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5496151154653603,
+      "grad_norm": 0.4085451573297374,
+      "learning_rate": 2.2217781811331885e-05,
+      "loss": 0.5534,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5497550734779566,
+      "grad_norm": 0.413762414731123,
+      "learning_rate": 2.22065191919631e-05,
+      "loss": 0.5493,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5498950314905529,
+      "grad_norm": 0.4148155757825788,
+      "learning_rate": 2.2195257146682975e-05,
+      "loss": 0.5439,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5500349895031491,
+      "grad_norm": 0.44638831807610185,
+      "learning_rate": 2.218399567780597e-05,
+      "loss": 0.5899,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5501749475157452,
+      "grad_norm": 0.41295788493334096,
+      "learning_rate": 2.217273478764642e-05,
+      "loss": 0.5524,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5503149055283415,
+      "grad_norm": 0.402717034759649,
+      "learning_rate": 2.216147447851855e-05,
+      "loss": 0.5534,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5504548635409378,
+      "grad_norm": 0.40126403258397014,
+      "learning_rate": 2.2150214752736488e-05,
+      "loss": 0.5227,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5505948215535339,
+      "grad_norm": 0.4081017029720488,
+      "learning_rate": 2.2138955612614207e-05,
+      "loss": 0.5316,
+      "step": 3934
+    },
+    {
+      "epoch": 0.5507347795661302,
+      "grad_norm": 0.3952776691289986,
+      "learning_rate": 2.2127697060465576e-05,
+      "loss": 0.54,
+      "step": 3935
+    },
+    {
+      "epoch": 0.5508747375787264,
+      "grad_norm": 0.4086382501258365,
+      "learning_rate": 2.211643909860433e-05,
+      "loss": 0.5376,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5510146955913227,
+      "grad_norm": 0.44788307946149386,
+      "learning_rate": 2.210518172934412e-05,
+      "loss": 0.5916,
+      "step": 3937
+    },
+    {
+      "epoch": 0.5511546536039188,
+      "grad_norm": 0.4092566815599357,
+      "learning_rate": 2.2093924954998438e-05,
+      "loss": 0.5282,
+      "step": 3938
+    },
+    {
+      "epoch": 0.551294611616515,
+      "grad_norm": 0.43673368860544104,
+      "learning_rate": 2.2082668777880653e-05,
+      "loss": 0.5458,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5514345696291113,
+      "grad_norm": 0.406082695925654,
+      "learning_rate": 2.2071413200304043e-05,
+      "loss": 0.5375,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5515745276417074,
+      "grad_norm": 0.3883615387736086,
+      "learning_rate": 2.206015822458174e-05,
+      "loss": 0.5221,
+      "step": 3941
+    },
+    {
+      "epoch": 0.5517144856543037,
+      "grad_norm": 0.4315662304861121,
+      "learning_rate": 2.2048903853026745e-05,
+      "loss": 0.5598,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5518544436669,
+      "grad_norm": 0.40829821883297823,
+      "learning_rate": 2.203765008795195e-05,
+      "loss": 0.5876,
+      "step": 3943
+    },
+    {
+      "epoch": 0.5519944016794962,
+      "grad_norm": 0.4263164757098248,
+      "learning_rate": 2.2026396931670125e-05,
+      "loss": 0.5506,
+      "step": 3944
+    },
+    {
+      "epoch": 0.5521343596920923,
+      "grad_norm": 0.4109132538637934,
+      "learning_rate": 2.2015144386493896e-05,
+      "loss": 0.5435,
+      "step": 3945
+    },
+    {
+      "epoch": 0.5522743177046886,
+      "grad_norm": 0.40656446743742547,
+      "learning_rate": 2.2003892454735786e-05,
+      "loss": 0.5169,
+      "step": 3946
+    },
+    {
+      "epoch": 0.5524142757172849,
+      "grad_norm": 0.4174968254449107,
+      "learning_rate": 2.1992641138708166e-05,
+      "loss": 0.5506,
+      "step": 3947
+    },
+    {
+      "epoch": 0.552554233729881,
+      "grad_norm": 0.41392476596527283,
+      "learning_rate": 2.1981390440723316e-05,
+      "loss": 0.5705,
+      "step": 3948
+    },
+    {
+      "epoch": 0.5526941917424772,
+      "grad_norm": 0.40768401049876407,
+      "learning_rate": 2.197014036309336e-05,
+      "loss": 0.5546,
+      "step": 3949
+    },
+    {
+      "epoch": 0.5528341497550735,
+      "grad_norm": 0.43659985592955725,
+      "learning_rate": 2.1958890908130288e-05,
+      "loss": 0.5448,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5529741077676698,
+      "grad_norm": 0.4191971534215795,
+      "learning_rate": 2.1947642078146004e-05,
+      "loss": 0.5442,
+      "step": 3951
+    },
+    {
+      "epoch": 0.5531140657802659,
+      "grad_norm": 0.43109773584435357,
+      "learning_rate": 2.193639387545224e-05,
+      "loss": 0.5506,
+      "step": 3952
+    },
+    {
+      "epoch": 0.5532540237928621,
+      "grad_norm": 0.42420382449551325,
+      "learning_rate": 2.1925146302360625e-05,
+      "loss": 0.5578,
+      "step": 3953
+    },
+    {
+      "epoch": 0.5533939818054584,
+      "grad_norm": 0.426621743894938,
+      "learning_rate": 2.1913899361182632e-05,
+      "loss": 0.5453,
+      "step": 3954
+    },
+    {
+      "epoch": 0.5535339398180545,
+      "grad_norm": 0.42952974584048115,
+      "learning_rate": 2.1902653054229642e-05,
+      "loss": 0.5514,
+      "step": 3955
+    },
+    {
+      "epoch": 0.5536738978306508,
+      "grad_norm": 0.42155187844422864,
+      "learning_rate": 2.189140738381288e-05,
+      "loss": 0.5622,
+      "step": 3956
+    },
+    {
+      "epoch": 0.553813855843247,
+      "grad_norm": 0.4260436696061489,
+      "learning_rate": 2.1880162352243425e-05,
+      "loss": 0.5845,
+      "step": 3957
+    },
+    {
+      "epoch": 0.5539538138558433,
+      "grad_norm": 0.4048895896225377,
+      "learning_rate": 2.1868917961832274e-05,
+      "loss": 0.5593,
+      "step": 3958
+    },
+    {
+      "epoch": 0.5540937718684394,
+      "grad_norm": 0.4088135684372084,
+      "learning_rate": 2.1857674214890254e-05,
+      "loss": 0.517,
+      "step": 3959
+    },
+    {
+      "epoch": 0.5542337298810357,
+      "grad_norm": 0.40953844739681516,
+      "learning_rate": 2.1846431113728064e-05,
+      "loss": 0.57,
+      "step": 3960
+    },
+    {
+      "epoch": 0.554373687893632,
+      "grad_norm": 0.41049014674652273,
+      "learning_rate": 2.1835188660656267e-05,
+      "loss": 0.582,
+      "step": 3961
+    },
+    {
+      "epoch": 0.5545136459062281,
+      "grad_norm": 0.44214177894488643,
+      "learning_rate": 2.1823946857985323e-05,
+      "loss": 0.5558,
+      "step": 3962
+    },
+    {
+      "epoch": 0.5546536039188243,
+      "grad_norm": 0.40688749353479037,
+      "learning_rate": 2.1812705708025526e-05,
+      "loss": 0.5647,
+      "step": 3963
+    },
+    {
+      "epoch": 0.5547935619314206,
+      "grad_norm": 0.43619186505830493,
+      "learning_rate": 2.1801465213087044e-05,
+      "loss": 0.5819,
+      "step": 3964
+    },
+    {
+      "epoch": 0.5549335199440167,
+      "grad_norm": 0.3873905618002296,
+      "learning_rate": 2.179022537547991e-05,
+      "loss": 0.5015,
+      "step": 3965
+    },
+    {
+      "epoch": 0.555073477956613,
+      "grad_norm": 0.40366380009710195,
+      "learning_rate": 2.1778986197514034e-05,
+      "loss": 0.5364,
+      "step": 3966
+    },
+    {
+      "epoch": 0.5552134359692092,
+      "grad_norm": 0.41225036663129866,
+      "learning_rate": 2.1767747681499176e-05,
+      "loss": 0.5395,
+      "step": 3967
+    },
+    {
+      "epoch": 0.5553533939818055,
+      "grad_norm": 0.41221185236407193,
+      "learning_rate": 2.1756509829744956e-05,
+      "loss": 0.5375,
+      "step": 3968
+    },
+    {
+      "epoch": 0.5554933519944016,
+      "grad_norm": 0.41637063641888467,
+      "learning_rate": 2.1745272644560885e-05,
+      "loss": 0.5695,
+      "step": 3969
+    },
+    {
+      "epoch": 0.5556333100069979,
+      "grad_norm": 0.4246669886827112,
+      "learning_rate": 2.173403612825631e-05,
+      "loss": 0.5226,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5557732680195941,
+      "grad_norm": 0.4399810838843672,
+      "learning_rate": 2.172280028314045e-05,
+      "loss": 0.5637,
+      "step": 3971
+    },
+    {
+      "epoch": 0.5559132260321903,
+      "grad_norm": 0.42264632884515463,
+      "learning_rate": 2.1711565111522372e-05,
+      "loss": 0.5715,
+      "step": 3972
+    },
+    {
+      "epoch": 0.5560531840447865,
+      "grad_norm": 0.40102465338699234,
+      "learning_rate": 2.170033061571104e-05,
+      "loss": 0.5468,
+      "step": 3973
+    },
+    {
+      "epoch": 0.5561931420573828,
+      "grad_norm": 0.4330504312787022,
+      "learning_rate": 2.1689096798015247e-05,
+      "loss": 0.5544,
+      "step": 3974
+    },
+    {
+      "epoch": 0.556333100069979,
+      "grad_norm": 0.41663293988065675,
+      "learning_rate": 2.167786366074365e-05,
+      "loss": 0.5855,
+      "step": 3975
+    },
+    {
+      "epoch": 0.5564730580825752,
+      "grad_norm": 0.40801980733280413,
+      "learning_rate": 2.1666631206204786e-05,
+      "loss": 0.5431,
+      "step": 3976
+    },
+    {
+      "epoch": 0.5566130160951714,
+      "grad_norm": 0.42663243125510747,
+      "learning_rate": 2.1655399436707026e-05,
+      "loss": 0.5574,
+      "step": 3977
+    },
+    {
+      "epoch": 0.5567529741077677,
+      "grad_norm": 0.4605067091274325,
+      "learning_rate": 2.164416835455862e-05,
+      "loss": 0.5592,
+      "step": 3978
+    },
+    {
+      "epoch": 0.5568929321203638,
+      "grad_norm": 0.42331401492783005,
+      "learning_rate": 2.1632937962067657e-05,
+      "loss": 0.5798,
+      "step": 3979
+    },
+    {
+      "epoch": 0.5570328901329601,
+      "grad_norm": 0.5978199091551641,
+      "learning_rate": 2.1621708261542116e-05,
+      "loss": 0.5474,
+      "step": 3980
+    },
+    {
+      "epoch": 0.5571728481455563,
+      "grad_norm": 0.39934848914020754,
+      "learning_rate": 2.161047925528981e-05,
+      "loss": 0.5362,
+      "step": 3981
+    },
+    {
+      "epoch": 0.5573128061581526,
+      "grad_norm": 0.41963116860550964,
+      "learning_rate": 2.1599250945618402e-05,
+      "loss": 0.5506,
+      "step": 3982
+    },
+    {
+      "epoch": 0.5574527641707487,
+      "grad_norm": 0.41948040388705365,
+      "learning_rate": 2.158802333483542e-05,
+      "loss": 0.5723,
+      "step": 3983
+    },
+    {
+      "epoch": 0.557592722183345,
+      "grad_norm": 0.4188462923542639,
+      "learning_rate": 2.157679642524828e-05,
+      "loss": 0.5542,
+      "step": 3984
+    },
+    {
+      "epoch": 0.5577326801959412,
+      "grad_norm": 0.4356609128042065,
+      "learning_rate": 2.15655702191642e-05,
+      "loss": 0.5568,
+      "step": 3985
+    },
+    {
+      "epoch": 0.5578726382085374,
+      "grad_norm": 0.4295054521048054,
+      "learning_rate": 2.1554344718890284e-05,
+      "loss": 0.6145,
+      "step": 3986
+    },
+    {
+      "epoch": 0.5580125962211336,
+      "grad_norm": 0.41247657463875265,
+      "learning_rate": 2.1543119926733495e-05,
+      "loss": 0.5561,
+      "step": 3987
+    },
+    {
+      "epoch": 0.5581525542337299,
+      "grad_norm": 0.41532771461538054,
+      "learning_rate": 2.1531895845000637e-05,
+      "loss": 0.5414,
+      "step": 3988
+    },
+    {
+      "epoch": 0.5582925122463261,
+      "grad_norm": 0.40582165725579067,
+      "learning_rate": 2.1520672475998373e-05,
+      "loss": 0.5509,
+      "step": 3989
+    },
+    {
+      "epoch": 0.5584324702589223,
+      "grad_norm": 0.4060285622816394,
+      "learning_rate": 2.1509449822033205e-05,
+      "loss": 0.5484,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5585724282715185,
+      "grad_norm": 0.4029440998090135,
+      "learning_rate": 2.1498227885411526e-05,
+      "loss": 0.585,
+      "step": 3991
+    },
+    {
+      "epoch": 0.5587123862841148,
+      "grad_norm": 0.4104149112850885,
+      "learning_rate": 2.148700666843955e-05,
+      "loss": 0.52,
+      "step": 3992
+    },
+    {
+      "epoch": 0.5588523442967109,
+      "grad_norm": 0.4171536565223565,
+      "learning_rate": 2.1475786173423335e-05,
+      "loss": 0.5623,
+      "step": 3993
+    },
+    {
+      "epoch": 0.5589923023093072,
+      "grad_norm": 0.3980539899980082,
+      "learning_rate": 2.146456640266883e-05,
+      "loss": 0.5812,
+      "step": 3994
+    },
+    {
+      "epoch": 0.5591322603219034,
+      "grad_norm": 0.41514890565668344,
+      "learning_rate": 2.14533473584818e-05,
+      "loss": 0.5826,
+      "step": 3995
+    },
+    {
+      "epoch": 0.5592722183344997,
+      "grad_norm": 0.4166753133049829,
+      "learning_rate": 2.1442129043167874e-05,
+      "loss": 0.5543,
+      "step": 3996
+    },
+    {
+      "epoch": 0.5594121763470958,
+      "grad_norm": 0.4114882654130403,
+      "learning_rate": 2.1430911459032526e-05,
+      "loss": 0.5741,
+      "step": 3997
+    },
+    {
+      "epoch": 0.5595521343596921,
+      "grad_norm": 0.4017370253857547,
+      "learning_rate": 2.1419694608381094e-05,
+      "loss": 0.5963,
+      "step": 3998
+    },
+    {
+      "epoch": 0.5596920923722883,
+      "grad_norm": 0.42955031856189274,
+      "learning_rate": 2.1408478493518742e-05,
+      "loss": 0.5896,
+      "step": 3999
+    },
+    {
+      "epoch": 0.5598320503848845,
+      "grad_norm": 0.40378298742736296,
+      "learning_rate": 2.1397263116750503e-05,
+      "loss": 0.5657,
+      "step": 4000
     }
   ],
   "logging_steps": 1,
@@ -7027,7 +28027,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 175094048915456.0,
+  "total_flos": 700001933443072.0,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null