diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.4325610316580605,
+  "epoch": 0.9996485441617778,
   "eval_steps": 500,
-  "global_step": 1000,
+  "global_step": 2311,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -7007,6 +7007,9183 @@
       "learning_rate": 1.1469816272965881e-05,
       "loss": 1.5889,
       "step": 1000
+    },
+    {
+      "epoch": 0.43299359268971854,
+      "grad_norm": 23.486230850219727,
+      "learning_rate": 1.1461067366579179e-05,
+      "loss": 1.7299,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4334261537213766,
+      "grad_norm": 21.794849395751953,
+      "learning_rate": 1.1452318460192478e-05,
+      "loss": 1.6118,
+      "step": 1002
+    },
+    {
+      "epoch": 0.43385871475303467,
+      "grad_norm": 20.80159568786621,
+      "learning_rate": 1.1443569553805776e-05,
+      "loss": 1.5558,
+      "step": 1003
+    },
+    {
+      "epoch": 0.43429127578469273,
+      "grad_norm": 21.411245346069336,
+      "learning_rate": 1.1434820647419073e-05,
+      "loss": 1.5028,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4347238368163508,
+      "grad_norm": 20.596715927124023,
+      "learning_rate": 1.1426071741032372e-05,
+      "loss": 1.5536,
+      "step": 1005
+    },
+    {
+      "epoch": 0.43515639784800886,
+      "grad_norm": 22.81339454650879,
+      "learning_rate": 1.141732283464567e-05,
+      "loss": 1.5613,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4355889588796669,
+      "grad_norm": 20.697265625,
+      "learning_rate": 1.1408573928258967e-05,
+      "loss": 1.5837,
+      "step": 1007
+    },
+    {
+      "epoch": 0.436021519911325,
+      "grad_norm": 18.886920928955078,
+      "learning_rate": 1.1399825021872267e-05,
+      "loss": 1.542,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43645408094298305,
+      "grad_norm": 22.665058135986328,
+      "learning_rate": 1.1391076115485564e-05,
+      "loss": 1.6007,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4368866419746411,
+      "grad_norm": 20.35870361328125,
+      "learning_rate": 1.1382327209098865e-05,
+      "loss": 1.5594,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4373192030062992,
+      "grad_norm": 19.82785987854004,
+      "learning_rate": 1.1373578302712163e-05,
+      "loss": 1.6128,
+      "step": 1011
+    },
+    {
+      "epoch": 0.43775176403795724,
+      "grad_norm": 21.4642276763916,
+      "learning_rate": 1.136482939632546e-05,
+      "loss": 1.647,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4381843250696153,
+      "grad_norm": 24.7077693939209,
+      "learning_rate": 1.135608048993876e-05,
+      "loss": 1.5577,
+      "step": 1013
+    },
+    {
+      "epoch": 0.43861688610127336,
+      "grad_norm": 21.480487823486328,
+      "learning_rate": 1.1347331583552057e-05,
+      "loss": 1.5856,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4390494471329314,
+      "grad_norm": 24.280683517456055,
+      "learning_rate": 1.1338582677165354e-05,
+      "loss": 1.5539,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4394820081645895,
+      "grad_norm": 22.40320587158203,
+      "learning_rate": 1.1329833770778653e-05,
+      "loss": 1.5617,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43991456919624755,
+      "grad_norm": 24.9234619140625,
+      "learning_rate": 1.1321084864391951e-05,
+      "loss": 1.5467,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4403471302279056,
+      "grad_norm": 20.776636123657227,
+      "learning_rate": 1.131233595800525e-05,
+      "loss": 1.56,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4407796912595637,
+      "grad_norm": 20.39991569519043,
+      "learning_rate": 1.1303587051618548e-05,
+      "loss": 1.557,
+      "step": 1019
+    },
+    {
+      "epoch": 0.44121225229122174,
+      "grad_norm": 19.826358795166016,
+      "learning_rate": 1.1294838145231849e-05,
+      "loss": 1.6674,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44164481332287975,
+      "grad_norm": 19.581579208374023,
+      "learning_rate": 1.1286089238845146e-05,
+      "loss": 1.5441,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4420773743545378,
+      "grad_norm": 22.186416625976562,
+      "learning_rate": 1.1277340332458444e-05,
+      "loss": 1.4218,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4425099353861959,
+      "grad_norm": 24.520288467407227,
+      "learning_rate": 1.1268591426071743e-05,
+      "loss": 1.6456,
+      "step": 1023
+    },
+    {
+      "epoch": 0.44294249641785394,
+      "grad_norm": 18.8167781829834,
+      "learning_rate": 1.125984251968504e-05,
+      "loss": 1.5243,
+      "step": 1024
+    },
+    {
+      "epoch": 0.443375057449512,
+      "grad_norm": 20.95041275024414,
+      "learning_rate": 1.1251093613298338e-05,
+      "loss": 1.6067,
+      "step": 1025
+    },
+    {
+      "epoch": 0.44380761848117006,
+      "grad_norm": 22.097684860229492,
+      "learning_rate": 1.1242344706911637e-05,
+      "loss": 1.4956,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4442401795128281,
+      "grad_norm": 21.358125686645508,
+      "learning_rate": 1.1233595800524935e-05,
+      "loss": 1.623,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4446727405444862,
+      "grad_norm": 19.22201919555664,
+      "learning_rate": 1.1224846894138232e-05,
+      "loss": 1.5917,
+      "step": 1028
+    },
+    {
+      "epoch": 0.44510530157614425,
+      "grad_norm": 20.375303268432617,
+      "learning_rate": 1.1216097987751533e-05,
+      "loss": 1.658,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4455378626078023,
+      "grad_norm": 22.127113342285156,
+      "learning_rate": 1.120734908136483e-05,
+      "loss": 1.5364,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4459704236394604,
+      "grad_norm": 23.640071868896484,
+      "learning_rate": 1.119860017497813e-05,
+      "loss": 1.4492,
+      "step": 1031
+    },
+    {
+      "epoch": 0.44640298467111844,
+      "grad_norm": 20.21621322631836,
+      "learning_rate": 1.1189851268591427e-05,
+      "loss": 1.5371,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4468355457027765,
+      "grad_norm": 18.869565963745117,
+      "learning_rate": 1.1181102362204725e-05,
+      "loss": 1.5825,
+      "step": 1033
+    },
+    {
+      "epoch": 0.44726810673443457,
+      "grad_norm": 19.211931228637695,
+      "learning_rate": 1.1172353455818024e-05,
+      "loss": 1.5023,
+      "step": 1034
+    },
+    {
+      "epoch": 0.44770066776609263,
+      "grad_norm": 20.949472427368164,
+      "learning_rate": 1.1163604549431321e-05,
+      "loss": 1.662,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4481332287977507,
+      "grad_norm": 19.43843650817871,
+      "learning_rate": 1.1154855643044619e-05,
+      "loss": 1.6401,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44856578982940876,
+      "grad_norm": 21.74172592163086,
+      "learning_rate": 1.1146106736657918e-05,
+      "loss": 1.6335,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4489983508610668,
+      "grad_norm": 23.03577423095703,
+      "learning_rate": 1.1137357830271216e-05,
+      "loss": 1.5985,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4494309118927249,
+      "grad_norm": 20.48150634765625,
+      "learning_rate": 1.1128608923884517e-05,
+      "loss": 1.6556,
+      "step": 1039
+    },
+    {
+      "epoch": 0.44986347292438295,
+      "grad_norm": 21.354633331298828,
+      "learning_rate": 1.1119860017497814e-05,
+      "loss": 1.5034,
+      "step": 1040
+    },
+    {
+      "epoch": 0.450296033956041,
+      "grad_norm": 22.22134017944336,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 1.5633,
+      "step": 1041
+    },
+    {
+      "epoch": 0.45072859498769907,
+      "grad_norm": 20.304677963256836,
+      "learning_rate": 1.1102362204724411e-05,
+      "loss": 1.5714,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4511611560193571,
+      "grad_norm": 19.70705795288086,
+      "learning_rate": 1.1093613298337708e-05,
+      "loss": 1.6236,
+      "step": 1043
+    },
+    {
+      "epoch": 0.45159371705101514,
+      "grad_norm": 20.76530647277832,
+      "learning_rate": 1.1084864391951008e-05,
+      "loss": 1.5905,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4520262780826732,
+      "grad_norm": 19.45387840270996,
+      "learning_rate": 1.1076115485564305e-05,
+      "loss": 1.5035,
+      "step": 1045
+    },
+    {
+      "epoch": 0.45245883911433127,
+      "grad_norm": 19.863208770751953,
+      "learning_rate": 1.1067366579177603e-05,
+      "loss": 1.5407,
+      "step": 1046
+    },
+    {
+      "epoch": 0.45289140014598933,
+      "grad_norm": 21.128826141357422,
+      "learning_rate": 1.1058617672790902e-05,
+      "loss": 1.5474,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4533239611776474,
+      "grad_norm": 21.49403953552246,
+      "learning_rate": 1.1049868766404201e-05,
+      "loss": 1.6163,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45375652220930546,
+      "grad_norm": 23.149555206298828,
+      "learning_rate": 1.10411198600175e-05,
+      "loss": 1.5622,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4541890832409635,
+      "grad_norm": 21.647279739379883,
+      "learning_rate": 1.1032370953630798e-05,
+      "loss": 1.6026,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4546216442726216,
+      "grad_norm": 21.786270141601562,
+      "learning_rate": 1.1023622047244095e-05,
+      "loss": 1.5098,
+      "step": 1051
+    },
+    {
+      "epoch": 0.45505420530427965,
+      "grad_norm": 23.140596389770508,
+      "learning_rate": 1.1014873140857394e-05,
+      "loss": 1.5424,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4554867663359377,
+      "grad_norm": 20.646238327026367,
+      "learning_rate": 1.1006124234470692e-05,
+      "loss": 1.5512,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4559193273675958,
+      "grad_norm": 21.41969871520996,
+      "learning_rate": 1.099737532808399e-05,
+      "loss": 1.5698,
+      "step": 1054
+    },
+    {
+      "epoch": 0.45635188839925384,
+      "grad_norm": 20.478525161743164,
+      "learning_rate": 1.0988626421697289e-05,
+      "loss": 1.5936,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4567844494309119,
+      "grad_norm": 21.19014549255371,
+      "learning_rate": 1.0979877515310586e-05,
+      "loss": 1.5705,
+      "step": 1056
+    },
+    {
+      "epoch": 0.45721701046256996,
+      "grad_norm": 21.32801055908203,
+      "learning_rate": 1.0971128608923884e-05,
+      "loss": 1.5308,
+      "step": 1057
+    },
+    {
+      "epoch": 0.457649571494228,
+      "grad_norm": 21.304298400878906,
+      "learning_rate": 1.0962379702537185e-05,
+      "loss": 1.5296,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4580821325258861,
+      "grad_norm": 19.55242156982422,
+      "learning_rate": 1.0953630796150482e-05,
+      "loss": 1.6441,
+      "step": 1059
+    },
+    {
+      "epoch": 0.45851469355754415,
+      "grad_norm": 19.598514556884766,
+      "learning_rate": 1.0944881889763781e-05,
+      "loss": 1.665,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4589472545892022,
+      "grad_norm": 21.648754119873047,
+      "learning_rate": 1.0936132983377079e-05,
+      "loss": 1.538,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4593798156208603,
+      "grad_norm": 20.485759735107422,
+      "learning_rate": 1.0927384076990376e-05,
+      "loss": 1.5405,
+      "step": 1062
+    },
+    {
+      "epoch": 0.45981237665251834,
+      "grad_norm": 21.35075569152832,
+      "learning_rate": 1.0918635170603676e-05,
+      "loss": 1.5241,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4602449376841764,
+      "grad_norm": 21.848602294921875,
+      "learning_rate": 1.0909886264216973e-05,
+      "loss": 1.5681,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4606774987158344,
+      "grad_norm": 23.42026710510254,
+      "learning_rate": 1.0901137357830272e-05,
+      "loss": 1.5778,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4611100597474925,
+      "grad_norm": 22.494203567504883,
+      "learning_rate": 1.089238845144357e-05,
+      "loss": 1.6443,
+      "step": 1066
+    },
+    {
+      "epoch": 0.46154262077915054,
+      "grad_norm": 20.862600326538086,
+      "learning_rate": 1.0883639545056867e-05,
+      "loss": 1.5278,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4619751818108086,
+      "grad_norm": 20.454578399658203,
+      "learning_rate": 1.0874890638670168e-05,
+      "loss": 1.6292,
+      "step": 1068
+    },
+    {
+      "epoch": 0.46240774284246666,
+      "grad_norm": 20.307090759277344,
+      "learning_rate": 1.0866141732283466e-05,
+      "loss": 1.665,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4628403038741247,
+      "grad_norm": 20.755657196044922,
+      "learning_rate": 1.0857392825896765e-05,
+      "loss": 1.6456,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4632728649057828,
+      "grad_norm": 22.096176147460938,
+      "learning_rate": 1.0848643919510062e-05,
+      "loss": 1.5195,
+      "step": 1071
+    },
+    {
+      "epoch": 0.46370542593744085,
+      "grad_norm": 23.748300552368164,
+      "learning_rate": 1.083989501312336e-05,
+      "loss": 1.5168,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4641379869690989,
+      "grad_norm": 20.105592727661133,
+      "learning_rate": 1.083114610673666e-05,
+      "loss": 1.5407,
+      "step": 1073
+    },
+    {
+      "epoch": 0.464570548000757,
+      "grad_norm": 19.950580596923828,
+      "learning_rate": 1.0822397200349957e-05,
+      "loss": 1.588,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46500310903241504,
+      "grad_norm": 20.73526954650879,
+      "learning_rate": 1.0813648293963254e-05,
+      "loss": 1.6316,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4654356700640731,
+      "grad_norm": 21.673297882080078,
+      "learning_rate": 1.0804899387576553e-05,
+      "loss": 1.5985,
+      "step": 1076
+    },
+    {
+      "epoch": 0.46586823109573117,
+      "grad_norm": 22.8228759765625,
+      "learning_rate": 1.0796150481189853e-05,
+      "loss": 1.5836,
+      "step": 1077
+    },
+    {
+      "epoch": 0.46630079212738923,
+      "grad_norm": 22.153406143188477,
+      "learning_rate": 1.0787401574803152e-05,
+      "loss": 1.6233,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4667333531590473,
+      "grad_norm": 21.604236602783203,
+      "learning_rate": 1.077865266841645e-05,
+      "loss": 1.5601,
+      "step": 1079
+    },
+    {
+      "epoch": 0.46716591419070536,
+      "grad_norm": 21.942169189453125,
+      "learning_rate": 1.0769903762029747e-05,
+      "loss": 1.5527,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4675984752223634,
+      "grad_norm": 21.710886001586914,
+      "learning_rate": 1.0761154855643046e-05,
+      "loss": 1.5555,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4680310362540215,
+      "grad_norm": 21.006336212158203,
+      "learning_rate": 1.0752405949256344e-05,
+      "loss": 1.6011,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46846359728567954,
+      "grad_norm": 21.56496810913086,
+      "learning_rate": 1.0743657042869641e-05,
+      "loss": 1.4629,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4688961583173376,
+      "grad_norm": 22.86446189880371,
+      "learning_rate": 1.073490813648294e-05,
+      "loss": 1.521,
+      "step": 1084
+    },
+    {
+      "epoch": 0.46932871934899567,
+      "grad_norm": 23.123199462890625,
+      "learning_rate": 1.0726159230096238e-05,
+      "loss": 1.5685,
+      "step": 1085
+    },
+    {
+      "epoch": 0.46976128038065373,
+      "grad_norm": 22.753873825073242,
+      "learning_rate": 1.0717410323709537e-05,
+      "loss": 1.668,
+      "step": 1086
+    },
+    {
+      "epoch": 0.47019384141231174,
+      "grad_norm": 22.325475692749023,
+      "learning_rate": 1.0708661417322836e-05,
+      "loss": 1.4948,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4706264024439698,
+      "grad_norm": 21.678844451904297,
+      "learning_rate": 1.0699912510936135e-05,
+      "loss": 1.5436,
+      "step": 1088
+    },
+    {
+      "epoch": 0.47105896347562787,
+      "grad_norm": 21.273897171020508,
+      "learning_rate": 1.0691163604549433e-05,
+      "loss": 1.6103,
+      "step": 1089
+    },
+    {
+      "epoch": 0.47149152450728593,
+      "grad_norm": 23.457490921020508,
+      "learning_rate": 1.068241469816273e-05,
+      "loss": 1.5893,
+      "step": 1090
+    },
+    {
+      "epoch": 0.471924085538944,
+      "grad_norm": 21.083839416503906,
+      "learning_rate": 1.067366579177603e-05,
+      "loss": 1.5555,
+      "step": 1091
+    },
+    {
+      "epoch": 0.47235664657060206,
+      "grad_norm": 25.2357120513916,
+      "learning_rate": 1.0664916885389327e-05,
+      "loss": 1.522,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4727892076022601,
+      "grad_norm": 23.40666961669922,
+      "learning_rate": 1.0656167979002625e-05,
+      "loss": 1.4986,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4732217686339182,
+      "grad_norm": 20.224884033203125,
+      "learning_rate": 1.0647419072615924e-05,
+      "loss": 1.6008,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47365432966557625,
+      "grad_norm": 19.231700897216797,
+      "learning_rate": 1.0638670166229221e-05,
+      "loss": 1.5874,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4740868906972343,
+      "grad_norm": 21.34067726135254,
+      "learning_rate": 1.0629921259842522e-05,
+      "loss": 1.5708,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47451945172889237,
+      "grad_norm": 20.3139705657959,
+      "learning_rate": 1.062117235345582e-05,
+      "loss": 1.5125,
+      "step": 1097
+    },
+    {
+      "epoch": 0.47495201276055043,
+      "grad_norm": 25.667116165161133,
+      "learning_rate": 1.0612423447069117e-05,
+      "loss": 1.5303,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4753845737922085,
+      "grad_norm": 22.950712203979492,
+      "learning_rate": 1.0603674540682417e-05,
+      "loss": 1.6314,
+      "step": 1099
+    },
+    {
+      "epoch": 0.47581713482386656,
+      "grad_norm": 20.83335304260254,
+      "learning_rate": 1.0594925634295714e-05,
+      "loss": 1.5334,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4762496958555246,
+      "grad_norm": 19.24706268310547,
+      "learning_rate": 1.0586176727909012e-05,
+      "loss": 1.6256,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4766822568871827,
+      "grad_norm": 20.814455032348633,
+      "learning_rate": 1.057742782152231e-05,
+      "loss": 1.5851,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47711481791884075,
+      "grad_norm": 21.661409378051758,
+      "learning_rate": 1.0568678915135608e-05,
+      "loss": 1.5247,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4775473789504988,
+      "grad_norm": 20.899145126342773,
+      "learning_rate": 1.0559930008748906e-05,
+      "loss": 1.5708,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4779799399821569,
+      "grad_norm": 24.386003494262695,
+      "learning_rate": 1.0551181102362205e-05,
+      "loss": 1.5527,
+      "step": 1105
+    },
+    {
+      "epoch": 0.47841250101381494,
+      "grad_norm": 25.34111785888672,
+      "learning_rate": 1.0542432195975504e-05,
+      "loss": 1.5851,
+      "step": 1106
+    },
+    {
+      "epoch": 0.478845062045473,
+      "grad_norm": 24.537324905395508,
+      "learning_rate": 1.0533683289588803e-05,
+      "loss": 1.596,
+      "step": 1107
+    },
+    {
+      "epoch": 0.47927762307713107,
+      "grad_norm": 21.804595947265625,
+      "learning_rate": 1.0524934383202101e-05,
+      "loss": 1.5322,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4797101841087891,
+      "grad_norm": 20.191699981689453,
+      "learning_rate": 1.0516185476815398e-05,
+      "loss": 1.5758,
+      "step": 1109
+    },
+    {
+      "epoch": 0.48014274514044714,
+      "grad_norm": 21.43497085571289,
+      "learning_rate": 1.0507436570428698e-05,
+      "loss": 1.508,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4805753061721052,
+      "grad_norm": 20.207677841186523,
+      "learning_rate": 1.0498687664041995e-05,
+      "loss": 1.6158,
+      "step": 1111
+    },
+    {
+      "epoch": 0.48100786720376326,
+      "grad_norm": 19.458316802978516,
+      "learning_rate": 1.0489938757655294e-05,
+      "loss": 1.5859,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4814404282354213,
+      "grad_norm": 24.825956344604492,
+      "learning_rate": 1.0481189851268592e-05,
+      "loss": 1.5391,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4818729892670794,
+      "grad_norm": 22.562938690185547,
+      "learning_rate": 1.047244094488189e-05,
+      "loss": 1.5431,
+      "step": 1114
+    },
+    {
+      "epoch": 0.48230555029873745,
+      "grad_norm": 22.471332550048828,
+      "learning_rate": 1.0463692038495189e-05,
+      "loss": 1.5292,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4827381113303955,
+      "grad_norm": 21.53379249572754,
+      "learning_rate": 1.0454943132108488e-05,
+      "loss": 1.5683,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4831706723620536,
+      "grad_norm": 19.84793472290039,
+      "learning_rate": 1.0446194225721787e-05,
+      "loss": 1.4873,
+      "step": 1117
+    },
+    {
+      "epoch": 0.48360323339371164,
+      "grad_norm": 21.779499053955078,
+      "learning_rate": 1.0437445319335085e-05,
+      "loss": 1.5814,
+      "step": 1118
+    },
+    {
+      "epoch": 0.4840357944253697,
+      "grad_norm": 20.556554794311523,
+      "learning_rate": 1.0428696412948382e-05,
+      "loss": 1.5547,
+      "step": 1119
+    },
+    {
+      "epoch": 0.48446835545702777,
+      "grad_norm": 20.402372360229492,
+      "learning_rate": 1.0419947506561681e-05,
+      "loss": 1.5654,
+      "step": 1120
+    },
+    {
+      "epoch": 0.48490091648868583,
+      "grad_norm": 22.942962646484375,
+      "learning_rate": 1.0411198600174979e-05,
+      "loss": 1.6105,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4853334775203439,
+      "grad_norm": 20.789451599121094,
+      "learning_rate": 1.0402449693788276e-05,
+      "loss": 1.5835,
+      "step": 1122
+    },
+    {
+      "epoch": 0.48576603855200196,
+      "grad_norm": 20.95735740661621,
+      "learning_rate": 1.0393700787401575e-05,
+      "loss": 1.6086,
+      "step": 1123
+    },
+    {
+      "epoch": 0.48619859958366,
+      "grad_norm": 24.3394775390625,
+      "learning_rate": 1.0384951881014873e-05,
+      "loss": 1.5987,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4866311606153181,
+      "grad_norm": 19.394668579101562,
+      "learning_rate": 1.0376202974628174e-05,
+      "loss": 1.5567,
+      "step": 1125
+    },
+    {
+      "epoch": 0.48706372164697614,
+      "grad_norm": 23.74590492248535,
+      "learning_rate": 1.0367454068241471e-05,
+      "loss": 1.5349,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4874962826786342,
+      "grad_norm": 19.871522903442383,
+      "learning_rate": 1.0358705161854769e-05,
+      "loss": 1.5797,
+      "step": 1127
+    },
+    {
+      "epoch": 0.48792884371029227,
+      "grad_norm": 20.489187240600586,
+      "learning_rate": 1.0349956255468068e-05,
+      "loss": 1.5805,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48836140474195033,
+      "grad_norm": 22.527732849121094,
+      "learning_rate": 1.0341207349081366e-05,
+      "loss": 1.5333,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4887939657736084,
+      "grad_norm": 23.267118453979492,
+      "learning_rate": 1.0332458442694663e-05,
+      "loss": 1.5872,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4892265268052664,
+      "grad_norm": 21.272428512573242,
+      "learning_rate": 1.0323709536307962e-05,
+      "loss": 1.5837,
+      "step": 1131
+    },
+    {
+      "epoch": 0.48965908783692447,
+      "grad_norm": 24.80011558532715,
+      "learning_rate": 1.031496062992126e-05,
+      "loss": 1.6086,
+      "step": 1132
+    },
+    {
+      "epoch": 0.49009164886858253,
+      "grad_norm": 21.551164627075195,
+      "learning_rate": 1.0306211723534559e-05,
+      "loss": 1.5909,
+      "step": 1133
+    },
+    {
+      "epoch": 0.4905242099002406,
+      "grad_norm": 19.916059494018555,
+      "learning_rate": 1.0297462817147857e-05,
+      "loss": 1.5726,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49095677093189866,
+      "grad_norm": 22.31676483154297,
+      "learning_rate": 1.0288713910761157e-05,
+      "loss": 1.5,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4913893319635567,
+      "grad_norm": 21.02925682067871,
+      "learning_rate": 1.0279965004374455e-05,
+      "loss": 1.5122,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4918218929952148,
+      "grad_norm": 20.82086181640625,
+      "learning_rate": 1.0271216097987753e-05,
+      "loss": 1.5684,
+      "step": 1137
+    },
+    {
+      "epoch": 0.49225445402687285,
+      "grad_norm": 20.781131744384766,
+      "learning_rate": 1.0262467191601052e-05,
+      "loss": 1.6631,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4926870150585309,
+      "grad_norm": 20.078981399536133,
+      "learning_rate": 1.025371828521435e-05,
+      "loss": 1.5923,
+      "step": 1139
+    },
+    {
+      "epoch": 0.49311957609018897,
+      "grad_norm": 21.633108139038086,
+      "learning_rate": 1.0244969378827647e-05,
+      "loss": 1.5812,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49355213712184703,
+      "grad_norm": 21.029897689819336,
+      "learning_rate": 1.0236220472440946e-05,
+      "loss": 1.5658,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4939846981535051,
+      "grad_norm": 20.9033203125,
+      "learning_rate": 1.0227471566054243e-05,
+      "loss": 1.5775,
+      "step": 1142
+    },
+    {
+      "epoch": 0.49441725918516316,
+      "grad_norm": 21.990711212158203,
+      "learning_rate": 1.0218722659667541e-05,
+      "loss": 1.6188,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4948498202168212,
+      "grad_norm": 21.208660125732422,
+      "learning_rate": 1.0209973753280842e-05,
+      "loss": 1.5635,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4952823812484793,
+      "grad_norm": 23.516277313232422,
+      "learning_rate": 1.020122484689414e-05,
+      "loss": 1.6485,
+      "step": 1145
+    },
+    {
+      "epoch": 0.49571494228013735,
+      "grad_norm": 24.508100509643555,
+      "learning_rate": 1.0192475940507439e-05,
+      "loss": 1.5996,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4961475033117954,
+      "grad_norm": 23.392810821533203,
+      "learning_rate": 1.0183727034120736e-05,
+      "loss": 1.5059,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4965800643434535,
+      "grad_norm": 20.714237213134766,
+      "learning_rate": 1.0174978127734034e-05,
+      "loss": 1.533,
+      "step": 1148
+    },
+    {
+      "epoch": 0.49701262537511154,
+      "grad_norm": 23.301856994628906,
+      "learning_rate": 1.0166229221347333e-05,
+      "loss": 1.5304,
+      "step": 1149
+    },
+    {
+      "epoch": 0.4974451864067696,
+      "grad_norm": 21.122617721557617,
+      "learning_rate": 1.015748031496063e-05,
+      "loss": 1.4702,
+      "step": 1150
+    },
+    {
+      "epoch": 0.49787774743842766,
+      "grad_norm": 24.017070770263672,
+      "learning_rate": 1.0148731408573928e-05,
+      "loss": 1.5159,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4983103084700857,
+      "grad_norm": 22.126039505004883,
+      "learning_rate": 1.0139982502187227e-05,
+      "loss": 1.5913,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49874286950174374,
+      "grad_norm": 21.73690414428711,
+      "learning_rate": 1.0131233595800525e-05,
+      "loss": 1.625,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4991754305334018,
+      "grad_norm": 21.988893508911133,
+      "learning_rate": 1.0122484689413825e-05,
+      "loss": 1.5914,
+      "step": 1154
+    },
+    {
+      "epoch": 0.49960799156505986,
+      "grad_norm": 20.745668411254883,
+      "learning_rate": 1.0113735783027123e-05,
+      "loss": 1.5551,
+      "step": 1155
+    },
+    {
+      "epoch": 0.500040552596718,
+      "grad_norm": 21.369321823120117,
+      "learning_rate": 1.010498687664042e-05,
+      "loss": 1.5585,
+      "step": 1156
+    },
+    {
+      "epoch": 0.500473113628376,
+      "grad_norm": 21.196489334106445,
+      "learning_rate": 1.009623797025372e-05,
+      "loss": 1.4679,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5009056746600341,
+      "grad_norm": 20.842437744140625,
+      "learning_rate": 1.0087489063867017e-05,
+      "loss": 1.5539,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5013382356916921,
+      "grad_norm": 22.951557159423828,
+      "learning_rate": 1.0078740157480316e-05,
+      "loss": 1.5833,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5017707967233502,
+      "grad_norm": 24.61578941345215,
+      "learning_rate": 1.0069991251093614e-05,
+      "loss": 1.4874,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5022033577550082,
+      "grad_norm": 20.68556785583496,
+      "learning_rate": 1.0061242344706911e-05,
+      "loss": 1.5937,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5026359187866664,
+      "grad_norm": 23.46632957458496,
+      "learning_rate": 1.005249343832021e-05,
+      "loss": 1.46,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5030684798183244,
+      "grad_norm": 21.403005599975586,
+      "learning_rate": 1.0043744531933508e-05,
+      "loss": 1.5609,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5035010408499824,
+      "grad_norm": 23.224864959716797,
+      "learning_rate": 1.0034995625546809e-05,
+      "loss": 1.5493,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5039336018816405,
+      "grad_norm": 22.939931869506836,
+      "learning_rate": 1.0026246719160107e-05,
+      "loss": 1.5497,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5043661629132985,
+      "grad_norm": 21.148677825927734,
+      "learning_rate": 1.0017497812773404e-05,
+      "loss": 1.6063,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5047987239449566,
+      "grad_norm": 24.93348503112793,
+      "learning_rate": 1.0008748906386703e-05,
+      "loss": 1.4764,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5052312849766146,
+      "grad_norm": 22.30760955810547,
+      "learning_rate": 1e-05,
+      "loss": 1.557,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5056638460082727,
+      "grad_norm": 23.62921142578125,
+      "learning_rate": 9.991251093613298e-06,
+      "loss": 1.6243,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5060964070399308,
+      "grad_norm": 23.39278221130371,
+      "learning_rate": 9.982502187226598e-06,
+      "loss": 1.5429,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5065289680715889,
+      "grad_norm": 21.39177703857422,
+      "learning_rate": 9.973753280839897e-06,
+      "loss": 1.5076,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5069615291032469,
+      "grad_norm": 20.722728729248047,
+      "learning_rate": 9.965004374453194e-06,
+      "loss": 1.4859,
+      "step": 1172
+    },
+    {
+      "epoch": 0.507394090134905,
+      "grad_norm": 21.878719329833984,
+      "learning_rate": 9.956255468066492e-06,
+      "loss": 1.5704,
+      "step": 1173
+    },
+    {
+      "epoch": 0.507826651166563,
+      "grad_norm": 21.7126407623291,
+      "learning_rate": 9.947506561679791e-06,
+      "loss": 1.5717,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5082592121982211,
+      "grad_norm": 24.65056037902832,
+      "learning_rate": 9.938757655293088e-06,
+      "loss": 1.5916,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5086917732298791,
+      "grad_norm": 22.16351318359375,
+      "learning_rate": 9.930008748906388e-06,
+      "loss": 1.6863,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5091243342615372,
+      "grad_norm": 21.595354080200195,
+      "learning_rate": 9.921259842519685e-06,
+      "loss": 1.5668,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5095568952931953,
+      "grad_norm": 21.123828887939453,
+      "learning_rate": 9.912510936132984e-06,
+      "loss": 1.6015,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5099894563248534,
+      "grad_norm": 19.92538833618164,
+      "learning_rate": 9.903762029746282e-06,
+      "loss": 1.5397,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5104220173565114,
+      "grad_norm": 19.72816276550293,
+      "learning_rate": 9.895013123359581e-06,
+      "loss": 1.5909,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5108545783881695,
+      "grad_norm": 23.017501831054688,
+      "learning_rate": 9.88626421697288e-06,
+      "loss": 1.5516,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5112871394198275,
+      "grad_norm": 20.519594192504883,
+      "learning_rate": 9.877515310586178e-06,
+      "loss": 1.6408,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5117197004514856,
+      "grad_norm": 23.874664306640625,
+      "learning_rate": 9.868766404199475e-06,
+      "loss": 1.5156,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5121522614831436,
+      "grad_norm": 22.534914016723633,
+      "learning_rate": 9.860017497812775e-06,
+      "loss": 1.5178,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5125848225148018,
+      "grad_norm": 23.634246826171875,
+      "learning_rate": 9.851268591426074e-06,
+      "loss": 1.5734,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5130173835464598,
+      "grad_norm": 21.05173683166504,
+      "learning_rate": 9.842519685039371e-06,
+      "loss": 1.5886,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5134499445781178,
+      "grad_norm": 22.97358512878418,
+      "learning_rate": 9.833770778652669e-06,
+      "loss": 1.5493,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5138825056097759,
+      "grad_norm": 23.60778045654297,
+      "learning_rate": 9.825021872265968e-06,
+      "loss": 1.5308,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5143150666414339,
+      "grad_norm": 19.591341018676758,
+      "learning_rate": 9.816272965879266e-06,
+      "loss": 1.6701,
+      "step": 1189
+    },
+    {
+      "epoch": 0.514747627673092,
+      "grad_norm": 22.59847640991211,
+      "learning_rate": 9.807524059492565e-06,
+      "loss": 1.7357,
+      "step": 1190
+    },
+    {
+      "epoch": 0.51518018870475,
+      "grad_norm": 21.31041145324707,
+      "learning_rate": 9.798775153105862e-06,
+      "loss": 1.6354,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5156127497364081,
+      "grad_norm": 26.626367568969727,
+      "learning_rate": 9.790026246719161e-06,
+      "loss": 1.5627,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5160453107680661,
+      "grad_norm": 20.857505798339844,
+      "learning_rate": 9.781277340332459e-06,
+      "loss": 1.5965,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5164778717997243,
+      "grad_norm": 20.571149826049805,
+      "learning_rate": 9.772528433945756e-06,
+      "loss": 1.5661,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5169104328313823,
+      "grad_norm": 23.989408493041992,
+      "learning_rate": 9.763779527559056e-06,
+      "loss": 1.5563,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5173429938630404,
+      "grad_norm": 21.287750244140625,
+      "learning_rate": 9.755030621172355e-06,
+      "loss": 1.5813,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5177755548946984,
+      "grad_norm": 19.994047164916992,
+      "learning_rate": 9.746281714785652e-06,
+      "loss": 1.4655,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5182081159263565,
+      "grad_norm": 22.351661682128906,
+      "learning_rate": 9.73753280839895e-06,
+      "loss": 1.5156,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5186406769580145,
+      "grad_norm": 21.73786163330078,
+      "learning_rate": 9.728783902012249e-06,
+      "loss": 1.5882,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5190732379896726,
+      "grad_norm": 24.997068405151367,
+      "learning_rate": 9.720034995625548e-06,
+      "loss": 1.5392,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5195057990213306,
+      "grad_norm": 24.620372772216797,
+      "learning_rate": 9.711286089238846e-06,
+      "loss": 1.6197,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5199383600529888,
+      "grad_norm": 20.77781105041504,
+      "learning_rate": 9.702537182852143e-06,
+      "loss": 1.5079,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5203709210846468,
+      "grad_norm": 23.355987548828125,
+      "learning_rate": 9.693788276465443e-06,
+      "loss": 1.5842,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5208034821163049,
+      "grad_norm": 19.493812561035156,
+      "learning_rate": 9.68503937007874e-06,
+      "loss": 1.4915,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5212360431479629,
+      "grad_norm": 20.286083221435547,
+      "learning_rate": 9.67629046369204e-06,
+      "loss": 1.5201,
+      "step": 1205
+    },
+    {
+      "epoch": 0.521668604179621,
+      "grad_norm": 19.314844131469727,
+      "learning_rate": 9.667541557305338e-06,
+      "loss": 1.6382,
+      "step": 1206
+    },
+    {
+      "epoch": 0.522101165211279,
+      "grad_norm": 24.922771453857422,
+      "learning_rate": 9.658792650918636e-06,
+      "loss": 1.5938,
+      "step": 1207
+    },
+    {
+      "epoch": 0.522533726242937,
+      "grad_norm": 23.530563354492188,
+      "learning_rate": 9.650043744531934e-06,
+      "loss": 1.5612,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5229662872745952,
+      "grad_norm": 23.83565902709961,
+      "learning_rate": 9.641294838145233e-06,
+      "loss": 1.6478,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5233988483062532,
+      "grad_norm": 22.341461181640625,
+      "learning_rate": 9.632545931758532e-06,
+      "loss": 1.5618,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5238314093379113,
+      "grad_norm": 20.392044067382812,
+      "learning_rate": 9.62379702537183e-06,
+      "loss": 1.5519,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5242639703695693,
+      "grad_norm": 22.426546096801758,
+      "learning_rate": 9.615048118985127e-06,
+      "loss": 1.5484,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5246965314012274,
+      "grad_norm": 23.25884437561035,
+      "learning_rate": 9.606299212598426e-06,
+      "loss": 1.5351,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5251290924328854,
+      "grad_norm": 20.078369140625,
+      "learning_rate": 9.597550306211725e-06,
+      "loss": 1.5788,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5255616534645435,
+      "grad_norm": 21.024873733520508,
+      "learning_rate": 9.588801399825023e-06,
+      "loss": 1.5521,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5259942144962015,
+      "grad_norm": 21.520021438598633,
+      "learning_rate": 9.58005249343832e-06,
+      "loss": 1.5283,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5264267755278597,
+      "grad_norm": 23.06103515625,
+      "learning_rate": 9.57130358705162e-06,
+      "loss": 1.5912,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5268593365595177,
+      "grad_norm": 23.135066986083984,
+      "learning_rate": 9.562554680664917e-06,
+      "loss": 1.5659,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5272918975911758,
+      "grad_norm": 22.494586944580078,
+      "learning_rate": 9.553805774278216e-06,
+      "loss": 1.5495,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5277244586228338,
+      "grad_norm": 24.494667053222656,
+      "learning_rate": 9.545056867891514e-06,
+      "loss": 1.4783,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5281570196544919,
+      "grad_norm": 22.232500076293945,
+      "learning_rate": 9.536307961504813e-06,
+      "loss": 1.6506,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5285895806861499,
+      "grad_norm": 20.534900665283203,
+      "learning_rate": 9.52755905511811e-06,
+      "loss": 1.4604,
+      "step": 1222
+    },
+    {
+      "epoch": 0.529022141717808,
+      "grad_norm": 22.456693649291992,
+      "learning_rate": 9.518810148731408e-06,
+      "loss": 1.5402,
+      "step": 1223
+    },
+    {
+      "epoch": 0.529454702749466,
+      "grad_norm": 24.285858154296875,
+      "learning_rate": 9.510061242344707e-06,
+      "loss": 1.6183,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5298872637811242,
+      "grad_norm": 24.39087677001953,
+      "learning_rate": 9.501312335958006e-06,
+      "loss": 1.5955,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5303198248127822,
+      "grad_norm": 23.711259841918945,
+      "learning_rate": 9.492563429571304e-06,
+      "loss": 1.4826,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5307523858444403,
+      "grad_norm": 23.221359252929688,
+      "learning_rate": 9.483814523184603e-06,
+      "loss": 1.541,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5311849468760983,
+      "grad_norm": 22.01033592224121,
+      "learning_rate": 9.4750656167979e-06,
+      "loss": 1.5117,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5316175079077564,
+      "grad_norm": 22.62164306640625,
+      "learning_rate": 9.4663167104112e-06,
+      "loss": 1.6338,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5320500689394144,
+      "grad_norm": 25.81968879699707,
+      "learning_rate": 9.457567804024497e-06,
+      "loss": 1.5945,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5324826299710724,
+      "grad_norm": 22.638782501220703,
+      "learning_rate": 9.448818897637797e-06,
+      "loss": 1.5698,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5329151910027305,
+      "grad_norm": 22.303646087646484,
+      "learning_rate": 9.440069991251094e-06,
+      "loss": 1.5867,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5333477520343886,
+      "grad_norm": 22.231616973876953,
+      "learning_rate": 9.431321084864393e-06,
+      "loss": 1.5598,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5337803130660467,
+      "grad_norm": 21.435476303100586,
+      "learning_rate": 9.422572178477691e-06,
+      "loss": 1.5487,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5342128740977047,
+      "grad_norm": 20.271909713745117,
+      "learning_rate": 9.41382327209099e-06,
+      "loss": 1.564,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5346454351293628,
+      "grad_norm": 21.8601016998291,
+      "learning_rate": 9.405074365704288e-06,
+      "loss": 1.5758,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5350779961610208,
+      "grad_norm": 18.850753784179688,
+      "learning_rate": 9.396325459317585e-06,
+      "loss": 1.6328,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5355105571926789,
+      "grad_norm": 20.55095672607422,
+      "learning_rate": 9.387576552930884e-06,
+      "loss": 1.5706,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5359431182243369,
+      "grad_norm": 22.207855224609375,
+      "learning_rate": 9.378827646544184e-06,
+      "loss": 1.5461,
+      "step": 1239
+    },
+    {
+      "epoch": 0.536375679255995,
+      "grad_norm": 22.02487564086914,
+      "learning_rate": 9.370078740157481e-06,
+      "loss": 1.5796,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5368082402876531,
+      "grad_norm": 22.791046142578125,
+      "learning_rate": 9.361329833770779e-06,
+      "loss": 1.5041,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5372408013193112,
+      "grad_norm": 21.210519790649414,
+      "learning_rate": 9.352580927384078e-06,
+      "loss": 1.541,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5376733623509692,
+      "grad_norm": 23.86467933654785,
+      "learning_rate": 9.343832020997377e-06,
+      "loss": 1.5611,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5381059233826273,
+      "grad_norm": 26.53850555419922,
+      "learning_rate": 9.335083114610674e-06,
+      "loss": 1.4797,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5385384844142853,
+      "grad_norm": 22.709075927734375,
+      "learning_rate": 9.326334208223972e-06,
+      "loss": 1.5384,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5389710454459434,
+      "grad_norm": 22.626386642456055,
+      "learning_rate": 9.317585301837271e-06,
+      "loss": 1.533,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5394036064776014,
+      "grad_norm": 22.51513671875,
+      "learning_rate": 9.308836395450569e-06,
+      "loss": 1.4891,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5398361675092596,
+      "grad_norm": 25.217557907104492,
+      "learning_rate": 9.300087489063868e-06,
+      "loss": 1.5106,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5402687285409176,
+      "grad_norm": 21.37566566467285,
+      "learning_rate": 9.291338582677165e-06,
+      "loss": 1.5115,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5407012895725757,
+      "grad_norm": 21.221803665161133,
+      "learning_rate": 9.282589676290465e-06,
+      "loss": 1.6049,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5411338506042337,
+      "grad_norm": 22.976816177368164,
+      "learning_rate": 9.273840769903762e-06,
+      "loss": 1.4837,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5415664116358917,
+      "grad_norm": 20.9698543548584,
+      "learning_rate": 9.265091863517061e-06,
+      "loss": 1.5605,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5419989726675498,
+      "grad_norm": 24.19113540649414,
+      "learning_rate": 9.25634295713036e-06,
+      "loss": 1.5536,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5424315336992078,
+      "grad_norm": 21.84588623046875,
+      "learning_rate": 9.247594050743658e-06,
+      "loss": 1.4345,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5428640947308659,
+      "grad_norm": 23.569164276123047,
+      "learning_rate": 9.238845144356956e-06,
+      "loss": 1.595,
+      "step": 1255
+    },
+    {
+      "epoch": 0.543296655762524,
+      "grad_norm": 24.85906982421875,
+      "learning_rate": 9.230096237970255e-06,
+      "loss": 1.5155,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5437292167941821,
+      "grad_norm": 19.473554611206055,
+      "learning_rate": 9.221347331583554e-06,
+      "loss": 1.5188,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5441617778258401,
+      "grad_norm": 23.21884536743164,
+      "learning_rate": 9.212598425196852e-06,
+      "loss": 1.506,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5445943388574982,
+      "grad_norm": 24.165271759033203,
+      "learning_rate": 9.203849518810149e-06,
+      "loss": 1.6034,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5450268998891562,
+      "grad_norm": 22.93230628967285,
+      "learning_rate": 9.195100612423448e-06,
+      "loss": 1.5442,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5454594609208143,
+      "grad_norm": 24.058223724365234,
+      "learning_rate": 9.186351706036746e-06,
+      "loss": 1.5162,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5458920219524723,
+      "grad_norm": 22.84027862548828,
+      "learning_rate": 9.177602799650045e-06,
+      "loss": 1.4725,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5463245829841304,
+      "grad_norm": 20.591529846191406,
+      "learning_rate": 9.168853893263342e-06,
+      "loss": 1.5473,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5467571440157885,
+      "grad_norm": 23.851757049560547,
+      "learning_rate": 9.160104986876642e-06,
+      "loss": 1.664,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5471897050474466,
+      "grad_norm": 21.493534088134766,
+      "learning_rate": 9.15135608048994e-06,
+      "loss": 1.5328,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5476222660791046,
+      "grad_norm": 20.198074340820312,
+      "learning_rate": 9.142607174103237e-06,
+      "loss": 1.5544,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5480548271107627,
+      "grad_norm": 22.380517959594727,
+      "learning_rate": 9.133858267716536e-06,
+      "loss": 1.4572,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5484873881424207,
+      "grad_norm": 22.481168746948242,
+      "learning_rate": 9.125109361329835e-06,
+      "loss": 1.5962,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5489199491740788,
+      "grad_norm": 22.247299194335938,
+      "learning_rate": 9.116360454943133e-06,
+      "loss": 1.5528,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5493525102057368,
+      "grad_norm": 19.119709014892578,
+      "learning_rate": 9.10761154855643e-06,
+      "loss": 1.4795,
+      "step": 1270
+    },
+    {
+      "epoch": 0.549785071237395,
+      "grad_norm": 23.676963806152344,
+      "learning_rate": 9.09886264216973e-06,
+      "loss": 1.6428,
+      "step": 1271
+    },
+    {
+      "epoch": 0.550217632269053,
+      "grad_norm": 22.227872848510742,
+      "learning_rate": 9.090113735783029e-06,
+      "loss": 1.5312,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5506501933007111,
+      "grad_norm": 22.329296112060547,
+      "learning_rate": 9.081364829396326e-06,
+      "loss": 1.5003,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5510827543323691,
+      "grad_norm": 24.662952423095703,
+      "learning_rate": 9.072615923009625e-06,
+      "loss": 1.5753,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5515153153640271,
+      "grad_norm": 23.11702537536621,
+      "learning_rate": 9.063867016622923e-06,
+      "loss": 1.535,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5519478763956852,
+      "grad_norm": 21.952640533447266,
+      "learning_rate": 9.05511811023622e-06,
+      "loss": 1.6321,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5523804374273432,
+      "grad_norm": 22.44942283630371,
+      "learning_rate": 9.04636920384952e-06,
+      "loss": 1.5539,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5528129984590013,
+      "grad_norm": 21.685388565063477,
+      "learning_rate": 9.037620297462819e-06,
+      "loss": 1.6138,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5532455594906593,
+      "grad_norm": 22.420637130737305,
+      "learning_rate": 9.028871391076116e-06,
+      "loss": 1.5373,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5536781205223175,
+      "grad_norm": 23.004444122314453,
+      "learning_rate": 9.020122484689414e-06,
+      "loss": 1.5647,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5541106815539755,
+      "grad_norm": 22.256057739257812,
+      "learning_rate": 9.011373578302713e-06,
+      "loss": 1.4894,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5545432425856336,
+      "grad_norm": 26.35243034362793,
+      "learning_rate": 9.002624671916012e-06,
+      "loss": 1.4642,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5549758036172916,
+      "grad_norm": 22.870492935180664,
+      "learning_rate": 8.99387576552931e-06,
+      "loss": 1.4622,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5554083646489497,
+      "grad_norm": 21.987051010131836,
+      "learning_rate": 8.985126859142607e-06,
+      "loss": 1.598,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5558409256806077,
+      "grad_norm": 21.95711898803711,
+      "learning_rate": 8.976377952755906e-06,
+      "loss": 1.557,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5562734867122658,
+      "grad_norm": 22.57794189453125,
+      "learning_rate": 8.967629046369206e-06,
+      "loss": 1.5029,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5567060477439238,
+      "grad_norm": 21.112510681152344,
+      "learning_rate": 8.958880139982503e-06,
+      "loss": 1.5741,
+      "step": 1287
+    },
+    {
+      "epoch": 0.557138608775582,
+      "grad_norm": 24.124717712402344,
+      "learning_rate": 8.9501312335958e-06,
+      "loss": 1.613,
+      "step": 1288
+    },
+    {
+      "epoch": 0.55757116980724,
+      "grad_norm": 31.32720375061035,
+      "learning_rate": 8.9413823272091e-06,
+      "loss": 1.6521,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5580037308388981,
+      "grad_norm": 23.76942253112793,
+      "learning_rate": 8.932633420822397e-06,
+      "loss": 1.5079,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5584362918705561,
+      "grad_norm": 22.088144302368164,
+      "learning_rate": 8.923884514435697e-06,
+      "loss": 1.5162,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5588688529022142,
+      "grad_norm": 21.93614387512207,
+      "learning_rate": 8.915135608048994e-06,
+      "loss": 1.4684,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5593014139338722,
+      "grad_norm": 23.71426773071289,
+      "learning_rate": 8.906386701662293e-06,
+      "loss": 1.6234,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5597339749655303,
+      "grad_norm": 23.15918731689453,
+      "learning_rate": 8.89763779527559e-06,
+      "loss": 1.5453,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5601665359971884,
+      "grad_norm": 24.360671997070312,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.5688,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5605990970288464,
+      "grad_norm": 22.142484664916992,
+      "learning_rate": 8.880139982502188e-06,
+      "loss": 1.5051,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5610316580605045,
+      "grad_norm": 23.395851135253906,
+      "learning_rate": 8.871391076115487e-06,
+      "loss": 1.6383,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5614642190921625,
+      "grad_norm": 22.678421020507812,
+      "learning_rate": 8.862642169728784e-06,
+      "loss": 1.5335,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5618967801238206,
+      "grad_norm": 23.081527709960938,
+      "learning_rate": 8.853893263342083e-06,
+      "loss": 1.6206,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5623293411554786,
+      "grad_norm": 22.97568702697754,
+      "learning_rate": 8.845144356955381e-06,
+      "loss": 1.4891,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5627619021871367,
+      "grad_norm": 22.3316707611084,
+      "learning_rate": 8.83639545056868e-06,
+      "loss": 1.6181,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5631944632187947,
+      "grad_norm": 21.807104110717773,
+      "learning_rate": 8.827646544181978e-06,
+      "loss": 1.5856,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5636270242504529,
+      "grad_norm": 23.432767868041992,
+      "learning_rate": 8.818897637795277e-06,
+      "loss": 1.5661,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5640595852821109,
+      "grad_norm": 23.371694564819336,
+      "learning_rate": 8.810148731408574e-06,
+      "loss": 1.5358,
+      "step": 1304
+    },
+    {
+      "epoch": 0.564492146313769,
+      "grad_norm": 23.72265625,
+      "learning_rate": 8.801399825021874e-06,
+      "loss": 1.5152,
+      "step": 1305
+    },
+    {
+      "epoch": 0.564924707345427,
+      "grad_norm": 20.524398803710938,
+      "learning_rate": 8.792650918635171e-06,
+      "loss": 1.4969,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5653572683770851,
+      "grad_norm": 24.223081588745117,
+      "learning_rate": 8.78390201224847e-06,
+      "loss": 1.5267,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5657898294087431,
+      "grad_norm": 21.053010940551758,
+      "learning_rate": 8.775153105861768e-06,
+      "loss": 1.5373,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5662223904404012,
+      "grad_norm": 24.919174194335938,
+      "learning_rate": 8.766404199475065e-06,
+      "loss": 1.5295,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5666549514720592,
+      "grad_norm": 20.551292419433594,
+      "learning_rate": 8.757655293088365e-06,
+      "loss": 1.5628,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5670875125037174,
+      "grad_norm": 22.388694763183594,
+      "learning_rate": 8.748906386701664e-06,
+      "loss": 1.5315,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5675200735353754,
+      "grad_norm": 22.377506256103516,
+      "learning_rate": 8.740157480314961e-06,
+      "loss": 1.5567,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5679526345670335,
+      "grad_norm": 21.797183990478516,
+      "learning_rate": 8.731408573928259e-06,
+      "loss": 1.524,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5683851955986915,
+      "grad_norm": 25.161802291870117,
+      "learning_rate": 8.722659667541558e-06,
+      "loss": 1.5708,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5688177566303496,
+      "grad_norm": 22.526390075683594,
+      "learning_rate": 8.713910761154857e-06,
+      "loss": 1.543,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5692503176620076,
+      "grad_norm": 22.93593978881836,
+      "learning_rate": 8.705161854768155e-06,
+      "loss": 1.5886,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5696828786936657,
+      "grad_norm": 21.80086326599121,
+      "learning_rate": 8.696412948381452e-06,
+      "loss": 1.5182,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5701154397253237,
+      "grad_norm": 26.52342987060547,
+      "learning_rate": 8.687664041994751e-06,
+      "loss": 1.5266,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5705480007569818,
+      "grad_norm": 23.737064361572266,
+      "learning_rate": 8.678915135608049e-06,
+      "loss": 1.5446,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5709805617886399,
+      "grad_norm": 23.06212615966797,
+      "learning_rate": 8.670166229221348e-06,
+      "loss": 1.5367,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5714131228202979,
+      "grad_norm": 22.22346305847168,
+      "learning_rate": 8.661417322834647e-06,
+      "loss": 1.5132,
+      "step": 1321
+    },
+    {
+      "epoch": 0.571845683851956,
+      "grad_norm": 23.046924591064453,
+      "learning_rate": 8.652668416447945e-06,
+      "loss": 1.6097,
+      "step": 1322
+    },
+    {
+      "epoch": 0.572278244883614,
+      "grad_norm": 23.361909866333008,
+      "learning_rate": 8.643919510061242e-06,
+      "loss": 1.5469,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5727108059152721,
+      "grad_norm": 23.01518440246582,
+      "learning_rate": 8.635170603674542e-06,
+      "loss": 1.5453,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5731433669469301,
+      "grad_norm": 22.828317642211914,
+      "learning_rate": 8.62642169728784e-06,
+      "loss": 1.4921,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5735759279785883,
+      "grad_norm": 27.066354751586914,
+      "learning_rate": 8.617672790901138e-06,
+      "loss": 1.4901,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5740084890102463,
+      "grad_norm": 23.641551971435547,
+      "learning_rate": 8.608923884514436e-06,
+      "loss": 1.5495,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5744410500419044,
+      "grad_norm": 22.956571578979492,
+      "learning_rate": 8.600174978127735e-06,
+      "loss": 1.5145,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5748736110735624,
+      "grad_norm": 21.42229461669922,
+      "learning_rate": 8.591426071741034e-06,
+      "loss": 1.4778,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5753061721052205,
+      "grad_norm": 23.688533782958984,
+      "learning_rate": 8.582677165354332e-06,
+      "loss": 1.4983,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5757387331368785,
+      "grad_norm": 24.6751708984375,
+      "learning_rate": 8.57392825896763e-06,
+      "loss": 1.4807,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5761712941685366,
+      "grad_norm": 25.89187240600586,
+      "learning_rate": 8.565179352580928e-06,
+      "loss": 1.6793,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5766038552001946,
+      "grad_norm": 21.923139572143555,
+      "learning_rate": 8.556430446194226e-06,
+      "loss": 1.4787,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5770364162318528,
+      "grad_norm": 27.02901840209961,
+      "learning_rate": 8.547681539807525e-06,
+      "loss": 1.5807,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5774689772635108,
+      "grad_norm": 24.652677536010742,
+      "learning_rate": 8.538932633420823e-06,
+      "loss": 1.5291,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5779015382951689,
+      "grad_norm": 22.28380584716797,
+      "learning_rate": 8.530183727034122e-06,
+      "loss": 1.5341,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5783340993268269,
+      "grad_norm": 27.593717575073242,
+      "learning_rate": 8.52143482064742e-06,
+      "loss": 1.5058,
+      "step": 1337
+    },
+    {
+      "epoch": 0.578766660358485,
+      "grad_norm": 23.606679916381836,
+      "learning_rate": 8.512685914260717e-06,
+      "loss": 1.5453,
+      "step": 1338
+    },
+    {
+      "epoch": 0.579199221390143,
+      "grad_norm": 19.93859100341797,
+      "learning_rate": 8.503937007874016e-06,
+      "loss": 1.6098,
+      "step": 1339
+    },
+    {
+      "epoch": 0.579631782421801,
+      "grad_norm": 21.376535415649414,
+      "learning_rate": 8.495188101487315e-06,
+      "loss": 1.6096,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5800643434534591,
+      "grad_norm": 22.244482040405273,
+      "learning_rate": 8.486439195100613e-06,
+      "loss": 1.5739,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5804969044851171,
+      "grad_norm": 21.95546531677246,
+      "learning_rate": 8.47769028871391e-06,
+      "loss": 1.5556,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5809294655167753,
+      "grad_norm": 23.818946838378906,
+      "learning_rate": 8.46894138232721e-06,
+      "loss": 1.6165,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5813620265484333,
+      "grad_norm": 24.244600296020508,
+      "learning_rate": 8.460192475940509e-06,
+      "loss": 1.5368,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5817945875800914,
+      "grad_norm": 23.75138282775879,
+      "learning_rate": 8.451443569553806e-06,
+      "loss": 1.5739,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5822271486117494,
+      "grad_norm": 21.759607315063477,
+      "learning_rate": 8.442694663167106e-06,
+      "loss": 1.5693,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5826597096434075,
+      "grad_norm": 21.369449615478516,
+      "learning_rate": 8.433945756780403e-06,
+      "loss": 1.509,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5830922706750655,
+      "grad_norm": 21.341304779052734,
+      "learning_rate": 8.4251968503937e-06,
+      "loss": 1.5313,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5835248317067236,
+      "grad_norm": 21.93410873413086,
+      "learning_rate": 8.416447944007e-06,
+      "loss": 1.4952,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5839573927383817,
+      "grad_norm": 21.560096740722656,
+      "learning_rate": 8.407699037620299e-06,
+      "loss": 1.587,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5843899537700398,
+      "grad_norm": 22.302555084228516,
+      "learning_rate": 8.398950131233596e-06,
+      "loss": 1.4969,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5848225148016978,
+      "grad_norm": 24.07124137878418,
+      "learning_rate": 8.390201224846894e-06,
+      "loss": 1.6227,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5852550758333559,
+      "grad_norm": 21.081422805786133,
+      "learning_rate": 8.381452318460193e-06,
+      "loss": 1.5545,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5856876368650139,
+      "grad_norm": 20.232650756835938,
+      "learning_rate": 8.372703412073492e-06,
+      "loss": 1.5344,
+      "step": 1354
+    },
+    {
+      "epoch": 0.586120197896672,
+      "grad_norm": 21.190690994262695,
+      "learning_rate": 8.36395450568679e-06,
+      "loss": 1.6478,
+      "step": 1355
+    },
+    {
+      "epoch": 0.58655275892833,
+      "grad_norm": 21.58733367919922,
+      "learning_rate": 8.355205599300087e-06,
+      "loss": 1.577,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5869853199599881,
+      "grad_norm": 21.453083038330078,
+      "learning_rate": 8.346456692913387e-06,
+      "loss": 1.5533,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5874178809916462,
+      "grad_norm": 20.776641845703125,
+      "learning_rate": 8.337707786526686e-06,
+      "loss": 1.5206,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5878504420233043,
+      "grad_norm": 24.146244049072266,
+      "learning_rate": 8.328958880139983e-06,
+      "loss": 1.5626,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5882830030549623,
+      "grad_norm": 19.88344383239746,
+      "learning_rate": 8.320209973753281e-06,
+      "loss": 1.5167,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5887155640866204,
+      "grad_norm": 21.336252212524414,
+      "learning_rate": 8.31146106736658e-06,
+      "loss": 1.5685,
+      "step": 1361
+    },
+    {
+      "epoch": 0.5891481251182784,
+      "grad_norm": 23.71305274963379,
+      "learning_rate": 8.302712160979878e-06,
+      "loss": 1.5559,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5895806861499364,
+      "grad_norm": 24.64626693725586,
+      "learning_rate": 8.293963254593177e-06,
+      "loss": 1.5773,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5900132471815945,
+      "grad_norm": 24.751689910888672,
+      "learning_rate": 8.285214348206474e-06,
+      "loss": 1.5788,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5904458082132525,
+      "grad_norm": 22.55054473876953,
+      "learning_rate": 8.276465441819773e-06,
+      "loss": 1.6049,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5908783692449107,
+      "grad_norm": 23.089393615722656,
+      "learning_rate": 8.267716535433071e-06,
+      "loss": 1.5444,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5913109302765687,
+      "grad_norm": 20.179182052612305,
+      "learning_rate": 8.25896762904637e-06,
+      "loss": 1.5167,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5917434913082268,
+      "grad_norm": 22.71483039855957,
+      "learning_rate": 8.25021872265967e-06,
+      "loss": 1.5535,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5921760523398848,
+      "grad_norm": 23.300128936767578,
+      "learning_rate": 8.241469816272967e-06,
+      "loss": 1.5337,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5926086133715429,
+      "grad_norm": 23.242700576782227,
+      "learning_rate": 8.232720909886264e-06,
+      "loss": 1.5624,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5930411744032009,
+      "grad_norm": 22.0483341217041,
+      "learning_rate": 8.223972003499564e-06,
+      "loss": 1.5575,
+      "step": 1371
+    },
+    {
+      "epoch": 0.593473735434859,
+      "grad_norm": 24.035541534423828,
+      "learning_rate": 8.215223097112861e-06,
+      "loss": 1.4662,
+      "step": 1372
+    },
+    {
+      "epoch": 0.593906296466517,
+      "grad_norm": 21.212360382080078,
+      "learning_rate": 8.20647419072616e-06,
+      "loss": 1.5534,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5943388574981752,
+      "grad_norm": 23.624034881591797,
+      "learning_rate": 8.197725284339458e-06,
+      "loss": 1.5738,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5947714185298332,
+      "grad_norm": 23.664756774902344,
+      "learning_rate": 8.188976377952757e-06,
+      "loss": 1.5385,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5952039795614913,
+      "grad_norm": 23.286773681640625,
+      "learning_rate": 8.180227471566055e-06,
+      "loss": 1.5215,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5956365405931493,
+      "grad_norm": 24.892528533935547,
+      "learning_rate": 8.171478565179354e-06,
+      "loss": 1.471,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5960691016248074,
+      "grad_norm": 25.03875160217285,
+      "learning_rate": 8.162729658792651e-06,
+      "loss": 1.6867,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5965016626564654,
+      "grad_norm": 24.291664123535156,
+      "learning_rate": 8.15398075240595e-06,
+      "loss": 1.5568,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5969342236881235,
+      "grad_norm": 21.74358367919922,
+      "learning_rate": 8.145231846019248e-06,
+      "loss": 1.4808,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5973667847197816,
+      "grad_norm": 20.80985450744629,
+      "learning_rate": 8.136482939632546e-06,
+      "loss": 1.6065,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5977993457514397,
+      "grad_norm": 23.85978126525879,
+      "learning_rate": 8.127734033245845e-06,
+      "loss": 1.5018,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5982319067830977,
+      "grad_norm": 23.80008316040039,
+      "learning_rate": 8.118985126859144e-06,
+      "loss": 1.4797,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5986644678147557,
+      "grad_norm": 21.72329330444336,
+      "learning_rate": 8.110236220472441e-06,
+      "loss": 1.5553,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5990970288464138,
+      "grad_norm": 23.285947799682617,
+      "learning_rate": 8.101487314085739e-06,
+      "loss": 1.4434,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5995295898780718,
+      "grad_norm": 24.382734298706055,
+      "learning_rate": 8.092738407699038e-06,
+      "loss": 1.6215,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5999621509097299,
+      "grad_norm": 22.842208862304688,
+      "learning_rate": 8.083989501312337e-06,
+      "loss": 1.5772,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6003947119413879,
+      "grad_norm": 22.464563369750977,
+      "learning_rate": 8.075240594925635e-06,
+      "loss": 1.5454,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6008272729730461,
+      "grad_norm": 23.2418155670166,
+      "learning_rate": 8.066491688538932e-06,
+      "loss": 1.4773,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6012598340047041,
+      "grad_norm": 24.3996639251709,
+      "learning_rate": 8.057742782152232e-06,
+      "loss": 1.4524,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6016923950363622,
+      "grad_norm": 21.880083084106445,
+      "learning_rate": 8.04899387576553e-06,
+      "loss": 1.5388,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6021249560680202,
+      "grad_norm": 21.42142105102539,
+      "learning_rate": 8.040244969378828e-06,
+      "loss": 1.5757,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6025575170996783,
+      "grad_norm": 24.134017944335938,
+      "learning_rate": 8.031496062992128e-06,
+      "loss": 1.5229,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6029900781313363,
+      "grad_norm": 23.24073600769043,
+      "learning_rate": 8.022747156605425e-06,
+      "loss": 1.5801,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6034226391629944,
+      "grad_norm": 20.904178619384766,
+      "learning_rate": 8.013998250218723e-06,
+      "loss": 1.5884,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6038552001946524,
+      "grad_norm": 23.131088256835938,
+      "learning_rate": 8.005249343832022e-06,
+      "loss": 1.489,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6042877612263106,
+      "grad_norm": 22.88324546813965,
+      "learning_rate": 7.996500437445321e-06,
+      "loss": 1.5062,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6047203222579686,
+      "grad_norm": 22.445697784423828,
+      "learning_rate": 7.987751531058619e-06,
+      "loss": 1.5516,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6051528832896267,
+      "grad_norm": 21.125619888305664,
+      "learning_rate": 7.979002624671916e-06,
+      "loss": 1.5458,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6055854443212847,
+      "grad_norm": 23.79015350341797,
+      "learning_rate": 7.970253718285215e-06,
+      "loss": 1.4498,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6060180053529428,
+      "grad_norm": 22.167564392089844,
+      "learning_rate": 7.961504811898514e-06,
+      "loss": 1.5495,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6064505663846008,
+      "grad_norm": 22.303564071655273,
+      "learning_rate": 7.952755905511812e-06,
+      "loss": 1.5881,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6068831274162589,
+      "grad_norm": 24.354480743408203,
+      "learning_rate": 7.94400699912511e-06,
+      "loss": 1.4995,
+      "step": 1403
+    },
+    {
+      "epoch": 0.607315688447917,
+      "grad_norm": 26.439773559570312,
+      "learning_rate": 7.935258092738409e-06,
+      "loss": 1.5314,
+      "step": 1404
+    },
+    {
+      "epoch": 0.607748249479575,
+      "grad_norm": 25.109752655029297,
+      "learning_rate": 7.926509186351706e-06,
+      "loss": 1.4808,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6081808105112331,
+      "grad_norm": 21.283794403076172,
+      "learning_rate": 7.917760279965005e-06,
+      "loss": 1.5175,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6086133715428911,
+      "grad_norm": 22.0447998046875,
+      "learning_rate": 7.909011373578303e-06,
+      "loss": 1.4999,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6090459325745492,
+      "grad_norm": 21.30337905883789,
+      "learning_rate": 7.900262467191602e-06,
+      "loss": 1.4887,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6094784936062072,
+      "grad_norm": 24.14295768737793,
+      "learning_rate": 7.8915135608049e-06,
+      "loss": 1.4664,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6099110546378653,
+      "grad_norm": 22.371164321899414,
+      "learning_rate": 7.882764654418197e-06,
+      "loss": 1.5782,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6103436156695233,
+      "grad_norm": 23.566246032714844,
+      "learning_rate": 7.874015748031496e-06,
+      "loss": 1.519,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6107761767011815,
+      "grad_norm": 21.215904235839844,
+      "learning_rate": 7.865266841644796e-06,
+      "loss": 1.5402,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6112087377328395,
+      "grad_norm": 21.46411895751953,
+      "learning_rate": 7.856517935258093e-06,
+      "loss": 1.4525,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6116412987644976,
+      "grad_norm": 23.8530216217041,
+      "learning_rate": 7.847769028871392e-06,
+      "loss": 1.4487,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6120738597961556,
+      "grad_norm": 23.55060386657715,
+      "learning_rate": 7.83902012248469e-06,
+      "loss": 1.5548,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6125064208278137,
+      "grad_norm": 31.455612182617188,
+      "learning_rate": 7.830271216097989e-06,
+      "loss": 1.6577,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6129389818594717,
+      "grad_norm": 26.714149475097656,
+      "learning_rate": 7.821522309711287e-06,
+      "loss": 1.5606,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6133715428911298,
+      "grad_norm": 24.780031204223633,
+      "learning_rate": 7.812773403324586e-06,
+      "loss": 1.5787,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6138041039227878,
+      "grad_norm": 23.39690589904785,
+      "learning_rate": 7.804024496937883e-06,
+      "loss": 1.5447,
+      "step": 1419
+    },
+    {
+      "epoch": 0.614236664954446,
+      "grad_norm": 21.379409790039062,
+      "learning_rate": 7.79527559055118e-06,
+      "loss": 1.4938,
+      "step": 1420
+    },
+    {
+      "epoch": 0.614669225986104,
+      "grad_norm": 22.49406623840332,
+      "learning_rate": 7.78652668416448e-06,
+      "loss": 1.4945,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6151017870177621,
+      "grad_norm": 21.45618438720703,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.6223,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6155343480494201,
+      "grad_norm": 21.71502113342285,
+      "learning_rate": 7.769028871391077e-06,
+      "loss": 1.594,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6159669090810782,
+      "grad_norm": 23.39321517944336,
+      "learning_rate": 7.760279965004374e-06,
+      "loss": 1.4467,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6163994701127362,
+      "grad_norm": 22.757183074951172,
+      "learning_rate": 7.751531058617673e-06,
+      "loss": 1.478,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6168320311443943,
+      "grad_norm": 34.57364273071289,
+      "learning_rate": 7.742782152230973e-06,
+      "loss": 1.4439,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6172645921760523,
+      "grad_norm": 24.04800796508789,
+      "learning_rate": 7.73403324584427e-06,
+      "loss": 1.5368,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6176971532077103,
+      "grad_norm": 25.531238555908203,
+      "learning_rate": 7.725284339457568e-06,
+      "loss": 1.6016,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6181297142393685,
+      "grad_norm": 25.783323287963867,
+      "learning_rate": 7.716535433070867e-06,
+      "loss": 1.5348,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6185622752710265,
+      "grad_norm": 24.652729034423828,
+      "learning_rate": 7.707786526684166e-06,
+      "loss": 1.5232,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6189948363026846,
+      "grad_norm": 24.11018180847168,
+      "learning_rate": 7.699037620297464e-06,
+      "loss": 1.5591,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6194273973343426,
+      "grad_norm": 24.16909408569336,
+      "learning_rate": 7.690288713910761e-06,
+      "loss": 1.6298,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6198599583660007,
+      "grad_norm": 23.931461334228516,
+      "learning_rate": 7.68153980752406e-06,
+      "loss": 1.4956,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6202925193976587,
+      "grad_norm": 25.672365188598633,
+      "learning_rate": 7.672790901137358e-06,
+      "loss": 1.5273,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6207250804293168,
+      "grad_norm": 22.84345817565918,
+      "learning_rate": 7.664041994750657e-06,
+      "loss": 1.5427,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6211576414609749,
+      "grad_norm": 22.920881271362305,
+      "learning_rate": 7.655293088363955e-06,
+      "loss": 1.549,
+      "step": 1436
+    },
+    {
+      "epoch": 0.621590202492633,
+      "grad_norm": 22.238073348999023,
+      "learning_rate": 7.646544181977254e-06,
+      "loss": 1.5617,
+      "step": 1437
+    },
+    {
+      "epoch": 0.622022763524291,
+      "grad_norm": 21.899429321289062,
+      "learning_rate": 7.637795275590551e-06,
+      "loss": 1.4983,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6224553245559491,
+      "grad_norm": 21.77754783630371,
+      "learning_rate": 7.62904636920385e-06,
+      "loss": 1.5032,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6228878855876071,
+      "grad_norm": 24.998987197875977,
+      "learning_rate": 7.620297462817149e-06,
+      "loss": 1.3883,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6233204466192652,
+      "grad_norm": 23.322284698486328,
+      "learning_rate": 7.611548556430447e-06,
+      "loss": 1.5162,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6237530076509232,
+      "grad_norm": 21.041311264038086,
+      "learning_rate": 7.602799650043745e-06,
+      "loss": 1.4762,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6241855686825813,
+      "grad_norm": 23.653440475463867,
+      "learning_rate": 7.594050743657043e-06,
+      "loss": 1.531,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6246181297142394,
+      "grad_norm": 20.211307525634766,
+      "learning_rate": 7.585301837270341e-06,
+      "loss": 1.5188,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6250506907458975,
+      "grad_norm": 25.027973175048828,
+      "learning_rate": 7.576552930883641e-06,
+      "loss": 1.5473,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6254832517775555,
+      "grad_norm": 23.346525192260742,
+      "learning_rate": 7.567804024496939e-06,
+      "loss": 1.4762,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6259158128092136,
+      "grad_norm": 24.292959213256836,
+      "learning_rate": 7.5590551181102365e-06,
+      "loss": 1.5621,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6263483738408716,
+      "grad_norm": 23.621158599853516,
+      "learning_rate": 7.550306211723535e-06,
+      "loss": 1.4602,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6267809348725296,
+      "grad_norm": 23.259824752807617,
+      "learning_rate": 7.541557305336834e-06,
+      "loss": 1.5399,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6272134959041877,
+      "grad_norm": 24.204967498779297,
+      "learning_rate": 7.532808398950132e-06,
+      "loss": 1.5119,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6276460569358457,
+      "grad_norm": 22.07662582397461,
+      "learning_rate": 7.52405949256343e-06,
+      "loss": 1.5115,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6280786179675039,
+      "grad_norm": 23.451147079467773,
+      "learning_rate": 7.515310586176728e-06,
+      "loss": 1.6045,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6285111789991619,
+      "grad_norm": 23.66141128540039,
+      "learning_rate": 7.506561679790027e-06,
+      "loss": 1.6329,
+      "step": 1453
+    },
+    {
+      "epoch": 0.62894374003082,
+      "grad_norm": 22.44936180114746,
+      "learning_rate": 7.497812773403326e-06,
+      "loss": 1.5992,
+      "step": 1454
+    },
+    {
+      "epoch": 0.629376301062478,
+      "grad_norm": 22.108919143676758,
+      "learning_rate": 7.489063867016624e-06,
+      "loss": 1.5392,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6298088620941361,
+      "grad_norm": 24.86928939819336,
+      "learning_rate": 7.480314960629922e-06,
+      "loss": 1.5499,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6302414231257941,
+      "grad_norm": 21.16976547241211,
+      "learning_rate": 7.47156605424322e-06,
+      "loss": 1.588,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6306739841574522,
+      "grad_norm": 21.789356231689453,
+      "learning_rate": 7.4628171478565184e-06,
+      "loss": 1.5114,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6311065451891102,
+      "grad_norm": 23.476760864257812,
+      "learning_rate": 7.454068241469818e-06,
+      "loss": 1.5496,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6315391062207684,
+      "grad_norm": 25.469806671142578,
+      "learning_rate": 7.445319335083115e-06,
+      "loss": 1.6105,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6319716672524264,
+      "grad_norm": 25.634347915649414,
+      "learning_rate": 7.4365704286964135e-06,
+      "loss": 1.5353,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6324042282840845,
+      "grad_norm": 27.508028030395508,
+      "learning_rate": 7.427821522309712e-06,
+      "loss": 1.5767,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6328367893157425,
+      "grad_norm": 23.841432571411133,
+      "learning_rate": 7.419072615923009e-06,
+      "loss": 1.6054,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6332693503474006,
+      "grad_norm": 22.372793197631836,
+      "learning_rate": 7.410323709536309e-06,
+      "loss": 1.5289,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6337019113790586,
+      "grad_norm": 23.19795036315918,
+      "learning_rate": 7.401574803149607e-06,
+      "loss": 1.5856,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6341344724107167,
+      "grad_norm": 22.361799240112305,
+      "learning_rate": 7.392825896762905e-06,
+      "loss": 1.5092,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6345670334423748,
+      "grad_norm": 22.069835662841797,
+      "learning_rate": 7.384076990376204e-06,
+      "loss": 1.6036,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6349995944740329,
+      "grad_norm": 23.586353302001953,
+      "learning_rate": 7.375328083989501e-06,
+      "loss": 1.518,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6354321555056909,
+      "grad_norm": 22.216527938842773,
+      "learning_rate": 7.3665791776028e-06,
+      "loss": 1.5903,
+      "step": 1469
+    },
+    {
+      "epoch": 0.635864716537349,
+      "grad_norm": 23.356054306030273,
+      "learning_rate": 7.357830271216099e-06,
+      "loss": 1.4951,
+      "step": 1470
+    },
+    {
+      "epoch": 0.636297277569007,
+      "grad_norm": 21.026004791259766,
+      "learning_rate": 7.349081364829397e-06,
+      "loss": 1.618,
+      "step": 1471
+    },
+    {
+      "epoch": 0.636729838600665,
+      "grad_norm": 23.06012535095215,
+      "learning_rate": 7.340332458442695e-06,
+      "loss": 1.5405,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6371623996323231,
+      "grad_norm": 23.857873916625977,
+      "learning_rate": 7.331583552055994e-06,
+      "loss": 1.5567,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6375949606639811,
+      "grad_norm": 23.048160552978516,
+      "learning_rate": 7.322834645669292e-06,
+      "loss": 1.5534,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6380275216956393,
+      "grad_norm": 24.181964874267578,
+      "learning_rate": 7.3140857392825906e-06,
+      "loss": 1.5192,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6384600827272973,
+      "grad_norm": 22.832210540771484,
+      "learning_rate": 7.305336832895888e-06,
+      "loss": 1.5114,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6388926437589554,
+      "grad_norm": 23.764835357666016,
+      "learning_rate": 7.2965879265091864e-06,
+      "loss": 1.4807,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6393252047906134,
+      "grad_norm": 24.419448852539062,
+      "learning_rate": 7.287839020122486e-06,
+      "loss": 1.4671,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6397577658222715,
+      "grad_norm": 25.90771484375,
+      "learning_rate": 7.279090113735784e-06,
+      "loss": 1.5565,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6401903268539295,
+      "grad_norm": 24.24135398864746,
+      "learning_rate": 7.270341207349082e-06,
+      "loss": 1.4725,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6406228878855876,
+      "grad_norm": 23.335277557373047,
+      "learning_rate": 7.26159230096238e-06,
+      "loss": 1.5215,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6410554489172456,
+      "grad_norm": 23.67999839782715,
+      "learning_rate": 7.252843394575678e-06,
+      "loss": 1.5626,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6414880099489038,
+      "grad_norm": 26.6694393157959,
+      "learning_rate": 7.2440944881889774e-06,
+      "loss": 1.4421,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6419205709805618,
+      "grad_norm": 23.231159210205078,
+      "learning_rate": 7.235345581802276e-06,
+      "loss": 1.5262,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6423531320122199,
+      "grad_norm": 24.250080108642578,
+      "learning_rate": 7.226596675415573e-06,
+      "loss": 1.5133,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6427856930438779,
+      "grad_norm": 21.954505920410156,
+      "learning_rate": 7.217847769028872e-06,
+      "loss": 1.5227,
+      "step": 1486
+    },
+    {
+      "epoch": 0.643218254075536,
+      "grad_norm": 22.040283203125,
+      "learning_rate": 7.20909886264217e-06,
+      "loss": 1.59,
+      "step": 1487
+    },
+    {
+      "epoch": 0.643650815107194,
+      "grad_norm": 23.680068969726562,
+      "learning_rate": 7.200349956255469e-06,
+      "loss": 1.5092,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6440833761388521,
+      "grad_norm": 25.096607208251953,
+      "learning_rate": 7.191601049868768e-06,
+      "loss": 1.468,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6445159371705101,
+      "grad_norm": 22.989320755004883,
+      "learning_rate": 7.182852143482065e-06,
+      "loss": 1.5895,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6449484982021683,
+      "grad_norm": 25.26317024230957,
+      "learning_rate": 7.1741032370953635e-06,
+      "loss": 1.4943,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6453810592338263,
+      "grad_norm": 25.946033477783203,
+      "learning_rate": 7.165354330708662e-06,
+      "loss": 1.635,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6458136202654843,
+      "grad_norm": 23.764036178588867,
+      "learning_rate": 7.156605424321961e-06,
+      "loss": 1.6082,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6462461812971424,
+      "grad_norm": 26.842212677001953,
+      "learning_rate": 7.1478565179352585e-06,
+      "loss": 1.5848,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6466787423288004,
+      "grad_norm": 24.90704917907715,
+      "learning_rate": 7.139107611548557e-06,
+      "loss": 1.5782,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6471113033604585,
+      "grad_norm": 26.029441833496094,
+      "learning_rate": 7.130358705161855e-06,
+      "loss": 1.5036,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6475438643921165,
+      "grad_norm": 25.079856872558594,
+      "learning_rate": 7.1216097987751545e-06,
+      "loss": 1.4376,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6479764254237746,
+      "grad_norm": 25.471174240112305,
+      "learning_rate": 7.112860892388452e-06,
+      "loss": 1.3823,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6484089864554327,
+      "grad_norm": 21.90082359313965,
+      "learning_rate": 7.10411198600175e-06,
+      "loss": 1.5895,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6488415474870908,
+      "grad_norm": 22.75554656982422,
+      "learning_rate": 7.095363079615049e-06,
+      "loss": 1.6006,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6492741085187488,
+      "grad_norm": 21.22007179260254,
+      "learning_rate": 7.086614173228347e-06,
+      "loss": 1.5194,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6497066695504069,
+      "grad_norm": 22.74795913696289,
+      "learning_rate": 7.077865266841646e-06,
+      "loss": 1.5761,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6501392305820649,
+      "grad_norm": 22.385860443115234,
+      "learning_rate": 7.069116360454944e-06,
+      "loss": 1.5468,
+      "step": 1503
+    },
+    {
+      "epoch": 0.650571791613723,
+      "grad_norm": 23.7584228515625,
+      "learning_rate": 7.060367454068242e-06,
+      "loss": 1.5242,
+      "step": 1504
+    },
+    {
+      "epoch": 0.651004352645381,
+      "grad_norm": 25.09479522705078,
+      "learning_rate": 7.0516185476815405e-06,
+      "loss": 1.5615,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6514369136770392,
+      "grad_norm": 24.615421295166016,
+      "learning_rate": 7.042869641294838e-06,
+      "loss": 1.5362,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6518694747086972,
+      "grad_norm": 21.421411514282227,
+      "learning_rate": 7.034120734908137e-06,
+      "loss": 1.5744,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6523020357403553,
+      "grad_norm": 23.396808624267578,
+      "learning_rate": 7.025371828521436e-06,
+      "loss": 1.5602,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6527345967720133,
+      "grad_norm": 22.569921493530273,
+      "learning_rate": 7.016622922134734e-06,
+      "loss": 1.5355,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6531671578036714,
+      "grad_norm": 24.40047264099121,
+      "learning_rate": 7.0078740157480315e-06,
+      "loss": 1.5158,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6535997188353294,
+      "grad_norm": 24.43023681640625,
+      "learning_rate": 6.99912510936133e-06,
+      "loss": 1.5551,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6540322798669875,
+      "grad_norm": 21.310428619384766,
+      "learning_rate": 6.990376202974629e-06,
+      "loss": 1.5114,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6544648408986455,
+      "grad_norm": 21.625455856323242,
+      "learning_rate": 6.981627296587927e-06,
+      "loss": 1.5002,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6548974019303037,
+      "grad_norm": 24.013227462768555,
+      "learning_rate": 6.972878390201226e-06,
+      "loss": 1.5115,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6553299629619617,
+      "grad_norm": 20.982900619506836,
+      "learning_rate": 6.964129483814523e-06,
+      "loss": 1.4118,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6557625239936197,
+      "grad_norm": 23.199033737182617,
+      "learning_rate": 6.955380577427822e-06,
+      "loss": 1.5651,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6561950850252778,
+      "grad_norm": 21.069751739501953,
+      "learning_rate": 6.946631671041121e-06,
+      "loss": 1.5955,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6566276460569358,
+      "grad_norm": 22.66398811340332,
+      "learning_rate": 6.937882764654419e-06,
+      "loss": 1.4432,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6570602070885939,
+      "grad_norm": 22.937976837158203,
+      "learning_rate": 6.929133858267717e-06,
+      "loss": 1.5522,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6574927681202519,
+      "grad_norm": 22.356679916381836,
+      "learning_rate": 6.920384951881015e-06,
+      "loss": 1.5067,
+      "step": 1520
+    },
+    {
+      "epoch": 0.65792532915191,
+      "grad_norm": 23.95052719116211,
+      "learning_rate": 6.911636045494314e-06,
+      "loss": 1.5553,
+      "step": 1521
+    },
+    {
+      "epoch": 0.658357890183568,
+      "grad_norm": 24.117918014526367,
+      "learning_rate": 6.902887139107613e-06,
+      "loss": 1.4865,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6587904512152262,
+      "grad_norm": 23.363407135009766,
+      "learning_rate": 6.89413823272091e-06,
+      "loss": 1.4588,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6592230122468842,
+      "grad_norm": 24.435625076293945,
+      "learning_rate": 6.8853893263342085e-06,
+      "loss": 1.4985,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6596555732785423,
+      "grad_norm": 25.14236831665039,
+      "learning_rate": 6.876640419947507e-06,
+      "loss": 1.5323,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6600881343102003,
+      "grad_norm": 22.674089431762695,
+      "learning_rate": 6.867891513560806e-06,
+      "loss": 1.5134,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6605206953418584,
+      "grad_norm": 24.138879776000977,
+      "learning_rate": 6.859142607174104e-06,
+      "loss": 1.5683,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6609532563735164,
+      "grad_norm": 24.909719467163086,
+      "learning_rate": 6.850393700787402e-06,
+      "loss": 1.5579,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6613858174051745,
+      "grad_norm": 23.420795440673828,
+      "learning_rate": 6.8416447944007e-06,
+      "loss": 1.4783,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6618183784368326,
+      "grad_norm": 22.931936264038086,
+      "learning_rate": 6.832895888013999e-06,
+      "loss": 1.4801,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6622509394684907,
+      "grad_norm": 21.55242156982422,
+      "learning_rate": 6.824146981627298e-06,
+      "loss": 1.4999,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6626835005001487,
+      "grad_norm": 21.650358200073242,
+      "learning_rate": 6.815398075240595e-06,
+      "loss": 1.5101,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6631160615318068,
+      "grad_norm": 23.8084659576416,
+      "learning_rate": 6.806649168853894e-06,
+      "loss": 1.4968,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6635486225634648,
+      "grad_norm": 25.13161849975586,
+      "learning_rate": 6.797900262467192e-06,
+      "loss": 1.4169,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6639811835951229,
+      "grad_norm": 22.81217384338379,
+      "learning_rate": 6.78915135608049e-06,
+      "loss": 1.4825,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6644137446267809,
+      "grad_norm": 25.360429763793945,
+      "learning_rate": 6.78040244969379e-06,
+      "loss": 1.6435,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6648463056584389,
+      "grad_norm": 22.05304527282715,
+      "learning_rate": 6.771653543307087e-06,
+      "loss": 1.5476,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6652788666900971,
+      "grad_norm": 24.10348892211914,
+      "learning_rate": 6.7629046369203855e-06,
+      "loss": 1.4382,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6657114277217551,
+      "grad_norm": 22.268739700317383,
+      "learning_rate": 6.754155730533684e-06,
+      "loss": 1.618,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6661439887534132,
+      "grad_norm": 21.813589096069336,
+      "learning_rate": 6.745406824146981e-06,
+      "loss": 1.5424,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6665765497850712,
+      "grad_norm": 24.29230308532715,
+      "learning_rate": 6.736657917760281e-06,
+      "loss": 1.5056,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6670091108167293,
+      "grad_norm": 21.526540756225586,
+      "learning_rate": 6.727909011373579e-06,
+      "loss": 1.531,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6674416718483873,
+      "grad_norm": 22.080156326293945,
+      "learning_rate": 6.719160104986877e-06,
+      "loss": 1.5371,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6678742328800454,
+      "grad_norm": 22.51006507873535,
+      "learning_rate": 6.710411198600175e-06,
+      "loss": 1.5069,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6683067939117034,
+      "grad_norm": 25.575986862182617,
+      "learning_rate": 6.701662292213474e-06,
+      "loss": 1.511,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6687393549433616,
+      "grad_norm": 22.93172836303711,
+      "learning_rate": 6.692913385826772e-06,
+      "loss": 1.501,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6691719159750196,
+      "grad_norm": 22.480873107910156,
+      "learning_rate": 6.684164479440071e-06,
+      "loss": 1.5077,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6696044770066777,
+      "grad_norm": 22.858264923095703,
+      "learning_rate": 6.675415573053369e-06,
+      "loss": 1.4774,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6700370380383357,
+      "grad_norm": 24.259767532348633,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.6461,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6704695990699938,
+      "grad_norm": 24.198890686035156,
+      "learning_rate": 6.657917760279966e-06,
+      "loss": 1.5393,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6709021601016518,
+      "grad_norm": 24.402128219604492,
+      "learning_rate": 6.649168853893264e-06,
+      "loss": 1.4286,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6713347211333099,
+      "grad_norm": 25.59145736694336,
+      "learning_rate": 6.6404199475065626e-06,
+      "loss": 1.4993,
+      "step": 1552
+    },
+    {
+      "epoch": 0.671767282164968,
+      "grad_norm": 23.374197006225586,
+      "learning_rate": 6.63167104111986e-06,
+      "loss": 1.5453,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6721998431966261,
+      "grad_norm": 26.37010955810547,
+      "learning_rate": 6.6229221347331584e-06,
+      "loss": 1.5741,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6726324042282841,
+      "grad_norm": 25.05316162109375,
+      "learning_rate": 6.614173228346458e-06,
+      "loss": 1.539,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6730649652599422,
+      "grad_norm": 23.657917022705078,
+      "learning_rate": 6.605424321959756e-06,
+      "loss": 1.5029,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6734975262916002,
+      "grad_norm": 22.25530433654785,
+      "learning_rate": 6.5966754155730535e-06,
+      "loss": 1.538,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6739300873232583,
+      "grad_norm": 26.930683135986328,
+      "learning_rate": 6.587926509186352e-06,
+      "loss": 1.5289,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6743626483549163,
+      "grad_norm": 23.665882110595703,
+      "learning_rate": 6.57917760279965e-06,
+      "loss": 1.5161,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6747952093865743,
+      "grad_norm": 22.633747100830078,
+      "learning_rate": 6.5704286964129495e-06,
+      "loss": 1.4837,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6752277704182325,
+      "grad_norm": 27.529674530029297,
+      "learning_rate": 6.561679790026248e-06,
+      "loss": 1.5903,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6756603314498905,
+      "grad_norm": 23.555315017700195,
+      "learning_rate": 6.552930883639545e-06,
+      "loss": 1.5309,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6760928924815486,
+      "grad_norm": 22.861183166503906,
+      "learning_rate": 6.544181977252844e-06,
+      "loss": 1.5248,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6765254535132066,
+      "grad_norm": 23.239778518676758,
+      "learning_rate": 6.535433070866142e-06,
+      "loss": 1.4878,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6769580145448647,
+      "grad_norm": 24.08327293395996,
+      "learning_rate": 6.526684164479441e-06,
+      "loss": 1.4779,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6773905755765227,
+      "grad_norm": 24.711755752563477,
+      "learning_rate": 6.517935258092739e-06,
+      "loss": 1.6168,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6778231366081808,
+      "grad_norm": 26.82582664489746,
+      "learning_rate": 6.509186351706037e-06,
+      "loss": 1.4592,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6782556976398388,
+      "grad_norm": 25.213279724121094,
+      "learning_rate": 6.5004374453193355e-06,
+      "loss": 1.5868,
+      "step": 1568
+    },
+    {
+      "epoch": 0.678688258671497,
+      "grad_norm": 22.981929779052734,
+      "learning_rate": 6.491688538932633e-06,
+      "loss": 1.6304,
+      "step": 1569
+    },
+    {
+      "epoch": 0.679120819703155,
+      "grad_norm": 22.704452514648438,
+      "learning_rate": 6.482939632545932e-06,
+      "loss": 1.4529,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6795533807348131,
+      "grad_norm": 25.932876586914062,
+      "learning_rate": 6.4741907261592306e-06,
+      "loss": 1.5983,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6799859417664711,
+      "grad_norm": 23.21229362487793,
+      "learning_rate": 6.465441819772529e-06,
+      "loss": 1.4824,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6804185027981292,
+      "grad_norm": 23.76874351501465,
+      "learning_rate": 6.456692913385827e-06,
+      "loss": 1.4459,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6808510638297872,
+      "grad_norm": 24.482866287231445,
+      "learning_rate": 6.4479440069991265e-06,
+      "loss": 1.5083,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6812836248614453,
+      "grad_norm": 23.35637855529785,
+      "learning_rate": 6.439195100612424e-06,
+      "loss": 1.5173,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6817161858931033,
+      "grad_norm": 22.728713989257812,
+      "learning_rate": 6.430446194225722e-06,
+      "loss": 1.5578,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6821487469247615,
+      "grad_norm": 27.757644653320312,
+      "learning_rate": 6.421697287839021e-06,
+      "loss": 1.557,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6825813079564195,
+      "grad_norm": 24.904579162597656,
+      "learning_rate": 6.412948381452318e-06,
+      "loss": 1.4613,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6830138689880776,
+      "grad_norm": 24.260875701904297,
+      "learning_rate": 6.4041994750656174e-06,
+      "loss": 1.5102,
+      "step": 1579
+    },
+    {
+      "epoch": 0.6834464300197356,
+      "grad_norm": 25.12516975402832,
+      "learning_rate": 6.395450568678916e-06,
+      "loss": 1.5472,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6838789910513936,
+      "grad_norm": 23.9464054107666,
+      "learning_rate": 6.386701662292214e-06,
+      "loss": 1.5611,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6843115520830517,
+      "grad_norm": 22.76926040649414,
+      "learning_rate": 6.3779527559055125e-06,
+      "loss": 1.5379,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6847441131147097,
+      "grad_norm": 24.24293327331543,
+      "learning_rate": 6.36920384951881e-06,
+      "loss": 1.5672,
+      "step": 1583
+    },
+    {
+      "epoch": 0.6851766741463678,
+      "grad_norm": 24.250621795654297,
+      "learning_rate": 6.360454943132109e-06,
+      "loss": 1.4307,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6856092351780259,
+      "grad_norm": 25.41415786743164,
+      "learning_rate": 6.351706036745408e-06,
+      "loss": 1.5174,
+      "step": 1585
+    },
+    {
+      "epoch": 0.686041796209684,
+      "grad_norm": 21.934829711914062,
+      "learning_rate": 6.342957130358706e-06,
+      "loss": 1.5329,
+      "step": 1586
+    },
+    {
+      "epoch": 0.686474357241342,
+      "grad_norm": 23.609012603759766,
+      "learning_rate": 6.3342082239720035e-06,
+      "loss": 1.5426,
+      "step": 1587
+    },
+    {
+      "epoch": 0.6869069182730001,
+      "grad_norm": 25.188135147094727,
+      "learning_rate": 6.325459317585302e-06,
+      "loss": 1.5221,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6873394793046581,
+      "grad_norm": 23.130638122558594,
+      "learning_rate": 6.316710411198601e-06,
+      "loss": 1.5202,
+      "step": 1589
+    },
+    {
+      "epoch": 0.6877720403363162,
+      "grad_norm": 24.72777557373047,
+      "learning_rate": 6.307961504811899e-06,
+      "loss": 1.5004,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6882046013679742,
+      "grad_norm": 23.27838134765625,
+      "learning_rate": 6.299212598425197e-06,
+      "loss": 1.5503,
+      "step": 1591
+    },
+    {
+      "epoch": 0.6886371623996324,
+      "grad_norm": 20.87567710876465,
+      "learning_rate": 6.290463692038495e-06,
+      "loss": 1.5856,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6890697234312904,
+      "grad_norm": 23.302906036376953,
+      "learning_rate": 6.281714785651794e-06,
+      "loss": 1.524,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6895022844629485,
+      "grad_norm": 24.006710052490234,
+      "learning_rate": 6.272965879265093e-06,
+      "loss": 1.4691,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6899348454946065,
+      "grad_norm": 24.12818145751953,
+      "learning_rate": 6.264216972878391e-06,
+      "loss": 1.5727,
+      "step": 1595
+    },
+    {
+      "epoch": 0.6903674065262646,
+      "grad_norm": 25.554643630981445,
+      "learning_rate": 6.255468066491689e-06,
+      "loss": 1.5754,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6907999675579226,
+      "grad_norm": 21.56632423400879,
+      "learning_rate": 6.246719160104987e-06,
+      "loss": 1.5766,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6912325285895807,
+      "grad_norm": 26.7813720703125,
+      "learning_rate": 6.237970253718286e-06,
+      "loss": 1.5778,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6916650896212387,
+      "grad_norm": 24.2366943359375,
+      "learning_rate": 6.229221347331585e-06,
+      "loss": 1.5162,
+      "step": 1599
+    },
+    {
+      "epoch": 0.6920976506528969,
+      "grad_norm": 24.078445434570312,
+      "learning_rate": 6.220472440944882e-06,
+      "loss": 1.4297,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6925302116845549,
+      "grad_norm": 26.070178985595703,
+      "learning_rate": 6.2117235345581805e-06,
+      "loss": 1.5393,
+      "step": 1601
+    },
+    {
+      "epoch": 0.692962772716213,
+      "grad_norm": 27.011350631713867,
+      "learning_rate": 6.202974628171479e-06,
+      "loss": 1.4795,
+      "step": 1602
+    },
+    {
+      "epoch": 0.693395333747871,
+      "grad_norm": 27.4875431060791,
+      "learning_rate": 6.194225721784778e-06,
+      "loss": 1.5485,
+      "step": 1603
+    },
+    {
+      "epoch": 0.693827894779529,
+      "grad_norm": 22.1238956451416,
+      "learning_rate": 6.185476815398076e-06,
+      "loss": 1.428,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6942604558111871,
+      "grad_norm": 23.28427505493164,
+      "learning_rate": 6.176727909011374e-06,
+      "loss": 1.6064,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6946930168428451,
+      "grad_norm": 24.525197982788086,
+      "learning_rate": 6.167979002624672e-06,
+      "loss": 1.5426,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6951255778745032,
+      "grad_norm": 26.526762008666992,
+      "learning_rate": 6.159230096237971e-06,
+      "loss": 1.4851,
+      "step": 1607
+    },
+    {
+      "epoch": 0.6955581389061612,
+      "grad_norm": 23.86712646484375,
+      "learning_rate": 6.15048118985127e-06,
+      "loss": 1.5635,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6959906999378194,
+      "grad_norm": 22.40746307373047,
+      "learning_rate": 6.141732283464567e-06,
+      "loss": 1.5988,
+      "step": 1609
+    },
+    {
+      "epoch": 0.6964232609694774,
+      "grad_norm": 22.69624137878418,
+      "learning_rate": 6.132983377077866e-06,
+      "loss": 1.444,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6968558220011355,
+      "grad_norm": 22.52001953125,
+      "learning_rate": 6.124234470691164e-06,
+      "loss": 1.4884,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6972883830327935,
+      "grad_norm": 25.607025146484375,
+      "learning_rate": 6.115485564304462e-06,
+      "loss": 1.4535,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6977209440644516,
+      "grad_norm": 22.519060134887695,
+      "learning_rate": 6.106736657917761e-06,
+      "loss": 1.6537,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6981535050961096,
+      "grad_norm": 22.56156349182129,
+      "learning_rate": 6.097987751531059e-06,
+      "loss": 1.5571,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6985860661277677,
+      "grad_norm": 22.961685180664062,
+      "learning_rate": 6.0892388451443576e-06,
+      "loss": 1.5163,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6990186271594258,
+      "grad_norm": 24.06474494934082,
+      "learning_rate": 6.080489938757655e-06,
+      "loss": 1.5118,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6994511881910839,
+      "grad_norm": 23.049427032470703,
+      "learning_rate": 6.0717410323709534e-06,
+      "loss": 1.5018,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6998837492227419,
+      "grad_norm": 25.137136459350586,
+      "learning_rate": 6.062992125984253e-06,
+      "loss": 1.5839,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7003163102544,
+      "grad_norm": 23.529890060424805,
+      "learning_rate": 6.054243219597551e-06,
+      "loss": 1.5558,
+      "step": 1619
+    },
+    {
+      "epoch": 0.700748871286058,
+      "grad_norm": 22.19614601135254,
+      "learning_rate": 6.045494313210849e-06,
+      "loss": 1.5381,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7011814323177161,
+      "grad_norm": 26.739168167114258,
+      "learning_rate": 6.036745406824147e-06,
+      "loss": 1.5358,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7016139933493741,
+      "grad_norm": 25.52427864074707,
+      "learning_rate": 6.027996500437446e-06,
+      "loss": 1.5421,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7020465543810323,
+      "grad_norm": 23.286373138427734,
+      "learning_rate": 6.0192475940507444e-06,
+      "loss": 1.4983,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7024791154126903,
+      "grad_norm": 25.0222225189209,
+      "learning_rate": 6.010498687664043e-06,
+      "loss": 1.4992,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7029116764443483,
+      "grad_norm": 26.569257736206055,
+      "learning_rate": 6.00174978127734e-06,
+      "loss": 1.5621,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7033442374760064,
+      "grad_norm": 23.545291900634766,
+      "learning_rate": 5.993000874890639e-06,
+      "loss": 1.5807,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7037767985076644,
+      "grad_norm": 25.80244255065918,
+      "learning_rate": 5.984251968503938e-06,
+      "loss": 1.5377,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7042093595393225,
+      "grad_norm": 24.305273056030273,
+      "learning_rate": 5.975503062117236e-06,
+      "loss": 1.5414,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7046419205709805,
+      "grad_norm": 24.8024845123291,
+      "learning_rate": 5.966754155730535e-06,
+      "loss": 1.5594,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7050744816026386,
+      "grad_norm": 22.716646194458008,
+      "learning_rate": 5.958005249343832e-06,
+      "loss": 1.5006,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7055070426342966,
+      "grad_norm": 25.36906623840332,
+      "learning_rate": 5.9492563429571305e-06,
+      "loss": 1.5169,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7059396036659548,
+      "grad_norm": 21.435344696044922,
+      "learning_rate": 5.94050743657043e-06,
+      "loss": 1.4967,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7063721646976128,
+      "grad_norm": 25.601346969604492,
+      "learning_rate": 5.931758530183728e-06,
+      "loss": 1.4927,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7068047257292709,
+      "grad_norm": 23.794708251953125,
+      "learning_rate": 5.9230096237970256e-06,
+      "loss": 1.4312,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7072372867609289,
+      "grad_norm": 26.20401954650879,
+      "learning_rate": 5.914260717410324e-06,
+      "loss": 1.488,
+      "step": 1635
+    },
+    {
+      "epoch": 0.707669847792587,
+      "grad_norm": 23.398365020751953,
+      "learning_rate": 5.905511811023622e-06,
+      "loss": 1.4526,
+      "step": 1636
+    },
+    {
+      "epoch": 0.708102408824245,
+      "grad_norm": 24.44811248779297,
+      "learning_rate": 5.8967629046369215e-06,
+      "loss": 1.5993,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7085349698559031,
+      "grad_norm": 24.23762321472168,
+      "learning_rate": 5.888013998250219e-06,
+      "loss": 1.5917,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7089675308875611,
+      "grad_norm": 27.343517303466797,
+      "learning_rate": 5.879265091863517e-06,
+      "loss": 1.4431,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7094000919192193,
+      "grad_norm": 26.040502548217773,
+      "learning_rate": 5.870516185476816e-06,
+      "loss": 1.5874,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7098326529508773,
+      "grad_norm": 25.721101760864258,
+      "learning_rate": 5.861767279090114e-06,
+      "loss": 1.5325,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7102652139825354,
+      "grad_norm": 24.612388610839844,
+      "learning_rate": 5.853018372703413e-06,
+      "loss": 1.4986,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7106977750141934,
+      "grad_norm": 24.61174964904785,
+      "learning_rate": 5.844269466316711e-06,
+      "loss": 1.5429,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7111303360458515,
+      "grad_norm": 31.583274841308594,
+      "learning_rate": 5.835520559930009e-06,
+      "loss": 1.5787,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7115628970775095,
+      "grad_norm": 23.573389053344727,
+      "learning_rate": 5.8267716535433075e-06,
+      "loss": 1.5006,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7119954581091676,
+      "grad_norm": 28.629196166992188,
+      "learning_rate": 5.818022747156607e-06,
+      "loss": 1.5066,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7124280191408257,
+      "grad_norm": 24.07172966003418,
+      "learning_rate": 5.809273840769904e-06,
+      "loss": 1.5167,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7128605801724837,
+      "grad_norm": 24.85234260559082,
+      "learning_rate": 5.800524934383203e-06,
+      "loss": 1.4958,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7132931412041418,
+      "grad_norm": 24.3563232421875,
+      "learning_rate": 5.791776027996501e-06,
+      "loss": 1.5272,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7137257022357998,
+      "grad_norm": 24.525150299072266,
+      "learning_rate": 5.7830271216097985e-06,
+      "loss": 1.4952,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7141582632674579,
+      "grad_norm": 26.782798767089844,
+      "learning_rate": 5.774278215223098e-06,
+      "loss": 1.5519,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7145908242991159,
+      "grad_norm": 22.02875518798828,
+      "learning_rate": 5.765529308836396e-06,
+      "loss": 1.5515,
+      "step": 1652
+    },
+    {
+      "epoch": 0.715023385330774,
+      "grad_norm": 24.18636703491211,
+      "learning_rate": 5.756780402449694e-06,
+      "loss": 1.6241,
+      "step": 1653
+    },
+    {
+      "epoch": 0.715455946362432,
+      "grad_norm": 20.57954216003418,
+      "learning_rate": 5.748031496062993e-06,
+      "loss": 1.4652,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7158885073940902,
+      "grad_norm": 24.09521484375,
+      "learning_rate": 5.73928258967629e-06,
+      "loss": 1.3606,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7163210684257482,
+      "grad_norm": 26.569869995117188,
+      "learning_rate": 5.7305336832895895e-06,
+      "loss": 1.5271,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7167536294574063,
+      "grad_norm": 24.32697296142578,
+      "learning_rate": 5.721784776902888e-06,
+      "loss": 1.5423,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7171861904890643,
+      "grad_norm": 24.233909606933594,
+      "learning_rate": 5.713035870516186e-06,
+      "loss": 1.6511,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7176187515207224,
+      "grad_norm": 26.0461483001709,
+      "learning_rate": 5.704286964129484e-06,
+      "loss": 1.4889,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7180513125523804,
+      "grad_norm": 24.910232543945312,
+      "learning_rate": 5.695538057742782e-06,
+      "loss": 1.5006,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7184838735840385,
+      "grad_norm": 25.941944122314453,
+      "learning_rate": 5.686789151356081e-06,
+      "loss": 1.4548,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7189164346156965,
+      "grad_norm": 21.430110931396484,
+      "learning_rate": 5.67804024496938e-06,
+      "loss": 1.4481,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7193489956473547,
+      "grad_norm": 27.890432357788086,
+      "learning_rate": 5.669291338582677e-06,
+      "loss": 1.4916,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7197815566790127,
+      "grad_norm": 25.04297637939453,
+      "learning_rate": 5.6605424321959755e-06,
+      "loss": 1.4469,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7202141177106708,
+      "grad_norm": 24.669153213500977,
+      "learning_rate": 5.651793525809274e-06,
+      "loss": 1.4939,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7206466787423288,
+      "grad_norm": 27.076704025268555,
+      "learning_rate": 5.643044619422573e-06,
+      "loss": 1.569,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7210792397739869,
+      "grad_norm": 22.596973419189453,
+      "learning_rate": 5.6342957130358714e-06,
+      "loss": 1.5492,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7215118008056449,
+      "grad_norm": 25.775602340698242,
+      "learning_rate": 5.625546806649169e-06,
+      "loss": 1.6141,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7219443618373029,
+      "grad_norm": 23.130680084228516,
+      "learning_rate": 5.616797900262467e-06,
+      "loss": 1.5137,
+      "step": 1669
+    },
+    {
+      "epoch": 0.722376922868961,
+      "grad_norm": 32.87105941772461,
+      "learning_rate": 5.6080489938757665e-06,
+      "loss": 1.5107,
+      "step": 1670
+    },
+    {
+      "epoch": 0.722809483900619,
+      "grad_norm": 22.43234634399414,
+      "learning_rate": 5.599300087489065e-06,
+      "loss": 1.5348,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7232420449322772,
+      "grad_norm": 24.693084716796875,
+      "learning_rate": 5.590551181102362e-06,
+      "loss": 1.5212,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7236746059639352,
+      "grad_norm": 27.20313835144043,
+      "learning_rate": 5.581802274715661e-06,
+      "loss": 1.5393,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7241071669955933,
+      "grad_norm": 23.293119430541992,
+      "learning_rate": 5.573053368328959e-06,
+      "loss": 1.4915,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7245397280272513,
+      "grad_norm": 26.227306365966797,
+      "learning_rate": 5.564304461942258e-06,
+      "loss": 1.513,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7249722890589094,
+      "grad_norm": 24.33916473388672,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 1.4589,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7254048500905674,
+      "grad_norm": 24.176958084106445,
+      "learning_rate": 5.546806649168854e-06,
+      "loss": 1.5841,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7258374111222256,
+      "grad_norm": 22.859975814819336,
+      "learning_rate": 5.5380577427821525e-06,
+      "loss": 1.5555,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7262699721538836,
+      "grad_norm": 24.679033279418945,
+      "learning_rate": 5.529308836395451e-06,
+      "loss": 1.544,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7267025331855417,
+      "grad_norm": 23.848819732666016,
+      "learning_rate": 5.52055993000875e-06,
+      "loss": 1.4953,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7271350942171997,
+      "grad_norm": 23.183744430541992,
+      "learning_rate": 5.511811023622048e-06,
+      "loss": 1.55,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7275676552488578,
+      "grad_norm": 25.85765838623047,
+      "learning_rate": 5.503062117235346e-06,
+      "loss": 1.543,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7280002162805158,
+      "grad_norm": 25.841501235961914,
+      "learning_rate": 5.494313210848644e-06,
+      "loss": 1.5411,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7284327773121739,
+      "grad_norm": 25.476608276367188,
+      "learning_rate": 5.485564304461942e-06,
+      "loss": 1.502,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7288653383438319,
+      "grad_norm": 24.201326370239258,
+      "learning_rate": 5.476815398075241e-06,
+      "loss": 1.5492,
+      "step": 1685
+    },
+    {
+      "epoch": 0.72929789937549,
+      "grad_norm": 26.180383682250977,
+      "learning_rate": 5.468066491688539e-06,
+      "loss": 1.5663,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7297304604071481,
+      "grad_norm": 23.471723556518555,
+      "learning_rate": 5.459317585301838e-06,
+      "loss": 1.4523,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7301630214388062,
+      "grad_norm": 28.88964080810547,
+      "learning_rate": 5.450568678915136e-06,
+      "loss": 1.5238,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7305955824704642,
+      "grad_norm": 24.187883377075195,
+      "learning_rate": 5.441819772528434e-06,
+      "loss": 1.6256,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7310281435021223,
+      "grad_norm": 23.76744270324707,
+      "learning_rate": 5.433070866141733e-06,
+      "loss": 1.5351,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7314607045337803,
+      "grad_norm": 25.365325927734375,
+      "learning_rate": 5.424321959755031e-06,
+      "loss": 1.4795,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7318932655654383,
+      "grad_norm": 27.68093490600586,
+      "learning_rate": 5.41557305336833e-06,
+      "loss": 1.6068,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7323258265970964,
+      "grad_norm": 26.40992546081543,
+      "learning_rate": 5.406824146981627e-06,
+      "loss": 1.497,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7327583876287544,
+      "grad_norm": 26.072214126586914,
+      "learning_rate": 5.398075240594926e-06,
+      "loss": 1.5387,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7331909486604126,
+      "grad_norm": 22.766809463500977,
+      "learning_rate": 5.389326334208225e-06,
+      "loss": 1.6233,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7336235096920706,
+      "grad_norm": 24.374536514282227,
+      "learning_rate": 5.380577427821523e-06,
+      "loss": 1.5207,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7340560707237287,
+      "grad_norm": 26.02762794494629,
+      "learning_rate": 5.3718285214348205e-06,
+      "loss": 1.5178,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7344886317553867,
+      "grad_norm": 24.251422882080078,
+      "learning_rate": 5.363079615048119e-06,
+      "loss": 1.5194,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7349211927870448,
+      "grad_norm": 28.283960342407227,
+      "learning_rate": 5.354330708661418e-06,
+      "loss": 1.5407,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7353537538187028,
+      "grad_norm": 24.403696060180664,
+      "learning_rate": 5.3455818022747165e-06,
+      "loss": 1.5313,
+      "step": 1700
+    },
+    {
+      "epoch": 0.735786314850361,
+      "grad_norm": 25.62765884399414,
+      "learning_rate": 5.336832895888015e-06,
+      "loss": 1.5413,
+      "step": 1701
+    },
+    {
+      "epoch": 0.736218875882019,
+      "grad_norm": 26.185192108154297,
+      "learning_rate": 5.328083989501312e-06,
+      "loss": 1.4735,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7366514369136771,
+      "grad_norm": 22.629499435424805,
+      "learning_rate": 5.319335083114611e-06,
+      "loss": 1.4404,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7370839979453351,
+      "grad_norm": 21.76143455505371,
+      "learning_rate": 5.31058617672791e-06,
+      "loss": 1.5389,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7375165589769932,
+      "grad_norm": 24.350290298461914,
+      "learning_rate": 5.301837270341208e-06,
+      "loss": 1.5272,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7379491200086512,
+      "grad_norm": 25.934980392456055,
+      "learning_rate": 5.293088363954506e-06,
+      "loss": 1.5434,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7383816810403093,
+      "grad_norm": 26.150161743164062,
+      "learning_rate": 5.284339457567804e-06,
+      "loss": 1.4915,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7388142420719673,
+      "grad_norm": 26.145160675048828,
+      "learning_rate": 5.2755905511811025e-06,
+      "loss": 1.4885,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7392468031036254,
+      "grad_norm": 27.158536911010742,
+      "learning_rate": 5.266841644794402e-06,
+      "loss": 1.4621,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7396793641352835,
+      "grad_norm": 27.394515991210938,
+      "learning_rate": 5.258092738407699e-06,
+      "loss": 1.5549,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7401119251669416,
+      "grad_norm": 24.41299819946289,
+      "learning_rate": 5.2493438320209976e-06,
+      "loss": 1.4812,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7405444861985996,
+      "grad_norm": 26.23542022705078,
+      "learning_rate": 5.240594925634296e-06,
+      "loss": 1.452,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7409770472302576,
+      "grad_norm": 23.603107452392578,
+      "learning_rate": 5.231846019247594e-06,
+      "loss": 1.5045,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7414096082619157,
+      "grad_norm": 25.918743133544922,
+      "learning_rate": 5.2230971128608935e-06,
+      "loss": 1.5866,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7418421692935737,
+      "grad_norm": 23.324562072753906,
+      "learning_rate": 5.214348206474191e-06,
+      "loss": 1.4572,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7422747303252318,
+      "grad_norm": 23.02033233642578,
+      "learning_rate": 5.205599300087489e-06,
+      "loss": 1.4901,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7427072913568898,
+      "grad_norm": 25.430767059326172,
+      "learning_rate": 5.196850393700788e-06,
+      "loss": 1.4732,
+      "step": 1717
+    },
+    {
+      "epoch": 0.743139852388548,
+      "grad_norm": 25.643844604492188,
+      "learning_rate": 5.188101487314087e-06,
+      "loss": 1.5468,
+      "step": 1718
+    },
+    {
+      "epoch": 0.743572413420206,
+      "grad_norm": 22.182497024536133,
+      "learning_rate": 5.1793525809273845e-06,
+      "loss": 1.5145,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7440049744518641,
+      "grad_norm": 25.15062713623047,
+      "learning_rate": 5.170603674540683e-06,
+      "loss": 1.5308,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7444375354835221,
+      "grad_norm": 27.65828514099121,
+      "learning_rate": 5.161854768153981e-06,
+      "loss": 1.5002,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7448700965151802,
+      "grad_norm": 23.610389709472656,
+      "learning_rate": 5.1531058617672795e-06,
+      "loss": 1.593,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7453026575468382,
+      "grad_norm": 22.645122528076172,
+      "learning_rate": 5.144356955380579e-06,
+      "loss": 1.5042,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7457352185784963,
+      "grad_norm": 26.383073806762695,
+      "learning_rate": 5.135608048993876e-06,
+      "loss": 1.5823,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7461677796101543,
+      "grad_norm": 24.08680534362793,
+      "learning_rate": 5.126859142607175e-06,
+      "loss": 1.4827,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7466003406418125,
+      "grad_norm": 28.29230308532715,
+      "learning_rate": 5.118110236220473e-06,
+      "loss": 1.5249,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7470329016734705,
+      "grad_norm": 24.57362937927246,
+      "learning_rate": 5.1093613298337705e-06,
+      "loss": 1.4491,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7474654627051286,
+      "grad_norm": 23.962614059448242,
+      "learning_rate": 5.10061242344707e-06,
+      "loss": 1.5894,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7478980237367866,
+      "grad_norm": 23.7990779876709,
+      "learning_rate": 5.091863517060368e-06,
+      "loss": 1.5585,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7483305847684447,
+      "grad_norm": 24.12856674194336,
+      "learning_rate": 5.083114610673666e-06,
+      "loss": 1.5773,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7487631458001027,
+      "grad_norm": 23.70834732055664,
+      "learning_rate": 5.074365704286964e-06,
+      "loss": 1.4753,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7491957068317608,
+      "grad_norm": 23.08877182006836,
+      "learning_rate": 5.065616797900262e-06,
+      "loss": 1.6544,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7496282678634189,
+      "grad_norm": 26.548851013183594,
+      "learning_rate": 5.0568678915135615e-06,
+      "loss": 1.5633,
+      "step": 1733
+    },
+    {
+      "epoch": 0.750060828895077,
+      "grad_norm": 24.939672470092773,
+      "learning_rate": 5.04811898512686e-06,
+      "loss": 1.5656,
+      "step": 1734
+    },
+    {
+      "epoch": 0.750493389926735,
+      "grad_norm": 23.20163345336914,
+      "learning_rate": 5.039370078740158e-06,
+      "loss": 1.5713,
+      "step": 1735
+    },
+    {
+      "epoch": 0.750925950958393,
+      "grad_norm": 22.919689178466797,
+      "learning_rate": 5.030621172353456e-06,
+      "loss": 1.6136,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7513585119900511,
+      "grad_norm": 24.276479721069336,
+      "learning_rate": 5.021872265966754e-06,
+      "loss": 1.5749,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7517910730217091,
+      "grad_norm": 23.846363067626953,
+      "learning_rate": 5.013123359580053e-06,
+      "loss": 1.4735,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7522236340533672,
+      "grad_norm": 25.617586135864258,
+      "learning_rate": 5.004374453193352e-06,
+      "loss": 1.5511,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7526561950850252,
+      "grad_norm": 22.6473445892334,
+      "learning_rate": 4.995625546806649e-06,
+      "loss": 1.504,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7530887561166834,
+      "grad_norm": 24.413610458374023,
+      "learning_rate": 4.986876640419948e-06,
+      "loss": 1.5588,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7535213171483414,
+      "grad_norm": 24.66851806640625,
+      "learning_rate": 4.978127734033246e-06,
+      "loss": 1.5151,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7539538781799995,
+      "grad_norm": 23.771665573120117,
+      "learning_rate": 4.969378827646544e-06,
+      "loss": 1.4847,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7543864392116575,
+      "grad_norm": 24.447561264038086,
+      "learning_rate": 4.960629921259843e-06,
+      "loss": 1.55,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7548190002433156,
+      "grad_norm": 22.75802230834961,
+      "learning_rate": 4.951881014873141e-06,
+      "loss": 1.5186,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7552515612749736,
+      "grad_norm": 26.083370208740234,
+      "learning_rate": 4.94313210848644e-06,
+      "loss": 1.4842,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7556841223066317,
+      "grad_norm": 29.3764705657959,
+      "learning_rate": 4.934383202099738e-06,
+      "loss": 1.4959,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7561166833382897,
+      "grad_norm": 21.326786041259766,
+      "learning_rate": 4.925634295713037e-06,
+      "loss": 1.5026,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7565492443699479,
+      "grad_norm": 23.675289154052734,
+      "learning_rate": 4.916885389326334e-06,
+      "loss": 1.5059,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7569818054016059,
+      "grad_norm": 25.60411834716797,
+      "learning_rate": 4.908136482939633e-06,
+      "loss": 1.5985,
+      "step": 1750
+    },
+    {
+      "epoch": 0.757414366433264,
+      "grad_norm": 24.007932662963867,
+      "learning_rate": 4.899387576552931e-06,
+      "loss": 1.4974,
+      "step": 1751
+    },
+    {
+      "epoch": 0.757846927464922,
+      "grad_norm": 28.738752365112305,
+      "learning_rate": 4.8906386701662295e-06,
+      "loss": 1.4758,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7582794884965801,
+      "grad_norm": 24.201730728149414,
+      "learning_rate": 4.881889763779528e-06,
+      "loss": 1.5549,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7587120495282381,
+      "grad_norm": 24.409114837646484,
+      "learning_rate": 4.873140857392826e-06,
+      "loss": 1.5193,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7591446105598962,
+      "grad_norm": 22.474170684814453,
+      "learning_rate": 4.8643919510061246e-06,
+      "loss": 1.6365,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7595771715915542,
+      "grad_norm": 22.524370193481445,
+      "learning_rate": 4.855643044619423e-06,
+      "loss": 1.6074,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7600097326232123,
+      "grad_norm": 26.538591384887695,
+      "learning_rate": 4.846894138232721e-06,
+      "loss": 1.5673,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7604422936548704,
+      "grad_norm": 25.404539108276367,
+      "learning_rate": 4.83814523184602e-06,
+      "loss": 1.5393,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7608748546865284,
+      "grad_norm": 23.63199806213379,
+      "learning_rate": 4.829396325459318e-06,
+      "loss": 1.5652,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7613074157181865,
+      "grad_norm": 26.102270126342773,
+      "learning_rate": 4.820647419072616e-06,
+      "loss": 1.5968,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7617399767498445,
+      "grad_norm": 27.475393295288086,
+      "learning_rate": 4.811898512685915e-06,
+      "loss": 1.5657,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7621725377815026,
+      "grad_norm": 24.71074676513672,
+      "learning_rate": 4.803149606299213e-06,
+      "loss": 1.5442,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7626050988131606,
+      "grad_norm": 25.49890899658203,
+      "learning_rate": 4.7944006999125114e-06,
+      "loss": 1.4888,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7630376598448187,
+      "grad_norm": 23.629859924316406,
+      "learning_rate": 4.78565179352581e-06,
+      "loss": 1.5163,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7634702208764768,
+      "grad_norm": 24.619083404541016,
+      "learning_rate": 4.776902887139108e-06,
+      "loss": 1.5097,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7639027819081349,
+      "grad_norm": 21.8817138671875,
+      "learning_rate": 4.7681539807524065e-06,
+      "loss": 1.5555,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7643353429397929,
+      "grad_norm": 23.76637840270996,
+      "learning_rate": 4.759405074365704e-06,
+      "loss": 1.4962,
+      "step": 1767
+    },
+    {
+      "epoch": 0.764767903971451,
+      "grad_norm": 24.75080108642578,
+      "learning_rate": 4.750656167979003e-06,
+      "loss": 1.5414,
+      "step": 1768
+    },
+    {
+      "epoch": 0.765200465003109,
+      "grad_norm": 24.92739486694336,
+      "learning_rate": 4.741907261592302e-06,
+      "loss": 1.5096,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7656330260347671,
+      "grad_norm": 24.84958839416504,
+      "learning_rate": 4.7331583552056e-06,
+      "loss": 1.5318,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7660655870664251,
+      "grad_norm": 24.568798065185547,
+      "learning_rate": 4.724409448818898e-06,
+      "loss": 1.5618,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7664981480980833,
+      "grad_norm": 23.898578643798828,
+      "learning_rate": 4.715660542432197e-06,
+      "loss": 1.58,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7669307091297413,
+      "grad_norm": 24.494558334350586,
+      "learning_rate": 4.706911636045495e-06,
+      "loss": 1.4997,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7673632701613994,
+      "grad_norm": 21.843421936035156,
+      "learning_rate": 4.6981627296587926e-06,
+      "loss": 1.5825,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7677958311930574,
+      "grad_norm": 25.82838249206543,
+      "learning_rate": 4.689413823272092e-06,
+      "loss": 1.5319,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7682283922247155,
+      "grad_norm": 24.003957748413086,
+      "learning_rate": 4.680664916885389e-06,
+      "loss": 1.4995,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7686609532563735,
+      "grad_norm": 24.655353546142578,
+      "learning_rate": 4.6719160104986885e-06,
+      "loss": 1.456,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7690935142880316,
+      "grad_norm": 24.714584350585938,
+      "learning_rate": 4.663167104111986e-06,
+      "loss": 1.4928,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7695260753196896,
+      "grad_norm": 26.724472045898438,
+      "learning_rate": 4.654418197725284e-06,
+      "loss": 1.5625,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7699586363513476,
+      "grad_norm": 24.749422073364258,
+      "learning_rate": 4.645669291338583e-06,
+      "loss": 1.5352,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7703911973830058,
+      "grad_norm": 25.769168853759766,
+      "learning_rate": 4.636920384951881e-06,
+      "loss": 1.525,
+      "step": 1781
+    },
+    {
+      "epoch": 0.7708237584146638,
+      "grad_norm": 27.238039016723633,
+      "learning_rate": 4.62817147856518e-06,
+      "loss": 1.4711,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7712563194463219,
+      "grad_norm": 28.644994735717773,
+      "learning_rate": 4.619422572178478e-06,
+      "loss": 1.5006,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7716888804779799,
+      "grad_norm": 27.155855178833008,
+      "learning_rate": 4.610673665791777e-06,
+      "loss": 1.5224,
+      "step": 1784
+    },
+    {
+      "epoch": 0.772121441509638,
+      "grad_norm": 24.1883487701416,
+      "learning_rate": 4.6019247594050745e-06,
+      "loss": 1.5639,
+      "step": 1785
+    },
+    {
+      "epoch": 0.772554002541296,
+      "grad_norm": 23.588848114013672,
+      "learning_rate": 4.593175853018373e-06,
+      "loss": 1.4729,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7729865635729541,
+      "grad_norm": 23.275426864624023,
+      "learning_rate": 4.584426946631671e-06,
+      "loss": 1.5298,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7734191246046122,
+      "grad_norm": 24.45703887939453,
+      "learning_rate": 4.57567804024497e-06,
+      "loss": 1.5589,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7738516856362703,
+      "grad_norm": 25.52694320678711,
+      "learning_rate": 4.566929133858268e-06,
+      "loss": 1.4867,
+      "step": 1789
+    },
+    {
+      "epoch": 0.7742842466679283,
+      "grad_norm": 24.09499740600586,
+      "learning_rate": 4.558180227471566e-06,
+      "loss": 1.5231,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7747168076995864,
+      "grad_norm": 24.523792266845703,
+      "learning_rate": 4.549431321084865e-06,
+      "loss": 1.5179,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7751493687312444,
+      "grad_norm": 23.561460494995117,
+      "learning_rate": 4.540682414698163e-06,
+      "loss": 1.5205,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7755819297629025,
+      "grad_norm": 26.378007888793945,
+      "learning_rate": 4.531933508311461e-06,
+      "loss": 1.5417,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7760144907945605,
+      "grad_norm": 23.561283111572266,
+      "learning_rate": 4.52318460192476e-06,
+      "loss": 1.589,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7764470518262186,
+      "grad_norm": 27.071557998657227,
+      "learning_rate": 4.514435695538058e-06,
+      "loss": 1.4608,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7768796128578767,
+      "grad_norm": 26.57014274597168,
+      "learning_rate": 4.5056867891513565e-06,
+      "loss": 1.4033,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7773121738895348,
+      "grad_norm": 24.375045776367188,
+      "learning_rate": 4.496937882764655e-06,
+      "loss": 1.5522,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7777447349211928,
+      "grad_norm": 25.525135040283203,
+      "learning_rate": 4.488188976377953e-06,
+      "loss": 1.6045,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7781772959528509,
+      "grad_norm": 28.482580184936523,
+      "learning_rate": 4.4794400699912516e-06,
+      "loss": 1.5458,
+      "step": 1799
+    },
+    {
+      "epoch": 0.7786098569845089,
+      "grad_norm": 25.094606399536133,
+      "learning_rate": 4.47069116360455e-06,
+      "loss": 1.4793,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7790424180161669,
+      "grad_norm": 26.066743850708008,
+      "learning_rate": 4.461942257217848e-06,
+      "loss": 1.4703,
+      "step": 1801
+    },
+    {
+      "epoch": 0.779474979047825,
+      "grad_norm": 25.871166229248047,
+      "learning_rate": 4.453193350831147e-06,
+      "loss": 1.5241,
+      "step": 1802
+    },
+    {
+      "epoch": 0.779907540079483,
+      "grad_norm": 24.948637008666992,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.517,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7803401011111412,
+      "grad_norm": 25.19872283935547,
+      "learning_rate": 4.435695538057743e-06,
+      "loss": 1.552,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7807726621427992,
+      "grad_norm": 24.39336395263672,
+      "learning_rate": 4.426946631671042e-06,
+      "loss": 1.3931,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7812052231744573,
+      "grad_norm": 24.73307228088379,
+      "learning_rate": 4.41819772528434e-06,
+      "loss": 1.5159,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7816377842061153,
+      "grad_norm": 24.064722061157227,
+      "learning_rate": 4.4094488188976384e-06,
+      "loss": 1.5075,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7820703452377734,
+      "grad_norm": 24.808164596557617,
+      "learning_rate": 4.400699912510937e-06,
+      "loss": 1.4825,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7825029062694314,
+      "grad_norm": 27.774396896362305,
+      "learning_rate": 4.391951006124235e-06,
+      "loss": 1.5035,
+      "step": 1809
+    },
+    {
+      "epoch": 0.7829354673010895,
+      "grad_norm": 27.53242301940918,
+      "learning_rate": 4.383202099737533e-06,
+      "loss": 1.5277,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7833680283327475,
+      "grad_norm": 24.68132781982422,
+      "learning_rate": 4.374453193350832e-06,
+      "loss": 1.4304,
+      "step": 1811
+    },
+    {
+      "epoch": 0.7838005893644057,
+      "grad_norm": 26.517210006713867,
+      "learning_rate": 4.365704286964129e-06,
+      "loss": 1.5441,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7842331503960637,
+      "grad_norm": 25.836271286010742,
+      "learning_rate": 4.356955380577429e-06,
+      "loss": 1.46,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7846657114277218,
+      "grad_norm": 22.400842666625977,
+      "learning_rate": 4.348206474190726e-06,
+      "loss": 1.5226,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7850982724593798,
+      "grad_norm": 23.194393157958984,
+      "learning_rate": 4.3394575678040245e-06,
+      "loss": 1.5745,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7855308334910379,
+      "grad_norm": 23.99626922607422,
+      "learning_rate": 4.330708661417324e-06,
+      "loss": 1.4476,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7859633945226959,
+      "grad_norm": 25.947378158569336,
+      "learning_rate": 4.321959755030621e-06,
+      "loss": 1.5264,
+      "step": 1817
+    },
+    {
+      "epoch": 0.786395955554354,
+      "grad_norm": 26.823701858520508,
+      "learning_rate": 4.31321084864392e-06,
+      "loss": 1.6003,
+      "step": 1818
+    },
+    {
+      "epoch": 0.786828516586012,
+      "grad_norm": 26.935449600219727,
+      "learning_rate": 4.304461942257218e-06,
+      "loss": 1.5052,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7872610776176702,
+      "grad_norm": 24.033824920654297,
+      "learning_rate": 4.295713035870517e-06,
+      "loss": 1.5205,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7876936386493282,
+      "grad_norm": 21.987449645996094,
+      "learning_rate": 4.286964129483815e-06,
+      "loss": 1.461,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7881261996809863,
+      "grad_norm": 28.15187644958496,
+      "learning_rate": 4.278215223097113e-06,
+      "loss": 1.4807,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7885587607126443,
+      "grad_norm": 23.641508102416992,
+      "learning_rate": 4.269466316710411e-06,
+      "loss": 1.5484,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7889913217443023,
+      "grad_norm": 24.58562469482422,
+      "learning_rate": 4.26071741032371e-06,
+      "loss": 1.4434,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7894238827759604,
+      "grad_norm": 23.53512954711914,
+      "learning_rate": 4.251968503937008e-06,
+      "loss": 1.5811,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7898564438076184,
+      "grad_norm": 23.88713264465332,
+      "learning_rate": 4.2432195975503064e-06,
+      "loss": 1.5309,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7902890048392766,
+      "grad_norm": 25.259674072265625,
+      "learning_rate": 4.234470691163605e-06,
+      "loss": 1.6166,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7907215658709346,
+      "grad_norm": 22.053977966308594,
+      "learning_rate": 4.225721784776903e-06,
+      "loss": 1.5886,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7911541269025927,
+      "grad_norm": 23.540071487426758,
+      "learning_rate": 4.2169728783902015e-06,
+      "loss": 1.5388,
+      "step": 1829
+    },
+    {
+      "epoch": 0.7915866879342507,
+      "grad_norm": 23.088363647460938,
+      "learning_rate": 4.2082239720035e-06,
+      "loss": 1.5588,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7920192489659088,
+      "grad_norm": 24.821250915527344,
+      "learning_rate": 4.199475065616798e-06,
+      "loss": 1.5518,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7924518099975668,
+      "grad_norm": 25.494277954101562,
+      "learning_rate": 4.190726159230097e-06,
+      "loss": 1.5499,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7928843710292249,
+      "grad_norm": 25.74853515625,
+      "learning_rate": 4.181977252843395e-06,
+      "loss": 1.5049,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7933169320608829,
+      "grad_norm": 24.494779586791992,
+      "learning_rate": 4.173228346456693e-06,
+      "loss": 1.4597,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7937494930925411,
+      "grad_norm": 23.52466583251953,
+      "learning_rate": 4.164479440069992e-06,
+      "loss": 1.4917,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7941820541241991,
+      "grad_norm": 24.15924072265625,
+      "learning_rate": 4.15573053368329e-06,
+      "loss": 1.5182,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7946146151558572,
+      "grad_norm": 24.49439239501953,
+      "learning_rate": 4.146981627296588e-06,
+      "loss": 1.5663,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7950471761875152,
+      "grad_norm": 25.374263763427734,
+      "learning_rate": 4.138232720909887e-06,
+      "loss": 1.4795,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7954797372191733,
+      "grad_norm": 25.588665008544922,
+      "learning_rate": 4.129483814523185e-06,
+      "loss": 1.531,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7959122982508313,
+      "grad_norm": 26.306419372558594,
+      "learning_rate": 4.1207349081364835e-06,
+      "loss": 1.5243,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7963448592824894,
+      "grad_norm": 24.638086318969727,
+      "learning_rate": 4.111986001749782e-06,
+      "loss": 1.4911,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7967774203141474,
+      "grad_norm": 25.27682113647461,
+      "learning_rate": 4.10323709536308e-06,
+      "loss": 1.4854,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7972099813458056,
+      "grad_norm": 28.576251983642578,
+      "learning_rate": 4.0944881889763785e-06,
+      "loss": 1.4369,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7976425423774636,
+      "grad_norm": 26.261287689208984,
+      "learning_rate": 4.085739282589677e-06,
+      "loss": 1.5401,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7980751034091216,
+      "grad_norm": 24.996610641479492,
+      "learning_rate": 4.076990376202975e-06,
+      "loss": 1.5306,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7985076644407797,
+      "grad_norm": 23.960554122924805,
+      "learning_rate": 4.068241469816273e-06,
+      "loss": 1.5065,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7989402254724377,
+      "grad_norm": 31.743310928344727,
+      "learning_rate": 4.059492563429572e-06,
+      "loss": 1.59,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7993727865040958,
+      "grad_norm": 23.445873260498047,
+      "learning_rate": 4.0507436570428695e-06,
+      "loss": 1.5049,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7998053475357538,
+      "grad_norm": 24.140777587890625,
+      "learning_rate": 4.041994750656169e-06,
+      "loss": 1.5721,
+      "step": 1849
+    },
+    {
+      "epoch": 0.800237908567412,
+      "grad_norm": 23.796100616455078,
+      "learning_rate": 4.033245844269466e-06,
+      "loss": 1.5244,
+      "step": 1850
+    },
+    {
+      "epoch": 0.80067046959907,
+      "grad_norm": 29.071531295776367,
+      "learning_rate": 4.024496937882765e-06,
+      "loss": 1.5071,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8011030306307281,
+      "grad_norm": 24.83522605895996,
+      "learning_rate": 4.015748031496064e-06,
+      "loss": 1.5093,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8015355916623861,
+      "grad_norm": 25.007980346679688,
+      "learning_rate": 4.006999125109361e-06,
+      "loss": 1.51,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8019681526940442,
+      "grad_norm": 23.728660583496094,
+      "learning_rate": 3.9982502187226605e-06,
+      "loss": 1.475,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8024007137257022,
+      "grad_norm": 26.348461151123047,
+      "learning_rate": 3.989501312335958e-06,
+      "loss": 1.5531,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8028332747573603,
+      "grad_norm": 25.381500244140625,
+      "learning_rate": 3.980752405949257e-06,
+      "loss": 1.4599,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8032658357890183,
+      "grad_norm": 24.115501403808594,
+      "learning_rate": 3.972003499562555e-06,
+      "loss": 1.5345,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8036983968206765,
+      "grad_norm": 24.597143173217773,
+      "learning_rate": 3.963254593175853e-06,
+      "loss": 1.4688,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8041309578523345,
+      "grad_norm": 24.78520965576172,
+      "learning_rate": 3.9545056867891515e-06,
+      "loss": 1.4823,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8045635188839926,
+      "grad_norm": 23.59719467163086,
+      "learning_rate": 3.94575678040245e-06,
+      "loss": 1.4657,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8049960799156506,
+      "grad_norm": 26.260265350341797,
+      "learning_rate": 3.937007874015748e-06,
+      "loss": 1.5238,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8054286409473087,
+      "grad_norm": 26.81787872314453,
+      "learning_rate": 3.9282589676290465e-06,
+      "loss": 1.5246,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8058612019789667,
+      "grad_norm": 24.150634765625,
+      "learning_rate": 3.919510061242345e-06,
+      "loss": 1.4304,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8062937630106248,
+      "grad_norm": 25.50913429260254,
+      "learning_rate": 3.910761154855643e-06,
+      "loss": 1.482,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8067263240422828,
+      "grad_norm": 27.578847885131836,
+      "learning_rate": 3.902012248468942e-06,
+      "loss": 1.4806,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8071588850739408,
+      "grad_norm": 28.148900985717773,
+      "learning_rate": 3.89326334208224e-06,
+      "loss": 1.5452,
+      "step": 1866
+    },
+    {
+      "epoch": 0.807591446105599,
+      "grad_norm": 25.29563331604004,
+      "learning_rate": 3.884514435695538e-06,
+      "loss": 1.4855,
+      "step": 1867
+    },
+    {
+      "epoch": 0.808024007137257,
+      "grad_norm": 26.5310001373291,
+      "learning_rate": 3.875765529308837e-06,
+      "loss": 1.4845,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8084565681689151,
+      "grad_norm": 24.8538818359375,
+      "learning_rate": 3.867016622922135e-06,
+      "loss": 1.5439,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8088891292005731,
+      "grad_norm": 24.38702964782715,
+      "learning_rate": 3.858267716535433e-06,
+      "loss": 1.524,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8093216902322312,
+      "grad_norm": 25.538375854492188,
+      "learning_rate": 3.849518810148732e-06,
+      "loss": 1.5586,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8097542512638892,
+      "grad_norm": 24.558469772338867,
+      "learning_rate": 3.84076990376203e-06,
+      "loss": 1.5687,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8101868122955473,
+      "grad_norm": 24.07853126525879,
+      "learning_rate": 3.8320209973753285e-06,
+      "loss": 1.5702,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8106193733272054,
+      "grad_norm": 23.100337982177734,
+      "learning_rate": 3.823272090988627e-06,
+      "loss": 1.5732,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8110519343588635,
+      "grad_norm": 24.077980041503906,
+      "learning_rate": 3.814523184601925e-06,
+      "loss": 1.4838,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8114844953905215,
+      "grad_norm": 24.32474708557129,
+      "learning_rate": 3.8057742782152236e-06,
+      "loss": 1.5323,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8119170564221796,
+      "grad_norm": 26.63454246520996,
+      "learning_rate": 3.7970253718285215e-06,
+      "loss": 1.581,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8123496174538376,
+      "grad_norm": 26.129480361938477,
+      "learning_rate": 3.7882764654418203e-06,
+      "loss": 1.5964,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8127821784854957,
+      "grad_norm": 25.31804847717285,
+      "learning_rate": 3.7795275590551182e-06,
+      "loss": 1.5183,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8132147395171537,
+      "grad_norm": 25.525976181030273,
+      "learning_rate": 3.770778652668417e-06,
+      "loss": 1.5983,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8136473005488118,
+      "grad_norm": 23.734786987304688,
+      "learning_rate": 3.762029746281715e-06,
+      "loss": 1.555,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8140798615804699,
+      "grad_norm": 23.824871063232422,
+      "learning_rate": 3.7532808398950133e-06,
+      "loss": 1.5189,
+      "step": 1882
+    },
+    {
+      "epoch": 0.814512422612128,
+      "grad_norm": 26.2841854095459,
+      "learning_rate": 3.744531933508312e-06,
+      "loss": 1.543,
+      "step": 1883
+    },
+    {
+      "epoch": 0.814944983643786,
+      "grad_norm": 27.476755142211914,
+      "learning_rate": 3.73578302712161e-06,
+      "loss": 1.5171,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8153775446754441,
+      "grad_norm": 24.633068084716797,
+      "learning_rate": 3.727034120734909e-06,
+      "loss": 1.5069,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8158101057071021,
+      "grad_norm": 26.17974281311035,
+      "learning_rate": 3.7182852143482068e-06,
+      "loss": 1.5122,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8162426667387602,
+      "grad_norm": 24.839109420776367,
+      "learning_rate": 3.7095363079615047e-06,
+      "loss": 1.5153,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8166752277704182,
+      "grad_norm": 21.22071647644043,
+      "learning_rate": 3.7007874015748035e-06,
+      "loss": 1.5247,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8171077888020762,
+      "grad_norm": 25.769725799560547,
+      "learning_rate": 3.692038495188102e-06,
+      "loss": 1.4744,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8175403498337344,
+      "grad_norm": 24.709821701049805,
+      "learning_rate": 3.6832895888014e-06,
+      "loss": 1.5333,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8179729108653924,
+      "grad_norm": 25.408071517944336,
+      "learning_rate": 3.6745406824146986e-06,
+      "loss": 1.5651,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8184054718970505,
+      "grad_norm": 24.688087463378906,
+      "learning_rate": 3.665791776027997e-06,
+      "loss": 1.4822,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8188380329287085,
+      "grad_norm": 24.41063117980957,
+      "learning_rate": 3.6570428696412953e-06,
+      "loss": 1.5928,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8192705939603666,
+      "grad_norm": 25.023773193359375,
+      "learning_rate": 3.6482939632545932e-06,
+      "loss": 1.5182,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8197031549920246,
+      "grad_norm": 26.05735969543457,
+      "learning_rate": 3.639545056867892e-06,
+      "loss": 1.5007,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8201357160236827,
+      "grad_norm": 27.816280364990234,
+      "learning_rate": 3.63079615048119e-06,
+      "loss": 1.4621,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8205682770553407,
+      "grad_norm": 27.703954696655273,
+      "learning_rate": 3.6220472440944887e-06,
+      "loss": 1.4638,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8210008380869989,
+      "grad_norm": 26.362079620361328,
+      "learning_rate": 3.6132983377077867e-06,
+      "loss": 1.5282,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8214333991186569,
+      "grad_norm": 25.99664306640625,
+      "learning_rate": 3.604549431321085e-06,
+      "loss": 1.4904,
+      "step": 1899
+    },
+    {
+      "epoch": 0.821865960150315,
+      "grad_norm": 24.36591911315918,
+      "learning_rate": 3.595800524934384e-06,
+      "loss": 1.555,
+      "step": 1900
+    },
+    {
+      "epoch": 0.822298521181973,
+      "grad_norm": 25.207164764404297,
+      "learning_rate": 3.5870516185476817e-06,
+      "loss": 1.5043,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8227310822136311,
+      "grad_norm": 27.927532196044922,
+      "learning_rate": 3.5783027121609805e-06,
+      "loss": 1.5988,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8231636432452891,
+      "grad_norm": 23.66588020324707,
+      "learning_rate": 3.5695538057742785e-06,
+      "loss": 1.5167,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8235962042769472,
+      "grad_norm": 25.874834060668945,
+      "learning_rate": 3.5608048993875772e-06,
+      "loss": 1.4365,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8240287653086052,
+      "grad_norm": 23.785438537597656,
+      "learning_rate": 3.552055993000875e-06,
+      "loss": 1.5468,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8244613263402634,
+      "grad_norm": 27.696264266967773,
+      "learning_rate": 3.5433070866141735e-06,
+      "loss": 1.5845,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8248938873719214,
+      "grad_norm": 24.913143157958984,
+      "learning_rate": 3.534558180227472e-06,
+      "loss": 1.6031,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8253264484035795,
+      "grad_norm": 25.623226165771484,
+      "learning_rate": 3.5258092738407703e-06,
+      "loss": 1.4782,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8257590094352375,
+      "grad_norm": 26.468833923339844,
+      "learning_rate": 3.5170603674540686e-06,
+      "loss": 1.5557,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8261915704668955,
+      "grad_norm": 25.889429092407227,
+      "learning_rate": 3.508311461067367e-06,
+      "loss": 1.4884,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8266241314985536,
+      "grad_norm": 23.6556453704834,
+      "learning_rate": 3.499562554680665e-06,
+      "loss": 1.5173,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8270566925302116,
+      "grad_norm": 26.38465690612793,
+      "learning_rate": 3.4908136482939637e-06,
+      "loss": 1.4916,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8274892535618698,
+      "grad_norm": 22.687150955200195,
+      "learning_rate": 3.4820647419072616e-06,
+      "loss": 1.4722,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8279218145935278,
+      "grad_norm": 23.410667419433594,
+      "learning_rate": 3.4733158355205604e-06,
+      "loss": 1.4913,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8283543756251859,
+      "grad_norm": 28.485891342163086,
+      "learning_rate": 3.4645669291338583e-06,
+      "loss": 1.4458,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8287869366568439,
+      "grad_norm": 25.982906341552734,
+      "learning_rate": 3.455818022747157e-06,
+      "loss": 1.4411,
+      "step": 1916
+    },
+    {
+      "epoch": 0.829219497688502,
+      "grad_norm": 25.231077194213867,
+      "learning_rate": 3.447069116360455e-06,
+      "loss": 1.4979,
+      "step": 1917
+    },
+    {
+      "epoch": 0.82965205872016,
+      "grad_norm": 23.16214370727539,
+      "learning_rate": 3.4383202099737534e-06,
+      "loss": 1.4828,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8300846197518181,
+      "grad_norm": 26.24925994873047,
+      "learning_rate": 3.429571303587052e-06,
+      "loss": 1.5102,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8305171807834761,
+      "grad_norm": 21.732011795043945,
+      "learning_rate": 3.42082239720035e-06,
+      "loss": 1.5264,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8309497418151343,
+      "grad_norm": 26.923465728759766,
+      "learning_rate": 3.412073490813649e-06,
+      "loss": 1.4386,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8313823028467923,
+      "grad_norm": 26.28173828125,
+      "learning_rate": 3.403324584426947e-06,
+      "loss": 1.5147,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8318148638784504,
+      "grad_norm": 27.070018768310547,
+      "learning_rate": 3.394575678040245e-06,
+      "loss": 1.4853,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8322474249101084,
+      "grad_norm": 23.679351806640625,
+      "learning_rate": 3.3858267716535436e-06,
+      "loss": 1.5418,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8326799859417665,
+      "grad_norm": 27.61638832092285,
+      "learning_rate": 3.377077865266842e-06,
+      "loss": 1.5201,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8331125469734245,
+      "grad_norm": 26.128068923950195,
+      "learning_rate": 3.3683289588801403e-06,
+      "loss": 1.5359,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8335451080050826,
+      "grad_norm": 24.252161026000977,
+      "learning_rate": 3.3595800524934387e-06,
+      "loss": 1.4833,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8339776690367406,
+      "grad_norm": 26.1506404876709,
+      "learning_rate": 3.350831146106737e-06,
+      "loss": 1.5333,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8344102300683988,
+      "grad_norm": 25.713539123535156,
+      "learning_rate": 3.3420822397200354e-06,
+      "loss": 1.5724,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8348427911000568,
+      "grad_norm": 24.772716522216797,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.5876,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8352753521317149,
+      "grad_norm": 26.96317481994629,
+      "learning_rate": 3.324584426946632e-06,
+      "loss": 1.6016,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8357079131633729,
+      "grad_norm": 25.352603912353516,
+      "learning_rate": 3.31583552055993e-06,
+      "loss": 1.4939,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8361404741950309,
+      "grad_norm": 24.081235885620117,
+      "learning_rate": 3.307086614173229e-06,
+      "loss": 1.4755,
+      "step": 1933
+    },
+    {
+      "epoch": 0.836573035226689,
+      "grad_norm": 26.306793212890625,
+      "learning_rate": 3.2983377077865268e-06,
+      "loss": 1.5863,
+      "step": 1934
+    },
+    {
+      "epoch": 0.837005596258347,
+      "grad_norm": 27.727188110351562,
+      "learning_rate": 3.289588801399825e-06,
+      "loss": 1.5823,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8374381572900051,
+      "grad_norm": 25.65025520324707,
+      "learning_rate": 3.280839895013124e-06,
+      "loss": 1.4594,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8378707183216632,
+      "grad_norm": 23.665359497070312,
+      "learning_rate": 3.272090988626422e-06,
+      "loss": 1.6313,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8383032793533213,
+      "grad_norm": 23.80894660949707,
+      "learning_rate": 3.2633420822397206e-06,
+      "loss": 1.5172,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8387358403849793,
+      "grad_norm": 26.22355842590332,
+      "learning_rate": 3.2545931758530186e-06,
+      "loss": 1.564,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8391684014166374,
+      "grad_norm": 24.528112411499023,
+      "learning_rate": 3.2458442694663165e-06,
+      "loss": 1.519,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8396009624482954,
+      "grad_norm": 26.274965286254883,
+      "learning_rate": 3.2370953630796153e-06,
+      "loss": 1.5012,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8400335234799535,
+      "grad_norm": 26.261157989501953,
+      "learning_rate": 3.2283464566929136e-06,
+      "loss": 1.481,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8404660845116115,
+      "grad_norm": 27.06733512878418,
+      "learning_rate": 3.219597550306212e-06,
+      "loss": 1.5262,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8408986455432697,
+      "grad_norm": 23.958738327026367,
+      "learning_rate": 3.2108486439195104e-06,
+      "loss": 1.5564,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8413312065749277,
+      "grad_norm": 24.236982345581055,
+      "learning_rate": 3.2020997375328087e-06,
+      "loss": 1.5062,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8417637676065858,
+      "grad_norm": 24.281795501708984,
+      "learning_rate": 3.193350831146107e-06,
+      "loss": 1.5618,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8421963286382438,
+      "grad_norm": 23.478912353515625,
+      "learning_rate": 3.184601924759405e-06,
+      "loss": 1.5643,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8426288896699019,
+      "grad_norm": 24.562108993530273,
+      "learning_rate": 3.175853018372704e-06,
+      "loss": 1.4998,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8430614507015599,
+      "grad_norm": 24.686317443847656,
+      "learning_rate": 3.1671041119860017e-06,
+      "loss": 1.4872,
+      "step": 1949
+    },
+    {
+      "epoch": 0.843494011733218,
+      "grad_norm": 25.728281021118164,
+      "learning_rate": 3.1583552055993005e-06,
+      "loss": 1.505,
+      "step": 1950
+    },
+    {
+      "epoch": 0.843926572764876,
+      "grad_norm": 26.217914581298828,
+      "learning_rate": 3.1496062992125985e-06,
+      "loss": 1.4911,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8443591337965342,
+      "grad_norm": 25.863449096679688,
+      "learning_rate": 3.140857392825897e-06,
+      "loss": 1.6666,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8447916948281922,
+      "grad_norm": 25.692018508911133,
+      "learning_rate": 3.1321084864391956e-06,
+      "loss": 1.5049,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8452242558598502,
+      "grad_norm": 25.147676467895508,
+      "learning_rate": 3.1233595800524935e-06,
+      "loss": 1.5504,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8456568168915083,
+      "grad_norm": 26.268417358398438,
+      "learning_rate": 3.1146106736657923e-06,
+      "loss": 1.4939,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8460893779231663,
+      "grad_norm": 22.62088394165039,
+      "learning_rate": 3.1058617672790903e-06,
+      "loss": 1.5196,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8465219389548244,
+      "grad_norm": 23.995553970336914,
+      "learning_rate": 3.097112860892389e-06,
+      "loss": 1.4776,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8469544999864824,
+      "grad_norm": 24.5943546295166,
+      "learning_rate": 3.088363954505687e-06,
+      "loss": 1.5271,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8473870610181405,
+      "grad_norm": 24.14809799194336,
+      "learning_rate": 3.0796150481189853e-06,
+      "loss": 1.5425,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8478196220497985,
+      "grad_norm": 22.18802833557129,
+      "learning_rate": 3.0708661417322837e-06,
+      "loss": 1.5071,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8482521830814567,
+      "grad_norm": 23.465879440307617,
+      "learning_rate": 3.062117235345582e-06,
+      "loss": 1.4322,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8486847441131147,
+      "grad_norm": 25.830198287963867,
+      "learning_rate": 3.0533683289588804e-06,
+      "loss": 1.5651,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8491173051447728,
+      "grad_norm": 28.602500915527344,
+      "learning_rate": 3.0446194225721788e-06,
+      "loss": 1.3828,
+      "step": 1963
+    },
+    {
+      "epoch": 0.8495498661764308,
+      "grad_norm": 24.01179313659668,
+      "learning_rate": 3.0358705161854767e-06,
+      "loss": 1.5234,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8499824272080889,
+      "grad_norm": 26.218347549438477,
+      "learning_rate": 3.0271216097987755e-06,
+      "loss": 1.4781,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8504149882397469,
+      "grad_norm": 26.33125877380371,
+      "learning_rate": 3.0183727034120734e-06,
+      "loss": 1.4859,
+      "step": 1966
+    },
+    {
+      "epoch": 0.850847549271405,
+      "grad_norm": 25.678081512451172,
+      "learning_rate": 3.0096237970253722e-06,
+      "loss": 1.4866,
+      "step": 1967
+    },
+    {
+      "epoch": 0.851280110303063,
+      "grad_norm": 29.468505859375,
+      "learning_rate": 3.00087489063867e-06,
+      "loss": 1.5086,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8517126713347212,
+      "grad_norm": 23.861677169799805,
+      "learning_rate": 2.992125984251969e-06,
+      "loss": 1.5575,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8521452323663792,
+      "grad_norm": 25.250822067260742,
+      "learning_rate": 2.9833770778652673e-06,
+      "loss": 1.4855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8525777933980373,
+      "grad_norm": 26.805891036987305,
+      "learning_rate": 2.9746281714785652e-06,
+      "loss": 1.4851,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8530103544296953,
+      "grad_norm": 23.66602897644043,
+      "learning_rate": 2.965879265091864e-06,
+      "loss": 1.5238,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8534429154613534,
+      "grad_norm": 25.112504959106445,
+      "learning_rate": 2.957130358705162e-06,
+      "loss": 1.4722,
+      "step": 1973
+    },
+    {
+      "epoch": 0.8538754764930114,
+      "grad_norm": 24.553518295288086,
+      "learning_rate": 2.9483814523184607e-06,
+      "loss": 1.5377,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8543080375246696,
+      "grad_norm": 27.337066650390625,
+      "learning_rate": 2.9396325459317587e-06,
+      "loss": 1.5673,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8547405985563276,
+      "grad_norm": 28.423484802246094,
+      "learning_rate": 2.930883639545057e-06,
+      "loss": 1.5152,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8551731595879856,
+      "grad_norm": 25.569091796875,
+      "learning_rate": 2.9221347331583554e-06,
+      "loss": 1.409,
+      "step": 1977
+    },
+    {
+      "epoch": 0.8556057206196437,
+      "grad_norm": 22.94341468811035,
+      "learning_rate": 2.9133858267716538e-06,
+      "loss": 1.4882,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8560382816513017,
+      "grad_norm": 23.691856384277344,
+      "learning_rate": 2.904636920384952e-06,
+      "loss": 1.4557,
+      "step": 1979
+    },
+    {
+      "epoch": 0.8564708426829598,
+      "grad_norm": 27.209903717041016,
+      "learning_rate": 2.8958880139982505e-06,
+      "loss": 1.51,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8569034037146178,
+      "grad_norm": 26.904014587402344,
+      "learning_rate": 2.887139107611549e-06,
+      "loss": 1.6156,
+      "step": 1981
+    },
+    {
+      "epoch": 0.8573359647462759,
+      "grad_norm": 24.399385452270508,
+      "learning_rate": 2.878390201224847e-06,
+      "loss": 1.5173,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8577685257779339,
+      "grad_norm": 23.10108184814453,
+      "learning_rate": 2.869641294838145e-06,
+      "loss": 1.4796,
+      "step": 1983
+    },
+    {
+      "epoch": 0.8582010868095921,
+      "grad_norm": 27.3771915435791,
+      "learning_rate": 2.860892388451444e-06,
+      "loss": 1.5105,
+      "step": 1984
+    },
+    {
+      "epoch": 0.8586336478412501,
+      "grad_norm": 24.214082717895508,
+      "learning_rate": 2.852143482064742e-06,
+      "loss": 1.5859,
+      "step": 1985
+    },
+    {
+      "epoch": 0.8590662088729082,
+      "grad_norm": 25.662439346313477,
+      "learning_rate": 2.8433945756780406e-06,
+      "loss": 1.4917,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8594987699045662,
+      "grad_norm": 26.082138061523438,
+      "learning_rate": 2.8346456692913386e-06,
+      "loss": 1.5926,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8599313309362243,
+      "grad_norm": 25.768564224243164,
+      "learning_rate": 2.825896762904637e-06,
+      "loss": 1.5071,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8603638919678823,
+      "grad_norm": 26.158370971679688,
+      "learning_rate": 2.8171478565179357e-06,
+      "loss": 1.4536,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8607964529995404,
+      "grad_norm": 25.917757034301758,
+      "learning_rate": 2.8083989501312337e-06,
+      "loss": 1.4675,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8612290140311984,
+      "grad_norm": 24.782541275024414,
+      "learning_rate": 2.7996500437445324e-06,
+      "loss": 1.4918,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8616615750628566,
+      "grad_norm": 24.238004684448242,
+      "learning_rate": 2.7909011373578304e-06,
+      "loss": 1.55,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8620941360945146,
+      "grad_norm": 28.761098861694336,
+      "learning_rate": 2.782152230971129e-06,
+      "loss": 1.5032,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8625266971261727,
+      "grad_norm": 28.440311431884766,
+      "learning_rate": 2.773403324584427e-06,
+      "loss": 1.4544,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8629592581578307,
+      "grad_norm": 24.347803115844727,
+      "learning_rate": 2.7646544181977255e-06,
+      "loss": 1.4856,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8633918191894888,
+      "grad_norm": 23.017044067382812,
+      "learning_rate": 2.755905511811024e-06,
+      "loss": 1.513,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8638243802211468,
+      "grad_norm": 24.277076721191406,
+      "learning_rate": 2.747156605424322e-06,
+      "loss": 1.5532,
+      "step": 1997
+    },
+    {
+      "epoch": 0.8642569412528048,
+      "grad_norm": 26.928024291992188,
+      "learning_rate": 2.7384076990376205e-06,
+      "loss": 1.5431,
+      "step": 1998
+    },
+    {
+      "epoch": 0.864689502284463,
+      "grad_norm": 25.287187576293945,
+      "learning_rate": 2.729658792650919e-06,
+      "loss": 1.5137,
+      "step": 1999
+    },
+    {
+      "epoch": 0.865122063316121,
+      "grad_norm": 23.46804428100586,
+      "learning_rate": 2.720909886264217e-06,
+      "loss": 1.5105,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8655546243477791,
+      "grad_norm": 26.317270278930664,
+      "learning_rate": 2.7121609798775156e-06,
+      "loss": 1.4855,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8659871853794371,
+      "grad_norm": 23.94223403930664,
+      "learning_rate": 2.7034120734908135e-06,
+      "loss": 1.5016,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8664197464110952,
+      "grad_norm": 25.011159896850586,
+      "learning_rate": 2.6946631671041123e-06,
+      "loss": 1.5259,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8668523074427532,
+      "grad_norm": 26.059602737426758,
+      "learning_rate": 2.6859142607174103e-06,
+      "loss": 1.4911,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8672848684744113,
+      "grad_norm": 25.560832977294922,
+      "learning_rate": 2.677165354330709e-06,
+      "loss": 1.5273,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8677174295060693,
+      "grad_norm": 24.12883186340332,
+      "learning_rate": 2.6684164479440074e-06,
+      "loss": 1.5653,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8681499905377275,
+      "grad_norm": 24.72055435180664,
+      "learning_rate": 2.6596675415573053e-06,
+      "loss": 1.532,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8685825515693855,
+      "grad_norm": 23.121793746948242,
+      "learning_rate": 2.650918635170604e-06,
+      "loss": 1.532,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8690151126010436,
+      "grad_norm": 27.0283145904541,
+      "learning_rate": 2.642169728783902e-06,
+      "loss": 1.4829,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8694476736327016,
+      "grad_norm": 25.010900497436523,
+      "learning_rate": 2.633420822397201e-06,
+      "loss": 1.5846,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8698802346643597,
+      "grad_norm": 26.372800827026367,
+      "learning_rate": 2.6246719160104988e-06,
+      "loss": 1.5037,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8703127956960177,
+      "grad_norm": 24.331981658935547,
+      "learning_rate": 2.615923009623797e-06,
+      "loss": 1.431,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8707453567276758,
+      "grad_norm": 24.65950584411621,
+      "learning_rate": 2.6071741032370955e-06,
+      "loss": 1.4707,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8711779177593338,
+      "grad_norm": 25.584592819213867,
+      "learning_rate": 2.598425196850394e-06,
+      "loss": 1.4639,
+      "step": 2014
+    },
+    {
+      "epoch": 0.871610478790992,
+      "grad_norm": 24.871679306030273,
+      "learning_rate": 2.5896762904636922e-06,
+      "loss": 1.4434,
+      "step": 2015
+    },
+    {
+      "epoch": 0.87204303982265,
+      "grad_norm": 28.186784744262695,
+      "learning_rate": 2.5809273840769906e-06,
+      "loss": 1.5004,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8724756008543081,
+      "grad_norm": 23.437673568725586,
+      "learning_rate": 2.5721784776902894e-06,
+      "loss": 1.5752,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8729081618859661,
+      "grad_norm": 24.34725570678711,
+      "learning_rate": 2.5634295713035873e-06,
+      "loss": 1.5573,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8733407229176242,
+      "grad_norm": 26.968994140625,
+      "learning_rate": 2.5546806649168852e-06,
+      "loss": 1.466,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8737732839492822,
+      "grad_norm": 23.812091827392578,
+      "learning_rate": 2.545931758530184e-06,
+      "loss": 1.4728,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8742058449809402,
+      "grad_norm": 26.316484451293945,
+      "learning_rate": 2.537182852143482e-06,
+      "loss": 1.495,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8746384060125983,
+      "grad_norm": 26.36723518371582,
+      "learning_rate": 2.5284339457567807e-06,
+      "loss": 1.4384,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8750709670442564,
+      "grad_norm": 27.980514526367188,
+      "learning_rate": 2.519685039370079e-06,
+      "loss": 1.5268,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8755035280759145,
+      "grad_norm": 25.934511184692383,
+      "learning_rate": 2.510936132983377e-06,
+      "loss": 1.5593,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8759360891075725,
+      "grad_norm": 24.022119522094727,
+      "learning_rate": 2.502187226596676e-06,
+      "loss": 1.5023,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8763686501392306,
+      "grad_norm": 24.290563583374023,
+      "learning_rate": 2.493438320209974e-06,
+      "loss": 1.4827,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8768012111708886,
+      "grad_norm": 24.21749496459961,
+      "learning_rate": 2.484689413823272e-06,
+      "loss": 1.425,
+      "step": 2027
+    },
+    {
+      "epoch": 0.8772337722025467,
+      "grad_norm": 26.60518455505371,
+      "learning_rate": 2.4759405074365705e-06,
+      "loss": 1.5017,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8776663332342047,
+      "grad_norm": 25.764251708984375,
+      "learning_rate": 2.467191601049869e-06,
+      "loss": 1.5647,
+      "step": 2029
+    },
+    {
+      "epoch": 0.8780988942658629,
+      "grad_norm": 24.295669555664062,
+      "learning_rate": 2.458442694663167e-06,
+      "loss": 1.4745,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8785314552975209,
+      "grad_norm": 28.384014129638672,
+      "learning_rate": 2.4496937882764656e-06,
+      "loss": 1.554,
+      "step": 2031
+    },
+    {
+      "epoch": 0.878964016329179,
+      "grad_norm": 23.786191940307617,
+      "learning_rate": 2.440944881889764e-06,
+      "loss": 1.5333,
+      "step": 2032
+    },
+    {
+      "epoch": 0.879396577360837,
+      "grad_norm": 25.375789642333984,
+      "learning_rate": 2.4321959755030623e-06,
+      "loss": 1.5183,
+      "step": 2033
+    },
+    {
+      "epoch": 0.8798291383924951,
+      "grad_norm": 24.33387565612793,
+      "learning_rate": 2.4234470691163606e-06,
+      "loss": 1.4466,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8802616994241531,
+      "grad_norm": 23.24261474609375,
+      "learning_rate": 2.414698162729659e-06,
+      "loss": 1.5362,
+      "step": 2035
+    },
+    {
+      "epoch": 0.8806942604558112,
+      "grad_norm": 24.91956329345703,
+      "learning_rate": 2.4059492563429574e-06,
+      "loss": 1.4217,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8811268214874692,
+      "grad_norm": 24.583356857299805,
+      "learning_rate": 2.3972003499562557e-06,
+      "loss": 1.4938,
+      "step": 2037
+    },
+    {
+      "epoch": 0.8815593825191274,
+      "grad_norm": 25.635778427124023,
+      "learning_rate": 2.388451443569554e-06,
+      "loss": 1.5102,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8819919435507854,
+      "grad_norm": 24.82320785522461,
+      "learning_rate": 2.379702537182852e-06,
+      "loss": 1.5763,
+      "step": 2039
+    },
+    {
+      "epoch": 0.8824245045824435,
+      "grad_norm": 24.23630714416504,
+      "learning_rate": 2.370953630796151e-06,
+      "loss": 1.5198,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8828570656141015,
+      "grad_norm": 22.778940200805664,
+      "learning_rate": 2.362204724409449e-06,
+      "loss": 1.5602,
+      "step": 2041
+    },
+    {
+      "epoch": 0.8832896266457595,
+      "grad_norm": 23.935001373291016,
+      "learning_rate": 2.3534558180227475e-06,
+      "loss": 1.4802,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8837221876774176,
+      "grad_norm": 27.67841148376465,
+      "learning_rate": 2.344706911636046e-06,
+      "loss": 1.5749,
+      "step": 2043
+    },
+    {
+      "epoch": 0.8841547487090756,
+      "grad_norm": 25.99941635131836,
+      "learning_rate": 2.3359580052493442e-06,
+      "loss": 1.4838,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8845873097407337,
+      "grad_norm": 26.464733123779297,
+      "learning_rate": 2.327209098862642e-06,
+      "loss": 1.4711,
+      "step": 2045
+    },
+    {
+      "epoch": 0.8850198707723917,
+      "grad_norm": 24.384105682373047,
+      "learning_rate": 2.3184601924759405e-06,
+      "loss": 1.5368,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8854524318040499,
+      "grad_norm": 26.473554611206055,
+      "learning_rate": 2.309711286089239e-06,
+      "loss": 1.5003,
+      "step": 2047
+    },
+    {
+      "epoch": 0.8858849928357079,
+      "grad_norm": 23.190519332885742,
+      "learning_rate": 2.3009623797025373e-06,
+      "loss": 1.4441,
+      "step": 2048
+    },
+    {
+      "epoch": 0.886317553867366,
+      "grad_norm": 25.446199417114258,
+      "learning_rate": 2.2922134733158356e-06,
+      "loss": 1.5295,
+      "step": 2049
+    },
+    {
+      "epoch": 0.886750114899024,
+      "grad_norm": 25.416261672973633,
+      "learning_rate": 2.283464566929134e-06,
+      "loss": 1.4706,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8871826759306821,
+      "grad_norm": 23.452497482299805,
+      "learning_rate": 2.2747156605424323e-06,
+      "loss": 1.54,
+      "step": 2051
+    },
+    {
+      "epoch": 0.8876152369623401,
+      "grad_norm": 26.014728546142578,
+      "learning_rate": 2.2659667541557307e-06,
+      "loss": 1.531,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8880477979939982,
+      "grad_norm": 28.156362533569336,
+      "learning_rate": 2.257217847769029e-06,
+      "loss": 1.6611,
+      "step": 2053
+    },
+    {
+      "epoch": 0.8884803590256563,
+      "grad_norm": 28.15045928955078,
+      "learning_rate": 2.2484689413823274e-06,
+      "loss": 1.4967,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8889129200573144,
+      "grad_norm": 26.710493087768555,
+      "learning_rate": 2.2397200349956258e-06,
+      "loss": 1.5291,
+      "step": 2055
+    },
+    {
+      "epoch": 0.8893454810889724,
+      "grad_norm": 26.1806640625,
+      "learning_rate": 2.230971128608924e-06,
+      "loss": 1.6146,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8897780421206305,
+      "grad_norm": 28.079967498779297,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.5331,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8902106031522885,
+      "grad_norm": 26.610944747924805,
+      "learning_rate": 2.213473315835521e-06,
+      "loss": 1.4869,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8906431641839466,
+      "grad_norm": 24.159992218017578,
+      "learning_rate": 2.2047244094488192e-06,
+      "loss": 1.5627,
+      "step": 2059
+    },
+    {
+      "epoch": 0.8910757252156046,
+      "grad_norm": 25.268817901611328,
+      "learning_rate": 2.1959755030621176e-06,
+      "loss": 1.513,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8915082862472627,
+      "grad_norm": 23.28951644897461,
+      "learning_rate": 2.187226596675416e-06,
+      "loss": 1.5356,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8919408472789208,
+      "grad_norm": 24.559032440185547,
+      "learning_rate": 2.1784776902887143e-06,
+      "loss": 1.5505,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8923734083105789,
+      "grad_norm": 22.86376190185547,
+      "learning_rate": 2.1697287839020122e-06,
+      "loss": 1.4958,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8928059693422369,
+      "grad_norm": 25.818510055541992,
+      "learning_rate": 2.1609798775153106e-06,
+      "loss": 1.5579,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8932385303738949,
+      "grad_norm": 23.554603576660156,
+      "learning_rate": 2.152230971128609e-06,
+      "loss": 1.5331,
+      "step": 2065
+    },
+    {
+      "epoch": 0.893671091405553,
+      "grad_norm": 22.988496780395508,
+      "learning_rate": 2.1434820647419073e-06,
+      "loss": 1.4828,
+      "step": 2066
+    },
+    {
+      "epoch": 0.894103652437211,
+      "grad_norm": 25.738300323486328,
+      "learning_rate": 2.1347331583552057e-06,
+      "loss": 1.5182,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8945362134688691,
+      "grad_norm": 26.744293212890625,
+      "learning_rate": 2.125984251968504e-06,
+      "loss": 1.4402,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8949687745005271,
+      "grad_norm": 26.04250144958496,
+      "learning_rate": 2.1172353455818024e-06,
+      "loss": 1.4826,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8954013355321853,
+      "grad_norm": 24.584280014038086,
+      "learning_rate": 2.1084864391951008e-06,
+      "loss": 1.5665,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8958338965638433,
+      "grad_norm": 28.59994888305664,
+      "learning_rate": 2.099737532808399e-06,
+      "loss": 1.4376,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8962664575955014,
+      "grad_norm": 24.084014892578125,
+      "learning_rate": 2.0909886264216975e-06,
+      "loss": 1.4553,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8966990186271594,
+      "grad_norm": 25.38064193725586,
+      "learning_rate": 2.082239720034996e-06,
+      "loss": 1.5537,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8971315796588175,
+      "grad_norm": 24.919994354248047,
+      "learning_rate": 2.073490813648294e-06,
+      "loss": 1.508,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8975641406904755,
+      "grad_norm": 24.744308471679688,
+      "learning_rate": 2.0647419072615926e-06,
+      "loss": 1.5346,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8979967017221336,
+      "grad_norm": 24.991436004638672,
+      "learning_rate": 2.055993000874891e-06,
+      "loss": 1.5086,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8984292627537916,
+      "grad_norm": 26.734596252441406,
+      "learning_rate": 2.0472440944881893e-06,
+      "loss": 1.4796,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8988618237854498,
+      "grad_norm": 24.347599029541016,
+      "learning_rate": 2.0384951881014876e-06,
+      "loss": 1.5146,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8992943848171078,
+      "grad_norm": 24.754878997802734,
+      "learning_rate": 2.029746281714786e-06,
+      "loss": 1.5116,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8997269458487659,
+      "grad_norm": 23.960243225097656,
+      "learning_rate": 2.0209973753280844e-06,
+      "loss": 1.5233,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9001595068804239,
+      "grad_norm": 24.335981369018555,
+      "learning_rate": 2.0122484689413823e-06,
+      "loss": 1.5004,
+      "step": 2081
+    },
+    {
+      "epoch": 0.900592067912082,
+      "grad_norm": 25.127904891967773,
+      "learning_rate": 2.0034995625546807e-06,
+      "loss": 1.4853,
+      "step": 2082
+    },
+    {
+      "epoch": 0.90102462894374,
+      "grad_norm": 27.572782516479492,
+      "learning_rate": 1.994750656167979e-06,
+      "loss": 1.4948,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9014571899753981,
+      "grad_norm": 25.628870010375977,
+      "learning_rate": 1.9860017497812774e-06,
+      "loss": 1.5031,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9018897510070562,
+      "grad_norm": 24.450708389282227,
+      "learning_rate": 1.9772528433945757e-06,
+      "loss": 1.4799,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9023223120387142,
+      "grad_norm": 23.836322784423828,
+      "learning_rate": 1.968503937007874e-06,
+      "loss": 1.571,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9027548730703723,
+      "grad_norm": 26.30670928955078,
+      "learning_rate": 1.9597550306211725e-06,
+      "loss": 1.6103,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9031874341020303,
+      "grad_norm": 25.850881576538086,
+      "learning_rate": 1.951006124234471e-06,
+      "loss": 1.5946,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9036199951336884,
+      "grad_norm": 21.866104125976562,
+      "learning_rate": 1.942257217847769e-06,
+      "loss": 1.5118,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9040525561653464,
+      "grad_norm": 24.595714569091797,
+      "learning_rate": 1.9335083114610675e-06,
+      "loss": 1.4713,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9044851171970045,
+      "grad_norm": 28.41547393798828,
+      "learning_rate": 1.924759405074366e-06,
+      "loss": 1.6027,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9049176782286625,
+      "grad_norm": 27.03108787536621,
+      "learning_rate": 1.9160104986876642e-06,
+      "loss": 1.5333,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9053502392603207,
+      "grad_norm": 24.61553955078125,
+      "learning_rate": 1.9072615923009624e-06,
+      "loss": 1.4534,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9057828002919787,
+      "grad_norm": 23.870311737060547,
+      "learning_rate": 1.8985126859142608e-06,
+      "loss": 1.5434,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9062153613236368,
+      "grad_norm": 23.868675231933594,
+      "learning_rate": 1.8897637795275591e-06,
+      "loss": 1.5228,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9066479223552948,
+      "grad_norm": 24.33563804626465,
+      "learning_rate": 1.8810148731408575e-06,
+      "loss": 1.5551,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9070804833869529,
+      "grad_norm": 24.130783081054688,
+      "learning_rate": 1.872265966754156e-06,
+      "loss": 1.5063,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9075130444186109,
+      "grad_norm": 22.190542221069336,
+      "learning_rate": 1.8635170603674544e-06,
+      "loss": 1.4745,
+      "step": 2098
+    },
+    {
+      "epoch": 0.907945605450269,
+      "grad_norm": 27.16103744506836,
+      "learning_rate": 1.8547681539807523e-06,
+      "loss": 1.4179,
+      "step": 2099
+    },
+    {
+      "epoch": 0.908378166481927,
+      "grad_norm": 23.1556339263916,
+      "learning_rate": 1.846019247594051e-06,
+      "loss": 1.5758,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9088107275135852,
+      "grad_norm": 23.74553871154785,
+      "learning_rate": 1.8372703412073493e-06,
+      "loss": 1.5291,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9092432885452432,
+      "grad_norm": 24.70638084411621,
+      "learning_rate": 1.8285214348206476e-06,
+      "loss": 1.5172,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9096758495769013,
+      "grad_norm": 28.004220962524414,
+      "learning_rate": 1.819772528433946e-06,
+      "loss": 1.4554,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9101084106085593,
+      "grad_norm": 25.194591522216797,
+      "learning_rate": 1.8110236220472444e-06,
+      "loss": 1.4265,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9105409716402174,
+      "grad_norm": 26.836763381958008,
+      "learning_rate": 1.8022747156605425e-06,
+      "loss": 1.4578,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9109735326718754,
+      "grad_norm": 22.51131248474121,
+      "learning_rate": 1.7935258092738409e-06,
+      "loss": 1.445,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9114060937035335,
+      "grad_norm": 23.413801193237305,
+      "learning_rate": 1.7847769028871392e-06,
+      "loss": 1.5159,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9118386547351915,
+      "grad_norm": 25.643709182739258,
+      "learning_rate": 1.7760279965004376e-06,
+      "loss": 1.4713,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9122712157668496,
+      "grad_norm": 26.521026611328125,
+      "learning_rate": 1.767279090113736e-06,
+      "loss": 1.5231,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9127037767985077,
+      "grad_norm": 26.597665786743164,
+      "learning_rate": 1.7585301837270343e-06,
+      "loss": 1.4962,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9131363378301657,
+      "grad_norm": 24.662185668945312,
+      "learning_rate": 1.7497812773403325e-06,
+      "loss": 1.4917,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9135688988618238,
+      "grad_norm": 25.956602096557617,
+      "learning_rate": 1.7410323709536308e-06,
+      "loss": 1.4278,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9140014598934818,
+      "grad_norm": 25.572887420654297,
+      "learning_rate": 1.7322834645669292e-06,
+      "loss": 1.5098,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9144340209251399,
+      "grad_norm": 26.01473045349121,
+      "learning_rate": 1.7235345581802275e-06,
+      "loss": 1.5062,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9148665819567979,
+      "grad_norm": 23.53647804260254,
+      "learning_rate": 1.714785651793526e-06,
+      "loss": 1.5157,
+      "step": 2115
+    },
+    {
+      "epoch": 0.915299142988456,
+      "grad_norm": 26.578235626220703,
+      "learning_rate": 1.7060367454068245e-06,
+      "loss": 1.5472,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9157317040201141,
+      "grad_norm": 24.396869659423828,
+      "learning_rate": 1.6972878390201224e-06,
+      "loss": 1.5047,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9161642650517722,
+      "grad_norm": 21.657743453979492,
+      "learning_rate": 1.688538932633421e-06,
+      "loss": 1.5496,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9165968260834302,
+      "grad_norm": 26.7194766998291,
+      "learning_rate": 1.6797900262467193e-06,
+      "loss": 1.5533,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9170293871150883,
+      "grad_norm": 28.471969604492188,
+      "learning_rate": 1.6710411198600177e-06,
+      "loss": 1.5192,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9174619481467463,
+      "grad_norm": 25.8294620513916,
+      "learning_rate": 1.662292213473316e-06,
+      "loss": 1.572,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9178945091784044,
+      "grad_norm": 24.569931030273438,
+      "learning_rate": 1.6535433070866144e-06,
+      "loss": 1.4803,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9183270702100624,
+      "grad_norm": 28.031354904174805,
+      "learning_rate": 1.6447944006999126e-06,
+      "loss": 1.402,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9187596312417206,
+      "grad_norm": 23.158405303955078,
+      "learning_rate": 1.636045494313211e-06,
+      "loss": 1.5053,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9191921922733786,
+      "grad_norm": 24.734914779663086,
+      "learning_rate": 1.6272965879265093e-06,
+      "loss": 1.4882,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9196247533050367,
+      "grad_norm": 25.3793888092041,
+      "learning_rate": 1.6185476815398076e-06,
+      "loss": 1.5247,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9200573143366947,
+      "grad_norm": 26.39560890197754,
+      "learning_rate": 1.609798775153106e-06,
+      "loss": 1.5502,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9204898753683528,
+      "grad_norm": 25.682659149169922,
+      "learning_rate": 1.6010498687664044e-06,
+      "loss": 1.4543,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9209224364000108,
+      "grad_norm": 25.748123168945312,
+      "learning_rate": 1.5923009623797025e-06,
+      "loss": 1.4761,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9213549974316688,
+      "grad_norm": 23.998931884765625,
+      "learning_rate": 1.5835520559930009e-06,
+      "loss": 1.4683,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9217875584633269,
+      "grad_norm": 31.975799560546875,
+      "learning_rate": 1.5748031496062992e-06,
+      "loss": 1.5012,
+      "step": 2131
+    },
+    {
+      "epoch": 0.922220119494985,
+      "grad_norm": 26.674463272094727,
+      "learning_rate": 1.5660542432195978e-06,
+      "loss": 1.4607,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9226526805266431,
+      "grad_norm": 22.74907684326172,
+      "learning_rate": 1.5573053368328962e-06,
+      "loss": 1.4775,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9230852415583011,
+      "grad_norm": 28.162485122680664,
+      "learning_rate": 1.5485564304461945e-06,
+      "loss": 1.4859,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9235178025899592,
+      "grad_norm": 25.198293685913086,
+      "learning_rate": 1.5398075240594927e-06,
+      "loss": 1.4851,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9239503636216172,
+      "grad_norm": 27.7397403717041,
+      "learning_rate": 1.531058617672791e-06,
+      "loss": 1.5154,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9243829246532753,
+      "grad_norm": 26.483898162841797,
+      "learning_rate": 1.5223097112860894e-06,
+      "loss": 1.5139,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9248154856849333,
+      "grad_norm": 23.793954849243164,
+      "learning_rate": 1.5135608048993877e-06,
+      "loss": 1.5656,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9252480467165914,
+      "grad_norm": 27.676443099975586,
+      "learning_rate": 1.5048118985126861e-06,
+      "loss": 1.5341,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9256806077482495,
+      "grad_norm": 27.942245483398438,
+      "learning_rate": 1.4960629921259845e-06,
+      "loss": 1.5453,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9261131687799076,
+      "grad_norm": 23.83132553100586,
+      "learning_rate": 1.4873140857392826e-06,
+      "loss": 1.5125,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9265457298115656,
+      "grad_norm": 25.71536636352539,
+      "learning_rate": 1.478565179352581e-06,
+      "loss": 1.4835,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9269782908432237,
+      "grad_norm": 22.844636917114258,
+      "learning_rate": 1.4698162729658793e-06,
+      "loss": 1.5344,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9274108518748817,
+      "grad_norm": 26.83110237121582,
+      "learning_rate": 1.4610673665791777e-06,
+      "loss": 1.4896,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9278434129065398,
+      "grad_norm": 23.741893768310547,
+      "learning_rate": 1.452318460192476e-06,
+      "loss": 1.5107,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9282759739381978,
+      "grad_norm": 29.269779205322266,
+      "learning_rate": 1.4435695538057744e-06,
+      "loss": 1.5069,
+      "step": 2146
+    },
+    {
+      "epoch": 0.928708534969856,
+      "grad_norm": 24.56169319152832,
+      "learning_rate": 1.4348206474190726e-06,
+      "loss": 1.4771,
+      "step": 2147
+    },
+    {
+      "epoch": 0.929141096001514,
+      "grad_norm": 26.20133399963379,
+      "learning_rate": 1.426071741032371e-06,
+      "loss": 1.5259,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9295736570331721,
+      "grad_norm": 25.875614166259766,
+      "learning_rate": 1.4173228346456693e-06,
+      "loss": 1.4957,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9300062180648301,
+      "grad_norm": 28.533090591430664,
+      "learning_rate": 1.4085739282589679e-06,
+      "loss": 1.5897,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9304387790964882,
+      "grad_norm": 26.25223731994629,
+      "learning_rate": 1.3998250218722662e-06,
+      "loss": 1.5214,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9308713401281462,
+      "grad_norm": 23.350339889526367,
+      "learning_rate": 1.3910761154855646e-06,
+      "loss": 1.5911,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9313039011598042,
+      "grad_norm": 25.860279083251953,
+      "learning_rate": 1.3823272090988627e-06,
+      "loss": 1.4472,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9317364621914623,
+      "grad_norm": 25.473173141479492,
+      "learning_rate": 1.373578302712161e-06,
+      "loss": 1.5506,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9321690232231203,
+      "grad_norm": 24.590129852294922,
+      "learning_rate": 1.3648293963254594e-06,
+      "loss": 1.5507,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9326015842547785,
+      "grad_norm": 26.397546768188477,
+      "learning_rate": 1.3560804899387578e-06,
+      "loss": 1.5377,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9330341452864365,
+      "grad_norm": 27.98323631286621,
+      "learning_rate": 1.3473315835520562e-06,
+      "loss": 1.5116,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9334667063180946,
+      "grad_norm": 24.400859832763672,
+      "learning_rate": 1.3385826771653545e-06,
+      "loss": 1.5775,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9338992673497526,
+      "grad_norm": 26.30158233642578,
+      "learning_rate": 1.3298337707786527e-06,
+      "loss": 1.4523,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9343318283814107,
+      "grad_norm": 23.896175384521484,
+      "learning_rate": 1.321084864391951e-06,
+      "loss": 1.4708,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9347643894130687,
+      "grad_norm": 24.987720489501953,
+      "learning_rate": 1.3123359580052494e-06,
+      "loss": 1.4932,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9351969504447268,
+      "grad_norm": 23.302629470825195,
+      "learning_rate": 1.3035870516185478e-06,
+      "loss": 1.463,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9356295114763848,
+      "grad_norm": 26.809837341308594,
+      "learning_rate": 1.2948381452318461e-06,
+      "loss": 1.5367,
+      "step": 2163
+    },
+    {
+      "epoch": 0.936062072508043,
+      "grad_norm": 23.170120239257812,
+      "learning_rate": 1.2860892388451447e-06,
+      "loss": 1.4773,
+      "step": 2164
+    },
+    {
+      "epoch": 0.936494633539701,
+      "grad_norm": 24.36250114440918,
+      "learning_rate": 1.2773403324584426e-06,
+      "loss": 1.5286,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9369271945713591,
+      "grad_norm": 26.31621551513672,
+      "learning_rate": 1.268591426071741e-06,
+      "loss": 1.4818,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9373597556030171,
+      "grad_norm": 27.618297576904297,
+      "learning_rate": 1.2598425196850396e-06,
+      "loss": 1.4788,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9377923166346752,
+      "grad_norm": 25.09479522705078,
+      "learning_rate": 1.251093613298338e-06,
+      "loss": 1.5178,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9382248776663332,
+      "grad_norm": 26.131237030029297,
+      "learning_rate": 1.242344706911636e-06,
+      "loss": 1.5872,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9386574386979913,
+      "grad_norm": 25.003849029541016,
+      "learning_rate": 1.2335958005249344e-06,
+      "loss": 1.5635,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9390899997296493,
+      "grad_norm": 26.515464782714844,
+      "learning_rate": 1.2248468941382328e-06,
+      "loss": 1.5046,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9395225607613075,
+      "grad_norm": 26.55370330810547,
+      "learning_rate": 1.2160979877515311e-06,
+      "loss": 1.5933,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9399551217929655,
+      "grad_norm": 24.8226375579834,
+      "learning_rate": 1.2073490813648295e-06,
+      "loss": 1.5164,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9403876828246235,
+      "grad_norm": 24.112558364868164,
+      "learning_rate": 1.1986001749781279e-06,
+      "loss": 1.5325,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9408202438562816,
+      "grad_norm": 25.355792999267578,
+      "learning_rate": 1.189851268591426e-06,
+      "loss": 1.5334,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9412528048879396,
+      "grad_norm": 26.321578979492188,
+      "learning_rate": 1.1811023622047246e-06,
+      "loss": 1.4804,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9416853659195977,
+      "grad_norm": 25.188034057617188,
+      "learning_rate": 1.172353455818023e-06,
+      "loss": 1.5346,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9421179269512557,
+      "grad_norm": 26.958105087280273,
+      "learning_rate": 1.163604549431321e-06,
+      "loss": 1.5058,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9425504879829139,
+      "grad_norm": 23.981430053710938,
+      "learning_rate": 1.1548556430446194e-06,
+      "loss": 1.5336,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9429830490145719,
+      "grad_norm": 24.413782119750977,
+      "learning_rate": 1.1461067366579178e-06,
+      "loss": 1.5212,
+      "step": 2180
+    },
+    {
+      "epoch": 0.94341561004623,
+      "grad_norm": 25.38534164428711,
+      "learning_rate": 1.1373578302712162e-06,
+      "loss": 1.4876,
+      "step": 2181
+    },
+    {
+      "epoch": 0.943848171077888,
+      "grad_norm": 27.410280227661133,
+      "learning_rate": 1.1286089238845145e-06,
+      "loss": 1.495,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9442807321095461,
+      "grad_norm": 24.01247787475586,
+      "learning_rate": 1.1198600174978129e-06,
+      "loss": 1.4992,
+      "step": 2183
+    },
+    {
+      "epoch": 0.9447132931412041,
+      "grad_norm": 28.51532554626465,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.4806,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9451458541728622,
+      "grad_norm": 24.75975227355957,
+      "learning_rate": 1.1023622047244096e-06,
+      "loss": 1.5124,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9455784152045202,
+      "grad_norm": 24.5733585357666,
+      "learning_rate": 1.093613298337708e-06,
+      "loss": 1.5007,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9460109762361784,
+      "grad_norm": 25.291410446166992,
+      "learning_rate": 1.0848643919510061e-06,
+      "loss": 1.4037,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9464435372678364,
+      "grad_norm": 22.40592384338379,
+      "learning_rate": 1.0761154855643045e-06,
+      "loss": 1.506,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9468760982994945,
+      "grad_norm": 25.94204330444336,
+      "learning_rate": 1.0673665791776028e-06,
+      "loss": 1.4871,
+      "step": 2189
+    },
+    {
+      "epoch": 0.9473086593311525,
+      "grad_norm": 23.761117935180664,
+      "learning_rate": 1.0586176727909012e-06,
+      "loss": 1.5323,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9477412203628106,
+      "grad_norm": 25.79743766784668,
+      "learning_rate": 1.0498687664041996e-06,
+      "loss": 1.5,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9481737813944686,
+      "grad_norm": 28.8542423248291,
+      "learning_rate": 1.041119860017498e-06,
+      "loss": 1.4627,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9486063424261267,
+      "grad_norm": 24.279600143432617,
+      "learning_rate": 1.0323709536307963e-06,
+      "loss": 1.5207,
+      "step": 2193
+    },
+    {
+      "epoch": 0.9490389034577847,
+      "grad_norm": 32.73613357543945,
+      "learning_rate": 1.0236220472440946e-06,
+      "loss": 1.5151,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9494714644894429,
+      "grad_norm": 25.603696823120117,
+      "learning_rate": 1.014873140857393e-06,
+      "loss": 1.4946,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9499040255211009,
+      "grad_norm": 29.79012680053711,
+      "learning_rate": 1.0061242344706911e-06,
+      "loss": 1.5321,
+      "step": 2196
+    },
+    {
+      "epoch": 0.9503365865527589,
+      "grad_norm": 25.095794677734375,
+      "learning_rate": 9.973753280839895e-07,
+      "loss": 1.5347,
+      "step": 2197
+    },
+    {
+      "epoch": 0.950769147584417,
+      "grad_norm": 24.920623779296875,
+      "learning_rate": 9.886264216972879e-07,
+      "loss": 1.4609,
+      "step": 2198
+    },
+    {
+      "epoch": 0.951201708616075,
+      "grad_norm": 24.81159019470215,
+      "learning_rate": 9.798775153105862e-07,
+      "loss": 1.4245,
+      "step": 2199
+    },
+    {
+      "epoch": 0.9516342696477331,
+      "grad_norm": 25.507144927978516,
+      "learning_rate": 9.711286089238846e-07,
+      "loss": 1.5234,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9520668306793911,
+      "grad_norm": 25.74333381652832,
+      "learning_rate": 9.62379702537183e-07,
+      "loss": 1.4334,
+      "step": 2201
+    },
+    {
+      "epoch": 0.9524993917110492,
+      "grad_norm": 24.9305477142334,
+      "learning_rate": 9.536307961504812e-07,
+      "loss": 1.4945,
+      "step": 2202
+    },
+    {
+      "epoch": 0.9529319527427073,
+      "grad_norm": 25.96180534362793,
+      "learning_rate": 9.448818897637796e-07,
+      "loss": 1.4937,
+      "step": 2203
+    },
+    {
+      "epoch": 0.9533645137743654,
+      "grad_norm": 27.482013702392578,
+      "learning_rate": 9.36132983377078e-07,
+      "loss": 1.5735,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9537970748060234,
+      "grad_norm": 27.310945510864258,
+      "learning_rate": 9.273840769903762e-07,
+      "loss": 1.6501,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9542296358376815,
+      "grad_norm": 23.459266662597656,
+      "learning_rate": 9.186351706036746e-07,
+      "loss": 1.4727,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9546621968693395,
+      "grad_norm": 23.028839111328125,
+      "learning_rate": 9.09886264216973e-07,
+      "loss": 1.5571,
+      "step": 2207
+    },
+    {
+      "epoch": 0.9550947579009976,
+      "grad_norm": 24.47319984436035,
+      "learning_rate": 9.011373578302713e-07,
+      "loss": 1.4652,
+      "step": 2208
+    },
+    {
+      "epoch": 0.9555273189326556,
+      "grad_norm": 27.27988052368164,
+      "learning_rate": 8.923884514435696e-07,
+      "loss": 1.6036,
+      "step": 2209
+    },
+    {
+      "epoch": 0.9559598799643138,
+      "grad_norm": 27.1883544921875,
+      "learning_rate": 8.83639545056868e-07,
+      "loss": 1.4815,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9563924409959718,
+      "grad_norm": 27.201946258544922,
+      "learning_rate": 8.748906386701662e-07,
+      "loss": 1.4541,
+      "step": 2211
+    },
+    {
+      "epoch": 0.9568250020276299,
+      "grad_norm": 27.259296417236328,
+      "learning_rate": 8.661417322834646e-07,
+      "loss": 1.5087,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9572575630592879,
+      "grad_norm": 25.546985626220703,
+      "learning_rate": 8.57392825896763e-07,
+      "loss": 1.5417,
+      "step": 2213
+    },
+    {
+      "epoch": 0.957690124090946,
+      "grad_norm": 26.166793823242188,
+      "learning_rate": 8.486439195100612e-07,
+      "loss": 1.4723,
+      "step": 2214
+    },
+    {
+      "epoch": 0.958122685122604,
+      "grad_norm": 26.597272872924805,
+      "learning_rate": 8.398950131233597e-07,
+      "loss": 1.5238,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9585552461542621,
+      "grad_norm": 24.605998992919922,
+      "learning_rate": 8.31146106736658e-07,
+      "loss": 1.5571,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9589878071859201,
+      "grad_norm": 24.375625610351562,
+      "learning_rate": 8.223972003499563e-07,
+      "loss": 1.4745,
+      "step": 2217
+    },
+    {
+      "epoch": 0.9594203682175781,
+      "grad_norm": 26.005847930908203,
+      "learning_rate": 8.136482939632546e-07,
+      "loss": 1.5447,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9598529292492363,
+      "grad_norm": 27.76963996887207,
+      "learning_rate": 8.04899387576553e-07,
+      "loss": 1.5089,
+      "step": 2219
+    },
+    {
+      "epoch": 0.9602854902808943,
+      "grad_norm": 25.300994873046875,
+      "learning_rate": 7.961504811898513e-07,
+      "loss": 1.5734,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9607180513125524,
+      "grad_norm": 26.90599822998047,
+      "learning_rate": 7.874015748031496e-07,
+      "loss": 1.5282,
+      "step": 2221
+    },
+    {
+      "epoch": 0.9611506123442104,
+      "grad_norm": 25.654613494873047,
+      "learning_rate": 7.786526684164481e-07,
+      "loss": 1.5522,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9615831733758685,
+      "grad_norm": 23.53044319152832,
+      "learning_rate": 7.699037620297463e-07,
+      "loss": 1.575,
+      "step": 2223
+    },
+    {
+      "epoch": 0.9620157344075265,
+      "grad_norm": 25.62672996520996,
+      "learning_rate": 7.611548556430447e-07,
+      "loss": 1.5661,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9624482954391846,
+      "grad_norm": 24.500530242919922,
+      "learning_rate": 7.524059492563431e-07,
+      "loss": 1.5151,
+      "step": 2225
+    },
+    {
+      "epoch": 0.9628808564708426,
+      "grad_norm": 25.327373504638672,
+      "learning_rate": 7.436570428696413e-07,
+      "loss": 1.5006,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9633134175025008,
+      "grad_norm": 28.519956588745117,
+      "learning_rate": 7.349081364829397e-07,
+      "loss": 1.5537,
+      "step": 2227
+    },
+    {
+      "epoch": 0.9637459785341588,
+      "grad_norm": 25.847875595092773,
+      "learning_rate": 7.26159230096238e-07,
+      "loss": 1.4917,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9641785395658169,
+      "grad_norm": 22.345125198364258,
+      "learning_rate": 7.174103237095363e-07,
+      "loss": 1.4354,
+      "step": 2229
+    },
+    {
+      "epoch": 0.9646111005974749,
+      "grad_norm": 24.792699813842773,
+      "learning_rate": 7.086614173228346e-07,
+      "loss": 1.4301,
+      "step": 2230
+    },
+    {
+      "epoch": 0.965043661629133,
+      "grad_norm": 25.965551376342773,
+      "learning_rate": 6.999125109361331e-07,
+      "loss": 1.5495,
+      "step": 2231
+    },
+    {
+      "epoch": 0.965476222660791,
+      "grad_norm": 26.518917083740234,
+      "learning_rate": 6.911636045494314e-07,
+      "loss": 1.5347,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9659087836924491,
+      "grad_norm": 22.920303344726562,
+      "learning_rate": 6.824146981627297e-07,
+      "loss": 1.4384,
+      "step": 2233
+    },
+    {
+      "epoch": 0.9663413447241072,
+      "grad_norm": 23.388620376586914,
+      "learning_rate": 6.736657917760281e-07,
+      "loss": 1.4741,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9667739057557653,
+      "grad_norm": 22.90789222717285,
+      "learning_rate": 6.649168853893263e-07,
+      "loss": 1.4389,
+      "step": 2235
+    },
+    {
+      "epoch": 0.9672064667874233,
+      "grad_norm": 27.07576560974121,
+      "learning_rate": 6.561679790026247e-07,
+      "loss": 1.5521,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9676390278190814,
+      "grad_norm": 25.001564025878906,
+      "learning_rate": 6.474190726159231e-07,
+      "loss": 1.5137,
+      "step": 2237
+    },
+    {
+      "epoch": 0.9680715888507394,
+      "grad_norm": 23.1292724609375,
+      "learning_rate": 6.386701662292213e-07,
+      "loss": 1.4152,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9685041498823975,
+      "grad_norm": 24.822486877441406,
+      "learning_rate": 6.299212598425198e-07,
+      "loss": 1.4223,
+      "step": 2239
+    },
+    {
+      "epoch": 0.9689367109140555,
+      "grad_norm": 23.44011116027832,
+      "learning_rate": 6.21172353455818e-07,
+      "loss": 1.5192,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9693692719457135,
+      "grad_norm": 25.917972564697266,
+      "learning_rate": 6.124234470691164e-07,
+      "loss": 1.5659,
+      "step": 2241
+    },
+    {
+      "epoch": 0.9698018329773717,
+      "grad_norm": 24.592409133911133,
+      "learning_rate": 6.036745406824148e-07,
+      "loss": 1.5472,
+      "step": 2242
+    },
+    {
+      "epoch": 0.9702343940090297,
+      "grad_norm": 25.710756301879883,
+      "learning_rate": 5.94925634295713e-07,
+      "loss": 1.4849,
+      "step": 2243
+    },
+    {
+      "epoch": 0.9706669550406878,
+      "grad_norm": 30.253639221191406,
+      "learning_rate": 5.861767279090115e-07,
+      "loss": 1.5312,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9710995160723458,
+      "grad_norm": 27.50519371032715,
+      "learning_rate": 5.774278215223097e-07,
+      "loss": 1.54,
+      "step": 2245
+    },
+    {
+      "epoch": 0.9715320771040039,
+      "grad_norm": 26.179685592651367,
+      "learning_rate": 5.686789151356081e-07,
+      "loss": 1.5439,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9719646381356619,
+      "grad_norm": 24.04256248474121,
+      "learning_rate": 5.599300087489064e-07,
+      "loss": 1.5165,
+      "step": 2247
+    },
+    {
+      "epoch": 0.97239719916732,
+      "grad_norm": 25.094362258911133,
+      "learning_rate": 5.511811023622048e-07,
+      "loss": 1.5274,
+      "step": 2248
+    },
+    {
+      "epoch": 0.972829760198978,
+      "grad_norm": 22.677230834960938,
+      "learning_rate": 5.424321959755031e-07,
+      "loss": 1.4912,
+      "step": 2249
+    },
+    {
+      "epoch": 0.9732623212306362,
+      "grad_norm": 27.18683624267578,
+      "learning_rate": 5.336832895888014e-07,
+      "loss": 1.4782,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9736948822622942,
+      "grad_norm": 25.23105239868164,
+      "learning_rate": 5.249343832020998e-07,
+      "loss": 1.5043,
+      "step": 2251
+    },
+    {
+      "epoch": 0.9741274432939523,
+      "grad_norm": 24.919349670410156,
+      "learning_rate": 5.161854768153981e-07,
+      "loss": 1.5297,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9745600043256103,
+      "grad_norm": 26.34563636779785,
+      "learning_rate": 5.074365704286965e-07,
+      "loss": 1.4319,
+      "step": 2253
+    },
+    {
+      "epoch": 0.9749925653572684,
+      "grad_norm": 24.52495002746582,
+      "learning_rate": 4.986876640419948e-07,
+      "loss": 1.6465,
+      "step": 2254
+    },
+    {
+      "epoch": 0.9754251263889264,
+      "grad_norm": 24.07614517211914,
+      "learning_rate": 4.899387576552931e-07,
+      "loss": 1.6281,
+      "step": 2255
+    },
+    {
+      "epoch": 0.9758576874205845,
+      "grad_norm": 26.345617294311523,
+      "learning_rate": 4.811898512685915e-07,
+      "loss": 1.5466,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9762902484522425,
+      "grad_norm": 27.2408390045166,
+      "learning_rate": 4.724409448818898e-07,
+      "loss": 1.5557,
+      "step": 2257
+    },
+    {
+      "epoch": 0.9767228094839007,
+      "grad_norm": 23.67082977294922,
+      "learning_rate": 4.636920384951881e-07,
+      "loss": 1.5431,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9771553705155587,
+      "grad_norm": 23.821247100830078,
+      "learning_rate": 4.549431321084865e-07,
+      "loss": 1.5224,
+      "step": 2259
+    },
+    {
+      "epoch": 0.9775879315472168,
+      "grad_norm": 24.50645637512207,
+      "learning_rate": 4.461942257217848e-07,
+      "loss": 1.4193,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9780204925788748,
+      "grad_norm": 24.018909454345703,
+      "learning_rate": 4.374453193350831e-07,
+      "loss": 1.5108,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9784530536105328,
+      "grad_norm": 24.982769012451172,
+      "learning_rate": 4.286964129483815e-07,
+      "loss": 1.4694,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9788856146421909,
+      "grad_norm": 26.493566513061523,
+      "learning_rate": 4.1994750656167983e-07,
+      "loss": 1.4441,
+      "step": 2263
+    },
+    {
+      "epoch": 0.9793181756738489,
+      "grad_norm": 22.410762786865234,
+      "learning_rate": 4.1119860017497814e-07,
+      "loss": 1.5065,
+      "step": 2264
+    },
+    {
+      "epoch": 0.979750736705507,
+      "grad_norm": 24.40163230895996,
+      "learning_rate": 4.024496937882765e-07,
+      "loss": 1.5389,
+      "step": 2265
+    },
+    {
+      "epoch": 0.9801832977371651,
+      "grad_norm": 26.319671630859375,
+      "learning_rate": 3.937007874015748e-07,
+      "loss": 1.4334,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9806158587688232,
+      "grad_norm": 22.440717697143555,
+      "learning_rate": 3.8495188101487317e-07,
+      "loss": 1.5234,
+      "step": 2267
+    },
+    {
+      "epoch": 0.9810484198004812,
+      "grad_norm": 23.569488525390625,
+      "learning_rate": 3.7620297462817153e-07,
+      "loss": 1.4278,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9814809808321393,
+      "grad_norm": 24.376298904418945,
+      "learning_rate": 3.6745406824146983e-07,
+      "loss": 1.5538,
+      "step": 2269
+    },
+    {
+      "epoch": 0.9819135418637973,
+      "grad_norm": 25.128238677978516,
+      "learning_rate": 3.5870516185476814e-07,
+      "loss": 1.5044,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9823461028954554,
+      "grad_norm": 24.355140686035156,
+      "learning_rate": 3.4995625546806655e-07,
+      "loss": 1.4869,
+      "step": 2271
+    },
+    {
+      "epoch": 0.9827786639271134,
+      "grad_norm": 26.226696014404297,
+      "learning_rate": 3.4120734908136486e-07,
+      "loss": 1.4047,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9832112249587716,
+      "grad_norm": 24.68053436279297,
+      "learning_rate": 3.3245844269466317e-07,
+      "loss": 1.4378,
+      "step": 2273
+    },
+    {
+      "epoch": 0.9836437859904296,
+      "grad_norm": 23.87020492553711,
+      "learning_rate": 3.2370953630796153e-07,
+      "loss": 1.4567,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9840763470220877,
+      "grad_norm": 26.111167907714844,
+      "learning_rate": 3.149606299212599e-07,
+      "loss": 1.4861,
+      "step": 2275
+    },
+    {
+      "epoch": 0.9845089080537457,
+      "grad_norm": 24.464523315429688,
+      "learning_rate": 3.062117235345582e-07,
+      "loss": 1.4386,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9849414690854038,
+      "grad_norm": 25.67591094970703,
+      "learning_rate": 2.974628171478565e-07,
+      "loss": 1.4829,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9853740301170618,
+      "grad_norm": 25.558279037475586,
+      "learning_rate": 2.8871391076115486e-07,
+      "loss": 1.487,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9858065911487199,
+      "grad_norm": 27.101078033447266,
+      "learning_rate": 2.799650043744532e-07,
+      "loss": 1.5293,
+      "step": 2279
+    },
+    {
+      "epoch": 0.9862391521803779,
+      "grad_norm": 23.139122009277344,
+      "learning_rate": 2.7121609798775153e-07,
+      "loss": 1.491,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9866717132120361,
+      "grad_norm": 24.572038650512695,
+      "learning_rate": 2.624671916010499e-07,
+      "loss": 1.4598,
+      "step": 2281
+    },
+    {
+      "epoch": 0.9871042742436941,
+      "grad_norm": 24.457639694213867,
+      "learning_rate": 2.5371828521434825e-07,
+      "loss": 1.5327,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9875368352753522,
+      "grad_norm": 26.919748306274414,
+      "learning_rate": 2.4496937882764656e-07,
+      "loss": 1.5191,
+      "step": 2283
+    },
+    {
+      "epoch": 0.9879693963070102,
+      "grad_norm": 24.509489059448242,
+      "learning_rate": 2.362204724409449e-07,
+      "loss": 1.5107,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9884019573386682,
+      "grad_norm": 24.133081436157227,
+      "learning_rate": 2.2747156605424325e-07,
+      "loss": 1.4597,
+      "step": 2285
+    },
+    {
+      "epoch": 0.9888345183703263,
+      "grad_norm": 27.456026077270508,
+      "learning_rate": 2.1872265966754156e-07,
+      "loss": 1.5723,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9892670794019843,
+      "grad_norm": 25.23147964477539,
+      "learning_rate": 2.0997375328083992e-07,
+      "loss": 1.472,
+      "step": 2287
+    },
+    {
+      "epoch": 0.9896996404336424,
+      "grad_norm": 25.515771865844727,
+      "learning_rate": 2.0122484689413825e-07,
+      "loss": 1.5353,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9901322014653005,
+      "grad_norm": 25.060243606567383,
+      "learning_rate": 1.9247594050743658e-07,
+      "loss": 1.4872,
+      "step": 2289
+    },
+    {
+      "epoch": 0.9905647624969586,
+      "grad_norm": 25.204627990722656,
+      "learning_rate": 1.8372703412073492e-07,
+      "loss": 1.5119,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9909973235286166,
+      "grad_norm": 26.539810180664062,
+      "learning_rate": 1.7497812773403328e-07,
+      "loss": 1.5112,
+      "step": 2291
+    },
+    {
+      "epoch": 0.9914298845602747,
+      "grad_norm": 25.467899322509766,
+      "learning_rate": 1.6622922134733158e-07,
+      "loss": 1.4897,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9918624455919327,
+      "grad_norm": 26.030517578125,
+      "learning_rate": 1.5748031496062994e-07,
+      "loss": 1.5333,
+      "step": 2293
+    },
+    {
+      "epoch": 0.9922950066235908,
+      "grad_norm": 26.485551834106445,
+      "learning_rate": 1.4873140857392825e-07,
+      "loss": 1.5446,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9927275676552488,
+      "grad_norm": 26.391727447509766,
+      "learning_rate": 1.399825021872266e-07,
+      "loss": 1.4711,
+      "step": 2295
+    },
+    {
+      "epoch": 0.993160128686907,
+      "grad_norm": 23.88796615600586,
+      "learning_rate": 1.3123359580052494e-07,
+      "loss": 1.4731,
+      "step": 2296
+    },
+    {
+      "epoch": 0.993592689718565,
+      "grad_norm": 24.405637741088867,
+      "learning_rate": 1.2248468941382328e-07,
+      "loss": 1.5611,
+      "step": 2297
+    },
+    {
+      "epoch": 0.9940252507502231,
+      "grad_norm": 26.79961585998535,
+      "learning_rate": 1.1373578302712162e-07,
+      "loss": 1.547,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9944578117818811,
+      "grad_norm": 27.189603805541992,
+      "learning_rate": 1.0498687664041996e-07,
+      "loss": 1.5443,
+      "step": 2299
+    },
+    {
+      "epoch": 0.9948903728135392,
+      "grad_norm": 27.315834045410156,
+      "learning_rate": 9.623797025371829e-08,
+      "loss": 1.3623,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9953229338451972,
+      "grad_norm": 27.901859283447266,
+      "learning_rate": 8.748906386701664e-08,
+      "loss": 1.6198,
+      "step": 2301
+    },
+    {
+      "epoch": 0.9957554948768553,
+      "grad_norm": 26.05302619934082,
+      "learning_rate": 7.874015748031497e-08,
+      "loss": 1.4631,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9961880559085133,
+      "grad_norm": 25.083433151245117,
+      "learning_rate": 6.99912510936133e-08,
+      "loss": 1.5496,
+      "step": 2303
+    },
+    {
+      "epoch": 0.9966206169401715,
+      "grad_norm": 25.853899002075195,
+      "learning_rate": 6.124234470691164e-08,
+      "loss": 1.5728,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9970531779718295,
+      "grad_norm": 24.412639617919922,
+      "learning_rate": 5.249343832020998e-08,
+      "loss": 1.5394,
+      "step": 2305
+    },
+    {
+      "epoch": 0.9974857390034875,
+      "grad_norm": 26.720706939697266,
+      "learning_rate": 4.374453193350832e-08,
+      "loss": 1.4967,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9979183000351456,
+      "grad_norm": 27.0684814453125,
+      "learning_rate": 3.499562554680665e-08,
+      "loss": 1.5429,
+      "step": 2307
+    },
+    {
+      "epoch": 0.9983508610668036,
+      "grad_norm": 24.51413917541504,
+      "learning_rate": 2.624671916010499e-08,
+      "loss": 1.5263,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9987834220984617,
+      "grad_norm": 27.829641342163086,
+      "learning_rate": 1.7497812773403326e-08,
+      "loss": 1.492,
+      "step": 2309
+    },
+    {
+      "epoch": 0.9992159831301197,
+      "grad_norm": 24.01323127746582,
+      "learning_rate": 8.748906386701663e-09,
+      "loss": 1.5018,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9996485441617778,
+      "grad_norm": 24.6170597076416,
+      "learning_rate": 0.0,
+      "loss": 1.533,
+      "step": 2311
     }
   ],
   "logging_steps": 1,
@@ -7021,12 +16198,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.2407280868764634e+17,
+  "total_flos": 2.868910410465792e+17,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null