diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,67370 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 9620,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 37.1042463148092,
+      "learning_rate": 3.4602076124567476e-08,
+      "loss": 3.1037,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.207250934654958,
+      "learning_rate": 6.920415224913495e-08,
+      "loss": 2.5937,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.4866252238902,
+      "learning_rate": 1.0380622837370243e-07,
+      "loss": 3.1256,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 37.287480505182415,
+      "learning_rate": 1.384083044982699e-07,
+      "loss": 3.208,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.298895147565904,
+      "learning_rate": 1.730103806228374e-07,
+      "loss": 3.1655,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.5099030032688,
+      "learning_rate": 2.0761245674740486e-07,
+      "loss": 2.9876,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.91252751714465,
+      "learning_rate": 2.422145328719723e-07,
+      "loss": 3.0633,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.07859194615419,
+      "learning_rate": 2.768166089965398e-07,
+      "loss": 3.0316,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.74310170641338,
+      "learning_rate": 3.114186851211073e-07,
+      "loss": 2.5585,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.18649162666028,
+      "learning_rate": 3.460207612456748e-07,
+      "loss": 2.9646,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.07862998050571,
+      "learning_rate": 3.8062283737024223e-07,
+      "loss": 2.9688,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.233677390485585,
+      "learning_rate": 4.152249134948097e-07,
+      "loss": 2.545,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.86301568731232,
+      "learning_rate": 4.498269896193772e-07,
+      "loss": 3.1589,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.297767276970973,
+      "learning_rate": 4.844290657439446e-07,
+      "loss": 2.9288,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.327661562085037,
+      "learning_rate": 5.190311418685121e-07,
+      "loss": 3.187,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.09228220848974,
+      "learning_rate": 5.536332179930796e-07,
+      "loss": 3.2806,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.693100834543692,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 2.8149,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.493063660463303,
+      "learning_rate": 6.228373702422146e-07,
+      "loss": 2.6128,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.455923474790353,
+      "learning_rate": 6.57439446366782e-07,
+      "loss": 2.473,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.804700164829907,
+      "learning_rate": 6.920415224913496e-07,
+      "loss": 2.8412,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.638497790578334,
+      "learning_rate": 7.266435986159171e-07,
+      "loss": 2.5866,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.04505500067805,
+      "learning_rate": 7.612456747404845e-07,
+      "loss": 2.3629,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.49490636193544,
+      "learning_rate": 7.958477508650519e-07,
+      "loss": 2.0194,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.347203073629032,
+      "learning_rate": 8.304498269896194e-07,
+      "loss": 1.9241,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.464360513749167,
+      "learning_rate": 8.650519031141868e-07,
+      "loss": 1.9576,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.683836407692967,
+      "learning_rate": 8.996539792387544e-07,
+      "loss": 1.8351,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.676400385932228,
+      "learning_rate": 9.342560553633219e-07,
+      "loss": 1.8349,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.392235449843433,
+      "learning_rate": 9.688581314878893e-07,
+      "loss": 1.6501,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.527415581012864,
+      "learning_rate": 1.0034602076124569e-06,
+      "loss": 1.8315,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.694825750097737,
+      "learning_rate": 1.0380622837370243e-06,
+      "loss": 1.5652,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.763414643263694,
+      "learning_rate": 1.0726643598615919e-06,
+      "loss": 1.6854,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.33865629455925,
+      "learning_rate": 1.1072664359861592e-06,
+      "loss": 1.5039,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.763246152388538,
+      "learning_rate": 1.1418685121107268e-06,
+      "loss": 1.3142,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.263903811995936,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 1.2849,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.439565470509954,
+      "learning_rate": 1.2110726643598616e-06,
+      "loss": 1.1053,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.952049399538764,
+      "learning_rate": 1.2456747404844292e-06,
+      "loss": 1.1444,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.01097935846218,
+      "learning_rate": 1.2802768166089966e-06,
+      "loss": 1.143,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.871471103375363,
+      "learning_rate": 1.314878892733564e-06,
+      "loss": 1.1158,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.728062914617196,
+      "learning_rate": 1.3494809688581318e-06,
+      "loss": 1.1664,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.892186888010498,
+      "learning_rate": 1.3840830449826992e-06,
+      "loss": 1.0177,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.409090504197081,
+      "learning_rate": 1.4186851211072665e-06,
+      "loss": 0.971,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.601033155317786,
+      "learning_rate": 1.4532871972318341e-06,
+      "loss": 0.8399,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.431069512628087,
+      "learning_rate": 1.4878892733564015e-06,
+      "loss": 0.8619,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.253547266802608,
+      "learning_rate": 1.522491349480969e-06,
+      "loss": 0.8158,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.451235861424146,
+      "learning_rate": 1.5570934256055365e-06,
+      "loss": 0.8109,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3351939398703987,
+      "learning_rate": 1.5916955017301039e-06,
+      "loss": 0.7258,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.942797536420241,
+      "learning_rate": 1.6262975778546713e-06,
+      "loss": 0.6719,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7074094109335363,
+      "learning_rate": 1.6608996539792389e-06,
+      "loss": 0.6014,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9317496409320096,
+      "learning_rate": 1.6955017301038063e-06,
+      "loss": 0.6469,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5351542200072026,
+      "learning_rate": 1.7301038062283736e-06,
+      "loss": 0.8268,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.393778283674769,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.6724,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.274905557406526,
+      "learning_rate": 1.7993079584775088e-06,
+      "loss": 0.6593,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8743291218680316,
+      "learning_rate": 1.8339100346020764e-06,
+      "loss": 0.7329,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5017134187558137,
+      "learning_rate": 1.8685121107266438e-06,
+      "loss": 0.7654,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9875397548331608,
+      "learning_rate": 1.9031141868512112e-06,
+      "loss": 0.7153,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3596053776617127,
+      "learning_rate": 1.9377162629757786e-06,
+      "loss": 0.7149,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.166401176661029,
+      "learning_rate": 1.972318339100346e-06,
+      "loss": 0.6804,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0951820308221527,
+      "learning_rate": 2.0069204152249138e-06,
+      "loss": 0.6972,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4657081002289796,
+      "learning_rate": 2.041522491349481e-06,
+      "loss": 0.6038,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.760273057855939,
+      "learning_rate": 2.0761245674740485e-06,
+      "loss": 0.7049,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.111128237734746,
+      "learning_rate": 2.110726643598616e-06,
+      "loss": 0.7256,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1541359357914462,
+      "learning_rate": 2.1453287197231837e-06,
+      "loss": 0.7687,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.169472440876441,
+      "learning_rate": 2.1799307958477513e-06,
+      "loss": 0.6799,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.02142080617264,
+      "learning_rate": 2.2145328719723185e-06,
+      "loss": 0.5648,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4144204497805015,
+      "learning_rate": 2.249134948096886e-06,
+      "loss": 0.7278,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4257419352048184,
+      "learning_rate": 2.2837370242214537e-06,
+      "loss": 0.691,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3058060047031423,
+      "learning_rate": 2.318339100346021e-06,
+      "loss": 0.7196,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.963077781184842,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.5515,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.312593059971846,
+      "learning_rate": 2.387543252595156e-06,
+      "loss": 0.7084,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2385262578581604,
+      "learning_rate": 2.4221453287197232e-06,
+      "loss": 0.7047,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1074433866202082,
+      "learning_rate": 2.456747404844291e-06,
+      "loss": 0.6752,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1381474039570216,
+      "learning_rate": 2.4913494809688584e-06,
+      "loss": 0.7712,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0078823847937706,
+      "learning_rate": 2.5259515570934256e-06,
+      "loss": 0.655,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.28321088070362,
+      "learning_rate": 2.560553633217993e-06,
+      "loss": 0.7052,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.310153415920439,
+      "learning_rate": 2.5951557093425604e-06,
+      "loss": 0.5941,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.165070016989384,
+      "learning_rate": 2.629757785467128e-06,
+      "loss": 0.6693,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0797734502814063,
+      "learning_rate": 2.6643598615916955e-06,
+      "loss": 0.6087,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2813786170710113,
+      "learning_rate": 2.6989619377162636e-06,
+      "loss": 0.7298,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3968138762934,
+      "learning_rate": 2.7335640138408307e-06,
+      "loss": 0.769,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.130423945439428,
+      "learning_rate": 2.7681660899653983e-06,
+      "loss": 0.6529,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3282052971504674,
+      "learning_rate": 2.802768166089966e-06,
+      "loss": 0.7926,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.282966381673334,
+      "learning_rate": 2.837370242214533e-06,
+      "loss": 0.6593,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.160804388058384,
+      "learning_rate": 2.8719723183391007e-06,
+      "loss": 0.7173,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.072121784879253,
+      "learning_rate": 2.9065743944636683e-06,
+      "loss": 0.6375,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1103173238484403,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.5965,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3387135179333542,
+      "learning_rate": 2.975778546712803e-06,
+      "loss": 0.6975,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3469544991445765,
+      "learning_rate": 3.0103806228373707e-06,
+      "loss": 0.6548,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4499716727822958,
+      "learning_rate": 3.044982698961938e-06,
+      "loss": 0.7056,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.395912867646452,
+      "learning_rate": 3.0795847750865054e-06,
+      "loss": 0.6031,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.124375425140296,
+      "learning_rate": 3.114186851211073e-06,
+      "loss": 0.6674,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4152158278140674,
+      "learning_rate": 3.14878892733564e-06,
+      "loss": 0.6561,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.645106503355264,
+      "learning_rate": 3.1833910034602078e-06,
+      "loss": 0.6203,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.170849019708104,
+      "learning_rate": 3.2179930795847754e-06,
+      "loss": 0.5372,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.085614854916659,
+      "learning_rate": 3.2525951557093425e-06,
+      "loss": 0.6429,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1899737577611824,
+      "learning_rate": 3.28719723183391e-06,
+      "loss": 0.663,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0251727260233188,
+      "learning_rate": 3.3217993079584777e-06,
+      "loss": 0.6597,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3645519837714564,
+      "learning_rate": 3.356401384083045e-06,
+      "loss": 0.6406,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4034145704451593,
+      "learning_rate": 3.3910034602076125e-06,
+      "loss": 0.7707,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.322115990647384,
+      "learning_rate": 3.42560553633218e-06,
+      "loss": 0.7373,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.326687603326004,
+      "learning_rate": 3.4602076124567473e-06,
+      "loss": 0.584,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3114109061072936,
+      "learning_rate": 3.4948096885813153e-06,
+      "loss": 0.6832,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.195809015277355,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.609,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1867017908477737,
+      "learning_rate": 3.5640138408304505e-06,
+      "loss": 0.5833,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0914140112949577,
+      "learning_rate": 3.5986159169550177e-06,
+      "loss": 0.6508,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3076122744482297,
+      "learning_rate": 3.6332179930795853e-06,
+      "loss": 0.6108,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1765304437298867,
+      "learning_rate": 3.667820069204153e-06,
+      "loss": 0.6641,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6803492350052185,
+      "learning_rate": 3.70242214532872e-06,
+      "loss": 0.6669,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.199931218934597,
+      "learning_rate": 3.7370242214532876e-06,
+      "loss": 0.705,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9358294457907268,
+      "learning_rate": 3.7716262975778552e-06,
+      "loss": 0.733,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3313246015906355,
+      "learning_rate": 3.8062283737024224e-06,
+      "loss": 0.6512,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7685678589026836,
+      "learning_rate": 3.84083044982699e-06,
+      "loss": 0.6651,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.207787027694181,
+      "learning_rate": 3.875432525951557e-06,
+      "loss": 0.6167,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.343240858434686,
+      "learning_rate": 3.910034602076125e-06,
+      "loss": 0.7268,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0089194459476643,
+      "learning_rate": 3.944636678200692e-06,
+      "loss": 0.5841,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.012684446361227,
+      "learning_rate": 3.9792387543252595e-06,
+      "loss": 0.5901,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.396831390118752,
+      "learning_rate": 4.0138408304498275e-06,
+      "loss": 0.706,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2855393661408883,
+      "learning_rate": 4.048442906574395e-06,
+      "loss": 0.8205,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.015682700463381,
+      "learning_rate": 4.083044982698962e-06,
+      "loss": 0.5668,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.181291526395132,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.6465,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2248428690994806,
+      "learning_rate": 4.152249134948097e-06,
+      "loss": 0.5714,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.202853322518201,
+      "learning_rate": 4.186851211072664e-06,
+      "loss": 0.6137,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9078978560651283,
+      "learning_rate": 4.221453287197232e-06,
+      "loss": 0.5781,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.283481882986506,
+      "learning_rate": 4.2560553633218e-06,
+      "loss": 0.7298,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2045034916450135,
+      "learning_rate": 4.2906574394463675e-06,
+      "loss": 0.6633,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.064486144309271,
+      "learning_rate": 4.325259515570935e-06,
+      "loss": 0.6187,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.191299849944456,
+      "learning_rate": 4.359861591695503e-06,
+      "loss": 0.6644,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.256115550781429,
+      "learning_rate": 4.39446366782007e-06,
+      "loss": 0.7002,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4171539822945998,
+      "learning_rate": 4.429065743944637e-06,
+      "loss": 0.642,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9676280086993452,
+      "learning_rate": 4.463667820069205e-06,
+      "loss": 0.6169,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.315131249667166,
+      "learning_rate": 4.498269896193772e-06,
+      "loss": 0.6301,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.016187038955443,
+      "learning_rate": 4.532871972318339e-06,
+      "loss": 0.6258,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1690571819984803,
+      "learning_rate": 4.567474048442907e-06,
+      "loss": 0.6108,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.228343012890392,
+      "learning_rate": 4.6020761245674745e-06,
+      "loss": 0.7412,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3230648080604444,
+      "learning_rate": 4.636678200692042e-06,
+      "loss": 0.7101,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2702336427350787,
+      "learning_rate": 4.67128027681661e-06,
+      "loss": 0.6535,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9953352017886066,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.5654,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1335665240811483,
+      "learning_rate": 4.740484429065744e-06,
+      "loss": 0.7365,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0081551952295067,
+      "learning_rate": 4.775086505190312e-06,
+      "loss": 0.5654,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.608564793222911,
+      "learning_rate": 4.809688581314879e-06,
+      "loss": 0.6156,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2903018042810865,
+      "learning_rate": 4.8442906574394464e-06,
+      "loss": 0.5923,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.606584727809493,
+      "learning_rate": 4.8788927335640145e-06,
+      "loss": 0.6669,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2676586545153983,
+      "learning_rate": 4.913494809688582e-06,
+      "loss": 0.5231,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.08733797395526,
+      "learning_rate": 4.948096885813149e-06,
+      "loss": 0.6447,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.280353041745458,
+      "learning_rate": 4.982698961937717e-06,
+      "loss": 0.7165,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2993638112411756,
+      "learning_rate": 5.017301038062284e-06,
+      "loss": 0.6171,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1155138048920445,
+      "learning_rate": 5.051903114186851e-06,
+      "loss": 0.678,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.092000766414222,
+      "learning_rate": 5.086505190311419e-06,
+      "loss": 0.5561,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4110387919086103,
+      "learning_rate": 5.121107266435986e-06,
+      "loss": 0.6236,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4295417190634003,
+      "learning_rate": 5.155709342560554e-06,
+      "loss": 0.7297,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1821328342305524,
+      "learning_rate": 5.190311418685121e-06,
+      "loss": 0.8451,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9821707599109635,
+      "learning_rate": 5.224913494809689e-06,
+      "loss": 0.5909,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.935677871558701,
+      "learning_rate": 5.259515570934256e-06,
+      "loss": 0.6249,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2167343078418633,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.6286,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.177970005050448,
+      "learning_rate": 5.328719723183391e-06,
+      "loss": 0.631,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2333962515280295,
+      "learning_rate": 5.363321799307959e-06,
+      "loss": 0.6793,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1950449323297736,
+      "learning_rate": 5.397923875432527e-06,
+      "loss": 0.6113,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.172906333435821,
+      "learning_rate": 5.4325259515570934e-06,
+      "loss": 0.6419,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.219433419070003,
+      "learning_rate": 5.4671280276816615e-06,
+      "loss": 0.6086,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1272302168820723,
+      "learning_rate": 5.501730103806229e-06,
+      "loss": 0.6471,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2725105069940783,
+      "learning_rate": 5.536332179930797e-06,
+      "loss": 0.5728,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3320622478808186,
+      "learning_rate": 5.570934256055364e-06,
+      "loss": 0.6638,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4359930813276236,
+      "learning_rate": 5.605536332179932e-06,
+      "loss": 0.6696,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.225861962560492,
+      "learning_rate": 5.640138408304498e-06,
+      "loss": 0.6584,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2669038485616375,
+      "learning_rate": 5.674740484429066e-06,
+      "loss": 0.6774,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3265153532039795,
+      "learning_rate": 5.709342560553633e-06,
+      "loss": 0.6741,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.419471267005024,
+      "learning_rate": 5.743944636678201e-06,
+      "loss": 0.6057,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.276964154368608,
+      "learning_rate": 5.7785467128027686e-06,
+      "loss": 0.5884,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.410507009541092,
+      "learning_rate": 5.8131487889273366e-06,
+      "loss": 0.6878,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.083503032279742,
+      "learning_rate": 5.847750865051903e-06,
+      "loss": 0.6552,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2509896017989117,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.6079,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1859289065791256,
+      "learning_rate": 5.916955017301038e-06,
+      "loss": 0.621,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.147850941536798,
+      "learning_rate": 5.951557093425606e-06,
+      "loss": 0.5711,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1062480954749314,
+      "learning_rate": 5.986159169550173e-06,
+      "loss": 0.6168,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.025096030131394,
+      "learning_rate": 6.020761245674741e-06,
+      "loss": 0.6314,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1106251684861093,
+      "learning_rate": 6.055363321799308e-06,
+      "loss": 0.6165,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9386258712780828,
+      "learning_rate": 6.089965397923876e-06,
+      "loss": 0.5944,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.439704310973515,
+      "learning_rate": 6.124567474048443e-06,
+      "loss": 0.7301,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.193053790974262,
+      "learning_rate": 6.159169550173011e-06,
+      "loss": 0.5184,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1972037628573364,
+      "learning_rate": 6.193771626297579e-06,
+      "loss": 0.6819,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.108660851248478,
+      "learning_rate": 6.228373702422146e-06,
+      "loss": 0.6435,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.36479259661094,
+      "learning_rate": 6.262975778546714e-06,
+      "loss": 0.5749,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0278184423083347,
+      "learning_rate": 6.29757785467128e-06,
+      "loss": 0.5742,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3224464327459984,
+      "learning_rate": 6.332179930795848e-06,
+      "loss": 0.6336,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3133190199193168,
+      "learning_rate": 6.3667820069204156e-06,
+      "loss": 0.6914,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.188171505795143,
+      "learning_rate": 6.401384083044984e-06,
+      "loss": 0.7221,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.58564757756167,
+      "learning_rate": 6.435986159169551e-06,
+      "loss": 0.6711,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.122268004241992,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.6429,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9690551649287422,
+      "learning_rate": 6.505190311418685e-06,
+      "loss": 0.6549,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3770166505133887,
+      "learning_rate": 6.539792387543253e-06,
+      "loss": 0.7386,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.332845067178479,
+      "learning_rate": 6.57439446366782e-06,
+      "loss": 0.7431,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.292052198271124,
+      "learning_rate": 6.608996539792388e-06,
+      "loss": 0.569,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2147634719443836,
+      "learning_rate": 6.6435986159169555e-06,
+      "loss": 0.7045,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.328089977091099,
+      "learning_rate": 6.6782006920415235e-06,
+      "loss": 0.6364,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.274054083618043,
+      "learning_rate": 6.71280276816609e-06,
+      "loss": 0.6056,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3784158616015643,
+      "learning_rate": 6.747404844290658e-06,
+      "loss": 0.6709,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0892411869210346,
+      "learning_rate": 6.782006920415225e-06,
+      "loss": 0.562,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1091681111983496,
+      "learning_rate": 6.816608996539793e-06,
+      "loss": 0.5607,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8902408651238167,
+      "learning_rate": 6.85121107266436e-06,
+      "loss": 0.707,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.561780143261129,
+      "learning_rate": 6.885813148788928e-06,
+      "loss": 0.6728,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0776566394374365,
+      "learning_rate": 6.9204152249134946e-06,
+      "loss": 0.6455,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.215431897046199,
+      "learning_rate": 6.9550173010380626e-06,
+      "loss": 0.5443,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0873081020830857,
+      "learning_rate": 6.989619377162631e-06,
+      "loss": 0.6858,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1124011856789098,
+      "learning_rate": 7.024221453287198e-06,
+      "loss": 0.5993,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4320988871766493,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.6647,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.035002393310255,
+      "learning_rate": 7.093425605536333e-06,
+      "loss": 0.619,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0914719636443317,
+      "learning_rate": 7.128027681660901e-06,
+      "loss": 0.7546,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.008874075320383,
+      "learning_rate": 7.162629757785467e-06,
+      "loss": 0.6114,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.193799041028231,
+      "learning_rate": 7.197231833910035e-06,
+      "loss": 0.7129,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.197495559283392,
+      "learning_rate": 7.2318339100346025e-06,
+      "loss": 0.7395,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8616873688955218,
+      "learning_rate": 7.2664359861591705e-06,
+      "loss": 0.5019,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.432719510249827,
+      "learning_rate": 7.301038062283738e-06,
+      "loss": 0.7087,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3669190124323127,
+      "learning_rate": 7.335640138408306e-06,
+      "loss": 0.6296,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1872694291480785,
+      "learning_rate": 7.370242214532872e-06,
+      "loss": 0.6005,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7828744622515829,
+      "learning_rate": 7.40484429065744e-06,
+      "loss": 0.6305,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0040108138250723,
+      "learning_rate": 7.439446366782007e-06,
+      "loss": 0.6873,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0796961366745226,
+      "learning_rate": 7.474048442906575e-06,
+      "loss": 0.6407,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1967248607492937,
+      "learning_rate": 7.508650519031142e-06,
+      "loss": 0.6707,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9175115859089473,
+      "learning_rate": 7.5432525951557104e-06,
+      "loss": 0.6526,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0514016284982213,
+      "learning_rate": 7.577854671280277e-06,
+      "loss": 0.6123,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5117956633223484,
+      "learning_rate": 7.612456747404845e-06,
+      "loss": 0.6014,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2156322226471796,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.6571,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2926948007219115,
+      "learning_rate": 7.68166089965398e-06,
+      "loss": 0.774,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9721980908588659,
+      "learning_rate": 7.716262975778547e-06,
+      "loss": 0.6152,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.111098976593361,
+      "learning_rate": 7.750865051903114e-06,
+      "loss": 0.6419,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.300877354238802,
+      "learning_rate": 7.785467128027683e-06,
+      "loss": 0.7069,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2866412650431434,
+      "learning_rate": 7.82006920415225e-06,
+      "loss": 0.6085,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0848079985085226,
+      "learning_rate": 7.854671280276818e-06,
+      "loss": 0.5879,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.201530251445851,
+      "learning_rate": 7.889273356401385e-06,
+      "loss": 0.7731,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0570052426900394,
+      "learning_rate": 7.923875432525952e-06,
+      "loss": 0.5881,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1400020806991757,
+      "learning_rate": 7.958477508650519e-06,
+      "loss": 0.694,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1227819097860645,
+      "learning_rate": 7.993079584775088e-06,
+      "loss": 0.7255,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.137417177560603,
+      "learning_rate": 8.027681660899655e-06,
+      "loss": 0.5723,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1108437635031136,
+      "learning_rate": 8.062283737024222e-06,
+      "loss": 0.6994,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.124414466611215,
+      "learning_rate": 8.09688581314879e-06,
+      "loss": 0.6547,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0520237300562854,
+      "learning_rate": 8.131487889273357e-06,
+      "loss": 0.6525,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9714185147383696,
+      "learning_rate": 8.166089965397924e-06,
+      "loss": 0.6038,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.939078039152394,
+      "learning_rate": 8.200692041522493e-06,
+      "loss": 0.711,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0670237054998393,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.614,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.242190530112608,
+      "learning_rate": 8.269896193771627e-06,
+      "loss": 0.6014,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1485542120755152,
+      "learning_rate": 8.304498269896194e-06,
+      "loss": 0.6358,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8482103878471823,
+      "learning_rate": 8.339100346020761e-06,
+      "loss": 0.5711,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9737201478872424,
+      "learning_rate": 8.373702422145328e-06,
+      "loss": 0.6359,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4163559055070887,
+      "learning_rate": 8.408304498269897e-06,
+      "loss": 0.7363,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.244452576938016,
+      "learning_rate": 8.442906574394465e-06,
+      "loss": 0.6451,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.086004638946662,
+      "learning_rate": 8.477508650519032e-06,
+      "loss": 0.7345,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.128590628871988,
+      "learning_rate": 8.5121107266436e-06,
+      "loss": 0.7326,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.235562449068478,
+      "learning_rate": 8.546712802768166e-06,
+      "loss": 0.7656,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2509200985615063,
+      "learning_rate": 8.581314878892735e-06,
+      "loss": 0.6271,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.295331842170622,
+      "learning_rate": 8.615916955017302e-06,
+      "loss": 0.6678,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.391429221795401,
+      "learning_rate": 8.65051903114187e-06,
+      "loss": 0.7425,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9529324621840602,
+      "learning_rate": 8.685121107266436e-06,
+      "loss": 0.5745,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.246043588101345,
+      "learning_rate": 8.719723183391005e-06,
+      "loss": 0.6345,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4904984344511076,
+      "learning_rate": 8.75432525951557e-06,
+      "loss": 0.6939,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1642783247420363,
+      "learning_rate": 8.78892733564014e-06,
+      "loss": 0.6032,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1940931042243257,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.7277,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8948954672045033,
+      "learning_rate": 8.858131487889274e-06,
+      "loss": 0.6893,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4582031636820525,
+      "learning_rate": 8.892733564013841e-06,
+      "loss": 0.7363,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3058632907633654,
+      "learning_rate": 8.92733564013841e-06,
+      "loss": 0.7746,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8228869043171054,
+      "learning_rate": 8.961937716262975e-06,
+      "loss": 0.487,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.219005308745244,
+      "learning_rate": 8.996539792387544e-06,
+      "loss": 0.7147,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9983314588691001,
+      "learning_rate": 9.031141868512112e-06,
+      "loss": 0.6375,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5810389232454507,
+      "learning_rate": 9.065743944636679e-06,
+      "loss": 0.8198,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.319366870420018,
+      "learning_rate": 9.100346020761246e-06,
+      "loss": 0.7102,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2528281665453656,
+      "learning_rate": 9.134948096885815e-06,
+      "loss": 0.6178,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.268866438407876,
+      "learning_rate": 9.16955017301038e-06,
+      "loss": 0.649,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1190340028998707,
+      "learning_rate": 9.204152249134949e-06,
+      "loss": 0.6279,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0714620681045663,
+      "learning_rate": 9.238754325259516e-06,
+      "loss": 0.6938,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.325221307942943,
+      "learning_rate": 9.273356401384083e-06,
+      "loss": 0.5867,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2470942028995986,
+      "learning_rate": 9.307958477508652e-06,
+      "loss": 0.6852,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.007572350176665,
+      "learning_rate": 9.34256055363322e-06,
+      "loss": 0.5481,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0257482560883275,
+      "learning_rate": 9.377162629757787e-06,
+      "loss": 0.6386,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0485510426599878,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.7489,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9623446612734317,
+      "learning_rate": 9.446366782006921e-06,
+      "loss": 0.5764,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.127396095227505,
+      "learning_rate": 9.480968858131488e-06,
+      "loss": 0.6338,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.215498491322338,
+      "learning_rate": 9.515570934256057e-06,
+      "loss": 0.758,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.08023339836091,
+      "learning_rate": 9.550173010380624e-06,
+      "loss": 0.6621,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9378232010147092,
+      "learning_rate": 9.584775086505191e-06,
+      "loss": 0.6002,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1412793080715358,
+      "learning_rate": 9.619377162629759e-06,
+      "loss": 0.6288,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.095326824165883,
+      "learning_rate": 9.653979238754326e-06,
+      "loss": 0.6168,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.093548208464301,
+      "learning_rate": 9.688581314878893e-06,
+      "loss": 0.6185,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0683217073155142,
+      "learning_rate": 9.723183391003462e-06,
+      "loss": 0.6579,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2050921361510256,
+      "learning_rate": 9.757785467128029e-06,
+      "loss": 0.6564,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0666754378724512,
+      "learning_rate": 9.792387543252596e-06,
+      "loss": 0.6938,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0255579239228676,
+      "learning_rate": 9.826989619377163e-06,
+      "loss": 0.5832,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2059232192312477,
+      "learning_rate": 9.86159169550173e-06,
+      "loss": 0.6408,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.227961737598211,
+      "learning_rate": 9.896193771626298e-06,
+      "loss": 0.6834,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.384395993502413,
+      "learning_rate": 9.930795847750866e-06,
+      "loss": 0.6507,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2259501520536786,
+      "learning_rate": 9.965397923875434e-06,
+      "loss": 0.8218,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2153438165060004,
+      "learning_rate": 1e-05,
+      "loss": 0.7089,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.090933521636028,
+      "learning_rate": 9.99999971661075e-06,
+      "loss": 0.6669,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2188852773516707,
+      "learning_rate": 9.999998866443035e-06,
+      "loss": 0.8329,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9643905919582283,
+      "learning_rate": 9.999997449496947e-06,
+      "loss": 0.604,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7302965204706007,
+      "learning_rate": 9.999995465772652e-06,
+      "loss": 0.5917,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.836246496796778,
+      "learning_rate": 9.99999291527037e-06,
+      "loss": 0.5272,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2820920836002587,
+      "learning_rate": 9.999989797990391e-06,
+      "loss": 0.7376,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2425304223811446,
+      "learning_rate": 9.999986113933071e-06,
+      "loss": 0.6333,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.021448989644418,
+      "learning_rate": 9.999981863098825e-06,
+      "loss": 0.5886,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1689623016635897,
+      "learning_rate": 9.999977045488135e-06,
+      "loss": 0.6014,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3336741309421347,
+      "learning_rate": 9.999971661101551e-06,
+      "loss": 0.7371,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1629679103619783,
+      "learning_rate": 9.999965709939679e-06,
+      "loss": 0.6577,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8935808966516685,
+      "learning_rate": 9.999959192003194e-06,
+      "loss": 0.5906,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1444391808967738,
+      "learning_rate": 9.999952107292836e-06,
+      "loss": 0.7357,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.022296007147585,
+      "learning_rate": 9.999944455809408e-06,
+      "loss": 0.574,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.063507571811703,
+      "learning_rate": 9.999936237553777e-06,
+      "loss": 0.6103,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1313448786683913,
+      "learning_rate": 9.999927452526877e-06,
+      "loss": 0.6104,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.112051884258629,
+      "learning_rate": 9.999918100729698e-06,
+      "loss": 0.657,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2604850689998757,
+      "learning_rate": 9.999908182163306e-06,
+      "loss": 0.7052,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1654923977851617,
+      "learning_rate": 9.999897696828822e-06,
+      "loss": 0.6801,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.996921654859269,
+      "learning_rate": 9.999886644727436e-06,
+      "loss": 0.6856,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.145453425352305,
+      "learning_rate": 9.999875025860401e-06,
+      "loss": 0.6513,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.234415993232513,
+      "learning_rate": 9.999862840229033e-06,
+      "loss": 0.6971,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0645659727143943,
+      "learning_rate": 9.999850087834715e-06,
+      "loss": 0.7306,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8744568621278508,
+      "learning_rate": 9.99983676867889e-06,
+      "loss": 0.638,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3222295206374604,
+      "learning_rate": 9.999822882763068e-06,
+      "loss": 0.6553,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.099484652851686,
+      "learning_rate": 9.999808430088826e-06,
+      "loss": 0.7374,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.228145088882417,
+      "learning_rate": 9.999793410657802e-06,
+      "loss": 0.8169,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1291708934516986,
+      "learning_rate": 9.999777824471694e-06,
+      "loss": 0.631,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.080530025890949,
+      "learning_rate": 9.999761671532273e-06,
+      "loss": 0.6407,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.119009430136004,
+      "learning_rate": 9.999744951841367e-06,
+      "loss": 0.6849,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8955703890753102,
+      "learning_rate": 9.999727665400876e-06,
+      "loss": 0.7005,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1038990841547496,
+      "learning_rate": 9.999709812212756e-06,
+      "loss": 0.6685,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2589247469878213,
+      "learning_rate": 9.99969139227903e-06,
+      "loss": 0.7754,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1825354390943943,
+      "learning_rate": 9.99967240560179e-06,
+      "loss": 0.7121,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.916129116293965,
+      "learning_rate": 9.999652852183184e-06,
+      "loss": 0.5697,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0643197115454033,
+      "learning_rate": 9.999632732025428e-06,
+      "loss": 0.6152,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0612939818120957,
+      "learning_rate": 9.99961204513081e-06,
+      "loss": 0.6136,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.216817383090462,
+      "learning_rate": 9.999590791501665e-06,
+      "loss": 0.5763,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0495658313065555,
+      "learning_rate": 9.999568971140409e-06,
+      "loss": 0.6226,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0958367292073277,
+      "learning_rate": 9.999546584049513e-06,
+      "loss": 0.7229,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0926501956699592,
+      "learning_rate": 9.999523630231516e-06,
+      "loss": 0.7376,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9068405784141171,
+      "learning_rate": 9.999500109689018e-06,
+      "loss": 0.5987,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0006081400254963,
+      "learning_rate": 9.999476022424688e-06,
+      "loss": 0.6085,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9085783103448943,
+      "learning_rate": 9.999451368441254e-06,
+      "loss": 0.564,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1115680818865235,
+      "learning_rate": 9.999426147741512e-06,
+      "loss": 0.6871,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1237154103762186,
+      "learning_rate": 9.999400360328318e-06,
+      "loss": 0.6901,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.929206892415265,
+      "learning_rate": 9.999374006204601e-06,
+      "loss": 0.6018,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0204061701835094,
+      "learning_rate": 9.999347085373343e-06,
+      "loss": 0.645,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9226738874188818,
+      "learning_rate": 9.999319597837599e-06,
+      "loss": 0.6821,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.107623603422912,
+      "learning_rate": 9.99929154360048e-06,
+      "loss": 0.6378,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8743119821525225,
+      "learning_rate": 9.999262922665172e-06,
+      "loss": 0.59,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8550368642600081,
+      "learning_rate": 9.999233735034916e-06,
+      "loss": 0.6298,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.079258305023673,
+      "learning_rate": 9.999203980713023e-06,
+      "loss": 0.7754,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.16644850862517,
+      "learning_rate": 9.999173659702863e-06,
+      "loss": 0.5724,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7680175701749203,
+      "learning_rate": 9.999142772007875e-06,
+      "loss": 0.6038,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9697469124412608,
+      "learning_rate": 9.999111317631559e-06,
+      "loss": 0.6786,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0214076901251707,
+      "learning_rate": 9.999079296577482e-06,
+      "loss": 0.7071,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1220150741444526,
+      "learning_rate": 9.99904670884927e-06,
+      "loss": 0.7243,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.023388262077159,
+      "learning_rate": 9.999013554450624e-06,
+      "loss": 0.8155,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.047321068044065,
+      "learning_rate": 9.998979833385296e-06,
+      "loss": 0.6606,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.090748901325982,
+      "learning_rate": 9.998945545657113e-06,
+      "loss": 0.6928,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.052795459098353,
+      "learning_rate": 9.998910691269957e-06,
+      "loss": 0.8328,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.402064265656104,
+      "learning_rate": 9.998875270227781e-06,
+      "loss": 0.672,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1481436238474636,
+      "learning_rate": 9.998839282534602e-06,
+      "loss": 0.6665,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2395539693277526,
+      "learning_rate": 9.998802728194496e-06,
+      "loss": 0.7083,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9776107871723176,
+      "learning_rate": 9.998765607211612e-06,
+      "loss": 0.6582,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.147090458916406,
+      "learning_rate": 9.99872791959015e-06,
+      "loss": 0.6026,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.034674606820822,
+      "learning_rate": 9.99868966533439e-06,
+      "loss": 0.6911,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0951350122740298,
+      "learning_rate": 9.998650844448663e-06,
+      "loss": 0.6416,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.089497478746895,
+      "learning_rate": 9.998611456937373e-06,
+      "loss": 0.6531,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0664793585495898,
+      "learning_rate": 9.998571502804982e-06,
+      "loss": 0.6725,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8814458336505542,
+      "learning_rate": 9.998530982056021e-06,
+      "loss": 0.6481,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.774755021585096,
+      "learning_rate": 9.99848989469508e-06,
+      "loss": 0.6176,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1620141766232686,
+      "learning_rate": 9.998448240726822e-06,
+      "loss": 0.6294,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.288599249402456,
+      "learning_rate": 9.998406020155964e-06,
+      "loss": 0.6562,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9908199779664653,
+      "learning_rate": 9.998363232987294e-06,
+      "loss": 0.687,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.139573273038414,
+      "learning_rate": 9.998319879225662e-06,
+      "loss": 0.591,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.951025573732339,
+      "learning_rate": 9.998275958875983e-06,
+      "loss": 0.7088,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.185154704796897,
+      "learning_rate": 9.998231471943234e-06,
+      "loss": 0.7013,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.093786252997598,
+      "learning_rate": 9.998186418432459e-06,
+      "loss": 0.7835,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2585179169164817,
+      "learning_rate": 9.998140798348766e-06,
+      "loss": 0.7529,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.086435056903612,
+      "learning_rate": 9.998094611697322e-06,
+      "loss": 0.783,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.076639780170476,
+      "learning_rate": 9.998047858483369e-06,
+      "loss": 0.7017,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1634960145020017,
+      "learning_rate": 9.9980005387122e-06,
+      "loss": 0.7516,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.359129077276345,
+      "learning_rate": 9.997952652389184e-06,
+      "loss": 0.7177,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.69710045388187,
+      "learning_rate": 9.997904199519748e-06,
+      "loss": 0.7553,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.932081602102441,
+      "learning_rate": 9.997855180109383e-06,
+      "loss": 0.7331,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.155197890696538,
+      "learning_rate": 9.997805594163646e-06,
+      "loss": 0.7416,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9314889268578335,
+      "learning_rate": 9.997755441688159e-06,
+      "loss": 0.7268,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0768241365813447,
+      "learning_rate": 9.997704722688607e-06,
+      "loss": 0.696,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1302481157387874,
+      "learning_rate": 9.997653437170739e-06,
+      "loss": 0.8132,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8926932428974472,
+      "learning_rate": 9.997601585140367e-06,
+      "loss": 0.5856,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1087789610541,
+      "learning_rate": 9.99754916660337e-06,
+      "loss": 0.7283,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0973004274036833,
+      "learning_rate": 9.997496181565691e-06,
+      "loss": 0.615,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9798386111356991,
+      "learning_rate": 9.997442630033333e-06,
+      "loss": 0.6571,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9119680312757281,
+      "learning_rate": 9.997388512012371e-06,
+      "loss": 0.6829,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.132931602798204,
+      "learning_rate": 9.997333827508936e-06,
+      "loss": 0.8136,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.926068179217276,
+      "learning_rate": 9.997278576529228e-06,
+      "loss": 0.6462,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8398296773499356,
+      "learning_rate": 9.997222759079507e-06,
+      "loss": 0.6632,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2236470134644355,
+      "learning_rate": 9.997166375166107e-06,
+      "loss": 0.7079,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3302822862993993,
+      "learning_rate": 9.997109424795415e-06,
+      "loss": 0.6415,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0538248803181034,
+      "learning_rate": 9.997051907973885e-06,
+      "loss": 0.6829,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.039125915190051,
+      "learning_rate": 9.99699382470804e-06,
+      "loss": 0.6388,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9562271717913728,
+      "learning_rate": 9.996935175004462e-06,
+      "loss": 0.5704,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4363968316295357,
+      "learning_rate": 9.996875958869803e-06,
+      "loss": 0.5985,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8544403613485045,
+      "learning_rate": 9.996816176310771e-06,
+      "loss": 0.5727,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8820693987504735,
+      "learning_rate": 9.996755827334145e-06,
+      "loss": 0.5675,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.008444476428537,
+      "learning_rate": 9.996694911946765e-06,
+      "loss": 0.8005,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3530880181617486,
+      "learning_rate": 9.996633430155537e-06,
+      "loss": 0.7435,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.325936114579622,
+      "learning_rate": 9.99657138196743e-06,
+      "loss": 0.6764,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.780809596195899,
+      "learning_rate": 9.996508767389477e-06,
+      "loss": 0.7005,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.915152329293021,
+      "learning_rate": 9.996445586428776e-06,
+      "loss": 0.6623,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.956582503534672,
+      "learning_rate": 9.996381839092489e-06,
+      "loss": 0.7644,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.432085078434551,
+      "learning_rate": 9.996317525387841e-06,
+      "loss": 0.7633,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2033123567972406,
+      "learning_rate": 9.996252645322124e-06,
+      "loss": 0.6479,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8373000696608504,
+      "learning_rate": 9.996187198902693e-06,
+      "loss": 0.605,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.115896017831422,
+      "learning_rate": 9.996121186136964e-06,
+      "loss": 0.7126,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9436621726914547,
+      "learning_rate": 9.996054607032425e-06,
+      "loss": 0.5966,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2683255229608346,
+      "learning_rate": 9.995987461596617e-06,
+      "loss": 0.7094,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1334958544798432,
+      "learning_rate": 9.995919749837154e-06,
+      "loss": 0.7025,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3033075412526345,
+      "learning_rate": 9.995851471761711e-06,
+      "loss": 0.7209,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.069310070603776,
+      "learning_rate": 9.99578262737803e-06,
+      "loss": 0.7111,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7124534446102473,
+      "learning_rate": 9.995713216693913e-06,
+      "loss": 0.7023,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.116142709972029,
+      "learning_rate": 9.995643239717228e-06,
+      "loss": 0.6724,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.04036069618067,
+      "learning_rate": 9.995572696455907e-06,
+      "loss": 0.7658,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.080715602390722,
+      "learning_rate": 9.995501586917949e-06,
+      "loss": 0.67,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.104421238433803,
+      "learning_rate": 9.99542991111141e-06,
+      "loss": 0.6032,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2301607740265124,
+      "learning_rate": 9.995357669044418e-06,
+      "loss": 0.616,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0462075925593584,
+      "learning_rate": 9.995284860725162e-06,
+      "loss": 0.6749,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.100344727358706,
+      "learning_rate": 9.995211486161896e-06,
+      "loss": 0.6815,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3266601471436794,
+      "learning_rate": 9.995137545362937e-06,
+      "loss": 0.8017,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9367072875868272,
+      "learning_rate": 9.995063038336663e-06,
+      "loss": 0.6028,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2292281685756214,
+      "learning_rate": 9.994987965091525e-06,
+      "loss": 0.6752,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.129402276054949,
+      "learning_rate": 9.994912325636029e-06,
+      "loss": 0.6371,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.070140533823007,
+      "learning_rate": 9.99483611997875e-06,
+      "loss": 0.8579,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.418866897472175,
+      "learning_rate": 9.994759348128331e-06,
+      "loss": 0.6827,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8482776838557704,
+      "learning_rate": 9.994682010093468e-06,
+      "loss": 0.6993,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2212930148340524,
+      "learning_rate": 9.99460410588293e-06,
+      "loss": 0.6308,
+      "step": 427
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.872863356848253,
+      "learning_rate": 9.99452563550555e-06,
+      "loss": 0.6843,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1550669400337763,
+      "learning_rate": 9.99444659897022e-06,
+      "loss": 0.6675,
+      "step": 429
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3214088233689716,
+      "learning_rate": 9.994366996285903e-06,
+      "loss": 0.5922,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9772960079757542,
+      "learning_rate": 9.994286827461616e-06,
+      "loss": 0.5801,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0146780220081135,
+      "learning_rate": 9.994206092506455e-06,
+      "loss": 0.6058,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.985534965943388,
+      "learning_rate": 9.994124791429565e-06,
+      "loss": 0.6692,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.010878212245351,
+      "learning_rate": 9.994042924240164e-06,
+      "loss": 0.6293,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2715551445991986,
+      "learning_rate": 9.993960490947533e-06,
+      "loss": 0.6041,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2311954165914183,
+      "learning_rate": 9.993877491561015e-06,
+      "loss": 0.7262,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2721565123934253,
+      "learning_rate": 9.99379392609002e-06,
+      "loss": 0.5909,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0679330089748333,
+      "learning_rate": 9.993709794544022e-06,
+      "loss": 0.6332,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.018072655643991,
+      "learning_rate": 9.993625096932552e-06,
+      "loss": 0.7004,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1394905686491663,
+      "learning_rate": 9.993539833265216e-06,
+      "loss": 0.6536,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.040527633692036,
+      "learning_rate": 9.993454003551676e-06,
+      "loss": 0.6955,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.890590599198691,
+      "learning_rate": 9.993367607801666e-06,
+      "loss": 0.5941,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9929644547312386,
+      "learning_rate": 9.993280646024975e-06,
+      "loss": 0.6812,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.961053725767758,
+      "learning_rate": 9.993193118231463e-06,
+      "loss": 0.6022,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9181967448391344,
+      "learning_rate": 9.993105024431049e-06,
+      "loss": 0.6078,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.968002625002131,
+      "learning_rate": 9.99301636463372e-06,
+      "loss": 0.6773,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4130336710151465,
+      "learning_rate": 9.99292713884953e-06,
+      "loss": 0.6979,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1316184739110313,
+      "learning_rate": 9.992837347088589e-06,
+      "loss": 0.6288,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.892151196066065,
+      "learning_rate": 9.992746989361075e-06,
+      "loss": 0.6294,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9575225798059654,
+      "learning_rate": 9.992656065677234e-06,
+      "loss": 0.7004,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2829375163660637,
+      "learning_rate": 9.99256457604737e-06,
+      "loss": 0.7471,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9971984927642012,
+      "learning_rate": 9.992472520481852e-06,
+      "loss": 0.6576,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2277902080354606,
+      "learning_rate": 9.99237989899112e-06,
+      "loss": 0.7368,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9400694676938235,
+      "learning_rate": 9.992286711585673e-06,
+      "loss": 0.7523,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.088769530783493,
+      "learning_rate": 9.992192958276068e-06,
+      "loss": 0.7252,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.932950291206959,
+      "learning_rate": 9.99209863907294e-06,
+      "loss": 0.7604,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.024874686283672,
+      "learning_rate": 9.992003753986976e-06,
+      "loss": 0.6524,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.87618503750641,
+      "learning_rate": 9.991908303028932e-06,
+      "loss": 0.6223,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.036129483108596,
+      "learning_rate": 9.99181228620963e-06,
+      "loss": 0.7189,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.785804649378539,
+      "learning_rate": 9.991715703539952e-06,
+      "loss": 0.6044,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.090553085564057,
+      "learning_rate": 9.991618555030848e-06,
+      "loss": 0.6913,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9098029143505888,
+      "learning_rate": 9.991520840693331e-06,
+      "loss": 0.7355,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1358723717085457,
+      "learning_rate": 9.991422560538475e-06,
+      "loss": 0.6728,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7182227416827212,
+      "learning_rate": 9.991323714577421e-06,
+      "loss": 0.6789,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9285440404571363,
+      "learning_rate": 9.991224302821374e-06,
+      "loss": 0.5603,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.317719534583608,
+      "learning_rate": 9.991124325281603e-06,
+      "loss": 0.8379,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8938461805189346,
+      "learning_rate": 9.991023781969442e-06,
+      "loss": 0.5413,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.843694757502963,
+      "learning_rate": 9.990922672896288e-06,
+      "loss": 0.6094,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.073935582799042,
+      "learning_rate": 9.9908209980736e-06,
+      "loss": 0.5987,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2169101162154323,
+      "learning_rate": 9.990718757512906e-06,
+      "loss": 0.6508,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.149787600797134,
+      "learning_rate": 9.990615951225797e-06,
+      "loss": 0.7176,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9433589280184018,
+      "learning_rate": 9.990512579223921e-06,
+      "loss": 0.7356,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0519977838803003,
+      "learning_rate": 9.990408641519e-06,
+      "loss": 0.6478,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9732250611772912,
+      "learning_rate": 9.990304138122818e-06,
+      "loss": 0.7158,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0862447871994005,
+      "learning_rate": 9.990199069047216e-06,
+      "loss": 0.7862,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9627513971284538,
+      "learning_rate": 9.990093434304104e-06,
+      "loss": 0.7475,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5440687784424054,
+      "learning_rate": 9.989987233905462e-06,
+      "loss": 0.6919,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.011594577111942,
+      "learning_rate": 9.989880467863323e-06,
+      "loss": 0.6319,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8685240663183704,
+      "learning_rate": 9.989773136189793e-06,
+      "loss": 0.6188,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.031643677974096,
+      "learning_rate": 9.989665238897036e-06,
+      "loss": 0.7435,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9163910387696532,
+      "learning_rate": 9.989556775997284e-06,
+      "loss": 0.6668,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.017766119421384,
+      "learning_rate": 9.989447747502834e-06,
+      "loss": 0.6212,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.145062820353988,
+      "learning_rate": 9.98933815342604e-06,
+      "loss": 0.8157,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9832549864486289,
+      "learning_rate": 9.989227993779332e-06,
+      "loss": 0.6679,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8769152114042982,
+      "learning_rate": 9.98911726857519e-06,
+      "loss": 0.6053,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.135665021547759,
+      "learning_rate": 9.98900597782617e-06,
+      "loss": 0.7071,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1442490339858016,
+      "learning_rate": 9.988894121544885e-06,
+      "loss": 0.671,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3109603092567896,
+      "learning_rate": 9.988781699744016e-06,
+      "loss": 0.6207,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.03188736806228,
+      "learning_rate": 9.988668712436306e-06,
+      "loss": 0.6313,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.019550529610631,
+      "learning_rate": 9.988555159634563e-06,
+      "loss": 0.6206,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0704169777481303,
+      "learning_rate": 9.98844104135166e-06,
+      "loss": 0.6363,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9685873032572212,
+      "learning_rate": 9.98832635760053e-06,
+      "loss": 0.6772,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8690658409387129,
+      "learning_rate": 9.988211108394177e-06,
+      "loss": 0.6678,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.192593402844829,
+      "learning_rate": 9.98809529374566e-06,
+      "loss": 0.5719,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.064493939887632,
+      "learning_rate": 9.987978913668112e-06,
+      "loss": 0.6047,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9756919384468792,
+      "learning_rate": 9.987861968174723e-06,
+      "loss": 0.6073,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9737506247163636,
+      "learning_rate": 9.987744457278753e-06,
+      "loss": 0.6947,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.904287026030052,
+      "learning_rate": 9.987626380993516e-06,
+      "loss": 0.7352,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0224472723257634,
+      "learning_rate": 9.987507739332401e-06,
+      "loss": 0.614,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0295872447401706,
+      "learning_rate": 9.987388532308858e-06,
+      "loss": 0.6972,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0149908366267906,
+      "learning_rate": 9.987268759936396e-06,
+      "loss": 0.7121,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0962870953861183,
+      "learning_rate": 9.987148422228591e-06,
+      "loss": 0.6515,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.984269566815456,
+      "learning_rate": 9.98702751919909e-06,
+      "loss": 0.7571,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1123697955385983,
+      "learning_rate": 9.986906050861595e-06,
+      "loss": 0.6053,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3344562773460167,
+      "learning_rate": 9.986784017229873e-06,
+      "loss": 0.7462,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9590396488402444,
+      "learning_rate": 9.986661418317759e-06,
+      "loss": 0.6522,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2031954129578954,
+      "learning_rate": 9.986538254139151e-06,
+      "loss": 0.742,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8872521125527797,
+      "learning_rate": 9.98641452470801e-06,
+      "loss": 0.5652,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0562007032310485,
+      "learning_rate": 9.986290230038359e-06,
+      "loss": 0.6684,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9331683041832113,
+      "learning_rate": 9.986165370144291e-06,
+      "loss": 0.5762,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9644374337511354,
+      "learning_rate": 9.986039945039959e-06,
+      "loss": 0.6082,
+      "step": 511
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8950957493260308,
+      "learning_rate": 9.985913954739577e-06,
+      "loss": 0.5627,
+      "step": 512
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.897071876769113,
+      "learning_rate": 9.985787399257431e-06,
+      "loss": 0.6015,
+      "step": 513
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2177366478263383,
+      "learning_rate": 9.985660278607865e-06,
+      "loss": 0.7003,
+      "step": 514
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.229245926344614,
+      "learning_rate": 9.985532592805289e-06,
+      "loss": 0.6129,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.312963580341358,
+      "learning_rate": 9.985404341864178e-06,
+      "loss": 0.849,
+      "step": 516
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.121250005175201,
+      "learning_rate": 9.985275525799069e-06,
+      "loss": 0.6587,
+      "step": 517
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.999789191521043,
+      "learning_rate": 9.985146144624563e-06,
+      "loss": 0.7018,
+      "step": 518
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6613885225525393,
+      "learning_rate": 9.985016198355328e-06,
+      "loss": 0.5701,
+      "step": 519
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.048330879977503,
+      "learning_rate": 9.984885687006093e-06,
+      "loss": 0.7511,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1663530153067065,
+      "learning_rate": 9.98475461059165e-06,
+      "loss": 0.5997,
+      "step": 521
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1390771494454555,
+      "learning_rate": 9.984622969126864e-06,
+      "loss": 0.7191,
+      "step": 522
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9328096420880079,
+      "learning_rate": 9.984490762626651e-06,
+      "loss": 0.6176,
+      "step": 523
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.850042905202259,
+      "learning_rate": 9.984357991105999e-06,
+      "loss": 0.7516,
+      "step": 524
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.125264947491382,
+      "learning_rate": 9.984224654579959e-06,
+      "loss": 0.6857,
+      "step": 525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.34692785182022,
+      "learning_rate": 9.984090753063647e-06,
+      "loss": 0.6944,
+      "step": 526
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0665039132353433,
+      "learning_rate": 9.983956286572238e-06,
+      "loss": 0.7515,
+      "step": 527
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.085772307929629,
+      "learning_rate": 9.983821255120977e-06,
+      "loss": 0.6996,
+      "step": 528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9510182753018945,
+      "learning_rate": 9.98368565872517e-06,
+      "loss": 0.7206,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.357416665846521,
+      "learning_rate": 9.983549497400187e-06,
+      "loss": 0.5892,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.108482481371756,
+      "learning_rate": 9.983412771161463e-06,
+      "loss": 0.605,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.118696322223107,
+      "learning_rate": 9.983275480024498e-06,
+      "loss": 0.5796,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9269095508935703,
+      "learning_rate": 9.983137624004851e-06,
+      "loss": 0.5309,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9783461770372632,
+      "learning_rate": 9.982999203118153e-06,
+      "loss": 0.68,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1354909936380326,
+      "learning_rate": 9.982860217380096e-06,
+      "loss": 0.6303,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1276687172019457,
+      "learning_rate": 9.982720666806427e-06,
+      "loss": 0.6457,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5725435712780516,
+      "learning_rate": 9.982580551412972e-06,
+      "loss": 0.6247,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0827610945764903,
+      "learning_rate": 9.982439871215612e-06,
+      "loss": 0.7351,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.928191390389645,
+      "learning_rate": 9.982298626230295e-06,
+      "loss": 0.6179,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0571831064927184,
+      "learning_rate": 9.982156816473029e-06,
+      "loss": 0.5946,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9831697955213374,
+      "learning_rate": 9.982014441959891e-06,
+      "loss": 0.7186,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8500259652367927,
+      "learning_rate": 9.98187150270702e-06,
+      "loss": 0.6823,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.026854564673327,
+      "learning_rate": 9.981727998730616e-06,
+      "loss": 0.7699,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9807028683174117,
+      "learning_rate": 9.98158393004695e-06,
+      "loss": 0.6984,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.589963467561242,
+      "learning_rate": 9.981439296672352e-06,
+      "loss": 0.822,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8283692667191949,
+      "learning_rate": 9.981294098623215e-06,
+      "loss": 0.668,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0929282423372433,
+      "learning_rate": 9.981148335916e-06,
+      "loss": 0.6257,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9273830056696597,
+      "learning_rate": 9.98100200856723e-06,
+      "loss": 0.6686,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1577089469497137,
+      "learning_rate": 9.980855116593494e-06,
+      "loss": 0.6959,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8461204532541409,
+      "learning_rate": 9.980707660011437e-06,
+      "loss": 0.6695,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0891244440911216,
+      "learning_rate": 9.980559638837778e-06,
+      "loss": 0.7155,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.60684984565562,
+      "learning_rate": 9.980411053089298e-06,
+      "loss": 0.7105,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.274419285519808,
+      "learning_rate": 9.980261902782835e-06,
+      "loss": 0.6461,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1995811698396714,
+      "learning_rate": 9.9801121879353e-06,
+      "loss": 0.6659,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.24104794740926,
+      "learning_rate": 9.979961908563663e-06,
+      "loss": 0.6843,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.027896058717004,
+      "learning_rate": 9.97981106468496e-06,
+      "loss": 0.6571,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8618556456848303,
+      "learning_rate": 9.979659656316288e-06,
+      "loss": 0.6097,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.011473620144725,
+      "learning_rate": 9.97950768347481e-06,
+      "loss": 0.5543,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8198767395429643,
+      "learning_rate": 9.979355146177754e-06,
+      "loss": 0.5747,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1523169045634907,
+      "learning_rate": 9.97920204444241e-06,
+      "loss": 0.5956,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1313458275799584,
+      "learning_rate": 9.979048378286134e-06,
+      "loss": 0.6804,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.972092650478314,
+      "learning_rate": 9.978894147726346e-06,
+      "loss": 0.685,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0429159532350414,
+      "learning_rate": 9.978739352780528e-06,
+      "loss": 0.5422,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4276858350172383,
+      "learning_rate": 9.978583993466224e-06,
+      "loss": 0.6217,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.117733238279458,
+      "learning_rate": 9.97842806980105e-06,
+      "loss": 0.8119,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0507740651286666,
+      "learning_rate": 9.978271581802677e-06,
+      "loss": 0.6132,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8381248701512136,
+      "learning_rate": 9.978114529488845e-06,
+      "loss": 0.6151,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.251649730917416,
+      "learning_rate": 9.977956912877356e-06,
+      "loss": 0.6969,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.085417466229146,
+      "learning_rate": 9.977798731986079e-06,
+      "loss": 0.6001,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1417707930582948,
+      "learning_rate": 9.977639986832943e-06,
+      "loss": 0.6477,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.002691296527279,
+      "learning_rate": 9.977480677435942e-06,
+      "loss": 0.6204,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3573137310537087,
+      "learning_rate": 9.977320803813137e-06,
+      "loss": 0.666,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.13178658494689,
+      "learning_rate": 9.977160365982647e-06,
+      "loss": 0.6276,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9299279931740343,
+      "learning_rate": 9.976999363962663e-06,
+      "loss": 0.6463,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1014604248299285,
+      "learning_rate": 9.97683779777143e-06,
+      "loss": 0.6793,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8767137873590285,
+      "learning_rate": 9.976675667427268e-06,
+      "loss": 0.5687,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9640671342301481,
+      "learning_rate": 9.976512972948553e-06,
+      "loss": 0.6154,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7884320005984804,
+      "learning_rate": 9.976349714353729e-06,
+      "loss": 0.6065,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.950622961463605,
+      "learning_rate": 9.976185891661296e-06,
+      "loss": 0.7064,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3083440495765513,
+      "learning_rate": 9.976021504889833e-06,
+      "loss": 0.6901,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.963846800423321,
+      "learning_rate": 9.975856554057968e-06,
+      "loss": 0.5613,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0211788630851535,
+      "learning_rate": 9.9756910391844e-06,
+      "loss": 0.6375,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8268256773710685,
+      "learning_rate": 9.975524960287895e-06,
+      "loss": 0.671,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7600000040813513,
+      "learning_rate": 9.975358317387277e-06,
+      "loss": 0.5318,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.049565771650243,
+      "learning_rate": 9.975191110501432e-06,
+      "loss": 0.6675,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.314117076668117,
+      "learning_rate": 9.975023339649317e-06,
+      "loss": 0.6742,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0701957129127018,
+      "learning_rate": 9.974855004849952e-06,
+      "loss": 0.7539,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.90648641172633,
+      "learning_rate": 9.974686106122415e-06,
+      "loss": 0.6037,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9875373685123054,
+      "learning_rate": 9.974516643485852e-06,
+      "loss": 0.7258,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2446293325637647,
+      "learning_rate": 9.974346616959476e-06,
+      "loss": 0.7915,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.106574799745109,
+      "learning_rate": 9.974176026562558e-06,
+      "loss": 0.6764,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.062901024543719,
+      "learning_rate": 9.974004872314435e-06,
+      "loss": 0.6992,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0645404013113877,
+      "learning_rate": 9.97383315423451e-06,
+      "loss": 0.7663,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1783453708050393,
+      "learning_rate": 9.973660872342244e-06,
+      "loss": 0.7534,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.040114380019622,
+      "learning_rate": 9.973488026657171e-06,
+      "loss": 0.7122,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.063422864247258,
+      "learning_rate": 9.973314617198881e-06,
+      "loss": 0.6104,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8912025270720374,
+      "learning_rate": 9.973140643987034e-06,
+      "loss": 0.611,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0682555441292996,
+      "learning_rate": 9.972966107041349e-06,
+      "loss": 0.748,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0628034068097745,
+      "learning_rate": 9.97279100638161e-06,
+      "loss": 0.7246,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.084889035523409,
+      "learning_rate": 9.972615342027667e-06,
+      "loss": 0.7109,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8611852818863357,
+      "learning_rate": 9.972439113999431e-06,
+      "loss": 0.6385,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0416032338007635,
+      "learning_rate": 9.97226232231688e-06,
+      "loss": 0.7019,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9483099939451032,
+      "learning_rate": 9.972084967000055e-06,
+      "loss": 0.7654,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7570415982383194,
+      "learning_rate": 9.971907048069058e-06,
+      "loss": 0.5443,
+      "step": 604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3140913687038647,
+      "learning_rate": 9.97172856554406e-06,
+      "loss": 0.6249,
+      "step": 605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.943667180111752,
+      "learning_rate": 9.971549519445288e-06,
+      "loss": 0.7652,
+      "step": 606
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2683476874649537,
+      "learning_rate": 9.971369909793043e-06,
+      "loss": 0.5836,
+      "step": 607
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9552541335662836,
+      "learning_rate": 9.971189736607681e-06,
+      "loss": 0.6996,
+      "step": 608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.054084324981065,
+      "learning_rate": 9.97100899990963e-06,
+      "loss": 0.7218,
+      "step": 609
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1292215277853184,
+      "learning_rate": 9.970827699719372e-06,
+      "loss": 0.7322,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2273055903437506,
+      "learning_rate": 9.970645836057464e-06,
+      "loss": 0.7003,
+      "step": 611
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9357280807082045,
+      "learning_rate": 9.97046340894452e-06,
+      "loss": 0.7023,
+      "step": 612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.197029414583496,
+      "learning_rate": 9.970280418401215e-06,
+      "loss": 0.68,
+      "step": 613
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0053051191734634,
+      "learning_rate": 9.970096864448296e-06,
+      "loss": 0.7367,
+      "step": 614
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1301040443505563,
+      "learning_rate": 9.96991274710657e-06,
+      "loss": 0.6358,
+      "step": 615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.419659091613593,
+      "learning_rate": 9.969728066396904e-06,
+      "loss": 0.7072,
+      "step": 616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9261589500687413,
+      "learning_rate": 9.969542822340238e-06,
+      "loss": 0.6603,
+      "step": 617
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1022673917783963,
+      "learning_rate": 9.969357014957564e-06,
+      "loss": 0.7254,
+      "step": 618
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.182516114339237,
+      "learning_rate": 9.96917064426995e-06,
+      "loss": 0.6666,
+      "step": 619
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3098245453776904,
+      "learning_rate": 9.968983710298522e-06,
+      "loss": 0.7824,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9539441765468966,
+      "learning_rate": 9.968796213064466e-06,
+      "loss": 0.7051,
+      "step": 621
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0138948888168176,
+      "learning_rate": 9.968608152589038e-06,
+      "loss": 0.7378,
+      "step": 622
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.175463284078723,
+      "learning_rate": 9.968419528893555e-06,
+      "loss": 0.6006,
+      "step": 623
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3105811086307466,
+      "learning_rate": 9.968230341999403e-06,
+      "loss": 0.763,
+      "step": 624
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.08502012437374,
+      "learning_rate": 9.96804059192802e-06,
+      "loss": 0.6397,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5120934211323087,
+      "learning_rate": 9.96785027870092e-06,
+      "loss": 0.6508,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.131440836614907,
+      "learning_rate": 9.967659402339677e-06,
+      "loss": 0.677,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2797304046639217,
+      "learning_rate": 9.967467962865925e-06,
+      "loss": 0.8344,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8568138944212114,
+      "learning_rate": 9.967275960301364e-06,
+      "loss": 0.6212,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9700358484439096,
+      "learning_rate": 9.967083394667763e-06,
+      "loss": 0.7429,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9885571768773675,
+      "learning_rate": 9.966890265986947e-06,
+      "loss": 0.6928,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8696680843723692,
+      "learning_rate": 9.966696574280808e-06,
+      "loss": 0.639,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9946364083173456,
+      "learning_rate": 9.966502319571303e-06,
+      "loss": 0.7326,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9618197771145713,
+      "learning_rate": 9.966307501880452e-06,
+      "loss": 0.6483,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9669889529206006,
+      "learning_rate": 9.966112121230341e-06,
+      "loss": 0.6443,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0924747286776855,
+      "learning_rate": 9.965916177643112e-06,
+      "loss": 0.738,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.935928912743703,
+      "learning_rate": 9.965719671140981e-06,
+      "loss": 0.776,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.047670358994524,
+      "learning_rate": 9.965522601746222e-06,
+      "loss": 0.767,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9296679711225158,
+      "learning_rate": 9.965324969481172e-06,
+      "loss": 0.6554,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1303563423399687,
+      "learning_rate": 9.965126774368237e-06,
+      "loss": 0.6197,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.074249841183784,
+      "learning_rate": 9.964928016429883e-06,
+      "loss": 0.5803,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9354294943036128,
+      "learning_rate": 9.964728695688635e-06,
+      "loss": 0.5228,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8052649810413786,
+      "learning_rate": 9.964528812167095e-06,
+      "loss": 0.7194,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.177670415548599,
+      "learning_rate": 9.964328365887917e-06,
+      "loss": 0.7194,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9997756821719956,
+      "learning_rate": 9.964127356873821e-06,
+      "loss": 0.6131,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8801154696599298,
+      "learning_rate": 9.963925785147595e-06,
+      "loss": 0.6044,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0091370827752275,
+      "learning_rate": 9.96372365073209e-06,
+      "loss": 0.7101,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.839460149960612,
+      "learning_rate": 9.963520953650214e-06,
+      "loss": 0.6756,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.036076135546113,
+      "learning_rate": 9.963317693924947e-06,
+      "loss": 0.6115,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.545887352696099,
+      "learning_rate": 9.963113871579332e-06,
+      "loss": 0.7111,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0933171239683985,
+      "learning_rate": 9.96290948663647e-06,
+      "loss": 0.6289,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9363189028685541,
+      "learning_rate": 9.962704539119528e-06,
+      "loss": 0.6259,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.064128197173807,
+      "learning_rate": 9.962499029051742e-06,
+      "loss": 0.6223,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8337003404299514,
+      "learning_rate": 9.962292956456405e-06,
+      "loss": 0.6341,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.094853295361747,
+      "learning_rate": 9.962086321356878e-06,
+      "loss": 0.6245,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.977672637587837,
+      "learning_rate": 9.961879123776584e-06,
+      "loss": 0.6979,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1712110101678403,
+      "learning_rate": 9.961671363739008e-06,
+      "loss": 0.6226,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.951254338928478,
+      "learning_rate": 9.961463041267703e-06,
+      "loss": 0.6741,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.170459383653736,
+      "learning_rate": 9.961254156386282e-06,
+      "loss": 0.6541,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1749433390231916,
+      "learning_rate": 9.961044709118425e-06,
+      "loss": 0.7235,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3401410783814383,
+      "learning_rate": 9.960834699487873e-06,
+      "loss": 0.6685,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9667835737612456,
+      "learning_rate": 9.960624127518432e-06,
+      "loss": 0.6615,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.928408867792227,
+      "learning_rate": 9.960412993233973e-06,
+      "loss": 0.6487,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1721273990729517,
+      "learning_rate": 9.960201296658425e-06,
+      "loss": 0.6058,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.391564010754368,
+      "learning_rate": 9.959989037815789e-06,
+      "loss": 0.5665,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.944892387838344,
+      "learning_rate": 9.959776216730125e-06,
+      "loss": 0.6924,
+      "step": 666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.236898508390765,
+      "learning_rate": 9.959562833425557e-06,
+      "loss": 0.6553,
+      "step": 667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9932574533230005,
+      "learning_rate": 9.959348887926274e-06,
+      "loss": 0.7432,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.027491358248483,
+      "learning_rate": 9.959134380256525e-06,
+      "loss": 0.6426,
+      "step": 669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0001475575651027,
+      "learning_rate": 9.95891931044063e-06,
+      "loss": 0.6808,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.820682926080947,
+      "learning_rate": 9.958703678502966e-06,
+      "loss": 0.6167,
+      "step": 671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.018955263359039,
+      "learning_rate": 9.958487484467976e-06,
+      "loss": 0.6538,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.040371696124914,
+      "learning_rate": 9.958270728360166e-06,
+      "loss": 0.6865,
+      "step": 673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1337195894507404,
+      "learning_rate": 9.95805341020411e-06,
+      "loss": 0.6423,
+      "step": 674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.01030394805898,
+      "learning_rate": 9.957835530024438e-06,
+      "loss": 0.6973,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.220551541879092,
+      "learning_rate": 9.95761708784585e-06,
+      "loss": 0.6665,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0803421961018715,
+      "learning_rate": 9.95739808369311e-06,
+      "loss": 0.711,
+      "step": 677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2906932549637853,
+      "learning_rate": 9.95717851759104e-06,
+      "loss": 0.7801,
+      "step": 678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.13204944489717,
+      "learning_rate": 9.956958389564528e-06,
+      "loss": 0.6244,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0425735599412174,
+      "learning_rate": 9.95673769963853e-06,
+      "loss": 0.7015,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9345199013727183,
+      "learning_rate": 9.956516447838063e-06,
+      "loss": 0.6477,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7219188066839115,
+      "learning_rate": 9.956294634188204e-06,
+      "loss": 0.5635,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2382443419248705,
+      "learning_rate": 9.956072258714097e-06,
+      "loss": 0.7586,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9477024443776676,
+      "learning_rate": 9.955849321440953e-06,
+      "loss": 0.728,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9813355258791023,
+      "learning_rate": 9.95562582239404e-06,
+      "loss": 0.6703,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1476591448584776,
+      "learning_rate": 9.955401761598693e-06,
+      "loss": 0.6842,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0908641901740936,
+      "learning_rate": 9.955177139080312e-06,
+      "loss": 0.7313,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.084194905379439,
+      "learning_rate": 9.954951954864361e-06,
+      "loss": 0.7477,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9635843996926026,
+      "learning_rate": 9.954726208976361e-06,
+      "loss": 0.6205,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9130869685081529,
+      "learning_rate": 9.954499901441905e-06,
+      "loss": 0.6619,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.037636079788371,
+      "learning_rate": 9.954273032286646e-06,
+      "loss": 0.6274,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9224440773396205,
+      "learning_rate": 9.9540456015363e-06,
+      "loss": 0.6808,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.134253881991058,
+      "learning_rate": 9.953817609216647e-06,
+      "loss": 0.7179,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.120477071210494,
+      "learning_rate": 9.953589055353534e-06,
+      "loss": 0.6869,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.05650737737351,
+      "learning_rate": 9.953359939972866e-06,
+      "loss": 0.6046,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.886773870673862,
+      "learning_rate": 9.953130263100615e-06,
+      "loss": 0.7154,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.036217797024567,
+      "learning_rate": 9.952900024762818e-06,
+      "loss": 0.761,
+      "step": 697
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2783625596602746,
+      "learning_rate": 9.952669224985572e-06,
+      "loss": 0.6307,
+      "step": 698
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9435212040471868,
+      "learning_rate": 9.95243786379504e-06,
+      "loss": 0.7085,
+      "step": 699
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.804310202195145,
+      "learning_rate": 9.952205941217449e-06,
+      "loss": 0.6781,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.06225387355469,
+      "learning_rate": 9.951973457279087e-06,
+      "loss": 0.7211,
+      "step": 701
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.076961953947264,
+      "learning_rate": 9.951740412006308e-06,
+      "loss": 0.6094,
+      "step": 702
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.985672552340446,
+      "learning_rate": 9.951506805425531e-06,
+      "loss": 0.7098,
+      "step": 703
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8527325941283108,
+      "learning_rate": 9.951272637563233e-06,
+      "loss": 0.5818,
+      "step": 704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9109812540751534,
+      "learning_rate": 9.951037908445961e-06,
+      "loss": 0.6898,
+      "step": 705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.020324053220724,
+      "learning_rate": 9.950802618100323e-06,
+      "loss": 0.7053,
+      "step": 706
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1678914704220067,
+      "learning_rate": 9.950566766552989e-06,
+      "loss": 0.6887,
+      "step": 707
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0424556490788475,
+      "learning_rate": 9.950330353830694e-06,
+      "loss": 0.7281,
+      "step": 708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9996565401479152,
+      "learning_rate": 9.950093379960238e-06,
+      "loss": 0.6922,
+      "step": 709
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9744601447827723,
+      "learning_rate": 9.949855844968484e-06,
+      "loss": 0.6319,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9385415848049423,
+      "learning_rate": 9.949617748882354e-06,
+      "loss": 0.6659,
+      "step": 711
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.078954845276994,
+      "learning_rate": 9.949379091728843e-06,
+      "loss": 0.6764,
+      "step": 712
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9193955270459089,
+      "learning_rate": 9.949139873535e-06,
+      "loss": 0.6107,
+      "step": 713
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8955909321628803,
+      "learning_rate": 9.948900094327943e-06,
+      "loss": 0.6845,
+      "step": 714
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.932089420380341,
+      "learning_rate": 9.948659754134852e-06,
+      "loss": 0.6588,
+      "step": 715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8948454482363528,
+      "learning_rate": 9.948418852982973e-06,
+      "loss": 0.5758,
+      "step": 716
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8286956041099895,
+      "learning_rate": 9.948177390899611e-06,
+      "loss": 0.6549,
+      "step": 717
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.894187924245138,
+      "learning_rate": 9.94793536791214e-06,
+      "loss": 0.5799,
+      "step": 718
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0176210263591576,
+      "learning_rate": 9.94769278404799e-06,
+      "loss": 0.6644,
+      "step": 719
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8158957435685823,
+      "learning_rate": 9.947449639334663e-06,
+      "loss": 0.7072,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8892811407727792,
+      "learning_rate": 9.94720593379972e-06,
+      "loss": 0.6623,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.201408847040572,
+      "learning_rate": 9.946961667470787e-06,
+      "loss": 0.6785,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8596998671238845,
+      "learning_rate": 9.946716840375552e-06,
+      "loss": 0.6325,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1929689551861618,
+      "learning_rate": 9.946471452541768e-06,
+      "loss": 0.7123,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8739563959458077,
+      "learning_rate": 9.94622550399725e-06,
+      "loss": 0.7399,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9897122706062023,
+      "learning_rate": 9.945978994769878e-06,
+      "loss": 0.7497,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9516686408380788,
+      "learning_rate": 9.945731924887598e-06,
+      "loss": 0.6577,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7720438709863602,
+      "learning_rate": 9.945484294378413e-06,
+      "loss": 0.6154,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.249334522149298,
+      "learning_rate": 9.945236103270395e-06,
+      "loss": 0.7274,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8705068155172135,
+      "learning_rate": 9.944987351591677e-06,
+      "loss": 0.5798,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8901122235028085,
+      "learning_rate": 9.944738039370458e-06,
+      "loss": 0.7301,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0901188573895166,
+      "learning_rate": 9.944488166635e-06,
+      "loss": 0.6661,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0716119906288104,
+      "learning_rate": 9.944237733413623e-06,
+      "loss": 0.7104,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0084060471211003,
+      "learning_rate": 9.943986739734718e-06,
+      "loss": 0.7807,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.126162332371512,
+      "learning_rate": 9.943735185626739e-06,
+      "loss": 0.7289,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.048734123772361,
+      "learning_rate": 9.943483071118197e-06,
+      "loss": 0.7006,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.232634046448412,
+      "learning_rate": 9.94323039623767e-06,
+      "loss": 0.7472,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0219318375835824,
+      "learning_rate": 9.942977161013802e-06,
+      "loss": 0.6657,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9114763004154225,
+      "learning_rate": 9.942723365475301e-06,
+      "loss": 0.7154,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2527373356434115,
+      "learning_rate": 9.942469009650933e-06,
+      "loss": 0.732,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2324878692245385,
+      "learning_rate": 9.942214093569534e-06,
+      "loss": 0.7158,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1365648202427594,
+      "learning_rate": 9.941958617259994e-06,
+      "loss": 0.722,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1565366949992963,
+      "learning_rate": 9.941702580751278e-06,
+      "loss": 0.7789,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9027629563938848,
+      "learning_rate": 9.941445984072408e-06,
+      "loss": 0.6342,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1215899854013403,
+      "learning_rate": 9.941188827252471e-06,
+      "loss": 0.6236,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.124700681765373,
+      "learning_rate": 9.940931110320615e-06,
+      "loss": 0.6907,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.894826319876532,
+      "learning_rate": 9.940672833306056e-06,
+      "loss": 0.648,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1683138583935646,
+      "learning_rate": 9.940413996238071e-06,
+      "loss": 0.6152,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.304827301613214,
+      "learning_rate": 9.940154599145998e-06,
+      "loss": 0.6609,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1667628822195653,
+      "learning_rate": 9.939894642059248e-06,
+      "loss": 0.773,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2670965613851357,
+      "learning_rate": 9.939634125007279e-06,
+      "loss": 0.749,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9105308819933733,
+      "learning_rate": 9.939373048019629e-06,
+      "loss": 0.5988,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.001684787938153,
+      "learning_rate": 9.93911141112589e-06,
+      "loss": 0.756,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9897912439886951,
+      "learning_rate": 9.938849214355722e-06,
+      "loss": 0.5951,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8260079718444404,
+      "learning_rate": 9.938586457738844e-06,
+      "loss": 0.6922,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0208943552103538,
+      "learning_rate": 9.938323141305042e-06,
+      "loss": 0.6881,
+      "step": 756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8892014681961153,
+      "learning_rate": 9.938059265084163e-06,
+      "loss": 0.6852,
+      "step": 757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.132937380951946,
+      "learning_rate": 9.937794829106122e-06,
+      "loss": 0.7,
+      "step": 758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0088322043327507,
+      "learning_rate": 9.937529833400892e-06,
+      "loss": 0.6808,
+      "step": 759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.041185119184499,
+      "learning_rate": 9.937264277998513e-06,
+      "loss": 0.5472,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8536310070000221,
+      "learning_rate": 9.936998162929086e-06,
+      "loss": 0.5643,
+      "step": 761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.001083354207652,
+      "learning_rate": 9.936731488222776e-06,
+      "loss": 0.6498,
+      "step": 762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0706938113208717,
+      "learning_rate": 9.936464253909817e-06,
+      "loss": 0.6487,
+      "step": 763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0794572945560916,
+      "learning_rate": 9.936196460020496e-06,
+      "loss": 0.6505,
+      "step": 764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1561768544194146,
+      "learning_rate": 9.93592810658517e-06,
+      "loss": 0.6906,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.008123268630148,
+      "learning_rate": 9.935659193634261e-06,
+      "loss": 0.6367,
+      "step": 766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.25399791489673,
+      "learning_rate": 9.935389721198249e-06,
+      "loss": 0.7417,
+      "step": 767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8352434635664747,
+      "learning_rate": 9.935119689307682e-06,
+      "loss": 0.6074,
+      "step": 768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.326860474330298,
+      "learning_rate": 9.934849097993168e-06,
+      "loss": 0.7976,
+      "step": 769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8668608391500925,
+      "learning_rate": 9.934577947285382e-06,
+      "loss": 0.7362,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.877597080458074,
+      "learning_rate": 9.934306237215057e-06,
+      "loss": 0.5724,
+      "step": 771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8322003762305377,
+      "learning_rate": 9.934033967812998e-06,
+      "loss": 0.6854,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7621008758677785,
+      "learning_rate": 9.933761139110065e-06,
+      "loss": 0.6506,
+      "step": 773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0299854137963975,
+      "learning_rate": 9.933487751137185e-06,
+      "loss": 0.6442,
+      "step": 774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8777959177763561,
+      "learning_rate": 9.93321380392535e-06,
+      "loss": 0.7335,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3950292493331333,
+      "learning_rate": 9.932939297505611e-06,
+      "loss": 0.7287,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.148775632603258,
+      "learning_rate": 9.932664231909087e-06,
+      "loss": 0.6965,
+      "step": 777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3030378873226254,
+      "learning_rate": 9.932388607166954e-06,
+      "loss": 0.5862,
+      "step": 778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8487026804028814,
+      "learning_rate": 9.93211242331046e-06,
+      "loss": 0.5834,
+      "step": 779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6659735801019722,
+      "learning_rate": 9.931835680370912e-06,
+      "loss": 0.6148,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0126658622092233,
+      "learning_rate": 9.931558378379677e-06,
+      "loss": 0.7435,
+      "step": 781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.067686271560952,
+      "learning_rate": 9.931280517368193e-06,
+      "loss": 0.7257,
+      "step": 782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1745154175639487,
+      "learning_rate": 9.931002097367954e-06,
+      "loss": 0.6712,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.045784496263562,
+      "learning_rate": 9.930723118410521e-06,
+      "loss": 0.6423,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.270602416768308,
+      "learning_rate": 9.93044358052752e-06,
+      "loss": 0.6607,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.845961952430096,
+      "learning_rate": 9.930163483750636e-06,
+      "loss": 0.6405,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.681941760116097,
+      "learning_rate": 9.929882828111619e-06,
+      "loss": 0.6073,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1011423694291844,
+      "learning_rate": 9.929601613642285e-06,
+      "loss": 0.715,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7766205903261256,
+      "learning_rate": 9.92931984037451e-06,
+      "loss": 0.5952,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.897836028769625,
+      "learning_rate": 9.929037508340234e-06,
+      "loss": 0.7663,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.85405318753218,
+      "learning_rate": 9.928754617571464e-06,
+      "loss": 0.7752,
+      "step": 791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0696560918392723,
+      "learning_rate": 9.928471168100264e-06,
+      "loss": 0.5531,
+      "step": 792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2262392903553256,
+      "learning_rate": 9.928187159958764e-06,
+      "loss": 0.7426,
+      "step": 793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1714035118411035,
+      "learning_rate": 9.927902593179163e-06,
+      "loss": 0.6794,
+      "step": 794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1031096079567733,
+      "learning_rate": 9.927617467793713e-06,
+      "loss": 0.7313,
+      "step": 795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.178237030703268,
+      "learning_rate": 9.927331783834737e-06,
+      "loss": 0.6139,
+      "step": 796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.006402940126805,
+      "learning_rate": 9.927045541334618e-06,
+      "loss": 0.7167,
+      "step": 797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7349004505955714,
+      "learning_rate": 9.926758740325803e-06,
+      "loss": 0.5405,
+      "step": 798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7514974917690225,
+      "learning_rate": 9.926471380840805e-06,
+      "loss": 0.6142,
+      "step": 799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.002352977374873,
+      "learning_rate": 9.926183462912196e-06,
+      "loss": 0.673,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.820390271795529,
+      "learning_rate": 9.92589498657261e-06,
+      "loss": 0.5349,
+      "step": 801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.061683813396355,
+      "learning_rate": 9.925605951854754e-06,
+      "loss": 0.7079,
+      "step": 802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9960408210755969,
+      "learning_rate": 9.925316358791388e-06,
+      "loss": 0.6418,
+      "step": 803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.014159080687005,
+      "learning_rate": 9.925026207415338e-06,
+      "loss": 0.5767,
+      "step": 804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7464740551614473,
+      "learning_rate": 9.924735497759497e-06,
+      "loss": 0.5434,
+      "step": 805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9006992994964862,
+      "learning_rate": 9.924444229856817e-06,
+      "loss": 0.6631,
+      "step": 806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8335712721819453,
+      "learning_rate": 9.924152403740315e-06,
+      "loss": 0.6265,
+      "step": 807
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9521569470070748,
+      "learning_rate": 9.92386001944307e-06,
+      "loss": 0.6922,
+      "step": 808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.102488618677563,
+      "learning_rate": 9.923567076998228e-06,
+      "loss": 0.6094,
+      "step": 809
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.150033884300534,
+      "learning_rate": 9.923273576438994e-06,
+      "loss": 0.5725,
+      "step": 810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.030112095841831,
+      "learning_rate": 9.92297951779864e-06,
+      "loss": 0.6741,
+      "step": 811
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9791624948612083,
+      "learning_rate": 9.922684901110496e-06,
+      "loss": 0.6707,
+      "step": 812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2930407423495707,
+      "learning_rate": 9.92238972640796e-06,
+      "loss": 0.6664,
+      "step": 813
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8906542205262489,
+      "learning_rate": 9.922093993724492e-06,
+      "loss": 0.6267,
+      "step": 814
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9261005637706718,
+      "learning_rate": 9.921797703093614e-06,
+      "loss": 0.6355,
+      "step": 815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8862503970508953,
+      "learning_rate": 9.921500854548916e-06,
+      "loss": 0.6027,
+      "step": 816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1505482544915844,
+      "learning_rate": 9.921203448124042e-06,
+      "loss": 0.6235,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.918340051848613,
+      "learning_rate": 9.920905483852708e-06,
+      "loss": 0.7459,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8334734949147398,
+      "learning_rate": 9.920606961768689e-06,
+      "loss": 0.628,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9869297816294038,
+      "learning_rate": 9.920307881905824e-06,
+      "loss": 0.6617,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9449685673150783,
+      "learning_rate": 9.920008244298016e-06,
+      "loss": 0.6223,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.03168416029326,
+      "learning_rate": 9.91970804897923e-06,
+      "loss": 0.6794,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.878625929884055,
+      "learning_rate": 9.919407295983496e-06,
+      "loss": 0.6658,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9650937939981765,
+      "learning_rate": 9.919105985344906e-06,
+      "loss": 0.5458,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0909683014256286,
+      "learning_rate": 9.918804117097612e-06,
+      "loss": 0.7894,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0613301951740173,
+      "learning_rate": 9.918501691275837e-06,
+      "loss": 0.6921,
+      "step": 826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0419520750382265,
+      "learning_rate": 9.918198707913861e-06,
+      "loss": 0.6937,
+      "step": 827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8581097977278722,
+      "learning_rate": 9.917895167046027e-06,
+      "loss": 0.6482,
+      "step": 828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8748461988863785,
+      "learning_rate": 9.917591068706747e-06,
+      "loss": 0.6154,
+      "step": 829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8925608747518752,
+      "learning_rate": 9.917286412930489e-06,
+      "loss": 0.6991,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8160060819942032,
+      "learning_rate": 9.916981199751789e-06,
+      "loss": 0.5783,
+      "step": 831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.871521248122172,
+      "learning_rate": 9.916675429205243e-06,
+      "loss": 0.7109,
+      "step": 832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1074885386758977,
+      "learning_rate": 9.916369101325514e-06,
+      "loss": 0.7108,
+      "step": 833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1192414719954025,
+      "learning_rate": 9.916062216147324e-06,
+      "loss": 0.6341,
+      "step": 834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.08239120677423,
+      "learning_rate": 9.915754773705461e-06,
+      "loss": 0.6375,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1316657605531093,
+      "learning_rate": 9.915446774034776e-06,
+      "loss": 0.7224,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.281939628017067,
+      "learning_rate": 9.915138217170184e-06,
+      "loss": 0.7129,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8741130808070126,
+      "learning_rate": 9.914829103146658e-06,
+      "loss": 0.6406,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9684244785118747,
+      "learning_rate": 9.91451943199924e-06,
+      "loss": 0.7195,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0475888837677045,
+      "learning_rate": 9.914209203763032e-06,
+      "loss": 0.6547,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.04921767979233,
+      "learning_rate": 9.9138984184732e-06,
+      "loss": 0.6901,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.172629056653797,
+      "learning_rate": 9.913587076164976e-06,
+      "loss": 0.7042,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.267660676241233,
+      "learning_rate": 9.91327517687365e-06,
+      "loss": 0.685,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7760733119676897,
+      "learning_rate": 9.912962720634575e-06,
+      "loss": 0.7247,
+      "step": 844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8798377763004066,
+      "learning_rate": 9.912649707483174e-06,
+      "loss": 0.6346,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.869284735547129,
+      "learning_rate": 9.91233613745493e-06,
+      "loss": 0.6542,
+      "step": 846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.926516930261669,
+      "learning_rate": 9.912022010585385e-06,
+      "loss": 0.5949,
+      "step": 847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.972750668754731,
+      "learning_rate": 9.911707326910145e-06,
+      "loss": 0.5647,
+      "step": 848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7730609944376714,
+      "learning_rate": 9.911392086464886e-06,
+      "loss": 0.69,
+      "step": 849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0654042895948033,
+      "learning_rate": 9.911076289285338e-06,
+      "loss": 0.6786,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9272137533364486,
+      "learning_rate": 9.910759935407301e-06,
+      "loss": 0.6813,
+      "step": 851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.076512024838693,
+      "learning_rate": 9.910443024866636e-06,
+      "loss": 0.7075,
+      "step": 852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9792218894356493,
+      "learning_rate": 9.910125557699266e-06,
+      "loss": 0.5437,
+      "step": 853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.96480203139793,
+      "learning_rate": 9.909807533941176e-06,
+      "loss": 0.6791,
+      "step": 854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.571498964350275,
+      "learning_rate": 9.909488953628416e-06,
+      "loss": 0.7562,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0928806202081334,
+      "learning_rate": 9.909169816797102e-06,
+      "loss": 0.6678,
+      "step": 856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1045247209467752,
+      "learning_rate": 9.908850123483406e-06,
+      "loss": 0.63,
+      "step": 857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0200550049787336,
+      "learning_rate": 9.908529873723571e-06,
+      "loss": 0.6311,
+      "step": 858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.152052863379161,
+      "learning_rate": 9.908209067553897e-06,
+      "loss": 0.6502,
+      "step": 859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8592974529643835,
+      "learning_rate": 9.907887705010748e-06,
+      "loss": 0.6613,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9787376817789002,
+      "learning_rate": 9.907565786130556e-06,
+      "loss": 0.6377,
+      "step": 861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1440337219046173,
+      "learning_rate": 9.907243310949806e-06,
+      "loss": 0.7986,
+      "step": 862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8768089367088299,
+      "learning_rate": 9.906920279505058e-06,
+      "loss": 0.6163,
+      "step": 863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.23557857803684,
+      "learning_rate": 9.90659669183293e-06,
+      "loss": 0.7218,
+      "step": 864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.739848221698496,
+      "learning_rate": 9.906272547970098e-06,
+      "loss": 0.6745,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.055645158811453,
+      "learning_rate": 9.90594784795331e-06,
+      "loss": 0.7725,
+      "step": 866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9873627014566904,
+      "learning_rate": 9.905622591819368e-06,
+      "loss": 0.6261,
+      "step": 867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2197485693258354,
+      "learning_rate": 9.905296779605144e-06,
+      "loss": 0.6184,
+      "step": 868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8826852954404683,
+      "learning_rate": 9.904970411347574e-06,
+      "loss": 0.5683,
+      "step": 869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8557977806798933,
+      "learning_rate": 9.904643487083648e-06,
+      "loss": 0.5471,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.871864907218477,
+      "learning_rate": 9.904316006850428e-06,
+      "loss": 0.6914,
+      "step": 871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.020931495944233,
+      "learning_rate": 9.903987970685034e-06,
+      "loss": 0.7881,
+      "step": 872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9461530000097567,
+      "learning_rate": 9.903659378624652e-06,
+      "loss": 0.6323,
+      "step": 873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9688203893811644,
+      "learning_rate": 9.903330230706529e-06,
+      "loss": 0.7003,
+      "step": 874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8043702789163878,
+      "learning_rate": 9.903000526967977e-06,
+      "loss": 0.6036,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.950517288648671,
+      "learning_rate": 9.90267026744637e-06,
+      "loss": 0.6606,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.053483126005183,
+      "learning_rate": 9.902339452179142e-06,
+      "loss": 0.6498,
+      "step": 877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8803225276792594,
+      "learning_rate": 9.902008081203796e-06,
+      "loss": 0.7156,
+      "step": 878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8191419159760944,
+      "learning_rate": 9.901676154557893e-06,
+      "loss": 0.6745,
+      "step": 879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8484125244211091,
+      "learning_rate": 9.90134367227906e-06,
+      "loss": 0.7064,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0590899215765233,
+      "learning_rate": 9.901010634404983e-06,
+      "loss": 0.656,
+      "step": 881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9746227071307032,
+      "learning_rate": 9.900677040973418e-06,
+      "loss": 0.7299,
+      "step": 882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.073087199185832,
+      "learning_rate": 9.900342892022176e-06,
+      "loss": 0.7405,
+      "step": 883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0167336165515475,
+      "learning_rate": 9.900008187589138e-06,
+      "loss": 0.6863,
+      "step": 884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.938178039873934,
+      "learning_rate": 9.899672927712242e-06,
+      "loss": 0.8384,
+      "step": 885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6977338968973716,
+      "learning_rate": 9.899337112429492e-06,
+      "loss": 0.6711,
+      "step": 886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9888290143305485,
+      "learning_rate": 9.899000741778956e-06,
+      "loss": 0.6642,
+      "step": 887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9917854977117735,
+      "learning_rate": 9.898663815798761e-06,
+      "loss": 0.6263,
+      "step": 888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0969486651793816,
+      "learning_rate": 9.898326334527102e-06,
+      "loss": 0.6614,
+      "step": 889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.026670466094795,
+      "learning_rate": 9.897988298002233e-06,
+      "loss": 0.6663,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9815021235689791,
+      "learning_rate": 9.897649706262474e-06,
+      "loss": 0.6395,
+      "step": 891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2613612523304845,
+      "learning_rate": 9.897310559346203e-06,
+      "loss": 0.801,
+      "step": 892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.821267540525475,
+      "learning_rate": 9.896970857291868e-06,
+      "loss": 0.6847,
+      "step": 893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0966029040749694,
+      "learning_rate": 9.896630600137974e-06,
+      "loss": 0.713,
+      "step": 894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.920981843170863,
+      "learning_rate": 9.896289787923092e-06,
+      "loss": 0.6597,
+      "step": 895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9392140861418414,
+      "learning_rate": 9.895948420685855e-06,
+      "loss": 0.7259,
+      "step": 896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1670282608670166,
+      "learning_rate": 9.895606498464956e-06,
+      "loss": 0.7213,
+      "step": 897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0848482577609,
+      "learning_rate": 9.895264021299158e-06,
+      "loss": 0.7159,
+      "step": 898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.716703867604487,
+      "learning_rate": 9.894920989227282e-06,
+      "loss": 0.5575,
+      "step": 899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7866881854548302,
+      "learning_rate": 9.89457740228821e-06,
+      "loss": 0.6121,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8671996440775374,
+      "learning_rate": 9.894233260520893e-06,
+      "loss": 0.5676,
+      "step": 901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9434408487574717,
+      "learning_rate": 9.89388856396434e-06,
+      "loss": 0.6724,
+      "step": 902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7936908863669354,
+      "learning_rate": 9.893543312657623e-06,
+      "loss": 0.6487,
+      "step": 903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.027708079476974,
+      "learning_rate": 9.89319750663988e-06,
+      "loss": 0.6754,
+      "step": 904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0114888734872203,
+      "learning_rate": 9.892851145950308e-06,
+      "loss": 0.6927,
+      "step": 905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.924424371946275,
+      "learning_rate": 9.89250423062817e-06,
+      "loss": 0.6884,
+      "step": 906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8961310755643144,
+      "learning_rate": 9.892156760712793e-06,
+      "loss": 0.6769,
+      "step": 907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9632794427660414,
+      "learning_rate": 9.891808736243563e-06,
+      "loss": 0.6999,
+      "step": 908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.495602941098346,
+      "learning_rate": 9.89146015725993e-06,
+      "loss": 0.7327,
+      "step": 909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3402932804708834,
+      "learning_rate": 9.891111023801405e-06,
+      "loss": 0.636,
+      "step": 910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9472235512569598,
+      "learning_rate": 9.89076133590757e-06,
+      "loss": 0.6924,
+      "step": 911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9751978430386354,
+      "learning_rate": 9.89041109361806e-06,
+      "loss": 0.6588,
+      "step": 912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9931745510211398,
+      "learning_rate": 9.89006029697258e-06,
+      "loss": 0.7378,
+      "step": 913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9711240247848647,
+      "learning_rate": 9.88970894601089e-06,
+      "loss": 0.7044,
+      "step": 914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.945758925356476,
+      "learning_rate": 9.889357040772822e-06,
+      "loss": 0.7053,
+      "step": 915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.903319178019049,
+      "learning_rate": 9.889004581298265e-06,
+      "loss": 0.6292,
+      "step": 916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9525247895967082,
+      "learning_rate": 9.888651567627173e-06,
+      "loss": 0.5822,
+      "step": 917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0574072003726425,
+      "learning_rate": 9.88829799979956e-06,
+      "loss": 0.6102,
+      "step": 918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.044284692228813,
+      "learning_rate": 9.887943877855505e-06,
+      "loss": 0.6571,
+      "step": 919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5942595611222257,
+      "learning_rate": 9.887589201835154e-06,
+      "loss": 0.5227,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8786375947915854,
+      "learning_rate": 9.88723397177871e-06,
+      "loss": 0.586,
+      "step": 921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8615389080365194,
+      "learning_rate": 9.886878187726435e-06,
+      "loss": 0.5845,
+      "step": 922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.064836397948735,
+      "learning_rate": 9.886521849718665e-06,
+      "loss": 0.614,
+      "step": 923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8259649534315854,
+      "learning_rate": 9.886164957795792e-06,
+      "loss": 0.5673,
+      "step": 924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7726574431040119,
+      "learning_rate": 9.885807511998269e-06,
+      "loss": 0.6505,
+      "step": 925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9798645528179286,
+      "learning_rate": 9.885449512366617e-06,
+      "loss": 0.7142,
+      "step": 926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7854622151133066,
+      "learning_rate": 9.885090958941416e-06,
+      "loss": 0.6493,
+      "step": 927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9520396817553052,
+      "learning_rate": 9.884731851763313e-06,
+      "loss": 0.5635,
+      "step": 928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.002592135377459,
+      "learning_rate": 9.884372190873011e-06,
+      "loss": 0.7452,
+      "step": 929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7836620496269653,
+      "learning_rate": 9.88401197631128e-06,
+      "loss": 0.6534,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8592328756550733,
+      "learning_rate": 9.883651208118956e-06,
+      "loss": 0.6744,
+      "step": 931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8997837483930826,
+      "learning_rate": 9.88328988633693e-06,
+      "loss": 0.7845,
+      "step": 932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1336391293428023,
+      "learning_rate": 9.882928011006163e-06,
+      "loss": 0.5957,
+      "step": 933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8537580965801874,
+      "learning_rate": 9.882565582167673e-06,
+      "loss": 0.6615,
+      "step": 934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.857686711318004,
+      "learning_rate": 9.882202599862545e-06,
+      "loss": 0.6753,
+      "step": 935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9365530494057581,
+      "learning_rate": 9.881839064131925e-06,
+      "loss": 0.6454,
+      "step": 936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0038963286408062,
+      "learning_rate": 9.88147497501702e-06,
+      "loss": 0.6664,
+      "step": 937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3477040250832326,
+      "learning_rate": 9.881110332559104e-06,
+      "loss": 0.7557,
+      "step": 938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0700186962715974,
+      "learning_rate": 9.88074513679951e-06,
+      "loss": 0.6809,
+      "step": 939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7916439421510102,
+      "learning_rate": 9.880379387779637e-06,
+      "loss": 0.5874,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0284864284873567,
+      "learning_rate": 9.880013085540942e-06,
+      "loss": 0.7315,
+      "step": 941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.079999840493714,
+      "learning_rate": 9.879646230124949e-06,
+      "loss": 0.7182,
+      "step": 942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.981513474191177,
+      "learning_rate": 9.879278821573241e-06,
+      "loss": 0.6527,
+      "step": 943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.082383292274721,
+      "learning_rate": 9.87891085992747e-06,
+      "loss": 0.6105,
+      "step": 944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.772693746109168,
+      "learning_rate": 9.878542345229342e-06,
+      "loss": 0.6322,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7969774735537083,
+      "learning_rate": 9.878173277520636e-06,
+      "loss": 0.7354,
+      "step": 946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7088315539160064,
+      "learning_rate": 9.877803656843182e-06,
+      "loss": 0.6111,
+      "step": 947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0378349193456065,
+      "learning_rate": 9.877433483238881e-06,
+      "loss": 0.6576,
+      "step": 948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9073383214298338,
+      "learning_rate": 9.877062756749694e-06,
+      "loss": 0.7225,
+      "step": 949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9049859401892608,
+      "learning_rate": 9.876691477417644e-06,
+      "loss": 0.6158,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8083745188555216,
+      "learning_rate": 9.876319645284821e-06,
+      "loss": 0.6074,
+      "step": 951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0656455996647165,
+      "learning_rate": 9.875947260393371e-06,
+      "loss": 0.6854,
+      "step": 952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1464444197411945,
+      "learning_rate": 9.875574322785508e-06,
+      "loss": 0.6364,
+      "step": 953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8324837012367032,
+      "learning_rate": 9.875200832503505e-06,
+      "loss": 0.6757,
+      "step": 954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9747983267467775,
+      "learning_rate": 9.8748267895897e-06,
+      "loss": 0.7113,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7988094223419764,
+      "learning_rate": 9.874452194086492e-06,
+      "loss": 0.5959,
+      "step": 956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.297513189742376,
+      "learning_rate": 9.874077046036345e-06,
+      "loss": 0.71,
+      "step": 957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0557038443885123,
+      "learning_rate": 9.873701345481784e-06,
+      "loss": 0.7535,
+      "step": 958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1326435760926987,
+      "learning_rate": 9.873325092465395e-06,
+      "loss": 0.7603,
+      "step": 959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.980369170639488,
+      "learning_rate": 9.872948287029833e-06,
+      "loss": 0.6861,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.004930978839084,
+      "learning_rate": 9.872570929217804e-06,
+      "loss": 0.6836,
+      "step": 961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0397373283212565,
+      "learning_rate": 9.87219301907209e-06,
+      "loss": 0.6653,
+      "step": 962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.932230567382412,
+      "learning_rate": 9.871814556635525e-06,
+      "loss": 0.674,
+      "step": 963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9015175721561417,
+      "learning_rate": 9.871435541951011e-06,
+      "loss": 0.6211,
+      "step": 964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8075315769084097,
+      "learning_rate": 9.871055975061514e-06,
+      "loss": 0.7,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.782486219766704,
+      "learning_rate": 9.870675856010058e-06,
+      "loss": 0.7312,
+      "step": 966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9553211188111437,
+      "learning_rate": 9.87029518483973e-06,
+      "loss": 0.6937,
+      "step": 967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.077651871787083,
+      "learning_rate": 9.869913961593685e-06,
+      "loss": 0.7183,
+      "step": 968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8307112275173123,
+      "learning_rate": 9.869532186315134e-06,
+      "loss": 0.6975,
+      "step": 969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1319077853591404,
+      "learning_rate": 9.869149859047355e-06,
+      "loss": 0.6904,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1805511045289396,
+      "learning_rate": 9.868766979833686e-06,
+      "loss": 0.6585,
+      "step": 971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9877639959778115,
+      "learning_rate": 9.868383548717529e-06,
+      "loss": 0.6813,
+      "step": 972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4334437231614343,
+      "learning_rate": 9.867999565742348e-06,
+      "loss": 0.7393,
+      "step": 973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7359878345550932,
+      "learning_rate": 9.867615030951668e-06,
+      "loss": 0.6564,
+      "step": 974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8375121930806249,
+      "learning_rate": 9.86722994438908e-06,
+      "loss": 0.7168,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7151473512488256,
+      "learning_rate": 9.866844306098238e-06,
+      "loss": 0.6273,
+      "step": 976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.830403049133302,
+      "learning_rate": 9.866458116122852e-06,
+      "loss": 0.6368,
+      "step": 977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.970209094089418,
+      "learning_rate": 9.866071374506701e-06,
+      "loss": 0.769,
+      "step": 978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.141400584070093,
+      "learning_rate": 9.865684081293624e-06,
+      "loss": 0.679,
+      "step": 979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.690775547376312,
+      "learning_rate": 9.865296236527523e-06,
+      "loss": 0.6908,
+      "step": 980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.043177681315642,
+      "learning_rate": 9.86490784025236e-06,
+      "loss": 0.6868,
+      "step": 981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.833251578843401,
+      "learning_rate": 9.864518892512167e-06,
+      "loss": 0.6162,
+      "step": 982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1456831160855403,
+      "learning_rate": 9.86412939335103e-06,
+      "loss": 0.7424,
+      "step": 983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0893900408271007,
+      "learning_rate": 9.8637393428131e-06,
+      "loss": 0.6559,
+      "step": 984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3334547598341455,
+      "learning_rate": 9.863348740942595e-06,
+      "loss": 0.6998,
+      "step": 985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9053267391294475,
+      "learning_rate": 9.86295758778379e-06,
+      "loss": 0.5793,
+      "step": 986
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.053265187490565,
+      "learning_rate": 9.862565883381025e-06,
+      "loss": 0.6087,
+      "step": 987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8738984376187984,
+      "learning_rate": 9.862173627778699e-06,
+      "loss": 0.6344,
+      "step": 988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9312070177429788,
+      "learning_rate": 9.861780821021282e-06,
+      "loss": 0.6113,
+      "step": 989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0840209825648373,
+      "learning_rate": 9.861387463153295e-06,
+      "loss": 0.7107,
+      "step": 990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.053479103718914,
+      "learning_rate": 9.860993554219333e-06,
+      "loss": 0.6901,
+      "step": 991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3013383389382795,
+      "learning_rate": 9.860599094264043e-06,
+      "loss": 0.6469,
+      "step": 992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8821636541147546,
+      "learning_rate": 9.860204083332142e-06,
+      "loss": 0.7016,
+      "step": 993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.871488491257933,
+      "learning_rate": 9.859808521468404e-06,
+      "loss": 0.6663,
+      "step": 994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8671302354097286,
+      "learning_rate": 9.859412408717672e-06,
+      "loss": 0.67,
+      "step": 995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1881384088375335,
+      "learning_rate": 9.859015745124844e-06,
+      "loss": 0.5869,
+      "step": 996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8839529894787386,
+      "learning_rate": 9.858618530734887e-06,
+      "loss": 0.6813,
+      "step": 997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.024864643623028,
+      "learning_rate": 9.858220765592828e-06,
+      "loss": 0.584,
+      "step": 998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8817794170694133,
+      "learning_rate": 9.857822449743752e-06,
+      "loss": 0.6694,
+      "step": 999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3342042069446327,
+      "learning_rate": 9.857423583232812e-06,
+      "loss": 0.8106,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7941891959216587,
+      "learning_rate": 9.857024166105224e-06,
+      "loss": 0.6335,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.181704718868215,
+      "learning_rate": 9.856624198406262e-06,
+      "loss": 0.6525,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.908208488414037,
+      "learning_rate": 9.856223680181267e-06,
+      "loss": 0.6317,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.517562370686754,
+      "learning_rate": 9.855822611475636e-06,
+      "loss": 0.6677,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.755038812115677,
+      "learning_rate": 9.855420992334836e-06,
+      "loss": 0.6095,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0790618780998917,
+      "learning_rate": 9.85501882280439e-06,
+      "loss": 0.7459,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9235236943259157,
+      "learning_rate": 9.85461610292989e-06,
+      "loss": 0.6692,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6253316092641337,
+      "learning_rate": 9.854212832756984e-06,
+      "loss": 0.6678,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.117501284290816,
+      "learning_rate": 9.853809012331384e-06,
+      "loss": 0.7446,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.156750435017462,
+      "learning_rate": 9.853404641698866e-06,
+      "loss": 0.6416,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7000772300008768,
+      "learning_rate": 9.85299972090527e-06,
+      "loss": 0.6848,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7492572101693626,
+      "learning_rate": 9.852594249996494e-06,
+      "loss": 0.5989,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9475380721456124,
+      "learning_rate": 9.852188229018502e-06,
+      "loss": 0.6094,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.923991064008745,
+      "learning_rate": 9.851781658017317e-06,
+      "loss": 0.6641,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2032915876519743,
+      "learning_rate": 9.851374537039027e-06,
+      "loss": 0.6064,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.112853405985047,
+      "learning_rate": 9.850966866129779e-06,
+      "loss": 0.6987,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.257447982728473,
+      "learning_rate": 9.85055864533579e-06,
+      "loss": 0.6616,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9407354984677891,
+      "learning_rate": 9.85014987470333e-06,
+      "loss": 0.618,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0091396799893797,
+      "learning_rate": 9.849740554278738e-06,
+      "loss": 0.7182,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0832354129855615,
+      "learning_rate": 9.849330684108409e-06,
+      "loss": 0.6517,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1550178340245205,
+      "learning_rate": 9.848920264238809e-06,
+      "loss": 0.8038,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8398166027965528,
+      "learning_rate": 9.84850929471646e-06,
+      "loss": 0.5977,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4712742132470282,
+      "learning_rate": 9.848097775587944e-06,
+      "loss": 0.7969,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9800854460786408,
+      "learning_rate": 9.847685706899913e-06,
+      "loss": 0.6424,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.160887630420676,
+      "learning_rate": 9.847273088699077e-06,
+      "loss": 0.5793,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7402490160067203,
+      "learning_rate": 9.846859921032207e-06,
+      "loss": 0.6306,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9858410317423765,
+      "learning_rate": 9.846446203946139e-06,
+      "loss": 0.6972,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7681585176033914,
+      "learning_rate": 9.84603193748777e-06,
+      "loss": 0.689,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8047938502571725,
+      "learning_rate": 9.84561712170406e-06,
+      "loss": 0.5521,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0028920908430434,
+      "learning_rate": 9.84520175664203e-06,
+      "loss": 0.693,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.074323617732677,
+      "learning_rate": 9.844785842348764e-06,
+      "loss": 0.7734,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7873634371638705,
+      "learning_rate": 9.844369378871409e-06,
+      "loss": 0.5996,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8578006878258289,
+      "learning_rate": 9.84395236625717e-06,
+      "loss": 0.5908,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9206170170554828,
+      "learning_rate": 9.843534804553323e-06,
+      "loss": 0.6535,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8725064621105443,
+      "learning_rate": 9.843116693807199e-06,
+      "loss": 0.5975,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2188161598759315,
+      "learning_rate": 9.842698034066192e-06,
+      "loss": 0.7365,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.921338516012938,
+      "learning_rate": 9.842278825377761e-06,
+      "loss": 0.7024,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2716226482375554,
+      "learning_rate": 9.841859067789425e-06,
+      "loss": 0.6117,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8260022112754144,
+      "learning_rate": 9.841438761348766e-06,
+      "loss": 0.5538,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8908219800095016,
+      "learning_rate": 9.841017906103427e-06,
+      "loss": 0.7356,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.021688449597609,
+      "learning_rate": 9.840596502101117e-06,
+      "loss": 0.6077,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1832877381909452,
+      "learning_rate": 9.840174549389603e-06,
+      "loss": 0.6493,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.174239711824139,
+      "learning_rate": 9.839752048016714e-06,
+      "loss": 0.633,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0734549390443435,
+      "learning_rate": 9.839328998030347e-06,
+      "loss": 0.7503,
+      "step": 1044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7587601682627316,
+      "learning_rate": 9.838905399478453e-06,
+      "loss": 0.6627,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.254845491574179,
+      "learning_rate": 9.838481252409053e-06,
+      "loss": 0.7206,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0682471001717357,
+      "learning_rate": 9.838056556870223e-06,
+      "loss": 0.6788,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9779991745810939,
+      "learning_rate": 9.837631312910107e-06,
+      "loss": 0.6749,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9718255224988028,
+      "learning_rate": 9.837205520576907e-06,
+      "loss": 0.6956,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8145469518157462,
+      "learning_rate": 9.836779179918891e-06,
+      "loss": 0.6854,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9375617890849886,
+      "learning_rate": 9.836352290984386e-06,
+      "loss": 0.6184,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9338031264153726,
+      "learning_rate": 9.835924853821783e-06,
+      "loss": 0.5955,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9522052631485587,
+      "learning_rate": 9.835496868479533e-06,
+      "loss": 0.5681,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8700014507217293,
+      "learning_rate": 9.835068335006153e-06,
+      "loss": 0.6794,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9556206221572114,
+      "learning_rate": 9.834639253450217e-06,
+      "loss": 0.5931,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1358836379045387,
+      "learning_rate": 9.834209623860367e-06,
+      "loss": 0.7265,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8328538254509168,
+      "learning_rate": 9.8337794462853e-06,
+      "loss": 0.6213,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9064305026210864,
+      "learning_rate": 9.833348720773782e-06,
+      "loss": 0.6486,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.887679929354486,
+      "learning_rate": 9.832917447374637e-06,
+      "loss": 0.7484,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.012238919175227,
+      "learning_rate": 9.832485626136751e-06,
+      "loss": 0.6625,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0410737340872145,
+      "learning_rate": 9.832053257109077e-06,
+      "loss": 0.7055,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7772590663935988,
+      "learning_rate": 9.831620340340626e-06,
+      "loss": 0.6214,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8966041179438153,
+      "learning_rate": 9.831186875880467e-06,
+      "loss": 0.6835,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8203185091064622,
+      "learning_rate": 9.830752863777741e-06,
+      "loss": 0.6473,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.993896863641108,
+      "learning_rate": 9.830318304081642e-06,
+      "loss": 0.7101,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2042609661267836,
+      "learning_rate": 9.829883196841433e-06,
+      "loss": 0.6174,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0253946730893886,
+      "learning_rate": 9.829447542106434e-06,
+      "loss": 0.7096,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8318754462273392,
+      "learning_rate": 9.829011339926028e-06,
+      "loss": 0.662,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9474284559392272,
+      "learning_rate": 9.828574590349662e-06,
+      "loss": 0.7439,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.07339427770848,
+      "learning_rate": 9.828137293426844e-06,
+      "loss": 0.7443,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9755027213686065,
+      "learning_rate": 9.827699449207147e-06,
+      "loss": 0.6352,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9897415411397834,
+      "learning_rate": 9.827261057740198e-06,
+      "loss": 0.7521,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.04504258074226,
+      "learning_rate": 9.826822119075694e-06,
+      "loss": 0.7339,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2651880336477728,
+      "learning_rate": 9.826382633263392e-06,
+      "loss": 0.7575,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0751191341447877,
+      "learning_rate": 9.825942600353107e-06,
+      "loss": 0.6432,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9471220241261464,
+      "learning_rate": 9.825502020394724e-06,
+      "loss": 0.6165,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0197119177571525,
+      "learning_rate": 9.82506089343818e-06,
+      "loss": 0.6898,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2111513000318053,
+      "learning_rate": 9.824619219533482e-06,
+      "loss": 0.8225,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.072255616698154,
+      "learning_rate": 9.824176998730698e-06,
+      "loss": 0.5451,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0203232264892574,
+      "learning_rate": 9.823734231079953e-06,
+      "loss": 0.6907,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9978646542749907,
+      "learning_rate": 9.823290916631438e-06,
+      "loss": 0.7582,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.997655399908216,
+      "learning_rate": 9.822847055435407e-06,
+      "loss": 0.6675,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.600826888649854,
+      "learning_rate": 9.822402647542173e-06,
+      "loss": 0.6995,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8108675661821911,
+      "learning_rate": 9.82195769300211e-06,
+      "loss": 0.7427,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0866686825001666,
+      "learning_rate": 9.821512191865662e-06,
+      "loss": 0.653,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9633193896611827,
+      "learning_rate": 9.821066144183322e-06,
+      "loss": 0.6388,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7284744819995272,
+      "learning_rate": 9.820619550005656e-06,
+      "loss": 0.5978,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9204126261783792,
+      "learning_rate": 9.820172409383288e-06,
+      "loss": 0.6652,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9950911805464604,
+      "learning_rate": 9.819724722366903e-06,
+      "loss": 0.5597,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9588977885654488,
+      "learning_rate": 9.81927648900725e-06,
+      "loss": 0.6817,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9645877951323167,
+      "learning_rate": 9.818827709355138e-06,
+      "loss": 0.6927,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9913277378900867,
+      "learning_rate": 9.818378383461438e-06,
+      "loss": 0.5648,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.167093446329645,
+      "learning_rate": 9.817928511377085e-06,
+      "loss": 0.721,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.006202235369134,
+      "learning_rate": 9.817478093153074e-06,
+      "loss": 0.6331,
+      "step": 1094
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8430866473192573,
+      "learning_rate": 9.817027128840462e-06,
+      "loss": 0.6632,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1072808477952307,
+      "learning_rate": 9.816575618490368e-06,
+      "loss": 0.677,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8674328585658524,
+      "learning_rate": 9.816123562153975e-06,
+      "loss": 0.6894,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9548566740415998,
+      "learning_rate": 9.815670959882526e-06,
+      "loss": 0.6354,
+      "step": 1098
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.958662754789633,
+      "learning_rate": 9.815217811727325e-06,
+      "loss": 0.6949,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0309136527346965,
+      "learning_rate": 9.814764117739737e-06,
+      "loss": 0.6162,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.866529266557843,
+      "learning_rate": 9.814309877971195e-06,
+      "loss": 0.7048,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7941888834949336,
+      "learning_rate": 9.813855092473189e-06,
+      "loss": 0.6341,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8740743463137512,
+      "learning_rate": 9.813399761297267e-06,
+      "loss": 0.6796,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9916876162956634,
+      "learning_rate": 9.81294388449505e-06,
+      "loss": 0.7041,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.328375158894037,
+      "learning_rate": 9.812487462118207e-06,
+      "loss": 0.7484,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0851750400015714,
+      "learning_rate": 9.812030494218484e-06,
+      "loss": 0.6148,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0721317087262685,
+      "learning_rate": 9.811572980847674e-06,
+      "loss": 0.6666,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3735592824344534,
+      "learning_rate": 9.811114922057642e-06,
+      "loss": 0.7982,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0309673267983404,
+      "learning_rate": 9.810656317900312e-06,
+      "loss": 0.5905,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9784198177751062,
+      "learning_rate": 9.810197168427667e-06,
+      "loss": 0.7343,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8042340469617641,
+      "learning_rate": 9.809737473691758e-06,
+      "loss": 0.6254,
+      "step": 1111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0968682452666974,
+      "learning_rate": 9.80927723374469e-06,
+      "loss": 0.6969,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.354904562448494,
+      "learning_rate": 9.808816448638636e-06,
+      "loss": 0.7833,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.163126833363332,
+      "learning_rate": 9.808355118425827e-06,
+      "loss": 0.6467,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0346429879262433,
+      "learning_rate": 9.807893243158562e-06,
+      "loss": 0.6335,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1759311495573788,
+      "learning_rate": 9.80743082288919e-06,
+      "loss": 0.7353,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8028313593352352,
+      "learning_rate": 9.806967857670135e-06,
+      "loss": 0.609,
+      "step": 1117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7536865880903942,
+      "learning_rate": 9.806504347553874e-06,
+      "loss": 0.6642,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9892546388260939,
+      "learning_rate": 9.80604029259295e-06,
+      "loss": 0.7288,
+      "step": 1119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9316832882932709,
+      "learning_rate": 9.805575692839964e-06,
+      "loss": 0.5975,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8501490970207397,
+      "learning_rate": 9.805110548347583e-06,
+      "loss": 0.7319,
+      "step": 1121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9680500445329732,
+      "learning_rate": 9.804644859168534e-06,
+      "loss": 0.697,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.109835278999002,
+      "learning_rate": 9.804178625355602e-06,
+      "loss": 0.6059,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.145342051858608,
+      "learning_rate": 9.803711846961641e-06,
+      "loss": 0.7171,
+      "step": 1124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.088705701522735,
+      "learning_rate": 9.803244524039564e-06,
+      "loss": 0.745,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9112491216751644,
+      "learning_rate": 9.802776656642341e-06,
+      "loss": 0.6755,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.317044885396738,
+      "learning_rate": 9.80230824482301e-06,
+      "loss": 0.7217,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.359531160259817,
+      "learning_rate": 9.801839288634664e-06,
+      "loss": 0.7032,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8376679215636302,
+      "learning_rate": 9.801369788130468e-06,
+      "loss": 0.5589,
+      "step": 1129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.038462127646075,
+      "learning_rate": 9.800899743363638e-06,
+      "loss": 0.5897,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.958505348413095,
+      "learning_rate": 9.80042915438746e-06,
+      "loss": 0.6568,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9417917565556433,
+      "learning_rate": 9.799958021255275e-06,
+      "loss": 0.7257,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9224737730859809,
+      "learning_rate": 9.799486344020488e-06,
+      "loss": 0.6253,
+      "step": 1133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8633516811392656,
+      "learning_rate": 9.79901412273657e-06,
+      "loss": 0.6349,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.020034769094531,
+      "learning_rate": 9.798541357457045e-06,
+      "loss": 0.5633,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9064146253836125,
+      "learning_rate": 9.79806804823551e-06,
+      "loss": 0.6964,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9919505999551952,
+      "learning_rate": 9.797594195125611e-06,
+      "loss": 0.5642,
+      "step": 1137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.09562120886189,
+      "learning_rate": 9.797119798181066e-06,
+      "loss": 0.6715,
+      "step": 1138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8738420808946492,
+      "learning_rate": 9.796644857455648e-06,
+      "loss": 0.6187,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.072085736447773,
+      "learning_rate": 9.796169373003194e-06,
+      "loss": 0.7333,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0711394615651577,
+      "learning_rate": 9.795693344877609e-06,
+      "loss": 0.697,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9046229443546778,
+      "learning_rate": 9.795216773132846e-06,
+      "loss": 0.7762,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9207903000528437,
+      "learning_rate": 9.794739657822929e-06,
+      "loss": 0.6793,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.782458495479854,
+      "learning_rate": 9.794261999001944e-06,
+      "loss": 0.7346,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7754725000419709,
+      "learning_rate": 9.793783796724033e-06,
+      "loss": 0.5992,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8592470716968033,
+      "learning_rate": 9.793305051043407e-06,
+      "loss": 0.7468,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.857487138917833,
+      "learning_rate": 9.792825762014333e-06,
+      "loss": 0.6613,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8265947454383122,
+      "learning_rate": 9.79234592969114e-06,
+      "loss": 0.6637,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9552167151750668,
+      "learning_rate": 9.79186555412822e-06,
+      "loss": 0.7084,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9545771112614327,
+      "learning_rate": 9.791384635380028e-06,
+      "loss": 0.7181,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.958640881269104,
+      "learning_rate": 9.790903173501075e-06,
+      "loss": 0.6016,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9827078690692417,
+      "learning_rate": 9.790421168545942e-06,
+      "loss": 0.7443,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.133727647342895,
+      "learning_rate": 9.789938620569265e-06,
+      "loss": 0.731,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2672150699453457,
+      "learning_rate": 9.789455529625743e-06,
+      "loss": 0.6824,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.91456235430189,
+      "learning_rate": 9.788971895770138e-06,
+      "loss": 0.6286,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7757643292986403,
+      "learning_rate": 9.788487719057273e-06,
+      "loss": 0.614,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1250419908716687,
+      "learning_rate": 9.78800299954203e-06,
+      "loss": 0.6005,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8486976022057549,
+      "learning_rate": 9.787517737279357e-06,
+      "loss": 0.6048,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3506355338274845,
+      "learning_rate": 9.787031932324262e-06,
+      "loss": 0.7821,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8385308824731086,
+      "learning_rate": 9.78654558473181e-06,
+      "loss": 0.7732,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8744460433610264,
+      "learning_rate": 9.786058694557136e-06,
+      "loss": 0.7004,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7990251826604449,
+      "learning_rate": 9.785571261855429e-06,
+      "loss": 0.6008,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0387781703421046,
+      "learning_rate": 9.78508328668194e-06,
+      "loss": 0.7354,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8379294386830753,
+      "learning_rate": 9.784594769091989e-06,
+      "loss": 0.6562,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.951858093945663,
+      "learning_rate": 9.78410570914095e-06,
+      "loss": 0.5601,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.963636595805986,
+      "learning_rate": 9.783616106884258e-06,
+      "loss": 0.7216,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0298983330028193,
+      "learning_rate": 9.783125962377416e-06,
+      "loss": 0.6501,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.273467968545706,
+      "learning_rate": 9.782635275675983e-06,
+      "loss": 0.7842,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.872811853225732,
+      "learning_rate": 9.78214404683558e-06,
+      "loss": 0.6591,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8397528593162067,
+      "learning_rate": 9.781652275911894e-06,
+      "loss": 0.6264,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8859285853928063,
+      "learning_rate": 9.781159962960667e-06,
+      "loss": 0.5893,
+      "step": 1171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.905973149180084,
+      "learning_rate": 9.780667108037706e-06,
+      "loss": 0.6312,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9716374922186313,
+      "learning_rate": 9.78017371119888e-06,
+      "loss": 0.6401,
+      "step": 1173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9769244317036858,
+      "learning_rate": 9.77967977250012e-06,
+      "loss": 0.6444,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0872679872136737,
+      "learning_rate": 9.779185291997412e-06,
+      "loss": 0.761,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9107214967319497,
+      "learning_rate": 9.778690269746811e-06,
+      "loss": 0.6832,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7386000757691096,
+      "learning_rate": 9.778194705804431e-06,
+      "loss": 0.619,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.181891137096523,
+      "learning_rate": 9.777698600226446e-06,
+      "loss": 0.7406,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.012111320499033,
+      "learning_rate": 9.777201953069092e-06,
+      "loss": 0.8356,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2266386752482403,
+      "learning_rate": 9.776704764388668e-06,
+      "loss": 0.7769,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0481630064459275,
+      "learning_rate": 9.776207034241534e-06,
+      "loss": 0.706,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1147866234960637,
+      "learning_rate": 9.775708762684106e-06,
+      "loss": 0.6764,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9980720033432766,
+      "learning_rate": 9.775209949772872e-06,
+      "loss": 0.5919,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9142784484222324,
+      "learning_rate": 9.774710595564371e-06,
+      "loss": 0.7484,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0104955568224896,
+      "learning_rate": 9.774210700115209e-06,
+      "loss": 0.6362,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.060715102831267,
+      "learning_rate": 9.773710263482053e-06,
+      "loss": 0.8285,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.375817415064652,
+      "learning_rate": 9.77320928572163e-06,
+      "loss": 0.6754,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9904053183850345,
+      "learning_rate": 9.772707766890726e-06,
+      "loss": 0.5919,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8189149884702953,
+      "learning_rate": 9.772205707046194e-06,
+      "loss": 0.6269,
+      "step": 1189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0893307362744165,
+      "learning_rate": 9.771703106244945e-06,
+      "loss": 0.6025,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.872217178188383,
+      "learning_rate": 9.77119996454395e-06,
+      "loss": 0.6917,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8516747196972398,
+      "learning_rate": 9.770696282000245e-06,
+      "loss": 0.6101,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7350140128893028,
+      "learning_rate": 9.770192058670924e-06,
+      "loss": 0.7053,
+      "step": 1193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.124522427032116,
+      "learning_rate": 9.769687294613143e-06,
+      "loss": 0.7225,
+      "step": 1194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6443356705537862,
+      "learning_rate": 9.769181989884123e-06,
+      "loss": 0.6327,
+      "step": 1195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8787937842514755,
+      "learning_rate": 9.76867614454114e-06,
+      "loss": 0.685,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.054491694078201,
+      "learning_rate": 9.768169758641535e-06,
+      "loss": 0.6185,
+      "step": 1197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0466021116932374,
+      "learning_rate": 9.767662832242711e-06,
+      "loss": 0.722,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9725740768686584,
+      "learning_rate": 9.76715536540213e-06,
+      "loss": 0.6419,
+      "step": 1199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8607471593989677,
+      "learning_rate": 9.766647358177317e-06,
+      "loss": 0.5834,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8176363699369578,
+      "learning_rate": 9.766138810625855e-06,
+      "loss": 0.5901,
+      "step": 1201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.902595364021203,
+      "learning_rate": 9.765629722805394e-06,
+      "loss": 0.642,
+      "step": 1202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9999598335834559,
+      "learning_rate": 9.765120094773641e-06,
+      "loss": 0.7298,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8506077363965423,
+      "learning_rate": 9.764609926588365e-06,
+      "loss": 0.6875,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0798876015732035,
+      "learning_rate": 9.764099218307396e-06,
+      "loss": 0.7775,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8982097105923708,
+      "learning_rate": 9.763587969988626e-06,
+      "loss": 0.6821,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.051160655228281,
+      "learning_rate": 9.763076181690008e-06,
+      "loss": 0.6225,
+      "step": 1207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7605857993358935,
+      "learning_rate": 9.762563853469557e-06,
+      "loss": 0.6409,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9105078969357518,
+      "learning_rate": 9.762050985385348e-06,
+      "loss": 0.5791,
+      "step": 1209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7956662172842133,
+      "learning_rate": 9.761537577495514e-06,
+      "loss": 0.6378,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9382003285098017,
+      "learning_rate": 9.761023629858258e-06,
+      "loss": 0.5976,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7644171198864156,
+      "learning_rate": 9.760509142531837e-06,
+      "loss": 0.6155,
+      "step": 1212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0937484763695258,
+      "learning_rate": 9.759994115574571e-06,
+      "loss": 0.7084,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.145729104645392,
+      "learning_rate": 9.759478549044839e-06,
+      "loss": 0.6385,
+      "step": 1214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7488025165240058,
+      "learning_rate": 9.758962443001086e-06,
+      "loss": 0.6194,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.008568905377275,
+      "learning_rate": 9.758445797501816e-06,
+      "loss": 0.5682,
+      "step": 1216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.998000158709324,
+      "learning_rate": 9.757928612605592e-06,
+      "loss": 0.6302,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.929368096839565,
+      "learning_rate": 9.75741088837104e-06,
+      "loss": 0.7117,
+      "step": 1218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7873401858196971,
+      "learning_rate": 9.756892624856848e-06,
+      "loss": 0.5734,
+      "step": 1219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8865122527560079,
+      "learning_rate": 9.756373822121762e-06,
+      "loss": 0.6212,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0688832226726075,
+      "learning_rate": 9.755854480224596e-06,
+      "loss": 0.6526,
+      "step": 1221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8863437615329601,
+      "learning_rate": 9.755334599224215e-06,
+      "loss": 0.6192,
+      "step": 1222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0972603354248753,
+      "learning_rate": 9.754814179179552e-06,
+      "loss": 0.6935,
+      "step": 1223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8672839080556047,
+      "learning_rate": 9.754293220149602e-06,
+      "loss": 0.6415,
+      "step": 1224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1056721713435014,
+      "learning_rate": 9.753771722193417e-06,
+      "loss": 0.6053,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9996835954094754,
+      "learning_rate": 9.753249685370112e-06,
+      "loss": 0.7002,
+      "step": 1226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.183109070932942,
+      "learning_rate": 9.752727109738859e-06,
+      "loss": 0.6372,
+      "step": 1227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9844165571120553,
+      "learning_rate": 9.752203995358902e-06,
+      "loss": 0.6659,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.910791627714843,
+      "learning_rate": 9.751680342289536e-06,
+      "loss": 0.6214,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9193510028750618,
+      "learning_rate": 9.751156150590117e-06,
+      "loss": 0.6494,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2006190540580146,
+      "learning_rate": 9.750631420320069e-06,
+      "loss": 0.6964,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8660242845944797,
+      "learning_rate": 9.750106151538871e-06,
+      "loss": 0.6691,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7139198365306583,
+      "learning_rate": 9.749580344306067e-06,
+      "loss": 0.5357,
+      "step": 1233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9794675828758101,
+      "learning_rate": 9.74905399868126e-06,
+      "loss": 0.6414,
+      "step": 1234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.053972284777646,
+      "learning_rate": 9.748527114724111e-06,
+      "loss": 0.6585,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8160635181241953,
+      "learning_rate": 9.74799969249435e-06,
+      "loss": 0.5822,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9594988518586947,
+      "learning_rate": 9.74747173205176e-06,
+      "loss": 0.6526,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1104409643045696,
+      "learning_rate": 9.74694323345619e-06,
+      "loss": 0.6936,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9216660825585377,
+      "learning_rate": 9.746414196767548e-06,
+      "loss": 0.6798,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1185106363995607,
+      "learning_rate": 9.745884622045803e-06,
+      "loss": 0.6708,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7835909517236361,
+      "learning_rate": 9.745354509350983e-06,
+      "loss": 0.7385,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3602417280460326,
+      "learning_rate": 9.744823858743186e-06,
+      "loss": 0.643,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.837677043941556,
+      "learning_rate": 9.744292670282557e-06,
+      "loss": 0.6202,
+      "step": 1243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9538243258295147,
+      "learning_rate": 9.743760944029313e-06,
+      "loss": 0.6141,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7094725933192672,
+      "learning_rate": 9.743228680043729e-06,
+      "loss": 0.5873,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.898722893037348,
+      "learning_rate": 9.742695878386136e-06,
+      "loss": 0.6622,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3475080247477265,
+      "learning_rate": 9.742162539116936e-06,
+      "loss": 0.6986,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8734682037675061,
+      "learning_rate": 9.74162866229658e-06,
+      "loss": 0.6931,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8891784901953723,
+      "learning_rate": 9.74109424798559e-06,
+      "loss": 0.6264,
+      "step": 1249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0263925697740683,
+      "learning_rate": 9.740559296244543e-06,
+      "loss": 0.6856,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2614934669273516,
+      "learning_rate": 9.74002380713408e-06,
+      "loss": 0.6658,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9590626287273527,
+      "learning_rate": 9.7394877807149e-06,
+      "loss": 0.6018,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8968171625714494,
+      "learning_rate": 9.738951217047767e-06,
+      "loss": 0.6378,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7508689880032957,
+      "learning_rate": 9.738414116193503e-06,
+      "loss": 0.7191,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1485846944757117,
+      "learning_rate": 9.737876478212989e-06,
+      "loss": 0.7119,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0333547103913663,
+      "learning_rate": 9.737338303167173e-06,
+      "loss": 0.6638,
+      "step": 1256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8862361580061715,
+      "learning_rate": 9.736799591117057e-06,
+      "loss": 0.7242,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.654552718644028,
+      "learning_rate": 9.73626034212371e-06,
+      "loss": 0.6576,
+      "step": 1258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8740868267530033,
+      "learning_rate": 9.735720556248256e-06,
+      "loss": 0.7337,
+      "step": 1259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9350463131930447,
+      "learning_rate": 9.735180233551884e-06,
+      "loss": 0.823,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0570724020864093,
+      "learning_rate": 9.734639374095845e-06,
+      "loss": 0.6597,
+      "step": 1261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8547501290708488,
+      "learning_rate": 9.734097977941446e-06,
+      "loss": 0.6362,
+      "step": 1262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9035543204587644,
+      "learning_rate": 9.733556045150057e-06,
+      "loss": 0.6648,
+      "step": 1263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0108959931171557,
+      "learning_rate": 9.73301357578311e-06,
+      "loss": 0.7228,
+      "step": 1264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.865013262710532,
+      "learning_rate": 9.732470569902097e-06,
+      "loss": 0.7365,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.217716624827927,
+      "learning_rate": 9.731927027568569e-06,
+      "loss": 0.7753,
+      "step": 1266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7979547620947482,
+      "learning_rate": 9.731382948844143e-06,
+      "loss": 0.5906,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7508977673648394,
+      "learning_rate": 9.730838333790493e-06,
+      "loss": 0.6117,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6893969959243762,
+      "learning_rate": 9.73029318246935e-06,
+      "loss": 0.6358,
+      "step": 1269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.166240330404318,
+      "learning_rate": 9.729747494942515e-06,
+      "loss": 0.6924,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.000608514191742,
+      "learning_rate": 9.729201271271842e-06,
+      "loss": 0.7347,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.103793789959187,
+      "learning_rate": 9.728654511519251e-06,
+      "loss": 0.7285,
+      "step": 1272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8409733568718099,
+      "learning_rate": 9.728107215746717e-06,
+      "loss": 0.692,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.963562475533344,
+      "learning_rate": 9.72755938401628e-06,
+      "loss": 0.6062,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8176381153691845,
+      "learning_rate": 9.727011016390044e-06,
+      "loss": 0.7648,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8358596377570153,
+      "learning_rate": 9.726462112930165e-06,
+      "loss": 0.6272,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8482133621035386,
+      "learning_rate": 9.725912673698863e-06,
+      "loss": 0.6437,
+      "step": 1277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7093243741213524,
+      "learning_rate": 9.725362698758425e-06,
+      "loss": 0.5643,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9313380353034446,
+      "learning_rate": 9.72481218817119e-06,
+      "loss": 0.7326,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0110578630440696,
+      "learning_rate": 9.724261141999564e-06,
+      "loss": 0.7809,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0212582494727056,
+      "learning_rate": 9.723709560306009e-06,
+      "loss": 0.7539,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.956126223537327,
+      "learning_rate": 9.723157443153053e-06,
+      "loss": 0.6653,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8182910112574044,
+      "learning_rate": 9.722604790603279e-06,
+      "loss": 0.7183,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.108210363465132,
+      "learning_rate": 9.722051602719333e-06,
+      "loss": 0.712,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9131048983229353,
+      "learning_rate": 9.721497879563924e-06,
+      "loss": 0.7338,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.949411903606035,
+      "learning_rate": 9.72094362119982e-06,
+      "loss": 0.6708,
+      "step": 1286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0746178969752416,
+      "learning_rate": 9.720388827689843e-06,
+      "loss": 0.563,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.190230314831356,
+      "learning_rate": 9.719833499096891e-06,
+      "loss": 0.6351,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8460044124762562,
+      "learning_rate": 9.71927763548391e-06,
+      "loss": 0.6958,
+      "step": 1289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6752039016736493,
+      "learning_rate": 9.718721236913909e-06,
+      "loss": 0.5744,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9927244031221076,
+      "learning_rate": 9.718164303449961e-06,
+      "loss": 0.7696,
+      "step": 1291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8995517099170962,
+      "learning_rate": 9.717606835155195e-06,
+      "loss": 0.6031,
+      "step": 1292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.940650266443028,
+      "learning_rate": 9.717048832092806e-06,
+      "loss": 0.6741,
+      "step": 1293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.578177156458987,
+      "learning_rate": 9.716490294326046e-06,
+      "loss": 0.5341,
+      "step": 1294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3682750170322975,
+      "learning_rate": 9.715931221918227e-06,
+      "loss": 0.6361,
+      "step": 1295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8435349334772175,
+      "learning_rate": 9.715371614932725e-06,
+      "loss": 0.6257,
+      "step": 1296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9264375579621833,
+      "learning_rate": 9.714811473432973e-06,
+      "loss": 0.7395,
+      "step": 1297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9461581602936884,
+      "learning_rate": 9.714250797482468e-06,
+      "loss": 0.6327,
+      "step": 1298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1144212881507425,
+      "learning_rate": 9.713689587144762e-06,
+      "loss": 0.6785,
+      "step": 1299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6928260272004876,
+      "learning_rate": 9.713127842483476e-06,
+      "loss": 0.5079,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3246045980360783,
+      "learning_rate": 9.712565563562286e-06,
+      "loss": 0.7336,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7227636255508532,
+      "learning_rate": 9.712002750444926e-06,
+      "loss": 0.7164,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8982343901775387,
+      "learning_rate": 9.7114394031952e-06,
+      "loss": 0.708,
+      "step": 1303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1258743453940028,
+      "learning_rate": 9.71087552187696e-06,
+      "loss": 0.7317,
+      "step": 1304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9721335267277356,
+      "learning_rate": 9.710311106554132e-06,
+      "loss": 0.6371,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7625782159118968,
+      "learning_rate": 9.70974615729069e-06,
+      "loss": 0.5839,
+      "step": 1306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4807222261088127,
+      "learning_rate": 9.709180674150677e-06,
+      "loss": 0.7081,
+      "step": 1307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9903239990967319,
+      "learning_rate": 9.708614657198194e-06,
+      "loss": 0.6491,
+      "step": 1308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.925530757651826,
+      "learning_rate": 9.7080481064974e-06,
+      "loss": 0.7496,
+      "step": 1309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9292533506628393,
+      "learning_rate": 9.70748102211252e-06,
+      "loss": 0.6853,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9690106352432073,
+      "learning_rate": 9.706913404107832e-06,
+      "loss": 0.5786,
+      "step": 1311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0319183887210044,
+      "learning_rate": 9.706345252547681e-06,
+      "loss": 0.6004,
+      "step": 1312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9599275382294734,
+      "learning_rate": 9.705776567496473e-06,
+      "loss": 0.7236,
+      "step": 1313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9230052599444754,
+      "learning_rate": 9.705207349018668e-06,
+      "loss": 0.708,
+      "step": 1314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8972194347454603,
+      "learning_rate": 9.70463759717879e-06,
+      "loss": 0.716,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.010675575041699,
+      "learning_rate": 9.704067312041426e-06,
+      "loss": 0.6905,
+      "step": 1316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2746498328993505,
+      "learning_rate": 9.703496493671219e-06,
+      "loss": 0.5917,
+      "step": 1317
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9221483056550144,
+      "learning_rate": 9.702925142132876e-06,
+      "loss": 0.6797,
+      "step": 1318
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0044204535629047,
+      "learning_rate": 9.702353257491162e-06,
+      "loss": 0.7663,
+      "step": 1319
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.133217438263875,
+      "learning_rate": 9.701780839810903e-06,
+      "loss": 0.7136,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.84335445808742,
+      "learning_rate": 9.701207889156989e-06,
+      "loss": 0.6814,
+      "step": 1321
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.984334565307144,
+      "learning_rate": 9.700634405594364e-06,
+      "loss": 0.6403,
+      "step": 1322
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.827908300013016,
+      "learning_rate": 9.700060389188035e-06,
+      "loss": 0.6725,
+      "step": 1323
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.009494835552701,
+      "learning_rate": 9.699485840003072e-06,
+      "loss": 0.6841,
+      "step": 1324
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.91704570730664,
+      "learning_rate": 9.698910758104603e-06,
+      "loss": 0.6762,
+      "step": 1325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.005127161184001,
+      "learning_rate": 9.698335143557818e-06,
+      "loss": 0.7461,
+      "step": 1326
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1535407683860095,
+      "learning_rate": 9.697758996427962e-06,
+      "loss": 0.7321,
+      "step": 1327
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9080003028880403,
+      "learning_rate": 9.69718231678035e-06,
+      "loss": 0.672,
+      "step": 1328
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9437619321838724,
+      "learning_rate": 9.696605104680348e-06,
+      "loss": 0.7428,
+      "step": 1329
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.037382120827923,
+      "learning_rate": 9.696027360193387e-06,
+      "loss": 0.7703,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8931689172830715,
+      "learning_rate": 9.69544908338496e-06,
+      "loss": 0.6491,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8790236237302318,
+      "learning_rate": 9.694870274320616e-06,
+      "loss": 0.6525,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.910244499985207,
+      "learning_rate": 9.694290933065966e-06,
+      "loss": 0.6577,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9579528999546667,
+      "learning_rate": 9.693711059686682e-06,
+      "loss": 0.6902,
+      "step": 1334
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9396330418699732,
+      "learning_rate": 9.693130654248497e-06,
+      "loss": 0.7542,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9042496696960487,
+      "learning_rate": 9.692549716817202e-06,
+      "loss": 0.5711,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8895202592925349,
+      "learning_rate": 9.691968247458648e-06,
+      "loss": 0.6255,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0965642202557544,
+      "learning_rate": 9.691386246238753e-06,
+      "loss": 0.7484,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8015674513354294,
+      "learning_rate": 9.690803713223485e-06,
+      "loss": 0.6473,
+      "step": 1339
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9184186707119182,
+      "learning_rate": 9.690220648478879e-06,
+      "loss": 0.5988,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9770192689507424,
+      "learning_rate": 9.689637052071031e-06,
+      "loss": 0.7,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0449862402255508,
+      "learning_rate": 9.689052924066091e-06,
+      "loss": 0.6675,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8859554632950104,
+      "learning_rate": 9.688468264530278e-06,
+      "loss": 0.676,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.006052287719628,
+      "learning_rate": 9.68788307352986e-06,
+      "loss": 0.7827,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8945746237944512,
+      "learning_rate": 9.687297351131179e-06,
+      "loss": 0.7134,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.02227847196895,
+      "learning_rate": 9.686711097400625e-06,
+      "loss": 0.7382,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.878484044957481,
+      "learning_rate": 9.686124312404656e-06,
+      "loss": 0.7726,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.630184078610918,
+      "learning_rate": 9.685536996209785e-06,
+      "loss": 0.5358,
+      "step": 1348
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.85620192643215,
+      "learning_rate": 9.68494914888259e-06,
+      "loss": 0.6286,
+      "step": 1349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0114268912492403,
+      "learning_rate": 9.684360770489704e-06,
+      "loss": 0.6904,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6746633261022315,
+      "learning_rate": 9.683771861097825e-06,
+      "loss": 0.6124,
+      "step": 1351
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9084737163323533,
+      "learning_rate": 9.68318242077371e-06,
+      "loss": 0.6784,
+      "step": 1352
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.053592952699433,
+      "learning_rate": 9.682592449584174e-06,
+      "loss": 0.7364,
+      "step": 1353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8759874911112053,
+      "learning_rate": 9.682001947596093e-06,
+      "loss": 0.7605,
+      "step": 1354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.051882469686672,
+      "learning_rate": 9.681410914876406e-06,
+      "loss": 0.6258,
+      "step": 1355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8668394817877236,
+      "learning_rate": 9.68081935149211e-06,
+      "loss": 0.6545,
+      "step": 1356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.151948478937282,
+      "learning_rate": 9.680227257510257e-06,
+      "loss": 0.7088,
+      "step": 1357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.896198321769729,
+      "learning_rate": 9.67963463299797e-06,
+      "loss": 0.6797,
+      "step": 1358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8333904019961873,
+      "learning_rate": 9.679041478022424e-06,
+      "loss": 0.6455,
+      "step": 1359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.920785330687945,
+      "learning_rate": 9.678447792650858e-06,
+      "loss": 0.7701,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9740522491158696,
+      "learning_rate": 9.677853576950567e-06,
+      "loss": 0.6122,
+      "step": 1361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0093433073912195,
+      "learning_rate": 9.677258830988911e-06,
+      "loss": 0.688,
+      "step": 1362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9899501677577536,
+      "learning_rate": 9.676663554833307e-06,
+      "loss": 0.6969,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0662149446869846,
+      "learning_rate": 9.676067748551232e-06,
+      "loss": 0.7624,
+      "step": 1364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8510578010237781,
+      "learning_rate": 9.675471412210225e-06,
+      "loss": 0.6667,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2586481507249565,
+      "learning_rate": 9.674874545877886e-06,
+      "loss": 0.7025,
+      "step": 1366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.774087072360693,
+      "learning_rate": 9.674277149621869e-06,
+      "loss": 0.7498,
+      "step": 1367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9300929758419567,
+      "learning_rate": 9.673679223509895e-06,
+      "loss": 0.7127,
+      "step": 1368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0352754562115294,
+      "learning_rate": 9.673080767609743e-06,
+      "loss": 0.6448,
+      "step": 1369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8977730151853256,
+      "learning_rate": 9.67248178198925e-06,
+      "loss": 0.6987,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.004055084209392,
+      "learning_rate": 9.671882266716315e-06,
+      "loss": 0.8061,
+      "step": 1371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9454168688791278,
+      "learning_rate": 9.671282221858897e-06,
+      "loss": 0.7274,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9198819052306686,
+      "learning_rate": 9.670681647485012e-06,
+      "loss": 0.6711,
+      "step": 1373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7097410217988573,
+      "learning_rate": 9.670080543662742e-06,
+      "loss": 0.6368,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.050598473467362,
+      "learning_rate": 9.669478910460222e-06,
+      "loss": 0.7401,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9465646412985562,
+      "learning_rate": 9.668876747945652e-06,
+      "loss": 0.6021,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9335106076774247,
+      "learning_rate": 9.668274056187293e-06,
+      "loss": 0.6118,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.022082750564337,
+      "learning_rate": 9.66767083525346e-06,
+      "loss": 0.6889,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.755623299169476,
+      "learning_rate": 9.667067085212533e-06,
+      "loss": 0.7057,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.928965446296919,
+      "learning_rate": 9.66646280613295e-06,
+      "loss": 0.5745,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8155202857113888,
+      "learning_rate": 9.665857998083212e-06,
+      "loss": 0.6229,
+      "step": 1381
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7813070676468963,
+      "learning_rate": 9.665252661131874e-06,
+      "loss": 0.5898,
+      "step": 1382
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9563053298185136,
+      "learning_rate": 9.664646795347556e-06,
+      "loss": 0.7274,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9349658516984791,
+      "learning_rate": 9.664040400798937e-06,
+      "loss": 0.6404,
+      "step": 1384
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.188084590268582,
+      "learning_rate": 9.663433477554753e-06,
+      "loss": 0.6934,
+      "step": 1385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2197372636124917,
+      "learning_rate": 9.662826025683805e-06,
+      "loss": 0.6701,
+      "step": 1386
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9856243079185785,
+      "learning_rate": 9.662218045254949e-06,
+      "loss": 0.667,
+      "step": 1387
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9225996019695955,
+      "learning_rate": 9.661609536337104e-06,
+      "loss": 0.6174,
+      "step": 1388
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0817089630736727,
+      "learning_rate": 9.661000498999248e-06,
+      "loss": 0.7104,
+      "step": 1389
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8238863336509912,
+      "learning_rate": 9.660390933310418e-06,
+      "loss": 0.612,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8431042408210576,
+      "learning_rate": 9.659780839339713e-06,
+      "loss": 0.6929,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.889104137450392,
+      "learning_rate": 9.65917021715629e-06,
+      "loss": 0.6759,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9497779511731066,
+      "learning_rate": 9.658559066829365e-06,
+      "loss": 0.6041,
+      "step": 1393
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.039428108567977,
+      "learning_rate": 9.65794738842822e-06,
+      "loss": 0.7342,
+      "step": 1394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0211962356651734,
+      "learning_rate": 9.657335182022187e-06,
+      "loss": 0.7797,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9990273830284,
+      "learning_rate": 9.656722447680664e-06,
+      "loss": 0.7011,
+      "step": 1396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.88149406805676,
+      "learning_rate": 9.65610918547311e-06,
+      "loss": 0.6511,
+      "step": 1397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6989907788026,
+      "learning_rate": 9.655495395469042e-06,
+      "loss": 0.5867,
+      "step": 1398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9148308369176725,
+      "learning_rate": 9.654881077738035e-06,
+      "loss": 0.577,
+      "step": 1399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8599257790893318,
+      "learning_rate": 9.654266232349727e-06,
+      "loss": 0.64,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0076623853739903,
+      "learning_rate": 9.653650859373811e-06,
+      "loss": 0.7321,
+      "step": 1401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8313666092285423,
+      "learning_rate": 9.653034958880045e-06,
+      "loss": 0.6341,
+      "step": 1402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8722180403613942,
+      "learning_rate": 9.652418530938248e-06,
+      "loss": 0.7881,
+      "step": 1403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.083375770806738,
+      "learning_rate": 9.651801575618289e-06,
+      "loss": 0.7394,
+      "step": 1404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6813245697465693,
+      "learning_rate": 9.651184092990109e-06,
+      "loss": 0.5917,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8348700669336107,
+      "learning_rate": 9.650566083123701e-06,
+      "loss": 0.5931,
+      "step": 1406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0187798705574207,
+      "learning_rate": 9.64994754608912e-06,
+      "loss": 0.7104,
+      "step": 1407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.949254115187299,
+      "learning_rate": 9.649328481956481e-06,
+      "loss": 0.7475,
+      "step": 1408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9312680985383277,
+      "learning_rate": 9.64870889079596e-06,
+      "loss": 0.6517,
+      "step": 1409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.28060400124159,
+      "learning_rate": 9.648088772677787e-06,
+      "loss": 0.6683,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9498068405147828,
+      "learning_rate": 9.647468127672258e-06,
+      "loss": 0.6945,
+      "step": 1411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3803787498905944,
+      "learning_rate": 9.64684695584973e-06,
+      "loss": 0.7755,
+      "step": 1412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7106971204478838,
+      "learning_rate": 9.64622525728061e-06,
+      "loss": 0.6462,
+      "step": 1413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.720138728717716,
+      "learning_rate": 9.645603032035375e-06,
+      "loss": 0.5997,
+      "step": 1414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8333700801882125,
+      "learning_rate": 9.644980280184559e-06,
+      "loss": 0.6211,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9341434824630561,
+      "learning_rate": 9.644357001798752e-06,
+      "loss": 0.5569,
+      "step": 1416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9348393525012055,
+      "learning_rate": 9.643733196948607e-06,
+      "loss": 0.5925,
+      "step": 1417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8537207494480576,
+      "learning_rate": 9.643108865704836e-06,
+      "loss": 0.686,
+      "step": 1418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9794781788094675,
+      "learning_rate": 9.64248400813821e-06,
+      "loss": 0.6331,
+      "step": 1419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8934282311552093,
+      "learning_rate": 9.64185862431956e-06,
+      "loss": 0.6109,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9403239182134335,
+      "learning_rate": 9.641232714319777e-06,
+      "loss": 0.7185,
+      "step": 1421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.157701765972057,
+      "learning_rate": 9.640606278209812e-06,
+      "loss": 0.7755,
+      "step": 1422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6740846643105078,
+      "learning_rate": 9.639979316060675e-06,
+      "loss": 0.6013,
+      "step": 1423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.12677231177225,
+      "learning_rate": 9.639351827943436e-06,
+      "loss": 0.6144,
+      "step": 1424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.118780104812153,
+      "learning_rate": 9.638723813929224e-06,
+      "loss": 0.7439,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2357617900959728,
+      "learning_rate": 9.638095274089226e-06,
+      "loss": 0.6893,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7246115215231097,
+      "learning_rate": 9.637466208494694e-06,
+      "loss": 0.7229,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.01688121939328,
+      "learning_rate": 9.636836617216934e-06,
+      "loss": 0.6563,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.916960546551675,
+      "learning_rate": 9.636206500327316e-06,
+      "loss": 0.734,
+      "step": 1429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9670238441866768,
+      "learning_rate": 9.635575857897264e-06,
+      "loss": 0.8007,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9222282752835942,
+      "learning_rate": 9.634944689998267e-06,
+      "loss": 0.6885,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8951313109327579,
+      "learning_rate": 9.63431299670187e-06,
+      "loss": 0.5511,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.087254254458319,
+      "learning_rate": 9.633680778079682e-06,
+      "loss": 0.7344,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8839664631491326,
+      "learning_rate": 9.633048034203365e-06,
+      "loss": 0.7775,
+      "step": 1434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.90221214075658,
+      "learning_rate": 9.632414765144646e-06,
+      "loss": 0.6708,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.078581620152747,
+      "learning_rate": 9.631780970975311e-06,
+      "loss": 0.6322,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.83487929881022,
+      "learning_rate": 9.631146651767202e-06,
+      "loss": 0.669,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.832923326521618,
+      "learning_rate": 9.630511807592224e-06,
+      "loss": 0.6101,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1141323443224436,
+      "learning_rate": 9.629876438522338e-06,
+      "loss": 0.7117,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.06022961270754,
+      "learning_rate": 9.62924054462957e-06,
+      "loss": 0.7348,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8267609596216754,
+      "learning_rate": 9.628604125985999e-06,
+      "loss": 0.7431,
+      "step": 1441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9004596571143735,
+      "learning_rate": 9.627967182663768e-06,
+      "loss": 0.7026,
+      "step": 1442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0016223705105727,
+      "learning_rate": 9.627329714735079e-06,
+      "loss": 0.6671,
+      "step": 1443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.930460301179286,
+      "learning_rate": 9.626691722272193e-06,
+      "loss": 0.6102,
+      "step": 1444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9009166256026644,
+      "learning_rate": 9.626053205347428e-06,
+      "loss": 0.7215,
+      "step": 1445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.18241662087263,
+      "learning_rate": 9.625414164033163e-06,
+      "loss": 0.7554,
+      "step": 1446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7751878116739175,
+      "learning_rate": 9.62477459840184e-06,
+      "loss": 0.6397,
+      "step": 1447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.946828033661672,
+      "learning_rate": 9.624134508525957e-06,
+      "loss": 0.6376,
+      "step": 1448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.096187878367154,
+      "learning_rate": 9.623493894478069e-06,
+      "loss": 0.782,
+      "step": 1449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6619296901186194,
+      "learning_rate": 9.622852756330797e-06,
+      "loss": 0.6232,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.130541226400497,
+      "learning_rate": 9.622211094156812e-06,
+      "loss": 0.6484,
+      "step": 1451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9178708685945536,
+      "learning_rate": 9.621568908028857e-06,
+      "loss": 0.7487,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9169807937008756,
+      "learning_rate": 9.620926198019724e-06,
+      "loss": 0.6049,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8835439671046228,
+      "learning_rate": 9.620282964202267e-06,
+      "loss": 0.7385,
+      "step": 1454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.92405276657743,
+      "learning_rate": 9.619639206649402e-06,
+      "loss": 0.646,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9741432328020077,
+      "learning_rate": 9.618994925434103e-06,
+      "loss": 0.6466,
+      "step": 1456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0886770963919425,
+      "learning_rate": 9.618350120629398e-06,
+      "loss": 0.614,
+      "step": 1457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8307278823762825,
+      "learning_rate": 9.617704792308387e-06,
+      "loss": 0.6774,
+      "step": 1458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.003204931637325,
+      "learning_rate": 9.617058940544216e-06,
+      "loss": 0.6881,
+      "step": 1459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.716683988327422,
+      "learning_rate": 9.616412565410097e-06,
+      "loss": 0.644,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9389630819810084,
+      "learning_rate": 9.615765666979302e-06,
+      "loss": 0.6889,
+      "step": 1461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.734722035180772,
+      "learning_rate": 9.615118245325161e-06,
+      "loss": 0.6487,
+      "step": 1462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9708045285573386,
+      "learning_rate": 9.61447030052106e-06,
+      "loss": 0.7563,
+      "step": 1463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.045152715335636,
+      "learning_rate": 9.613821832640448e-06,
+      "loss": 0.6307,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8477983077439863,
+      "learning_rate": 9.613172841756835e-06,
+      "loss": 0.5452,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9003093700116294,
+      "learning_rate": 9.612523327943786e-06,
+      "loss": 0.6483,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8638599615140454,
+      "learning_rate": 9.611873291274927e-06,
+      "loss": 0.7114,
+      "step": 1467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8520477349826896,
+      "learning_rate": 9.611222731823944e-06,
+      "loss": 0.6058,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9085000538619996,
+      "learning_rate": 9.61057164966458e-06,
+      "loss": 0.6045,
+      "step": 1469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8006432662990515,
+      "learning_rate": 9.609920044870643e-06,
+      "loss": 0.6029,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9179518990616973,
+      "learning_rate": 9.60926791751599e-06,
+      "loss": 0.7078,
+      "step": 1471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.042582586101166,
+      "learning_rate": 9.608615267674548e-06,
+      "loss": 0.662,
+      "step": 1472
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7613901736843571,
+      "learning_rate": 9.607962095420297e-06,
+      "loss": 0.6698,
+      "step": 1473
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0988777706402897,
+      "learning_rate": 9.607308400827277e-06,
+      "loss": 0.6381,
+      "step": 1474
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.378994760615408,
+      "learning_rate": 9.606654183969591e-06,
+      "loss": 0.7589,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8172655352330878,
+      "learning_rate": 9.605999444921394e-06,
+      "loss": 0.6398,
+      "step": 1476
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.037892899762544,
+      "learning_rate": 9.605344183756908e-06,
+      "loss": 0.7299,
+      "step": 1477
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7584598825730333,
+      "learning_rate": 9.60468840055041e-06,
+      "loss": 0.7002,
+      "step": 1478
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.743347054484532,
+      "learning_rate": 9.604032095376234e-06,
+      "loss": 0.6489,
+      "step": 1479
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8804620415715747,
+      "learning_rate": 9.603375268308779e-06,
+      "loss": 0.6259,
+      "step": 1480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.170002273856052,
+      "learning_rate": 9.602717919422499e-06,
+      "loss": 0.7196,
+      "step": 1481
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9421409150574973,
+      "learning_rate": 9.602060048791908e-06,
+      "loss": 0.6924,
+      "step": 1482
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9717758470116444,
+      "learning_rate": 9.60140165649158e-06,
+      "loss": 0.7171,
+      "step": 1483
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4743252790664467,
+      "learning_rate": 9.600742742596146e-06,
+      "loss": 0.7644,
+      "step": 1484
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.03725397790848,
+      "learning_rate": 9.6000833071803e-06,
+      "loss": 0.6426,
+      "step": 1485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.004957380913632,
+      "learning_rate": 9.599423350318791e-06,
+      "loss": 0.7055,
+      "step": 1486
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8582757642070944,
+      "learning_rate": 9.598762872086428e-06,
+      "loss": 0.6467,
+      "step": 1487
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0838491494548603,
+      "learning_rate": 9.598101872558085e-06,
+      "loss": 0.6039,
+      "step": 1488
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8164477514953552,
+      "learning_rate": 9.597440351808684e-06,
+      "loss": 0.6671,
+      "step": 1489
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9808384701216841,
+      "learning_rate": 9.596778309913215e-06,
+      "loss": 0.6528,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.003749745611465,
+      "learning_rate": 9.596115746946723e-06,
+      "loss": 0.7215,
+      "step": 1491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9103354885895816,
+      "learning_rate": 9.595452662984314e-06,
+      "loss": 0.6408,
+      "step": 1492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.056675445688021,
+      "learning_rate": 9.594789058101154e-06,
+      "loss": 0.729,
+      "step": 1493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.008591475324911,
+      "learning_rate": 9.594124932372465e-06,
+      "loss": 0.6412,
+      "step": 1494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.091130750278084,
+      "learning_rate": 9.593460285873528e-06,
+      "loss": 0.6649,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.920035920120745,
+      "learning_rate": 9.592795118679686e-06,
+      "loss": 0.6084,
+      "step": 1496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8093080186617532,
+      "learning_rate": 9.59212943086634e-06,
+      "loss": 0.6288,
+      "step": 1497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8862928362176083,
+      "learning_rate": 9.591463222508947e-06,
+      "loss": 0.6367,
+      "step": 1498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8277193723517655,
+      "learning_rate": 9.590796493683028e-06,
+      "loss": 0.7249,
+      "step": 1499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.19331615809431,
+      "learning_rate": 9.59012924446416e-06,
+      "loss": 0.7191,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6401164370086485,
+      "learning_rate": 9.589461474927979e-06,
+      "loss": 0.6349,
+      "step": 1501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.357420255966235,
+      "learning_rate": 9.588793185150182e-06,
+      "loss": 0.6767,
+      "step": 1502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9435062989464436,
+      "learning_rate": 9.58812437520652e-06,
+      "loss": 0.6545,
+      "step": 1503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.927844968000663,
+      "learning_rate": 9.58745504517281e-06,
+      "loss": 0.6596,
+      "step": 1504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7570483201394154,
+      "learning_rate": 9.586785195124924e-06,
+      "loss": 0.673,
+      "step": 1505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8155004596147315,
+      "learning_rate": 9.586114825138792e-06,
+      "loss": 0.683,
+      "step": 1506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8583467053252918,
+      "learning_rate": 9.585443935290403e-06,
+      "loss": 0.6367,
+      "step": 1507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9030635673277672,
+      "learning_rate": 9.58477252565581e-06,
+      "loss": 0.6275,
+      "step": 1508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.788718336973656,
+      "learning_rate": 9.584100596311117e-06,
+      "loss": 0.6945,
+      "step": 1509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7351275011949883,
+      "learning_rate": 9.583428147332493e-06,
+      "loss": 0.5619,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8450518135634049,
+      "learning_rate": 9.582755178796164e-06,
+      "loss": 0.6165,
+      "step": 1511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.972540427042877,
+      "learning_rate": 9.582081690778415e-06,
+      "loss": 0.5639,
+      "step": 1512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9050637486054645,
+      "learning_rate": 9.58140768335559e-06,
+      "loss": 0.6104,
+      "step": 1513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0985873225265284,
+      "learning_rate": 9.58073315660409e-06,
+      "loss": 0.717,
+      "step": 1514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3097437422761256,
+      "learning_rate": 9.580058110600377e-06,
+      "loss": 0.617,
+      "step": 1515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9690754280591791,
+      "learning_rate": 9.579382545420972e-06,
+      "loss": 0.6871,
+      "step": 1516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.782107200466709,
+      "learning_rate": 9.578706461142455e-06,
+      "loss": 0.7152,
+      "step": 1517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8507836163162588,
+      "learning_rate": 9.578029857841462e-06,
+      "loss": 0.673,
+      "step": 1518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.890334013229442,
+      "learning_rate": 9.577352735594692e-06,
+      "loss": 0.7108,
+      "step": 1519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8323036417412149,
+      "learning_rate": 9.576675094478898e-06,
+      "loss": 0.5764,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.088690846097015,
+      "learning_rate": 9.575996934570896e-06,
+      "loss": 0.7671,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.896412011489828,
+      "learning_rate": 9.575318255947558e-06,
+      "loss": 0.6737,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8886169195982683,
+      "learning_rate": 9.57463905868582e-06,
+      "loss": 0.6384,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.219113889815762,
+      "learning_rate": 9.573959342862667e-06,
+      "loss": 0.6598,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0301247271789857,
+      "learning_rate": 9.573279108555154e-06,
+      "loss": 0.6046,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9843830977227843,
+      "learning_rate": 9.572598355840386e-06,
+      "loss": 0.6969,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8464680774728668,
+      "learning_rate": 9.571917084795532e-06,
+      "loss": 0.5952,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7925551511247095,
+      "learning_rate": 9.571235295497818e-06,
+      "loss": 0.6387,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7944467845098961,
+      "learning_rate": 9.570552988024527e-06,
+      "loss": 0.6061,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1431104616084493,
+      "learning_rate": 9.569870162453004e-06,
+      "loss": 0.7645,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9799001726970609,
+      "learning_rate": 9.569186818860652e-06,
+      "loss": 0.8162,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.029920054940862,
+      "learning_rate": 9.568502957324928e-06,
+      "loss": 0.6148,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2030292163042304,
+      "learning_rate": 9.567818577923356e-06,
+      "loss": 0.7205,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9233321976022704,
+      "learning_rate": 9.567133680733512e-06,
+      "loss": 0.6518,
+      "step": 1534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7910034855382688,
+      "learning_rate": 9.566448265833034e-06,
+      "loss": 0.6862,
+      "step": 1535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0613102871389777,
+      "learning_rate": 9.565762333299616e-06,
+      "loss": 0.7762,
+      "step": 1536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1927507319579074,
+      "learning_rate": 9.565075883211015e-06,
+      "loss": 0.7046,
+      "step": 1537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.833817329880263,
+      "learning_rate": 9.564388915645042e-06,
+      "loss": 0.6771,
+      "step": 1538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9954731003848947,
+      "learning_rate": 9.563701430679568e-06,
+      "loss": 0.7152,
+      "step": 1539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9271801113839422,
+      "learning_rate": 9.563013428392528e-06,
+      "loss": 0.7178,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9666689051116464,
+      "learning_rate": 9.562324908861904e-06,
+      "loss": 0.5967,
+      "step": 1541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.020042739074386,
+      "learning_rate": 9.561635872165747e-06,
+      "loss": 0.6152,
+      "step": 1542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7887557940007968,
+      "learning_rate": 9.560946318382166e-06,
+      "loss": 0.6381,
+      "step": 1543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1556497061967823,
+      "learning_rate": 9.560256247589321e-06,
+      "loss": 0.5674,
+      "step": 1544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7794594694424544,
+      "learning_rate": 9.559565659865439e-06,
+      "loss": 0.6416,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9390939469813973,
+      "learning_rate": 9.558874555288801e-06,
+      "loss": 0.7169,
+      "step": 1546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8786485630013083,
+      "learning_rate": 9.558182933937747e-06,
+      "loss": 0.5799,
+      "step": 1547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.856557490530216,
+      "learning_rate": 9.557490795890679e-06,
+      "loss": 0.5993,
+      "step": 1548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.974287937611842,
+      "learning_rate": 9.55679814122605e-06,
+      "loss": 0.7003,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9169969331494634,
+      "learning_rate": 9.556104970022378e-06,
+      "loss": 0.7669,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.978215919977572,
+      "learning_rate": 9.55541128235824e-06,
+      "loss": 0.6096,
+      "step": 1551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8521148274637858,
+      "learning_rate": 9.554717078312269e-06,
+      "loss": 0.7282,
+      "step": 1552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0251613160942914,
+      "learning_rate": 9.554022357963153e-06,
+      "loss": 0.7232,
+      "step": 1553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0117740878267436,
+      "learning_rate": 9.553327121389648e-06,
+      "loss": 0.7499,
+      "step": 1554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8316021424004985,
+      "learning_rate": 9.552631368670562e-06,
+      "loss": 0.6253,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7375512663566117,
+      "learning_rate": 9.55193509988476e-06,
+      "loss": 0.5862,
+      "step": 1556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.226876317475117,
+      "learning_rate": 9.55123831511117e-06,
+      "loss": 0.6528,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.280983795980818,
+      "learning_rate": 9.550541014428773e-06,
+      "loss": 0.7361,
+      "step": 1558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.949008109587866,
+      "learning_rate": 9.549843197916616e-06,
+      "loss": 0.6629,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.876199206288432,
+      "learning_rate": 9.5491448656538e-06,
+      "loss": 0.5759,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0055504701332407,
+      "learning_rate": 9.548446017719484e-06,
+      "loss": 0.6555,
+      "step": 1561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8950761565227399,
+      "learning_rate": 9.547746654192887e-06,
+      "loss": 0.6419,
+      "step": 1562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.147644898855912,
+      "learning_rate": 9.547046775153285e-06,
+      "loss": 0.7729,
+      "step": 1563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9930761501595675,
+      "learning_rate": 9.546346380680015e-06,
+      "loss": 0.757,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.03876848772226,
+      "learning_rate": 9.545645470852466e-06,
+      "loss": 0.7823,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0341781603182385,
+      "learning_rate": 9.544944045750097e-06,
+      "loss": 0.6754,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9935660269354207,
+      "learning_rate": 9.544242105452414e-06,
+      "loss": 0.6485,
+      "step": 1567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2846043842839254,
+      "learning_rate": 9.543539650038987e-06,
+      "loss": 0.6906,
+      "step": 1568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1741854085761876,
+      "learning_rate": 9.542836679589443e-06,
+      "loss": 0.6777,
+      "step": 1569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8573681266861257,
+      "learning_rate": 9.542133194183469e-06,
+      "loss": 0.6227,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.768231953559204,
+      "learning_rate": 9.541429193900808e-06,
+      "loss": 0.6056,
+      "step": 1571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8561209820293387,
+      "learning_rate": 9.540724678821261e-06,
+      "loss": 0.7023,
+      "step": 1572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.769977024069318,
+      "learning_rate": 9.540019649024692e-06,
+      "loss": 0.6391,
+      "step": 1573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.913878231055899,
+      "learning_rate": 9.539314104591019e-06,
+      "loss": 0.6818,
+      "step": 1574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0615080474215635,
+      "learning_rate": 9.538608045600218e-06,
+      "loss": 0.6203,
+      "step": 1575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.872307597705844,
+      "learning_rate": 9.537901472132324e-06,
+      "loss": 0.6047,
+      "step": 1576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.001299949883912,
+      "learning_rate": 9.537194384267436e-06,
+      "loss": 0.6852,
+      "step": 1577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9756233996839685,
+      "learning_rate": 9.536486782085703e-06,
+      "loss": 0.666,
+      "step": 1578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7817231162157137,
+      "learning_rate": 9.535778665667334e-06,
+      "loss": 0.7383,
+      "step": 1579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8093842277276013,
+      "learning_rate": 9.535070035092603e-06,
+      "loss": 0.7256,
+      "step": 1580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7685386576759308,
+      "learning_rate": 9.534360890441833e-06,
+      "loss": 0.5198,
+      "step": 1581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1297343383103717,
+      "learning_rate": 9.533651231795412e-06,
+      "loss": 0.706,
+      "step": 1582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.059039551713341,
+      "learning_rate": 9.532941059233782e-06,
+      "loss": 0.7166,
+      "step": 1583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8142204295763857,
+      "learning_rate": 9.532230372837446e-06,
+      "loss": 0.7129,
+      "step": 1584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.039237252791234,
+      "learning_rate": 9.531519172686964e-06,
+      "loss": 0.7075,
+      "step": 1585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8287133514681926,
+      "learning_rate": 9.530807458862956e-06,
+      "loss": 0.6532,
+      "step": 1586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7667190625126445,
+      "learning_rate": 9.530095231446096e-06,
+      "loss": 0.5883,
+      "step": 1587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8291627106185575,
+      "learning_rate": 9.529382490517123e-06,
+      "loss": 0.6794,
+      "step": 1588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9608081910222792,
+      "learning_rate": 9.528669236156827e-06,
+      "loss": 0.6054,
+      "step": 1589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9924170196955373,
+      "learning_rate": 9.52795546844606e-06,
+      "loss": 0.7171,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6524851266305414,
+      "learning_rate": 9.527241187465735e-06,
+      "loss": 0.6372,
+      "step": 1591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.160203220354208,
+      "learning_rate": 9.526526393296814e-06,
+      "loss": 0.7238,
+      "step": 1592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9441794717429497,
+      "learning_rate": 9.525811086020327e-06,
+      "loss": 0.639,
+      "step": 1593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0185775044915775,
+      "learning_rate": 9.525095265717357e-06,
+      "loss": 0.7046,
+      "step": 1594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9035968009066015,
+      "learning_rate": 9.524378932469045e-06,
+      "loss": 0.7046,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.616356917467379,
+      "learning_rate": 9.523662086356596e-06,
+      "loss": 0.6115,
+      "step": 1596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.175508595395853,
+      "learning_rate": 9.522944727461264e-06,
+      "loss": 0.6567,
+      "step": 1597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1150702390919536,
+      "learning_rate": 9.522226855864366e-06,
+      "loss": 0.7182,
+      "step": 1598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7610105376084912,
+      "learning_rate": 9.52150847164728e-06,
+      "loss": 0.7062,
+      "step": 1599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8587370870268596,
+      "learning_rate": 9.520789574891436e-06,
+      "loss": 0.6725,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.246019651567058,
+      "learning_rate": 9.520070165678325e-06,
+      "loss": 0.6444,
+      "step": 1601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0514198751421873,
+      "learning_rate": 9.5193502440895e-06,
+      "loss": 0.7817,
+      "step": 1602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7478983922533107,
+      "learning_rate": 9.518629810206564e-06,
+      "loss": 0.5888,
+      "step": 1603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.809367012317806,
+      "learning_rate": 9.517908864111182e-06,
+      "loss": 0.6072,
+      "step": 1604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8021793691886772,
+      "learning_rate": 9.517187405885082e-06,
+      "loss": 0.653,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9639447106313237,
+      "learning_rate": 9.516465435610041e-06,
+      "loss": 0.7472,
+      "step": 1606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7756152134174483,
+      "learning_rate": 9.515742953367899e-06,
+      "loss": 0.5715,
+      "step": 1607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.910782270315552,
+      "learning_rate": 9.515019959240555e-06,
+      "loss": 0.7156,
+      "step": 1608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9990951548136033,
+      "learning_rate": 9.514296453309965e-06,
+      "loss": 0.7261,
+      "step": 1609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6970077590281472,
+      "learning_rate": 9.51357243565814e-06,
+      "loss": 0.5351,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1610796647883306,
+      "learning_rate": 9.512847906367153e-06,
+      "loss": 0.7213,
+      "step": 1611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0053270299753994,
+      "learning_rate": 9.512122865519135e-06,
+      "loss": 0.6144,
+      "step": 1612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.246635639941283,
+      "learning_rate": 9.51139731319627e-06,
+      "loss": 0.6258,
+      "step": 1613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0140091119514767,
+      "learning_rate": 9.510671249480806e-06,
+      "loss": 0.6814,
+      "step": 1614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0477698552459644,
+      "learning_rate": 9.509944674455047e-06,
+      "loss": 0.7852,
+      "step": 1615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0320102733793246,
+      "learning_rate": 9.509217588201351e-06,
+      "loss": 0.7089,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8191152766376675,
+      "learning_rate": 9.508489990802142e-06,
+      "loss": 0.6528,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.797405031205996,
+      "learning_rate": 9.507761882339895e-06,
+      "loss": 0.6468,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9825745171869442,
+      "learning_rate": 9.507033262897142e-06,
+      "loss": 0.706,
+      "step": 1619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.052359939793622,
+      "learning_rate": 9.506304132556484e-06,
+      "loss": 0.7695,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9565939487072954,
+      "learning_rate": 9.505574491400564e-06,
+      "loss": 0.7543,
+      "step": 1621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.819843813882747,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.5962,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9030655320222276,
+      "learning_rate": 9.504113676973846e-06,
+      "loss": 0.7255,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0056629019007954,
+      "learning_rate": 9.503382503868637e-06,
+      "loss": 0.8137,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9704662860200837,
+      "learning_rate": 9.502650820279354e-06,
+      "loss": 0.6582,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3756436822304345,
+      "learning_rate": 9.501918626288935e-06,
+      "loss": 0.7373,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7756355257502683,
+      "learning_rate": 9.50118592198038e-06,
+      "loss": 0.5558,
+      "step": 1627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8277201108402616,
+      "learning_rate": 9.500452707436744e-06,
+      "loss": 0.6353,
+      "step": 1628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8042213079432121,
+      "learning_rate": 9.499718982741143e-06,
+      "loss": 0.758,
+      "step": 1629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8420878240688283,
+      "learning_rate": 9.498984747976747e-06,
+      "loss": 0.5722,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.045825387663684,
+      "learning_rate": 9.498250003226787e-06,
+      "loss": 0.6005,
+      "step": 1631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8251107667602975,
+      "learning_rate": 9.49751474857455e-06,
+      "loss": 0.6692,
+      "step": 1632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8065273338239225,
+      "learning_rate": 9.496778984103381e-06,
+      "loss": 0.709,
+      "step": 1633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8042150998335234,
+      "learning_rate": 9.496042709896684e-06,
+      "loss": 0.6163,
+      "step": 1634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.898311108212609,
+      "learning_rate": 9.495305926037918e-06,
+      "loss": 0.647,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9689807722120047,
+      "learning_rate": 9.494568632610603e-06,
+      "loss": 0.7747,
+      "step": 1636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8341428636567332,
+      "learning_rate": 9.493830829698317e-06,
+      "loss": 0.5898,
+      "step": 1637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1033082614827414,
+      "learning_rate": 9.493092517384692e-06,
+      "loss": 0.6607,
+      "step": 1638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.829421971966054,
+      "learning_rate": 9.49235369575342e-06,
+      "loss": 0.7291,
+      "step": 1639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8996716409526258,
+      "learning_rate": 9.49161436488825e-06,
+      "loss": 0.7602,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1680339164477433,
+      "learning_rate": 9.49087452487299e-06,
+      "loss": 0.6938,
+      "step": 1641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6508009166489983,
+      "learning_rate": 9.490134175791507e-06,
+      "loss": 0.6141,
+      "step": 1642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7767239402836295,
+      "learning_rate": 9.489393317727724e-06,
+      "loss": 0.618,
+      "step": 1643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.038443432035079,
+      "learning_rate": 9.488651950765617e-06,
+      "loss": 0.6579,
+      "step": 1644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.741168133888333,
+      "learning_rate": 9.487910074989228e-06,
+      "loss": 0.6494,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.857138742618028,
+      "learning_rate": 9.48716769048265e-06,
+      "loss": 0.5755,
+      "step": 1646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1591345481710036,
+      "learning_rate": 9.48642479733004e-06,
+      "loss": 0.6383,
+      "step": 1647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9364895503622959,
+      "learning_rate": 9.485681395615607e-06,
+      "loss": 0.6839,
+      "step": 1648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.031828191064373,
+      "learning_rate": 9.484937485423622e-06,
+      "loss": 0.6035,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.754566896444676,
+      "learning_rate": 9.484193066838408e-06,
+      "loss": 0.6424,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9246941194236256,
+      "learning_rate": 9.48344813994435e-06,
+      "loss": 0.6513,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8684221297861972,
+      "learning_rate": 9.482702704825892e-06,
+      "loss": 0.7858,
+      "step": 1652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9414626199310694,
+      "learning_rate": 9.481956761567531e-06,
+      "loss": 0.6826,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0316286317178727,
+      "learning_rate": 9.481210310253826e-06,
+      "loss": 0.7082,
+      "step": 1654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7661332172941508,
+      "learning_rate": 9.480463350969388e-06,
+      "loss": 0.6982,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9932281813194153,
+      "learning_rate": 9.479715883798895e-06,
+      "loss": 0.64,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0257008912279417,
+      "learning_rate": 9.47896790882707e-06,
+      "loss": 0.6237,
+      "step": 1657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7559027656428055,
+      "learning_rate": 9.478219426138703e-06,
+      "loss": 0.6753,
+      "step": 1658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8541314724538382,
+      "learning_rate": 9.477470435818641e-06,
+      "loss": 0.7334,
+      "step": 1659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9262745014920255,
+      "learning_rate": 9.476720937951785e-06,
+      "loss": 0.6096,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.00672485707719,
+      "learning_rate": 9.47597093262309e-06,
+      "loss": 0.6902,
+      "step": 1661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8392421439660025,
+      "learning_rate": 9.475220419917581e-06,
+      "loss": 0.7256,
+      "step": 1662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9341160956399497,
+      "learning_rate": 9.47446939992033e-06,
+      "loss": 0.7523,
+      "step": 1663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8634455622742643,
+      "learning_rate": 9.473717872716469e-06,
+      "loss": 0.7325,
+      "step": 1664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8513564589120213,
+      "learning_rate": 9.472965838391187e-06,
+      "loss": 0.6888,
+      "step": 1665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7884451440018407,
+      "learning_rate": 9.47221329702973e-06,
+      "loss": 0.5837,
+      "step": 1666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9881558844202416,
+      "learning_rate": 9.471460248717406e-06,
+      "loss": 0.6877,
+      "step": 1667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8388210711744746,
+      "learning_rate": 9.470706693539578e-06,
+      "loss": 0.6638,
+      "step": 1668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0549800910188534,
+      "learning_rate": 9.469952631581663e-06,
+      "loss": 0.7381,
+      "step": 1669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9226671283551982,
+      "learning_rate": 9.469198062929139e-06,
+      "loss": 0.6649,
+      "step": 1670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0660185079140896,
+      "learning_rate": 9.468442987667542e-06,
+      "loss": 0.7564,
+      "step": 1671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.903836190555682,
+      "learning_rate": 9.467687405882463e-06,
+      "loss": 0.6271,
+      "step": 1672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0450685561813824,
+      "learning_rate": 9.466931317659551e-06,
+      "loss": 0.7239,
+      "step": 1673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.005334746057944,
+      "learning_rate": 9.466174723084514e-06,
+      "loss": 0.7554,
+      "step": 1674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6452748276607783,
+      "learning_rate": 9.465417622243116e-06,
+      "loss": 0.6588,
+      "step": 1675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9042511113167548,
+      "learning_rate": 9.464660015221177e-06,
+      "loss": 0.5927,
+      "step": 1676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8411751279551711,
+      "learning_rate": 9.463901902104579e-06,
+      "loss": 0.6422,
+      "step": 1677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9047976740761143,
+      "learning_rate": 9.463143282979258e-06,
+      "loss": 0.7052,
+      "step": 1678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0239698176707055,
+      "learning_rate": 9.462384157931203e-06,
+      "loss": 0.6387,
+      "step": 1679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.71917711556842,
+      "learning_rate": 9.461624527046472e-06,
+      "loss": 0.6492,
+      "step": 1680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7473641280277281,
+      "learning_rate": 9.46086439041117e-06,
+      "loss": 0.6081,
+      "step": 1681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.038864253926127,
+      "learning_rate": 9.460103748111462e-06,
+      "loss": 0.6386,
+      "step": 1682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9706571596391416,
+      "learning_rate": 9.459342600233575e-06,
+      "loss": 0.7428,
+      "step": 1683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.954883981835221,
+      "learning_rate": 9.458580946863784e-06,
+      "loss": 0.6086,
+      "step": 1684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0798834277695146,
+      "learning_rate": 9.457818788088431e-06,
+      "loss": 0.6907,
+      "step": 1685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.854504688334079,
+      "learning_rate": 9.45705612399391e-06,
+      "loss": 0.6769,
+      "step": 1686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.940980581413887,
+      "learning_rate": 9.456292954666673e-06,
+      "loss": 0.6356,
+      "step": 1687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.010496752053228,
+      "learning_rate": 9.45552928019323e-06,
+      "loss": 0.6778,
+      "step": 1688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.357313247435537,
+      "learning_rate": 9.454765100660144e-06,
+      "loss": 0.7522,
+      "step": 1689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.084001616032715,
+      "learning_rate": 9.454000416154046e-06,
+      "loss": 0.7403,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.038729610751833,
+      "learning_rate": 9.453235226761613e-06,
+      "loss": 0.6874,
+      "step": 1691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9823748346716907,
+      "learning_rate": 9.452469532569585e-06,
+      "loss": 0.6065,
+      "step": 1692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.998207692487797,
+      "learning_rate": 9.451703333664756e-06,
+      "loss": 0.6758,
+      "step": 1693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8763531828895923,
+      "learning_rate": 9.45093663013398e-06,
+      "loss": 0.6963,
+      "step": 1694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6424440865727896,
+      "learning_rate": 9.45016942206417e-06,
+      "loss": 0.6559,
+      "step": 1695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1430067998803564,
+      "learning_rate": 9.449401709542289e-06,
+      "loss": 0.7204,
+      "step": 1696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0047992594876627,
+      "learning_rate": 9.448633492655363e-06,
+      "loss": 0.7242,
+      "step": 1697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9198849998196943,
+      "learning_rate": 9.447864771490476e-06,
+      "loss": 0.743,
+      "step": 1698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8212504164264498,
+      "learning_rate": 9.447095546134763e-06,
+      "loss": 0.5829,
+      "step": 1699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8677748324266148,
+      "learning_rate": 9.446325816675423e-06,
+      "loss": 0.6371,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0007664737650885,
+      "learning_rate": 9.445555583199711e-06,
+      "loss": 0.6646,
+      "step": 1701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8964503663253003,
+      "learning_rate": 9.444784845794932e-06,
+      "loss": 0.7622,
+      "step": 1702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9525551773329504,
+      "learning_rate": 9.444013604548457e-06,
+      "loss": 0.5986,
+      "step": 1703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.357983873653802,
+      "learning_rate": 9.44324185954771e-06,
+      "loss": 0.5866,
+      "step": 1704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.974929397878334,
+      "learning_rate": 9.442469610880172e-06,
+      "loss": 0.7261,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.422100306321085,
+      "learning_rate": 9.441696858633382e-06,
+      "loss": 0.7537,
+      "step": 1706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7334732730860278,
+      "learning_rate": 9.440923602894937e-06,
+      "loss": 0.6426,
+      "step": 1707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.085845934820001,
+      "learning_rate": 9.44014984375249e-06,
+      "loss": 0.7491,
+      "step": 1708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.957209376001055,
+      "learning_rate": 9.439375581293751e-06,
+      "loss": 0.6847,
+      "step": 1709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.850758181214728,
+      "learning_rate": 9.438600815606483e-06,
+      "loss": 0.6593,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9311284970137075,
+      "learning_rate": 9.437825546778517e-06,
+      "loss": 0.6886,
+      "step": 1711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7095134979680247,
+      "learning_rate": 9.437049774897728e-06,
+      "loss": 0.6179,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.048187229194672,
+      "learning_rate": 9.436273500052056e-06,
+      "loss": 0.7139,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8280668981290464,
+      "learning_rate": 9.4354967223295e-06,
+      "loss": 0.6537,
+      "step": 1714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7321260549808468,
+      "learning_rate": 9.434719441818106e-06,
+      "loss": 0.5556,
+      "step": 1715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.224063985682685,
+      "learning_rate": 9.433941658605987e-06,
+      "loss": 0.6958,
+      "step": 1716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.244857086101658,
+      "learning_rate": 9.43316337278131e-06,
+      "loss": 0.6439,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1301665791756372,
+      "learning_rate": 9.432384584432294e-06,
+      "loss": 0.7263,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.204264658265296,
+      "learning_rate": 9.431605293647224e-06,
+      "loss": 0.6973,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9886649316067186,
+      "learning_rate": 9.430825500514433e-06,
+      "loss": 0.5987,
+      "step": 1720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9015353627790044,
+      "learning_rate": 9.430045205122317e-06,
+      "loss": 0.6792,
+      "step": 1721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.746055037768087,
+      "learning_rate": 9.429264407559328e-06,
+      "loss": 0.6747,
+      "step": 1722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8539817079671905,
+      "learning_rate": 9.42848310791397e-06,
+      "loss": 0.6928,
+      "step": 1723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1077148459976724,
+      "learning_rate": 9.427701306274812e-06,
+      "loss": 0.6944,
+      "step": 1724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2949928578619594,
+      "learning_rate": 9.426919002730473e-06,
+      "loss": 0.6411,
+      "step": 1725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3089363293156584,
+      "learning_rate": 9.426136197369633e-06,
+      "loss": 0.7011,
+      "step": 1726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3775684030564825,
+      "learning_rate": 9.425352890281028e-06,
+      "loss": 0.6686,
+      "step": 1727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3000666445379516,
+      "learning_rate": 9.424569081553447e-06,
+      "loss": 0.576,
+      "step": 1728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2271722236338998,
+      "learning_rate": 9.423784771275744e-06,
+      "loss": 0.6952,
+      "step": 1729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9171166819419831,
+      "learning_rate": 9.422999959536819e-06,
+      "loss": 0.8124,
+      "step": 1730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9396569476898367,
+      "learning_rate": 9.422214646425641e-06,
+      "loss": 0.5715,
+      "step": 1731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1983267960374264,
+      "learning_rate": 9.421428832031226e-06,
+      "loss": 0.6675,
+      "step": 1732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2057794317489123,
+      "learning_rate": 9.420642516442652e-06,
+      "loss": 0.6696,
+      "step": 1733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.719469875329039,
+      "learning_rate": 9.41985569974905e-06,
+      "loss": 0.6127,
+      "step": 1734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.083556833840758,
+      "learning_rate": 9.419068382039615e-06,
+      "loss": 0.7165,
+      "step": 1735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1664458577953467,
+      "learning_rate": 9.41828056340359e-06,
+      "loss": 0.7521,
+      "step": 1736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9227176468266851,
+      "learning_rate": 9.41749224393028e-06,
+      "loss": 0.7552,
+      "step": 1737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9964207302492,
+      "learning_rate": 9.416703423709044e-06,
+      "loss": 0.6417,
+      "step": 1738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8401124624218785,
+      "learning_rate": 9.415914102829302e-06,
+      "loss": 0.6845,
+      "step": 1739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9982912657871517,
+      "learning_rate": 9.415124281380525e-06,
+      "loss": 0.7257,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.931542910740269,
+      "learning_rate": 9.414333959452247e-06,
+      "loss": 0.7005,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0180011776973115,
+      "learning_rate": 9.413543137134053e-06,
+      "loss": 0.7342,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8253225407310454,
+      "learning_rate": 9.412751814515588e-06,
+      "loss": 0.697,
+      "step": 1743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1203442175782348,
+      "learning_rate": 9.411959991686554e-06,
+      "loss": 0.7561,
+      "step": 1744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9259566109424078,
+      "learning_rate": 9.411167668736707e-06,
+      "loss": 0.5717,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.95383111394422,
+      "learning_rate": 9.410374845755862e-06,
+      "loss": 0.6709,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4064921390288863,
+      "learning_rate": 9.409581522833889e-06,
+      "loss": 0.6546,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9152540024539662,
+      "learning_rate": 9.408787700060718e-06,
+      "loss": 0.6543,
+      "step": 1748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1151996633398142,
+      "learning_rate": 9.40799337752633e-06,
+      "loss": 0.6748,
+      "step": 1749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8568166821966514,
+      "learning_rate": 9.40719855532077e-06,
+      "loss": 0.7002,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0767917401632983,
+      "learning_rate": 9.406403233534134e-06,
+      "loss": 0.6867,
+      "step": 1751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.143310125483721,
+      "learning_rate": 9.405607412256573e-06,
+      "loss": 0.7443,
+      "step": 1752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.02257859190377,
+      "learning_rate": 9.404811091578302e-06,
+      "loss": 0.6455,
+      "step": 1753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9046396189464536,
+      "learning_rate": 9.404014271589588e-06,
+      "loss": 0.7036,
+      "step": 1754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7750985666179104,
+      "learning_rate": 9.403216952380755e-06,
+      "loss": 0.6842,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9874375531893456,
+      "learning_rate": 9.40241913404218e-06,
+      "loss": 0.6559,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.886971108135804,
+      "learning_rate": 9.401620816664305e-06,
+      "loss": 0.6458,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9427562819969475,
+      "learning_rate": 9.400822000337622e-06,
+      "loss": 0.7032,
+      "step": 1758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8186870029716242,
+      "learning_rate": 9.400022685152683e-06,
+      "loss": 0.6731,
+      "step": 1759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.855438322366915,
+      "learning_rate": 9.399222871200091e-06,
+      "loss": 0.6359,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.23468285760032,
+      "learning_rate": 9.398422558570512e-06,
+      "loss": 0.7936,
+      "step": 1761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9850677319168017,
+      "learning_rate": 9.397621747354666e-06,
+      "loss": 0.5443,
+      "step": 1762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0215518938615653,
+      "learning_rate": 9.39682043764333e-06,
+      "loss": 0.7142,
+      "step": 1763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1960036942141676,
+      "learning_rate": 9.396018629527336e-06,
+      "loss": 0.6934,
+      "step": 1764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9005397240127722,
+      "learning_rate": 9.395216323097573e-06,
+      "loss": 0.7108,
+      "step": 1765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7738157201021318,
+      "learning_rate": 9.394413518444989e-06,
+      "loss": 0.6096,
+      "step": 1766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.936537266975992,
+      "learning_rate": 9.393610215660585e-06,
+      "loss": 0.6309,
+      "step": 1767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9371372960509754,
+      "learning_rate": 9.392806414835422e-06,
+      "loss": 0.7551,
+      "step": 1768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9333032902331897,
+      "learning_rate": 9.392002116060612e-06,
+      "loss": 0.6436,
+      "step": 1769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.027033584198372,
+      "learning_rate": 9.391197319427328e-06,
+      "loss": 0.6055,
+      "step": 1770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.196135164479654,
+      "learning_rate": 9.3903920250268e-06,
+      "loss": 0.7347,
+      "step": 1771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.731083189320279,
+      "learning_rate": 9.389586232950312e-06,
+      "loss": 0.5823,
+      "step": 1772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9930184050076185,
+      "learning_rate": 9.388779943289204e-06,
+      "loss": 0.6757,
+      "step": 1773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.933884457327214,
+      "learning_rate": 9.387973156134872e-06,
+      "loss": 0.6756,
+      "step": 1774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.391449358670151,
+      "learning_rate": 9.387165871578774e-06,
+      "loss": 0.6467,
+      "step": 1775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0315456443703153,
+      "learning_rate": 9.38635808971242e-06,
+      "loss": 0.629,
+      "step": 1776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.051657440819527,
+      "learning_rate": 9.385549810627374e-06,
+      "loss": 0.6349,
+      "step": 1777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.208159799053304,
+      "learning_rate": 9.38474103441526e-06,
+      "loss": 0.7332,
+      "step": 1778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.003368483790586,
+      "learning_rate": 9.383931761167757e-06,
+      "loss": 0.6808,
+      "step": 1779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9662724949489288,
+      "learning_rate": 9.383121990976602e-06,
+      "loss": 0.7002,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.046593510996878,
+      "learning_rate": 9.382311723933586e-06,
+      "loss": 0.6752,
+      "step": 1781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7582074425722836,
+      "learning_rate": 9.381500960130558e-06,
+      "loss": 0.6253,
+      "step": 1782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8022736464724076,
+      "learning_rate": 9.380689699659423e-06,
+      "loss": 0.6269,
+      "step": 1783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.850845380128415,
+      "learning_rate": 9.37987794261214e-06,
+      "loss": 0.6402,
+      "step": 1784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8174325837103,
+      "learning_rate": 9.379065689080729e-06,
+      "loss": 0.6395,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.965486494511778,
+      "learning_rate": 9.378252939157262e-06,
+      "loss": 0.6842,
+      "step": 1786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7872036813111714,
+      "learning_rate": 9.377439692933869e-06,
+      "loss": 0.702,
+      "step": 1787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.056730874252554,
+      "learning_rate": 9.376625950502736e-06,
+      "loss": 0.6803,
+      "step": 1788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7395674020136729,
+      "learning_rate": 9.375811711956106e-06,
+      "loss": 0.6639,
+      "step": 1789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7736550746022077,
+      "learning_rate": 9.374996977386276e-06,
+      "loss": 0.6338,
+      "step": 1790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8660714665304707,
+      "learning_rate": 9.374181746885603e-06,
+      "loss": 0.5861,
+      "step": 1791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9014869405672314,
+      "learning_rate": 9.373366020546498e-06,
+      "loss": 0.7147,
+      "step": 1792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7393522079638204,
+      "learning_rate": 9.372549798461426e-06,
+      "loss": 0.5423,
+      "step": 1793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8812361480908122,
+      "learning_rate": 9.371733080722911e-06,
+      "loss": 0.7266,
+      "step": 1794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.808933104084343,
+      "learning_rate": 9.370915867423534e-06,
+      "loss": 0.7232,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7975636897079652,
+      "learning_rate": 9.37009815865593e-06,
+      "loss": 0.638,
+      "step": 1796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9655333154002042,
+      "learning_rate": 9.369279954512791e-06,
+      "loss": 0.81,
+      "step": 1797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1093837615576163,
+      "learning_rate": 9.368461255086866e-06,
+      "loss": 0.6529,
+      "step": 1798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8458410099324793,
+      "learning_rate": 9.367642060470958e-06,
+      "loss": 0.5708,
+      "step": 1799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7681980524691754,
+      "learning_rate": 9.366822370757927e-06,
+      "loss": 0.6468,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0440408178376956,
+      "learning_rate": 9.36600218604069e-06,
+      "loss": 0.6947,
+      "step": 1801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.709238322425783,
+      "learning_rate": 9.36518150641222e-06,
+      "loss": 0.6153,
+      "step": 1802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8525472802523686,
+      "learning_rate": 9.364360331965545e-06,
+      "loss": 0.6379,
+      "step": 1803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.040402922812123,
+      "learning_rate": 9.363538662793752e-06,
+      "loss": 0.73,
+      "step": 1804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9562526941554614,
+      "learning_rate": 9.362716498989979e-06,
+      "loss": 0.6633,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.102318885795319,
+      "learning_rate": 9.361893840647425e-06,
+      "loss": 0.6032,
+      "step": 1806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.934206711167693,
+      "learning_rate": 9.361070687859341e-06,
+      "loss": 0.6323,
+      "step": 1807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0505371589582104,
+      "learning_rate": 9.36024704071904e-06,
+      "loss": 0.7097,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9380575372950994,
+      "learning_rate": 9.359422899319882e-06,
+      "loss": 0.7334,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9703923692444525,
+      "learning_rate": 9.358598263755289e-06,
+      "loss": 0.6529,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0041847525050547,
+      "learning_rate": 9.357773134118743e-06,
+      "loss": 0.7106,
+      "step": 1811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8824885667479772,
+      "learning_rate": 9.356947510503771e-06,
+      "loss": 0.6663,
+      "step": 1812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9126377503709864,
+      "learning_rate": 9.356121393003968e-06,
+      "loss": 0.6731,
+      "step": 1813
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1234974613692432,
+      "learning_rate": 9.355294781712974e-06,
+      "loss": 0.668,
+      "step": 1814
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.427177580084952,
+      "learning_rate": 9.354467676724491e-06,
+      "loss": 0.6898,
+      "step": 1815
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7188594239543076,
+      "learning_rate": 9.35364007813228e-06,
+      "loss": 0.6387,
+      "step": 1816
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1711651881047604,
+      "learning_rate": 9.352811986030147e-06,
+      "loss": 0.7751,
+      "step": 1817
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.004661231210757,
+      "learning_rate": 9.35198340051197e-06,
+      "loss": 0.7885,
+      "step": 1818
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.202027654675994,
+      "learning_rate": 9.351154321671667e-06,
+      "loss": 0.714,
+      "step": 1819
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.108135480448231,
+      "learning_rate": 9.350324749603221e-06,
+      "loss": 0.6352,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0603342012079517,
+      "learning_rate": 9.34949468440067e-06,
+      "loss": 0.6355,
+      "step": 1821
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.014812774582301,
+      "learning_rate": 9.348664126158103e-06,
+      "loss": 0.6858,
+      "step": 1822
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.967906011677732,
+      "learning_rate": 9.34783307496967e-06,
+      "loss": 0.7113,
+      "step": 1823
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0691130215812827,
+      "learning_rate": 9.347001530929579e-06,
+      "loss": 0.7028,
+      "step": 1824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9448319235913356,
+      "learning_rate": 9.346169494132086e-06,
+      "loss": 0.6687,
+      "step": 1825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.858179669232822,
+      "learning_rate": 9.345336964671507e-06,
+      "loss": 0.5961,
+      "step": 1826
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6801584205129731,
+      "learning_rate": 9.344503942642218e-06,
+      "loss": 0.5703,
+      "step": 1827
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8900232563367874,
+      "learning_rate": 9.343670428138644e-06,
+      "loss": 0.6756,
+      "step": 1828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8460935273114019,
+      "learning_rate": 9.342836421255268e-06,
+      "loss": 0.6328,
+      "step": 1829
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8418240477099301,
+      "learning_rate": 9.34200192208663e-06,
+      "loss": 0.6728,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8254277059194648,
+      "learning_rate": 9.341166930727326e-06,
+      "loss": 0.7245,
+      "step": 1831
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.88091998086299,
+      "learning_rate": 9.340331447272006e-06,
+      "loss": 0.6205,
+      "step": 1832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.822431851717622,
+      "learning_rate": 9.339495471815379e-06,
+      "loss": 0.6293,
+      "step": 1833
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8356744862913454,
+      "learning_rate": 9.338659004452204e-06,
+      "loss": 0.6785,
+      "step": 1834
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.853780372536723,
+      "learning_rate": 9.337822045277303e-06,
+      "loss": 0.6948,
+      "step": 1835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8247544720172622,
+      "learning_rate": 9.336984594385547e-06,
+      "loss": 0.546,
+      "step": 1836
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7757233288010839,
+      "learning_rate": 9.33614665187187e-06,
+      "loss": 0.7182,
+      "step": 1837
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1038639116244946,
+      "learning_rate": 9.335308217831252e-06,
+      "loss": 0.7425,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9022465803573188,
+      "learning_rate": 9.334469292358736e-06,
+      "loss": 0.7418,
+      "step": 1839
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.982372166611791,
+      "learning_rate": 9.333629875549424e-06,
+      "loss": 0.6076,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.866068994191335,
+      "learning_rate": 9.33278996749846e-06,
+      "loss": 0.6356,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9918446912193064,
+      "learning_rate": 9.331949568301062e-06,
+      "loss": 0.643,
+      "step": 1842
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9290363357787772,
+      "learning_rate": 9.331108678052485e-06,
+      "loss": 0.6967,
+      "step": 1843
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.944946068496725,
+      "learning_rate": 9.330267296848056e-06,
+      "loss": 0.6828,
+      "step": 1844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2391804890671683,
+      "learning_rate": 9.329425424783145e-06,
+      "loss": 0.66,
+      "step": 1845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0239600614768927,
+      "learning_rate": 9.328583061953187e-06,
+      "loss": 0.6193,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7541080391109118,
+      "learning_rate": 9.327740208453666e-06,
+      "loss": 0.6133,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0824969747832474,
+      "learning_rate": 9.326896864380124e-06,
+      "loss": 0.7947,
+      "step": 1848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9952093471075816,
+      "learning_rate": 9.326053029828162e-06,
+      "loss": 0.6064,
+      "step": 1849
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8993269296443138,
+      "learning_rate": 9.32520870489343e-06,
+      "loss": 0.6876,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8474386957130455,
+      "learning_rate": 9.32436388967164e-06,
+      "loss": 0.6825,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.873580562276867,
+      "learning_rate": 9.323518584258554e-06,
+      "loss": 0.7735,
+      "step": 1852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.923520628246674,
+      "learning_rate": 9.32267278874999e-06,
+      "loss": 0.678,
+      "step": 1853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7349404592036248,
+      "learning_rate": 9.32182650324183e-06,
+      "loss": 0.6314,
+      "step": 1854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.983232010650299,
+      "learning_rate": 9.320979727830004e-06,
+      "loss": 0.663,
+      "step": 1855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.116098848893127,
+      "learning_rate": 9.320132462610495e-06,
+      "loss": 0.6062,
+      "step": 1856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.954496745764822,
+      "learning_rate": 9.319284707679348e-06,
+      "loss": 0.6566,
+      "step": 1857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.944786321375141,
+      "learning_rate": 9.318436463132661e-06,
+      "loss": 0.7398,
+      "step": 1858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9358948718792182,
+      "learning_rate": 9.317587729066585e-06,
+      "loss": 0.6646,
+      "step": 1859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9699922116768356,
+      "learning_rate": 9.316738505577331e-06,
+      "loss": 0.7234,
+      "step": 1860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.093839616518822,
+      "learning_rate": 9.315888792761163e-06,
+      "loss": 0.659,
+      "step": 1861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.925384133731945,
+      "learning_rate": 9.3150385907144e-06,
+      "loss": 0.6902,
+      "step": 1862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.808062266664772,
+      "learning_rate": 9.31418789953342e-06,
+      "loss": 0.576,
+      "step": 1863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.073959612645908,
+      "learning_rate": 9.31333671931465e-06,
+      "loss": 0.7181,
+      "step": 1864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8375303473816842,
+      "learning_rate": 9.312485050154578e-06,
+      "loss": 0.6194,
+      "step": 1865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.026136079077223,
+      "learning_rate": 9.311632892149744e-06,
+      "loss": 0.6778,
+      "step": 1866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0988523898716376,
+      "learning_rate": 9.310780245396747e-06,
+      "loss": 0.7,
+      "step": 1867
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9595698067023832,
+      "learning_rate": 9.309927109992239e-06,
+      "loss": 0.5943,
+      "step": 1868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.395683206373414,
+      "learning_rate": 9.309073486032926e-06,
+      "loss": 0.6852,
+      "step": 1869
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8437940606159533,
+      "learning_rate": 9.308219373615574e-06,
+      "loss": 0.5946,
+      "step": 1870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.23571475537344,
+      "learning_rate": 9.307364772837e-06,
+      "loss": 0.7415,
+      "step": 1871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.038379272307801,
+      "learning_rate": 9.306509683794077e-06,
+      "loss": 0.7228,
+      "step": 1872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9487725835663436,
+      "learning_rate": 9.305654106583735e-06,
+      "loss": 0.7507,
+      "step": 1873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.887653621297847,
+      "learning_rate": 9.304798041302959e-06,
+      "loss": 0.6303,
+      "step": 1874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.884397519021714,
+      "learning_rate": 9.30394148804879e-06,
+      "loss": 0.7304,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9021671805670946,
+      "learning_rate": 9.303084446918317e-06,
+      "loss": 0.6402,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8668810936120013,
+      "learning_rate": 9.302226918008699e-06,
+      "loss": 0.6011,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.793475054066034,
+      "learning_rate": 9.301368901417138e-06,
+      "loss": 0.6594,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.120762189756595,
+      "learning_rate": 9.300510397240894e-06,
+      "loss": 0.7769,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.300340831850511,
+      "learning_rate": 9.299651405577286e-06,
+      "loss": 0.7338,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9583934032957513,
+      "learning_rate": 9.298791926523683e-06,
+      "loss": 0.6799,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9638398719688912,
+      "learning_rate": 9.29793196017751e-06,
+      "loss": 0.6121,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.078735718161527,
+      "learning_rate": 9.297071506636256e-06,
+      "loss": 0.7302,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8398638850551559,
+      "learning_rate": 9.296210565997453e-06,
+      "loss": 0.6358,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7991017595645746,
+      "learning_rate": 9.295349138358693e-06,
+      "loss": 0.6302,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.738381217294171,
+      "learning_rate": 9.294487223817628e-06,
+      "loss": 0.5517,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7996019958686276,
+      "learning_rate": 9.293624822471957e-06,
+      "loss": 0.7795,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.04078306349116,
+      "learning_rate": 9.29276193441944e-06,
+      "loss": 0.7368,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.91957856399538,
+      "learning_rate": 9.29189855975789e-06,
+      "loss": 0.6685,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8755599618267553,
+      "learning_rate": 9.291034698585174e-06,
+      "loss": 0.6246,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.874547678759945,
+      "learning_rate": 9.290170350999217e-06,
+      "loss": 0.6128,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8915199997141332,
+      "learning_rate": 9.289305517098e-06,
+      "loss": 0.6731,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6748872030367143,
+      "learning_rate": 9.288440196979552e-06,
+      "loss": 0.5941,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7222395068337952,
+      "learning_rate": 9.287574390741965e-06,
+      "loss": 0.626,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.162350680248037,
+      "learning_rate": 9.286708098483383e-06,
+      "loss": 0.6518,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8336812096087574,
+      "learning_rate": 9.285841320302005e-06,
+      "loss": 0.5188,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2796155535184344,
+      "learning_rate": 9.284974056296084e-06,
+      "loss": 0.6926,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7071784585343275,
+      "learning_rate": 9.284106306563929e-06,
+      "loss": 0.6714,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6506461809111215,
+      "learning_rate": 9.283238071203907e-06,
+      "loss": 0.4856,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8781568684820324,
+      "learning_rate": 9.282369350314436e-06,
+      "loss": 0.6134,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9914978185733683,
+      "learning_rate": 9.28150014399399e-06,
+      "loss": 0.669,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8618014748000549,
+      "learning_rate": 9.280630452341099e-06,
+      "loss": 0.7035,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0645327853298525,
+      "learning_rate": 9.279760275454348e-06,
+      "loss": 0.6232,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7147310105966709,
+      "learning_rate": 9.278889613432375e-06,
+      "loss": 0.571,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9961133817036107,
+      "learning_rate": 9.278018466373877e-06,
+      "loss": 0.7313,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7602136103564725,
+      "learning_rate": 9.277146834377601e-06,
+      "loss": 0.6825,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8653246475670127,
+      "learning_rate": 9.276274717542352e-06,
+      "loss": 0.5975,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9630113986010014,
+      "learning_rate": 9.275402115966991e-06,
+      "loss": 0.7427,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.959845410358148,
+      "learning_rate": 9.27452902975043e-06,
+      "loss": 0.6803,
+      "step": 1909
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9040361049898615,
+      "learning_rate": 9.273655458991639e-06,
+      "loss": 0.5902,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0371368293310326,
+      "learning_rate": 9.272781403789644e-06,
+      "loss": 0.77,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8827798471230117,
+      "learning_rate": 9.271906864243523e-06,
+      "loss": 0.7044,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7981269024380084,
+      "learning_rate": 9.271031840452409e-06,
+      "loss": 0.6282,
+      "step": 1913
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.927440464619091,
+      "learning_rate": 9.270156332515492e-06,
+      "loss": 0.6634,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.946925561163631,
+      "learning_rate": 9.269280340532015e-06,
+      "loss": 0.6884,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7885164788997072,
+      "learning_rate": 9.268403864601279e-06,
+      "loss": 0.6771,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.740210348011536,
+      "learning_rate": 9.267526904822634e-06,
+      "loss": 0.5771,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8515761495261118,
+      "learning_rate": 9.266649461295491e-06,
+      "loss": 0.5456,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7858501985347042,
+      "learning_rate": 9.265771534119313e-06,
+      "loss": 0.6791,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7471660234340287,
+      "learning_rate": 9.264893123393618e-06,
+      "loss": 0.6445,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0050209490973847,
+      "learning_rate": 9.264014229217978e-06,
+      "loss": 0.6816,
+      "step": 1921
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7946185867470288,
+      "learning_rate": 9.263134851692019e-06,
+      "loss": 0.6995,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9338460880907744,
+      "learning_rate": 9.262254990915427e-06,
+      "loss": 0.7106,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1389806917513488,
+      "learning_rate": 9.261374646987939e-06,
+      "loss": 0.73,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5347394068987805,
+      "learning_rate": 9.260493820009343e-06,
+      "loss": 0.7632,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9525928567638877,
+      "learning_rate": 9.259612510079492e-06,
+      "loss": 0.6774,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0893340477312625,
+      "learning_rate": 9.258730717298281e-06,
+      "loss": 0.6821,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0650730951225564,
+      "learning_rate": 9.25784844176567e-06,
+      "loss": 0.7423,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7756513259013247,
+      "learning_rate": 9.256965683581669e-06,
+      "loss": 0.6076,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8897211314293623,
+      "learning_rate": 9.256082442846347e-06,
+      "loss": 0.651,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8280216607096946,
+      "learning_rate": 9.255198719659818e-06,
+      "loss": 0.7047,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0086394653794373,
+      "learning_rate": 9.25431451412226e-06,
+      "loss": 0.7658,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9127465298594584,
+      "learning_rate": 9.253429826333904e-06,
+      "loss": 0.618,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.826301070624459,
+      "learning_rate": 9.252544656395033e-06,
+      "loss": 0.7363,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9814272544927,
+      "learning_rate": 9.251659004405987e-06,
+      "loss": 0.6725,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9539602624207804,
+      "learning_rate": 9.25077287046716e-06,
+      "loss": 0.6789,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0032162419200046,
+      "learning_rate": 9.249886254678996e-06,
+      "loss": 0.6548,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7474671664321944,
+      "learning_rate": 9.248999157142006e-06,
+      "loss": 0.6904,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8925018918591514,
+      "learning_rate": 9.24811157795674e-06,
+      "loss": 0.6627,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.102589642676861,
+      "learning_rate": 9.247223517223816e-06,
+      "loss": 0.6378,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9593907278097766,
+      "learning_rate": 9.246334975043896e-06,
+      "loss": 0.7016,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8473768840344285,
+      "learning_rate": 9.245445951517705e-06,
+      "loss": 0.6621,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0796609104335326,
+      "learning_rate": 9.244556446746014e-06,
+      "loss": 0.7231,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9370198559859941,
+      "learning_rate": 9.243666460829659e-06,
+      "loss": 0.7223,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.053469252910239,
+      "learning_rate": 9.242775993869521e-06,
+      "loss": 0.7472,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9508358716738758,
+      "learning_rate": 9.241885045966543e-06,
+      "loss": 0.7247,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2450279545330183,
+      "learning_rate": 9.240993617221717e-06,
+      "loss": 0.7591,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.98183928572626,
+      "learning_rate": 9.24010170773609e-06,
+      "loss": 0.5902,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8675456918185938,
+      "learning_rate": 9.239209317610766e-06,
+      "loss": 0.6223,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1191395908530715,
+      "learning_rate": 9.238316446946907e-06,
+      "loss": 0.5797,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.875641872294237,
+      "learning_rate": 9.237423095845719e-06,
+      "loss": 0.7013,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7475234677089921,
+      "learning_rate": 9.23652926440847e-06,
+      "loss": 0.717,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.167896282874592,
+      "learning_rate": 9.23563495273648e-06,
+      "loss": 0.7803,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7286001629951508,
+      "learning_rate": 9.234740160931128e-06,
+      "loss": 0.573,
+      "step": 1954
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.266229733508544,
+      "learning_rate": 9.233844889093842e-06,
+      "loss": 0.5986,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.032639531476447,
+      "learning_rate": 9.232949137326104e-06,
+      "loss": 0.6202,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8892463261859485,
+      "learning_rate": 9.232052905729455e-06,
+      "loss": 0.7258,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.046805895646811,
+      "learning_rate": 9.231156194405488e-06,
+      "loss": 0.6311,
+      "step": 1958
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.125870271699572,
+      "learning_rate": 9.230259003455849e-06,
+      "loss": 0.7484,
+      "step": 1959
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9802725052113708,
+      "learning_rate": 9.229361332982241e-06,
+      "loss": 0.7112,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8447348073498075,
+      "learning_rate": 9.228463183086417e-06,
+      "loss": 0.6662,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8752305247433687,
+      "learning_rate": 9.227564553870192e-06,
+      "loss": 0.6881,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8026957544784308,
+      "learning_rate": 9.226665445435428e-06,
+      "loss": 0.5558,
+      "step": 1963
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9002089055531495,
+      "learning_rate": 9.225765857884044e-06,
+      "loss": 0.6797,
+      "step": 1964
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.81457632840884,
+      "learning_rate": 9.224865791318015e-06,
+      "loss": 0.6085,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7070620723972068,
+      "learning_rate": 9.223965245839367e-06,
+      "loss": 0.6723,
+      "step": 1966
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.105875693810757,
+      "learning_rate": 9.223064221550183e-06,
+      "loss": 0.7318,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0605135978619633,
+      "learning_rate": 9.222162718552598e-06,
+      "loss": 0.8245,
+      "step": 1968
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.238718777820951,
+      "learning_rate": 9.221260736948803e-06,
+      "loss": 0.7279,
+      "step": 1969
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.106801695108455,
+      "learning_rate": 9.220358276841044e-06,
+      "loss": 0.6773,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.665299995213328,
+      "learning_rate": 9.21945533833162e-06,
+      "loss": 0.6073,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7703299324370325,
+      "learning_rate": 9.218551921522882e-06,
+      "loss": 0.6295,
+      "step": 1972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9740036444553057,
+      "learning_rate": 9.217648026517238e-06,
+      "loss": 0.7084,
+      "step": 1973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.950547982837811,
+      "learning_rate": 9.216743653417154e-06,
+      "loss": 0.6489,
+      "step": 1974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8744414611735833,
+      "learning_rate": 9.215838802325139e-06,
+      "loss": 0.6199,
+      "step": 1975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.821711632746101,
+      "learning_rate": 9.214933473343765e-06,
+      "loss": 0.6074,
+      "step": 1976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.105716529248528,
+      "learning_rate": 9.21402766657566e-06,
+      "loss": 0.6654,
+      "step": 1977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8437304299013715,
+      "learning_rate": 9.2131213821235e-06,
+      "loss": 0.5944,
+      "step": 1978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9858141668801854,
+      "learning_rate": 9.212214620090016e-06,
+      "loss": 0.7049,
+      "step": 1979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.010065081274525,
+      "learning_rate": 9.211307380577997e-06,
+      "loss": 0.6696,
+      "step": 1980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8158239557845401,
+      "learning_rate": 9.210399663690282e-06,
+      "loss": 0.5995,
+      "step": 1981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.81249302739702,
+      "learning_rate": 9.209491469529767e-06,
+      "loss": 0.5868,
+      "step": 1982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.786339404329761,
+      "learning_rate": 9.208582798199402e-06,
+      "loss": 0.5719,
+      "step": 1983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9842702283554294,
+      "learning_rate": 9.207673649802187e-06,
+      "loss": 0.6408,
+      "step": 1984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5838760747233813,
+      "learning_rate": 9.206764024441181e-06,
+      "loss": 0.699,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9062861026868705,
+      "learning_rate": 9.205853922219494e-06,
+      "loss": 0.5787,
+      "step": 1986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7433682125475243,
+      "learning_rate": 9.204943343240293e-06,
+      "loss": 0.6315,
+      "step": 1987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.776540630355498,
+      "learning_rate": 9.2040322876068e-06,
+      "loss": 0.6349,
+      "step": 1988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8949934734507197,
+      "learning_rate": 9.203120755422282e-06,
+      "loss": 0.621,
+      "step": 1989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.013754218571031,
+      "learning_rate": 9.202208746790069e-06,
+      "loss": 0.6994,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6216738548604914,
+      "learning_rate": 9.201296261813542e-06,
+      "loss": 0.5664,
+      "step": 1991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.746950760201878,
+      "learning_rate": 9.20038330059614e-06,
+      "loss": 0.591,
+      "step": 1992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.044019666809054,
+      "learning_rate": 9.199469863241349e-06,
+      "loss": 0.6995,
+      "step": 1993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8783054256116338,
+      "learning_rate": 9.19855594985271e-06,
+      "loss": 0.6323,
+      "step": 1994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8318861488098133,
+      "learning_rate": 9.197641560533826e-06,
+      "loss": 0.7148,
+      "step": 1995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1884926195782684,
+      "learning_rate": 9.196726695388345e-06,
+      "loss": 0.6877,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.805667438158848,
+      "learning_rate": 9.19581135451997e-06,
+      "loss": 0.5927,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6125439114431472,
+      "learning_rate": 9.194895538032464e-06,
+      "loss": 0.5711,
+      "step": 1998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8686776317461966,
+      "learning_rate": 9.19397924602964e-06,
+      "loss": 0.6378,
+      "step": 1999
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.726924867450233,
+      "learning_rate": 9.193062478615363e-06,
+      "loss": 0.6463,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.146670051321336,
+      "learning_rate": 9.192145235893554e-06,
+      "loss": 0.6873,
+      "step": 2001
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.902666731160066,
+      "learning_rate": 9.191227517968189e-06,
+      "loss": 0.6114,
+      "step": 2002
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0687284258365057,
+      "learning_rate": 9.190309324943294e-06,
+      "loss": 0.712,
+      "step": 2003
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9026767117952792,
+      "learning_rate": 9.189390656922955e-06,
+      "loss": 0.6617,
+      "step": 2004
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8653739251485049,
+      "learning_rate": 9.188471514011304e-06,
+      "loss": 0.6892,
+      "step": 2005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.954084908851961,
+      "learning_rate": 9.187551896312536e-06,
+      "loss": 0.6613,
+      "step": 2006
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.855601522830974,
+      "learning_rate": 9.18663180393089e-06,
+      "loss": 0.5557,
+      "step": 2007
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0927778065386633,
+      "learning_rate": 9.185711236970666e-06,
+      "loss": 0.7094,
+      "step": 2008
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8476031345890764,
+      "learning_rate": 9.184790195536217e-06,
+      "loss": 0.6551,
+      "step": 2009
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9500926248954888,
+      "learning_rate": 9.183868679731947e-06,
+      "loss": 0.7358,
+      "step": 2010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8204670833802736,
+      "learning_rate": 9.182946689662314e-06,
+      "loss": 0.5735,
+      "step": 2011
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8122772993948097,
+      "learning_rate": 9.182024225431832e-06,
+      "loss": 0.5838,
+      "step": 2012
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0100838581998874,
+      "learning_rate": 9.181101287145067e-06,
+      "loss": 0.7309,
+      "step": 2013
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0174504680138416,
+      "learning_rate": 9.180177874906641e-06,
+      "loss": 0.7442,
+      "step": 2014
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8246358548978066,
+      "learning_rate": 9.179253988821225e-06,
+      "loss": 0.6281,
+      "step": 2015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.894892646996983,
+      "learning_rate": 9.17832962899355e-06,
+      "loss": 0.7488,
+      "step": 2016
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9804467894690987,
+      "learning_rate": 9.177404795528395e-06,
+      "loss": 0.5651,
+      "step": 2017
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9679539263706014,
+      "learning_rate": 9.176479488530594e-06,
+      "loss": 0.7255,
+      "step": 2018
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8442427437055013,
+      "learning_rate": 9.17555370810504e-06,
+      "loss": 0.6942,
+      "step": 2019
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.757489364681599,
+      "learning_rate": 9.174627454356673e-06,
+      "loss": 0.6402,
+      "step": 2020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0898518246682896,
+      "learning_rate": 9.17370072739049e-06,
+      "loss": 0.7267,
+      "step": 2021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.620345040398183,
+      "learning_rate": 9.172773527311541e-06,
+      "loss": 0.5773,
+      "step": 2022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8525961488620617,
+      "learning_rate": 9.171845854224925e-06,
+      "loss": 0.6378,
+      "step": 2023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7575000021739289,
+      "learning_rate": 9.170917708235806e-06,
+      "loss": 0.7192,
+      "step": 2024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.693136988470001,
+      "learning_rate": 9.16998908944939e-06,
+      "loss": 0.6771,
+      "step": 2025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9844607982204887,
+      "learning_rate": 9.169059997970943e-06,
+      "loss": 0.7001,
+      "step": 2026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0172894851733534,
+      "learning_rate": 9.168130433905783e-06,
+      "loss": 0.6599,
+      "step": 2027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.721838081002367,
+      "learning_rate": 9.167200397359279e-06,
+      "loss": 0.6509,
+      "step": 2028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6521500094494894,
+      "learning_rate": 9.16626988843686e-06,
+      "loss": 0.575,
+      "step": 2029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8866447487668851,
+      "learning_rate": 9.165338907244e-06,
+      "loss": 0.6215,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8011492071948654,
+      "learning_rate": 9.164407453886234e-06,
+      "loss": 0.7151,
+      "step": 2031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0920355496930485,
+      "learning_rate": 9.163475528469148e-06,
+      "loss": 0.6656,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.070222708565936,
+      "learning_rate": 9.16254313109838e-06,
+      "loss": 0.6258,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.123867145449913,
+      "learning_rate": 9.16161026187962e-06,
+      "loss": 0.66,
+      "step": 2034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8609378519738486,
+      "learning_rate": 9.160676920918618e-06,
+      "loss": 0.6742,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9956472499189537,
+      "learning_rate": 9.159743108321173e-06,
+      "loss": 0.7092,
+      "step": 2036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.308642760350203,
+      "learning_rate": 9.158808824193135e-06,
+      "loss": 0.7591,
+      "step": 2037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9751497331037764,
+      "learning_rate": 9.157874068640414e-06,
+      "loss": 0.6626,
+      "step": 2038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.172657081296383,
+      "learning_rate": 9.156938841768965e-06,
+      "loss": 0.6358,
+      "step": 2039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7053858901126682,
+      "learning_rate": 9.156003143684808e-06,
+      "loss": 0.5645,
+      "step": 2040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.669173862779289,
+      "learning_rate": 9.155066974494005e-06,
+      "loss": 0.5918,
+      "step": 2041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.887929904534194,
+      "learning_rate": 9.154130334302677e-06,
+      "loss": 0.6318,
+      "step": 2042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8383647204550129,
+      "learning_rate": 9.153193223216998e-06,
+      "loss": 0.5643,
+      "step": 2043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8251294761789811,
+      "learning_rate": 9.152255641343196e-06,
+      "loss": 0.5895,
+      "step": 2044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8955577533945258,
+      "learning_rate": 9.151317588787546e-06,
+      "loss": 0.6413,
+      "step": 2045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6992665218554355,
+      "learning_rate": 9.150379065656389e-06,
+      "loss": 0.6495,
+      "step": 2046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.267187657771848,
+      "learning_rate": 9.149440072056109e-06,
+      "loss": 0.686,
+      "step": 2047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8039609054222494,
+      "learning_rate": 9.148500608093144e-06,
+      "loss": 0.7414,
+      "step": 2048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7381527742569263,
+      "learning_rate": 9.147560673873991e-06,
+      "loss": 0.6521,
+      "step": 2049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9806331929152141,
+      "learning_rate": 9.146620269505194e-06,
+      "loss": 0.6065,
+      "step": 2050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8251796761712276,
+      "learning_rate": 9.145679395093357e-06,
+      "loss": 0.6152,
+      "step": 2051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9744130556131152,
+      "learning_rate": 9.144738050745129e-06,
+      "loss": 0.6398,
+      "step": 2052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8608757900689004,
+      "learning_rate": 9.143796236567218e-06,
+      "loss": 0.6509,
+      "step": 2053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0005118742578807,
+      "learning_rate": 9.142853952666385e-06,
+      "loss": 0.6495,
+      "step": 2054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8723374871113159,
+      "learning_rate": 9.141911199149443e-06,
+      "loss": 0.7244,
+      "step": 2055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9810936161123067,
+      "learning_rate": 9.14096797612326e-06,
+      "loss": 0.6901,
+      "step": 2056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7755395689456877,
+      "learning_rate": 9.140024283694752e-06,
+      "loss": 0.6726,
+      "step": 2057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0113646493029913,
+      "learning_rate": 9.139080121970895e-06,
+      "loss": 0.728,
+      "step": 2058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9395275583029608,
+      "learning_rate": 9.138135491058715e-06,
+      "loss": 0.6925,
+      "step": 2059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9086311561274867,
+      "learning_rate": 9.13719039106529e-06,
+      "loss": 0.6337,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0135221869102757,
+      "learning_rate": 9.136244822097754e-06,
+      "loss": 0.7061,
+      "step": 2061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7953338697723547,
+      "learning_rate": 9.13529878426329e-06,
+      "loss": 0.6007,
+      "step": 2062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9043752154264741,
+      "learning_rate": 9.134352277669139e-06,
+      "loss": 0.7779,
+      "step": 2063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8572959317924327,
+      "learning_rate": 9.13340530242259e-06,
+      "loss": 0.6555,
+      "step": 2064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.901098228455696,
+      "learning_rate": 9.132457858630993e-06,
+      "loss": 0.6443,
+      "step": 2065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8820392432706698,
+      "learning_rate": 9.131509946401744e-06,
+      "loss": 0.6921,
+      "step": 2066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.63515390188287,
+      "learning_rate": 9.130561565842293e-06,
+      "loss": 0.6493,
+      "step": 2067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8004550658092822,
+      "learning_rate": 9.129612717060145e-06,
+      "loss": 0.6021,
+      "step": 2068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8135533498191938,
+      "learning_rate": 9.128663400162859e-06,
+      "loss": 0.6516,
+      "step": 2069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9883782219408912,
+      "learning_rate": 9.12771361525804e-06,
+      "loss": 0.5804,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9267639414496214,
+      "learning_rate": 9.12676336245336e-06,
+      "loss": 0.7172,
+      "step": 2071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.241296242570306,
+      "learning_rate": 9.125812641856529e-06,
+      "loss": 0.6466,
+      "step": 2072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8465439639348022,
+      "learning_rate": 9.124861453575318e-06,
+      "loss": 0.5546,
+      "step": 2073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8408819912841634,
+      "learning_rate": 9.123909797717551e-06,
+      "loss": 0.6116,
+      "step": 2074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.076263596235832,
+      "learning_rate": 9.122957674391103e-06,
+      "loss": 0.7471,
+      "step": 2075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.161931389146741,
+      "learning_rate": 9.122005083703901e-06,
+      "loss": 0.7757,
+      "step": 2076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.832263381482899,
+      "learning_rate": 9.12105202576393e-06,
+      "loss": 0.6342,
+      "step": 2077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9116255589770395,
+      "learning_rate": 9.120098500679222e-06,
+      "loss": 0.5791,
+      "step": 2078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7512087408271144,
+      "learning_rate": 9.119144508557867e-06,
+      "loss": 0.6596,
+      "step": 2079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.316242858851651,
+      "learning_rate": 9.118190049508003e-06,
+      "loss": 0.7156,
+      "step": 2080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.098398279289849,
+      "learning_rate": 9.117235123637822e-06,
+      "loss": 0.7091,
+      "step": 2081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.005617652197642,
+      "learning_rate": 9.116279731055574e-06,
+      "loss": 0.7734,
+      "step": 2082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3181625277963342,
+      "learning_rate": 9.115323871869554e-06,
+      "loss": 0.7269,
+      "step": 2083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.09049499201878,
+      "learning_rate": 9.11436754618812e-06,
+      "loss": 0.6962,
+      "step": 2084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8233048319574325,
+      "learning_rate": 9.113410754119671e-06,
+      "loss": 0.657,
+      "step": 2085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7611399694198617,
+      "learning_rate": 9.112453495772668e-06,
+      "loss": 0.7066,
+      "step": 2086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.30757117904449,
+      "learning_rate": 9.111495771255623e-06,
+      "loss": 0.7161,
+      "step": 2087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7222869131586123,
+      "learning_rate": 9.110537580677094e-06,
+      "loss": 0.6955,
+      "step": 2088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.777824577605693,
+      "learning_rate": 9.109578924145705e-06,
+      "loss": 0.6405,
+      "step": 2089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.867004091367024,
+      "learning_rate": 9.108619801770117e-06,
+      "loss": 0.7337,
+      "step": 2090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8341940643841907,
+      "learning_rate": 9.10766021365906e-06,
+      "loss": 0.6397,
+      "step": 2091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7108974210893473,
+      "learning_rate": 9.106700159921301e-06,
+      "loss": 0.6176,
+      "step": 2092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.763586609221222,
+      "learning_rate": 9.105739640665675e-06,
+      "loss": 0.7577,
+      "step": 2093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8179769534117416,
+      "learning_rate": 9.104778656001057e-06,
+      "loss": 0.6602,
+      "step": 2094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0710656099544944,
+      "learning_rate": 9.103817206036383e-06,
+      "loss": 0.6258,
+      "step": 2095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9636515939076291,
+      "learning_rate": 9.102855290880637e-06,
+      "loss": 0.6908,
+      "step": 2096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7697645394492914,
+      "learning_rate": 9.101892910642858e-06,
+      "loss": 0.5772,
+      "step": 2097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8229562492894091,
+      "learning_rate": 9.100930065432136e-06,
+      "loss": 0.6319,
+      "step": 2098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9665961596361246,
+      "learning_rate": 9.09996675535762e-06,
+      "loss": 0.6321,
+      "step": 2099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9153659580729374,
+      "learning_rate": 9.099002980528502e-06,
+      "loss": 0.7386,
+      "step": 2100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8929607008969653,
+      "learning_rate": 9.098038741054032e-06,
+      "loss": 0.6038,
+      "step": 2101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.122236111686674,
+      "learning_rate": 9.097074037043512e-06,
+      "loss": 0.7902,
+      "step": 2102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8063786343678705,
+      "learning_rate": 9.0961088686063e-06,
+      "loss": 0.6154,
+      "step": 2103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8859714107413783,
+      "learning_rate": 9.095143235851797e-06,
+      "loss": 0.625,
+      "step": 2104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8899963332781613,
+      "learning_rate": 9.094177138889468e-06,
+      "loss": 0.6285,
+      "step": 2105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.823065246187759,
+      "learning_rate": 9.093210577828826e-06,
+      "loss": 0.5222,
+      "step": 2106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8409605037443102,
+      "learning_rate": 9.092243552779434e-06,
+      "loss": 0.545,
+      "step": 2107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7950723144319713,
+      "learning_rate": 9.091276063850909e-06,
+      "loss": 0.7386,
+      "step": 2108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8619428941975487,
+      "learning_rate": 9.090308111152924e-06,
+      "loss": 0.6613,
+      "step": 2109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8760132631684212,
+      "learning_rate": 9.0893396947952e-06,
+      "loss": 0.6567,
+      "step": 2110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5393612059524457,
+      "learning_rate": 9.088370814887512e-06,
+      "loss": 0.6192,
+      "step": 2111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1997633108009746,
+      "learning_rate": 9.08740147153969e-06,
+      "loss": 0.7291,
+      "step": 2112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0003762945667263,
+      "learning_rate": 9.086431664861615e-06,
+      "loss": 0.6385,
+      "step": 2113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9280568210847042,
+      "learning_rate": 9.085461394963218e-06,
+      "loss": 0.7926,
+      "step": 2114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0839385743826337,
+      "learning_rate": 9.084490661954487e-06,
+      "loss": 0.7176,
+      "step": 2115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.751926913173965,
+      "learning_rate": 9.083519465945456e-06,
+      "loss": 0.6207,
+      "step": 2116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6790647212307659,
+      "learning_rate": 9.082547807046218e-06,
+      "loss": 0.5995,
+      "step": 2117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9768132587469367,
+      "learning_rate": 9.081575685366919e-06,
+      "loss": 0.6638,
+      "step": 2118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.873044803611899,
+      "learning_rate": 9.080603101017751e-06,
+      "loss": 0.6287,
+      "step": 2119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9026675442854966,
+      "learning_rate": 9.079630054108962e-06,
+      "loss": 0.7277,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.786052227401782,
+      "learning_rate": 9.078656544750854e-06,
+      "loss": 0.5283,
+      "step": 2121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.835617801364485,
+      "learning_rate": 9.07768257305378e-06,
+      "loss": 0.6734,
+      "step": 2122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9059077413421042,
+      "learning_rate": 9.07670813912814e-06,
+      "loss": 0.5662,
+      "step": 2123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8430365616658317,
+      "learning_rate": 9.0757332430844e-06,
+      "loss": 0.6786,
+      "step": 2124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0912152253587233,
+      "learning_rate": 9.074757885033065e-06,
+      "loss": 0.6699,
+      "step": 2125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8326045285025117,
+      "learning_rate": 9.073782065084699e-06,
+      "loss": 0.6516,
+      "step": 2126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1141153087285747,
+      "learning_rate": 9.072805783349916e-06,
+      "loss": 0.6504,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7665606289047024,
+      "learning_rate": 9.071829039939382e-06,
+      "loss": 0.5843,
+      "step": 2128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.965086242675083,
+      "learning_rate": 9.070851834963818e-06,
+      "loss": 0.606,
+      "step": 2129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8212064883638712,
+      "learning_rate": 9.069874168533996e-06,
+      "loss": 0.5638,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.914061308706072,
+      "learning_rate": 9.06889604076074e-06,
+      "loss": 0.7702,
+      "step": 2131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9281649798903766,
+      "learning_rate": 9.067917451754926e-06,
+      "loss": 0.6979,
+      "step": 2132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9724035219322607,
+      "learning_rate": 9.06693840162748e-06,
+      "loss": 0.6326,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.926538052376226,
+      "learning_rate": 9.065958890489388e-06,
+      "loss": 0.6708,
+      "step": 2134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8135660948677115,
+      "learning_rate": 9.06497891845168e-06,
+      "loss": 0.6908,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0359781123918546,
+      "learning_rate": 9.063998485625442e-06,
+      "loss": 0.8177,
+      "step": 2136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8070490443578695,
+      "learning_rate": 9.063017592121812e-06,
+      "loss": 0.6308,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1534040200827858,
+      "learning_rate": 9.062036238051978e-06,
+      "loss": 0.7568,
+      "step": 2138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9932070854185635,
+      "learning_rate": 9.061054423527185e-06,
+      "loss": 0.6454,
+      "step": 2139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.06303057621518,
+      "learning_rate": 9.060072148658726e-06,
+      "loss": 0.7897,
+      "step": 2140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8541768464221198,
+      "learning_rate": 9.059089413557946e-06,
+      "loss": 0.7146,
+      "step": 2141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.131577767008987,
+      "learning_rate": 9.058106218336244e-06,
+      "loss": 0.64,
+      "step": 2142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.957951302391662,
+      "learning_rate": 9.057122563105074e-06,
+      "loss": 0.6772,
+      "step": 2143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8796034764382545,
+      "learning_rate": 9.056138447975936e-06,
+      "loss": 0.5828,
+      "step": 2144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9984362920448873,
+      "learning_rate": 9.055153873060387e-06,
+      "loss": 0.6582,
+      "step": 2145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.818826471290181,
+      "learning_rate": 9.05416883847003e-06,
+      "loss": 0.6103,
+      "step": 2146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9010961716566377,
+      "learning_rate": 9.05318334431653e-06,
+      "loss": 0.6502,
+      "step": 2147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6452104619968555,
+      "learning_rate": 9.052197390711594e-06,
+      "loss": 0.6073,
+      "step": 2148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0593507671841054,
+      "learning_rate": 9.051210977766987e-06,
+      "loss": 0.7939,
+      "step": 2149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8947339812978212,
+      "learning_rate": 9.050224105594525e-06,
+      "loss": 0.7403,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7701582298112364,
+      "learning_rate": 9.049236774306073e-06,
+      "loss": 0.7139,
+      "step": 2151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.769772295770342,
+      "learning_rate": 9.048248984013557e-06,
+      "loss": 0.6318,
+      "step": 2152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8668708776556142,
+      "learning_rate": 9.04726073482894e-06,
+      "loss": 0.6786,
+      "step": 2153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8599052090254813,
+      "learning_rate": 9.046272026864253e-06,
+      "loss": 0.6792,
+      "step": 2154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.306307988225551,
+      "learning_rate": 9.045282860231567e-06,
+      "loss": 0.8036,
+      "step": 2155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0248004612699835,
+      "learning_rate": 9.044293235043013e-06,
+      "loss": 0.5483,
+      "step": 2156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0040416841239312,
+      "learning_rate": 9.04330315141077e-06,
+      "loss": 0.5838,
+      "step": 2157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8934644260258988,
+      "learning_rate": 9.042312609447066e-06,
+      "loss": 0.7741,
+      "step": 2158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8369307032598101,
+      "learning_rate": 9.041321609264189e-06,
+      "loss": 0.6881,
+      "step": 2159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8734585208709953,
+      "learning_rate": 9.040330150974472e-06,
+      "loss": 0.7302,
+      "step": 2160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.274845630281377,
+      "learning_rate": 9.039338234690304e-06,
+      "loss": 0.7115,
+      "step": 2161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9120910239340116,
+      "learning_rate": 9.038345860524123e-06,
+      "loss": 0.6324,
+      "step": 2162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9268141093838203,
+      "learning_rate": 9.037353028588421e-06,
+      "loss": 0.6863,
+      "step": 2163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7626090493830637,
+      "learning_rate": 9.036359738995741e-06,
+      "loss": 0.7137,
+      "step": 2164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9785455080305587,
+      "learning_rate": 9.035365991858679e-06,
+      "loss": 0.6363,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7537820647478297,
+      "learning_rate": 9.034371787289879e-06,
+      "loss": 0.6999,
+      "step": 2166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8144705683970634,
+      "learning_rate": 9.033377125402045e-06,
+      "loss": 0.6111,
+      "step": 2167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0784711520480683,
+      "learning_rate": 9.032382006307923e-06,
+      "loss": 0.7317,
+      "step": 2168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.055321365353015,
+      "learning_rate": 9.031386430120315e-06,
+      "loss": 0.6697,
+      "step": 2169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9973752694136535,
+      "learning_rate": 9.030390396952077e-06,
+      "loss": 0.6653,
+      "step": 2170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1908786377684004,
+      "learning_rate": 9.029393906916118e-06,
+      "loss": 0.7126,
+      "step": 2171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2196682641106045,
+      "learning_rate": 9.028396960125392e-06,
+      "loss": 0.6975,
+      "step": 2172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.153676140622691,
+      "learning_rate": 9.02739955669291e-06,
+      "loss": 0.7135,
+      "step": 2173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7443698256344395,
+      "learning_rate": 9.02640169673173e-06,
+      "loss": 0.6252,
+      "step": 2174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8900431422297443,
+      "learning_rate": 9.02540338035497e-06,
+      "loss": 0.733,
+      "step": 2175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7568087601555695,
+      "learning_rate": 9.024404607675792e-06,
+      "loss": 0.673,
+      "step": 2176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.898316533786659,
+      "learning_rate": 9.023405378807413e-06,
+      "loss": 0.6297,
+      "step": 2177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.977725542941323,
+      "learning_rate": 9.022405693863102e-06,
+      "loss": 0.7268,
+      "step": 2178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1481357518729243,
+      "learning_rate": 9.02140555295618e-06,
+      "loss": 0.6296,
+      "step": 2179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9978345929549044,
+      "learning_rate": 9.020404956200016e-06,
+      "loss": 0.5896,
+      "step": 2180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.84587899229959,
+      "learning_rate": 9.019403903708036e-06,
+      "loss": 0.6713,
+      "step": 2181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8092614755951855,
+      "learning_rate": 9.018402395593711e-06,
+      "loss": 0.6312,
+      "step": 2182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0435447786587178,
+      "learning_rate": 9.017400431970572e-06,
+      "loss": 0.6963,
+      "step": 2183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7252658369683462,
+      "learning_rate": 9.016398012952196e-06,
+      "loss": 0.5749,
+      "step": 2184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7329348782754614,
+      "learning_rate": 9.015395138652212e-06,
+      "loss": 0.6343,
+      "step": 2185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0542689022507004,
+      "learning_rate": 9.014391809184302e-06,
+      "loss": 0.754,
+      "step": 2186
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0538719175252464,
+      "learning_rate": 9.013388024662199e-06,
+      "loss": 0.7492,
+      "step": 2187
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.194772716927874,
+      "learning_rate": 9.012383785199688e-06,
+      "loss": 0.6571,
+      "step": 2188
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7314238687560968,
+      "learning_rate": 9.011379090910605e-06,
+      "loss": 0.5858,
+      "step": 2189
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8309501359614229,
+      "learning_rate": 9.010373941908839e-06,
+      "loss": 0.561,
+      "step": 2190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9401424308724535,
+      "learning_rate": 9.009368338308328e-06,
+      "loss": 0.6818,
+      "step": 2191
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0913313231859534,
+      "learning_rate": 9.008362280223062e-06,
+      "loss": 0.7318,
+      "step": 2192
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.01779899943987,
+      "learning_rate": 9.007355767767085e-06,
+      "loss": 0.6818,
+      "step": 2193
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8462281597980958,
+      "learning_rate": 9.006348801054491e-06,
+      "loss": 0.6496,
+      "step": 2194
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9187146206986834,
+      "learning_rate": 9.005341380199426e-06,
+      "loss": 0.6457,
+      "step": 2195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9328520519266845,
+      "learning_rate": 9.004333505316085e-06,
+      "loss": 0.7879,
+      "step": 2196
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.708215185089118,
+      "learning_rate": 9.003325176518718e-06,
+      "loss": 0.5946,
+      "step": 2197
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9026124558937254,
+      "learning_rate": 9.002316393921623e-06,
+      "loss": 0.7187,
+      "step": 2198
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.002549571319192,
+      "learning_rate": 9.001307157639153e-06,
+      "loss": 0.711,
+      "step": 2199
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7359598802500287,
+      "learning_rate": 9.000297467785708e-06,
+      "loss": 0.7036,
+      "step": 2200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9435004256949489,
+      "learning_rate": 8.999287324475745e-06,
+      "loss": 0.6485,
+      "step": 2201
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9218778678631865,
+      "learning_rate": 8.998276727823769e-06,
+      "loss": 0.6781,
+      "step": 2202
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8933013319011063,
+      "learning_rate": 8.997265677944336e-06,
+      "loss": 0.6569,
+      "step": 2203
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9531255465327648,
+      "learning_rate": 8.996254174952056e-06,
+      "loss": 0.7081,
+      "step": 2204
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0631797487414527,
+      "learning_rate": 8.995242218961586e-06,
+      "loss": 0.6542,
+      "step": 2205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0026963960183264,
+      "learning_rate": 8.99422981008764e-06,
+      "loss": 0.6282,
+      "step": 2206
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9931938931672886,
+      "learning_rate": 8.993216948444978e-06,
+      "loss": 0.6685,
+      "step": 2207
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.03677992763092,
+      "learning_rate": 8.992203634148412e-06,
+      "loss": 0.5603,
+      "step": 2208
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9775804447589407,
+      "learning_rate": 8.991189867312813e-06,
+      "loss": 0.7024,
+      "step": 2209
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9936308622181156,
+      "learning_rate": 8.990175648053093e-06,
+      "loss": 0.7432,
+      "step": 2210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7983836508704376,
+      "learning_rate": 8.989160976484218e-06,
+      "loss": 0.6267,
+      "step": 2211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7341209015462455,
+      "learning_rate": 8.98814585272121e-06,
+      "loss": 0.6546,
+      "step": 2212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7535109689213118,
+      "learning_rate": 8.987130276879137e-06,
+      "loss": 0.6133,
+      "step": 2213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.926070930858834,
+      "learning_rate": 8.986114249073122e-06,
+      "loss": 0.7027,
+      "step": 2214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9392805940904663,
+      "learning_rate": 8.985097769418337e-06,
+      "loss": 0.6536,
+      "step": 2215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0611644129881763,
+      "learning_rate": 8.984080838030005e-06,
+      "loss": 0.8039,
+      "step": 2216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.142950425462867,
+      "learning_rate": 8.983063455023402e-06,
+      "loss": 0.7716,
+      "step": 2217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.054058905566524,
+      "learning_rate": 8.982045620513855e-06,
+      "loss": 0.7707,
+      "step": 2218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.889097548556331,
+      "learning_rate": 8.981027334616737e-06,
+      "loss": 0.5825,
+      "step": 2219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8795725986969594,
+      "learning_rate": 8.98000859744748e-06,
+      "loss": 0.6551,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.882519672163551,
+      "learning_rate": 8.978989409121565e-06,
+      "loss": 0.682,
+      "step": 2221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0537589705875603,
+      "learning_rate": 8.97796976975452e-06,
+      "loss": 0.7304,
+      "step": 2222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0767498129779574,
+      "learning_rate": 8.976949679461928e-06,
+      "loss": 0.7041,
+      "step": 2223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.110118698229804,
+      "learning_rate": 8.975929138359423e-06,
+      "loss": 0.6485,
+      "step": 2224
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9109900497230985,
+      "learning_rate": 8.974908146562686e-06,
+      "loss": 0.7517,
+      "step": 2225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.015281369277223,
+      "learning_rate": 8.973886704187457e-06,
+      "loss": 0.6787,
+      "step": 2226
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8262978970480288,
+      "learning_rate": 8.972864811349518e-06,
+      "loss": 0.6435,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.210288335596177,
+      "learning_rate": 8.97184246816471e-06,
+      "loss": 0.7442,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8288254991676658,
+      "learning_rate": 8.970819674748917e-06,
+      "loss": 0.6972,
+      "step": 2229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7051979804580626,
+      "learning_rate": 8.969796431218081e-06,
+      "loss": 0.5441,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9169044631050145,
+      "learning_rate": 8.968772737688193e-06,
+      "loss": 0.6199,
+      "step": 2231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0655732597769925,
+      "learning_rate": 8.967748594275294e-06,
+      "loss": 0.766,
+      "step": 2232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.978801046277295,
+      "learning_rate": 8.966724001095477e-06,
+      "loss": 0.6957,
+      "step": 2233
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9521332143429797,
+      "learning_rate": 8.965698958264883e-06,
+      "loss": 0.6472,
+      "step": 2234
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.83147289535707,
+      "learning_rate": 8.96467346589971e-06,
+      "loss": 0.6152,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9956163284753403,
+      "learning_rate": 8.963647524116202e-06,
+      "loss": 0.6321,
+      "step": 2236
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.133422804155597,
+      "learning_rate": 8.962621133030655e-06,
+      "loss": 0.6059,
+      "step": 2237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9380345874646834,
+      "learning_rate": 8.961594292759416e-06,
+      "loss": 0.789,
+      "step": 2238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8865426268162737,
+      "learning_rate": 8.960567003418882e-06,
+      "loss": 0.6757,
+      "step": 2239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.973482306954926,
+      "learning_rate": 8.959539265125507e-06,
+      "loss": 0.6109,
+      "step": 2240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8719939069545541,
+      "learning_rate": 8.958511077995786e-06,
+      "loss": 0.6642,
+      "step": 2241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.027942589354299,
+      "learning_rate": 8.957482442146271e-06,
+      "loss": 0.6152,
+      "step": 2242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9377797775087606,
+      "learning_rate": 8.956453357693565e-06,
+      "loss": 0.6306,
+      "step": 2243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1397437040204927,
+      "learning_rate": 8.955423824754319e-06,
+      "loss": 0.6442,
+      "step": 2244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.850879857064714,
+      "learning_rate": 8.954393843445239e-06,
+      "loss": 0.7046,
+      "step": 2245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9391040464860019,
+      "learning_rate": 8.953363413883077e-06,
+      "loss": 0.633,
+      "step": 2246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7458294438697912,
+      "learning_rate": 8.952332536184639e-06,
+      "loss": 0.6093,
+      "step": 2247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7462764140457245,
+      "learning_rate": 8.951301210466779e-06,
+      "loss": 0.6411,
+      "step": 2248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0005982387947276,
+      "learning_rate": 8.950269436846405e-06,
+      "loss": 0.6312,
+      "step": 2249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9155274879580744,
+      "learning_rate": 8.949237215440476e-06,
+      "loss": 0.6798,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.855490598220143,
+      "learning_rate": 8.948204546365996e-06,
+      "loss": 0.648,
+      "step": 2251
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.189347602052192,
+      "learning_rate": 8.94717142974003e-06,
+      "loss": 0.629,
+      "step": 2252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6403712428379003,
+      "learning_rate": 8.946137865679683e-06,
+      "loss": 0.5136,
+      "step": 2253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9275044475114163,
+      "learning_rate": 8.945103854302118e-06,
+      "loss": 0.6351,
+      "step": 2254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.813897721890047,
+      "learning_rate": 8.944069395724541e-06,
+      "loss": 0.7123,
+      "step": 2255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0580374034292044,
+      "learning_rate": 8.943034490064222e-06,
+      "loss": 0.6749,
+      "step": 2256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0017399104314246,
+      "learning_rate": 8.941999137438466e-06,
+      "loss": 0.695,
+      "step": 2257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9378953715759386,
+      "learning_rate": 8.940963337964642e-06,
+      "loss": 0.6528,
+      "step": 2258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.041326591727269,
+      "learning_rate": 8.93992709176016e-06,
+      "loss": 0.7131,
+      "step": 2259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9285147868178183,
+      "learning_rate": 8.938890398942482e-06,
+      "loss": 0.7492,
+      "step": 2260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8238045339749287,
+      "learning_rate": 8.93785325962913e-06,
+      "loss": 0.6304,
+      "step": 2261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0053183644082613,
+      "learning_rate": 8.936815673937665e-06,
+      "loss": 0.6946,
+      "step": 2262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.056091891002725,
+      "learning_rate": 8.935777641985704e-06,
+      "loss": 0.7323,
+      "step": 2263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8523458813916833,
+      "learning_rate": 8.934739163890914e-06,
+      "loss": 0.6583,
+      "step": 2264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9742354499979657,
+      "learning_rate": 8.933700239771013e-06,
+      "loss": 0.6349,
+      "step": 2265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9663161560770108,
+      "learning_rate": 8.932660869743766e-06,
+      "loss": 0.6875,
+      "step": 2266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7729449620047906,
+      "learning_rate": 8.931621053926998e-06,
+      "loss": 0.5993,
+      "step": 2267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.776396253549101,
+      "learning_rate": 8.930580792438571e-06,
+      "loss": 0.6088,
+      "step": 2268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.643385454952664,
+      "learning_rate": 8.929540085396409e-06,
+      "loss": 0.6575,
+      "step": 2269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.946490262609231,
+      "learning_rate": 8.92849893291848e-06,
+      "loss": 0.6845,
+      "step": 2270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9913037797693673,
+      "learning_rate": 8.927457335122807e-06,
+      "loss": 0.6021,
+      "step": 2271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9935906721529877,
+      "learning_rate": 8.926415292127458e-06,
+      "loss": 0.6359,
+      "step": 2272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9874498733381067,
+      "learning_rate": 8.925372804050554e-06,
+      "loss": 0.8141,
+      "step": 2273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8641504075702195,
+      "learning_rate": 8.924329871010271e-06,
+      "loss": 0.5803,
+      "step": 2274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0248361904003906,
+      "learning_rate": 8.92328649312483e-06,
+      "loss": 0.7611,
+      "step": 2275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8881478352562464,
+      "learning_rate": 8.922242670512501e-06,
+      "loss": 0.7352,
+      "step": 2276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.643021574441532,
+      "learning_rate": 8.92119840329161e-06,
+      "loss": 0.5815,
+      "step": 2277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8384891810153987,
+      "learning_rate": 8.92015369158053e-06,
+      "loss": 0.6841,
+      "step": 2278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8084240539765695,
+      "learning_rate": 8.919108535497684e-06,
+      "loss": 0.6603,
+      "step": 2279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.78225246310456,
+      "learning_rate": 8.91806293516155e-06,
+      "loss": 0.6892,
+      "step": 2280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0298859406349905,
+      "learning_rate": 8.917016890690648e-06,
+      "loss": 0.7486,
+      "step": 2281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0279639350940566,
+      "learning_rate": 8.915970402203555e-06,
+      "loss": 0.6696,
+      "step": 2282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.994856285908655,
+      "learning_rate": 8.914923469818897e-06,
+      "loss": 0.6637,
+      "step": 2283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8271612294370347,
+      "learning_rate": 8.913876093655351e-06,
+      "loss": 0.6463,
+      "step": 2284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.784844135622662,
+      "learning_rate": 8.912828273831639e-06,
+      "loss": 0.6222,
+      "step": 2285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7882797824424788,
+      "learning_rate": 8.911780010466542e-06,
+      "loss": 0.745,
+      "step": 2286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.087910231596911,
+      "learning_rate": 8.910731303678881e-06,
+      "loss": 0.6256,
+      "step": 2287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9216932425265978,
+      "learning_rate": 8.90968215358754e-06,
+      "loss": 0.6616,
+      "step": 2288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9486809117012627,
+      "learning_rate": 8.908632560311441e-06,
+      "loss": 0.6927,
+      "step": 2289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8170629926411348,
+      "learning_rate": 8.907582523969562e-06,
+      "loss": 0.679,
+      "step": 2290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.761269885787969,
+      "learning_rate": 8.906532044680933e-06,
+      "loss": 0.6878,
+      "step": 2291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8632205027463271,
+      "learning_rate": 8.905481122564628e-06,
+      "loss": 0.6873,
+      "step": 2292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9046793180822075,
+      "learning_rate": 8.90442975773978e-06,
+      "loss": 0.6389,
+      "step": 2293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9488081969903266,
+      "learning_rate": 8.903377950325563e-06,
+      "loss": 0.6598,
+      "step": 2294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1749846253223257,
+      "learning_rate": 8.902325700441207e-06,
+      "loss": 0.6878,
+      "step": 2295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7711244804453699,
+      "learning_rate": 8.901273008205991e-06,
+      "loss": 0.5985,
+      "step": 2296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8428616471257615,
+      "learning_rate": 8.900219873739242e-06,
+      "loss": 0.7583,
+      "step": 2297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2720506937274974,
+      "learning_rate": 8.89916629716034e-06,
+      "loss": 0.6609,
+      "step": 2298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.124027712091581,
+      "learning_rate": 8.898112278588713e-06,
+      "loss": 0.6003,
+      "step": 2299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7844632084108558,
+      "learning_rate": 8.897057818143842e-06,
+      "loss": 0.6297,
+      "step": 2300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.902743216888347,
+      "learning_rate": 8.896002915945254e-06,
+      "loss": 0.6839,
+      "step": 2301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7514502795812927,
+      "learning_rate": 8.89494757211253e-06,
+      "loss": 0.6339,
+      "step": 2302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.648375420162822,
+      "learning_rate": 8.893891786765298e-06,
+      "loss": 0.5742,
+      "step": 2303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.87874832019259,
+      "learning_rate": 8.892835560023236e-06,
+      "loss": 0.686,
+      "step": 2304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.119276599077235,
+      "learning_rate": 8.891778892006077e-06,
+      "loss": 0.7002,
+      "step": 2305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8114144842978803,
+      "learning_rate": 8.890721782833596e-06,
+      "loss": 0.6562,
+      "step": 2306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9513465793448206,
+      "learning_rate": 8.889664232625626e-06,
+      "loss": 0.6201,
+      "step": 2307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9660798153237917,
+      "learning_rate": 8.888606241502044e-06,
+      "loss": 0.6004,
+      "step": 2308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9794487766731317,
+      "learning_rate": 8.88754780958278e-06,
+      "loss": 0.7013,
+      "step": 2309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7888928185017705,
+      "learning_rate": 8.886488936987817e-06,
+      "loss": 0.6845,
+      "step": 2310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7762297353081091,
+      "learning_rate": 8.885429623837178e-06,
+      "loss": 0.6759,
+      "step": 2311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.96582047776657,
+      "learning_rate": 8.884369870250945e-06,
+      "loss": 0.6531,
+      "step": 2312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8501988880713138,
+      "learning_rate": 8.883309676349247e-06,
+      "loss": 0.6091,
+      "step": 2313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.003392735286015,
+      "learning_rate": 8.882249042252262e-06,
+      "loss": 0.7573,
+      "step": 2314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9134386167380948,
+      "learning_rate": 8.881187968080222e-06,
+      "loss": 0.6455,
+      "step": 2315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.870256379320453,
+      "learning_rate": 8.880126453953403e-06,
+      "loss": 0.655,
+      "step": 2316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.759883021241307,
+      "learning_rate": 8.879064499992133e-06,
+      "loss": 0.7253,
+      "step": 2317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.05074006883249,
+      "learning_rate": 8.878002106316795e-06,
+      "loss": 0.7489,
+      "step": 2318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.367048592554034,
+      "learning_rate": 8.876939273047813e-06,
+      "loss": 0.6746,
+      "step": 2319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.951477514551727,
+      "learning_rate": 8.875876000305666e-06,
+      "loss": 0.7181,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.839227616535758,
+      "learning_rate": 8.874812288210883e-06,
+      "loss": 0.6724,
+      "step": 2321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7580491487314909,
+      "learning_rate": 8.87374813688404e-06,
+      "loss": 0.7564,
+      "step": 2322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.082563087261863,
+      "learning_rate": 8.872683546445768e-06,
+      "loss": 0.7732,
+      "step": 2323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.947319163291198,
+      "learning_rate": 8.871618517016742e-06,
+      "loss": 0.6136,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8749719303586856,
+      "learning_rate": 8.870553048717689e-06,
+      "loss": 0.6375,
+      "step": 2325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8189093114890664,
+      "learning_rate": 8.86948714166939e-06,
+      "loss": 0.6291,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7890150891185206,
+      "learning_rate": 8.868420795992662e-06,
+      "loss": 0.5662,
+      "step": 2327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.841854935089076,
+      "learning_rate": 8.867354011808391e-06,
+      "loss": 0.6774,
+      "step": 2328
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9138052891691462,
+      "learning_rate": 8.866286789237499e-06,
+      "loss": 0.6585,
+      "step": 2329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9134802569985976,
+      "learning_rate": 8.865219128400964e-06,
+      "loss": 0.6528,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1272177907438614,
+      "learning_rate": 8.864151029419807e-06,
+      "loss": 0.6907,
+      "step": 2331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8186258535110367,
+      "learning_rate": 8.863082492415107e-06,
+      "loss": 0.6954,
+      "step": 2332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.65082051016591,
+      "learning_rate": 8.86201351750799e-06,
+      "loss": 0.5637,
+      "step": 2333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8293006727742915,
+      "learning_rate": 8.860944104819625e-06,
+      "loss": 0.719,
+      "step": 2334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8125833581391186,
+      "learning_rate": 8.85987425447124e-06,
+      "loss": 0.5463,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8212561193038788,
+      "learning_rate": 8.858803966584108e-06,
+      "loss": 0.6743,
+      "step": 2336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8575482610539216,
+      "learning_rate": 8.857733241279551e-06,
+      "loss": 0.6718,
+      "step": 2337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1513939629386445,
+      "learning_rate": 8.856662078678944e-06,
+      "loss": 0.6537,
+      "step": 2338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8719378153923891,
+      "learning_rate": 8.855590478903707e-06,
+      "loss": 0.644,
+      "step": 2339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1098349957847162,
+      "learning_rate": 8.854518442075313e-06,
+      "loss": 0.7041,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.140614121778238,
+      "learning_rate": 8.853445968315286e-06,
+      "loss": 0.7882,
+      "step": 2341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.79779569759238,
+      "learning_rate": 8.852373057745192e-06,
+      "loss": 0.6862,
+      "step": 2342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8913245263343401,
+      "learning_rate": 8.851299710486655e-06,
+      "loss": 0.6563,
+      "step": 2343
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7848657576070133,
+      "learning_rate": 8.850225926661344e-06,
+      "loss": 0.5715,
+      "step": 2344
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.042072907360112,
+      "learning_rate": 8.84915170639098e-06,
+      "loss": 0.6642,
+      "step": 2345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.882590587542723,
+      "learning_rate": 8.848077049797327e-06,
+      "loss": 0.5457,
+      "step": 2346
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0577242147645443,
+      "learning_rate": 8.847001957002211e-06,
+      "loss": 0.6979,
+      "step": 2347
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8692862797256795,
+      "learning_rate": 8.845926428127493e-06,
+      "loss": 0.678,
+      "step": 2348
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.780826018929952,
+      "learning_rate": 8.844850463295096e-06,
+      "loss": 0.6109,
+      "step": 2349
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.718886412604249,
+      "learning_rate": 8.843774062626982e-06,
+      "loss": 0.6187,
+      "step": 2350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0780976828794624,
+      "learning_rate": 8.842697226245171e-06,
+      "loss": 0.6609,
+      "step": 2351
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1857009596687407,
+      "learning_rate": 8.841619954271725e-06,
+      "loss": 0.6179,
+      "step": 2352
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9946623483439587,
+      "learning_rate": 8.840542246828763e-06,
+      "loss": 0.5932,
+      "step": 2353
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9732400040300602,
+      "learning_rate": 8.839464104038445e-06,
+      "loss": 0.719,
+      "step": 2354
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9466418799081235,
+      "learning_rate": 8.838385526022989e-06,
+      "loss": 0.6472,
+      "step": 2355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8709285020035453,
+      "learning_rate": 8.83730651290465e-06,
+      "loss": 0.6885,
+      "step": 2356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8144321685258162,
+      "learning_rate": 8.836227064805751e-06,
+      "loss": 0.6892,
+      "step": 2357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.92972926188176,
+      "learning_rate": 8.835147181848646e-06,
+      "loss": 0.6428,
+      "step": 2358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.832178705416403,
+      "learning_rate": 8.83406686415575e-06,
+      "loss": 0.6024,
+      "step": 2359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7708427249652576,
+      "learning_rate": 8.832986111849522e-06,
+      "loss": 0.5662,
+      "step": 2360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7190953214028308,
+      "learning_rate": 8.831904925052468e-06,
+      "loss": 0.5733,
+      "step": 2361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5950322378285189,
+      "learning_rate": 8.830823303887152e-06,
+      "loss": 0.5816,
+      "step": 2362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9181816926545536,
+      "learning_rate": 8.829741248476178e-06,
+      "loss": 0.6566,
+      "step": 2363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.250978660572423,
+      "learning_rate": 8.828658758942206e-06,
+      "loss": 0.7142,
+      "step": 2364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9235552583017665,
+      "learning_rate": 8.827575835407942e-06,
+      "loss": 0.6098,
+      "step": 2365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6808897773713956,
+      "learning_rate": 8.826492477996138e-06,
+      "loss": 0.6614,
+      "step": 2366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6939820996787394,
+      "learning_rate": 8.825408686829602e-06,
+      "loss": 0.6475,
+      "step": 2367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.005566835841052,
+      "learning_rate": 8.824324462031189e-06,
+      "loss": 0.7744,
+      "step": 2368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8165929910866654,
+      "learning_rate": 8.823239803723799e-06,
+      "loss": 0.7205,
+      "step": 2369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8947184852325152,
+      "learning_rate": 8.822154712030386e-06,
+      "loss": 0.6283,
+      "step": 2370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.847875985087468,
+      "learning_rate": 8.82106918707395e-06,
+      "loss": 0.6647,
+      "step": 2371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7369026204140432,
+      "learning_rate": 8.819983228977543e-06,
+      "loss": 0.6001,
+      "step": 2372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7589257105558311,
+      "learning_rate": 8.818896837864263e-06,
+      "loss": 0.6021,
+      "step": 2373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.782473528288657,
+      "learning_rate": 8.81781001385726e-06,
+      "loss": 0.6308,
+      "step": 2374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.24267074124853,
+      "learning_rate": 8.81672275707973e-06,
+      "loss": 0.6647,
+      "step": 2375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7755566946531254,
+      "learning_rate": 8.815635067654924e-06,
+      "loss": 0.6279,
+      "step": 2376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7420803857728735,
+      "learning_rate": 8.814546945706132e-06,
+      "loss": 0.6252,
+      "step": 2377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.00014949073347,
+      "learning_rate": 8.813458391356702e-06,
+      "loss": 0.6809,
+      "step": 2378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.988415369348472,
+      "learning_rate": 8.812369404730027e-06,
+      "loss": 0.6487,
+      "step": 2379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9412075336525594,
+      "learning_rate": 8.811279985949551e-06,
+      "loss": 0.6755,
+      "step": 2380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9139607368009248,
+      "learning_rate": 8.810190135138765e-06,
+      "loss": 0.722,
+      "step": 2381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9545338507069623,
+      "learning_rate": 8.80909985242121e-06,
+      "loss": 0.6374,
+      "step": 2382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.748644293222741,
+      "learning_rate": 8.808009137920475e-06,
+      "loss": 0.6028,
+      "step": 2383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0414545333900227,
+      "learning_rate": 8.8069179917602e-06,
+      "loss": 0.7474,
+      "step": 2384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.06277922678675,
+      "learning_rate": 8.805826414064071e-06,
+      "loss": 0.6891,
+      "step": 2385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0083336554857874,
+      "learning_rate": 8.804734404955825e-06,
+      "loss": 0.656,
+      "step": 2386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.660482284498147,
+      "learning_rate": 8.80364196455925e-06,
+      "loss": 0.5254,
+      "step": 2387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8909219127501342,
+      "learning_rate": 8.802549092998176e-06,
+      "loss": 0.6555,
+      "step": 2388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9055724159754508,
+      "learning_rate": 8.80145579039649e-06,
+      "loss": 0.6571,
+      "step": 2389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1750249859853925,
+      "learning_rate": 8.800362056878123e-06,
+      "loss": 0.7014,
+      "step": 2390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.17662140669081,
+      "learning_rate": 8.799267892567054e-06,
+      "loss": 0.6947,
+      "step": 2391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8554578649297455,
+      "learning_rate": 8.798173297587316e-06,
+      "loss": 0.5301,
+      "step": 2392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8693214266194902,
+      "learning_rate": 8.797078272062984e-06,
+      "loss": 0.6241,
+      "step": 2393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.116068708891031,
+      "learning_rate": 8.795982816118189e-06,
+      "loss": 0.7266,
+      "step": 2394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8822655074150916,
+      "learning_rate": 8.794886929877104e-06,
+      "loss": 0.6358,
+      "step": 2395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.753515161356571,
+      "learning_rate": 8.793790613463956e-06,
+      "loss": 0.6498,
+      "step": 2396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1198668025700362,
+      "learning_rate": 8.792693867003017e-06,
+      "loss": 0.8171,
+      "step": 2397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9558213690977169,
+      "learning_rate": 8.791596690618611e-06,
+      "loss": 0.8011,
+      "step": 2398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.071651445112021,
+      "learning_rate": 8.79049908443511e-06,
+      "loss": 0.6883,
+      "step": 2399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8405970903726203,
+      "learning_rate": 8.789401048576932e-06,
+      "loss": 0.7141,
+      "step": 2400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8874216666737584,
+      "learning_rate": 8.788302583168546e-06,
+      "loss": 0.6946,
+      "step": 2401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0114898090891806,
+      "learning_rate": 8.78720368833447e-06,
+      "loss": 0.5675,
+      "step": 2402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9083054611199246,
+      "learning_rate": 8.78610436419927e-06,
+      "loss": 0.6291,
+      "step": 2403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9970989061556155,
+      "learning_rate": 8.785004610887559e-06,
+      "loss": 0.6855,
+      "step": 2404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7311843615333404,
+      "learning_rate": 8.783904428524002e-06,
+      "loss": 0.5969,
+      "step": 2405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7874300778174452,
+      "learning_rate": 8.782803817233312e-06,
+      "loss": 0.6618,
+      "step": 2406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.745175178235986,
+      "learning_rate": 8.781702777140245e-06,
+      "loss": 0.6895,
+      "step": 2407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9717874839652878,
+      "learning_rate": 8.780601308369615e-06,
+      "loss": 0.7072,
+      "step": 2408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.286414310695891,
+      "learning_rate": 8.779499411046279e-06,
+      "loss": 0.6198,
+      "step": 2409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8423429664253486,
+      "learning_rate": 8.778397085295141e-06,
+      "loss": 0.5958,
+      "step": 2410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7508724073956576,
+      "learning_rate": 8.777294331241157e-06,
+      "loss": 0.6431,
+      "step": 2411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9141251716681176,
+      "learning_rate": 8.77619114900933e-06,
+      "loss": 0.6575,
+      "step": 2412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9832722148946702,
+      "learning_rate": 8.775087538724714e-06,
+      "loss": 0.6805,
+      "step": 2413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7600986147301019,
+      "learning_rate": 8.773983500512408e-06,
+      "loss": 0.7129,
+      "step": 2414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9150612390096766,
+      "learning_rate": 8.772879034497561e-06,
+      "loss": 0.8497,
+      "step": 2415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.812339592854274,
+      "learning_rate": 8.771774140805372e-06,
+      "loss": 0.668,
+      "step": 2416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9986647398148865,
+      "learning_rate": 8.770668819561085e-06,
+      "loss": 0.7162,
+      "step": 2417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9456747118199362,
+      "learning_rate": 8.769563070889995e-06,
+      "loss": 0.7052,
+      "step": 2418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.965657407221823,
+      "learning_rate": 8.768456894917445e-06,
+      "loss": 0.7759,
+      "step": 2419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.949060659018235,
+      "learning_rate": 8.767350291768827e-06,
+      "loss": 0.7126,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8428955759377683,
+      "learning_rate": 8.76624326156958e-06,
+      "loss": 0.6419,
+      "step": 2421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9139856823393193,
+      "learning_rate": 8.765135804445192e-06,
+      "loss": 0.6822,
+      "step": 2422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.12635949355531,
+      "learning_rate": 8.7640279205212e-06,
+      "loss": 0.7338,
+      "step": 2423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.871367384050979,
+      "learning_rate": 8.76291960992319e-06,
+      "loss": 0.6935,
+      "step": 2424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9112013632683538,
+      "learning_rate": 8.761810872776793e-06,
+      "loss": 0.6928,
+      "step": 2425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.986560427331108,
+      "learning_rate": 8.760701709207693e-06,
+      "loss": 0.6297,
+      "step": 2426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8219563126246536,
+      "learning_rate": 8.759592119341618e-06,
+      "loss": 0.6592,
+      "step": 2427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.063821372703502,
+      "learning_rate": 8.758482103304348e-06,
+      "loss": 0.6741,
+      "step": 2428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.834418853775193,
+      "learning_rate": 8.757371661221709e-06,
+      "loss": 0.5968,
+      "step": 2429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8513650814208955,
+      "learning_rate": 8.756260793219575e-06,
+      "loss": 0.6706,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.205372225245357,
+      "learning_rate": 8.755149499423871e-06,
+      "loss": 0.7245,
+      "step": 2431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8523502602927382,
+      "learning_rate": 8.754037779960566e-06,
+      "loss": 0.6928,
+      "step": 2432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.164253711937142,
+      "learning_rate": 8.752925634955685e-06,
+      "loss": 0.6091,
+      "step": 2433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7726757578138437,
+      "learning_rate": 8.751813064535288e-06,
+      "loss": 0.5928,
+      "step": 2434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9559306596077315,
+      "learning_rate": 8.750700068825499e-06,
+      "loss": 0.6752,
+      "step": 2435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7822366987620128,
+      "learning_rate": 8.749586647952478e-06,
+      "loss": 0.6989,
+      "step": 2436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2000460835818703,
+      "learning_rate": 8.748472802042438e-06,
+      "loss": 0.6844,
+      "step": 2437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0813206807667015,
+      "learning_rate": 8.74735853122164e-06,
+      "loss": 0.7764,
+      "step": 2438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.810316239654606,
+      "learning_rate": 8.746243835616392e-06,
+      "loss": 0.6687,
+      "step": 2439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2509528243743127,
+      "learning_rate": 8.745128715353055e-06,
+      "loss": 0.576,
+      "step": 2440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.073047694712885,
+      "learning_rate": 8.74401317055803e-06,
+      "loss": 0.7311,
+      "step": 2441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9258060000172623,
+      "learning_rate": 8.742897201357772e-06,
+      "loss": 0.6671,
+      "step": 2442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8935603943634967,
+      "learning_rate": 8.741780807878783e-06,
+      "loss": 0.7318,
+      "step": 2443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.803740494965822,
+      "learning_rate": 8.740663990247612e-06,
+      "loss": 0.6143,
+      "step": 2444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8225712920487915,
+      "learning_rate": 8.739546748590857e-06,
+      "loss": 0.6634,
+      "step": 2445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.79407696453161,
+      "learning_rate": 8.738429083035162e-06,
+      "loss": 0.6252,
+      "step": 2446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7820233409510757,
+      "learning_rate": 8.737310993707225e-06,
+      "loss": 0.6748,
+      "step": 2447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8736454414922534,
+      "learning_rate": 8.736192480733782e-06,
+      "loss": 0.6461,
+      "step": 2448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8092214496718138,
+      "learning_rate": 8.735073544241627e-06,
+      "loss": 0.5736,
+      "step": 2449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8901610250011938,
+      "learning_rate": 8.733954184357596e-06,
+      "loss": 0.5879,
+      "step": 2450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9058356933274532,
+      "learning_rate": 8.732834401208575e-06,
+      "loss": 0.7635,
+      "step": 2451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.145236956293225,
+      "learning_rate": 8.731714194921498e-06,
+      "loss": 0.7627,
+      "step": 2452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9661727069908106,
+      "learning_rate": 8.730593565623349e-06,
+      "loss": 0.6266,
+      "step": 2453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9383169325757479,
+      "learning_rate": 8.729472513441152e-06,
+      "loss": 0.6724,
+      "step": 2454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0297056180550634,
+      "learning_rate": 8.728351038501991e-06,
+      "loss": 0.7136,
+      "step": 2455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2711745650680903,
+      "learning_rate": 8.727229140932988e-06,
+      "loss": 0.6877,
+      "step": 2456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8425469528169693,
+      "learning_rate": 8.726106820861319e-06,
+      "loss": 0.645,
+      "step": 2457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9820413892321256,
+      "learning_rate": 8.724984078414202e-06,
+      "loss": 0.6962,
+      "step": 2458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.943380318626068,
+      "learning_rate": 8.72386091371891e-06,
+      "loss": 0.6888,
+      "step": 2459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8212199805904132,
+      "learning_rate": 8.722737326902757e-06,
+      "loss": 0.6929,
+      "step": 2460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8554451720860514,
+      "learning_rate": 8.72161331809311e-06,
+      "loss": 0.7531,
+      "step": 2461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9310112538498863,
+      "learning_rate": 8.720488887417379e-06,
+      "loss": 0.6156,
+      "step": 2462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8574398933023148,
+      "learning_rate": 8.719364035003028e-06,
+      "loss": 0.6742,
+      "step": 2463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9700711516872065,
+      "learning_rate": 8.718238760977562e-06,
+      "loss": 0.7087,
+      "step": 2464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.896331167359135,
+      "learning_rate": 8.71711306546854e-06,
+      "loss": 0.6402,
+      "step": 2465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5871072247915825,
+      "learning_rate": 8.715986948603566e-06,
+      "loss": 0.5402,
+      "step": 2466
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9043062384767355,
+      "learning_rate": 8.71486041051029e-06,
+      "loss": 0.7504,
+      "step": 2467
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9304941660777277,
+      "learning_rate": 8.713733451316415e-06,
+      "loss": 0.7583,
+      "step": 2468
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9562908438044804,
+      "learning_rate": 8.712606071149683e-06,
+      "loss": 0.6654,
+      "step": 2469
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.955293052258211,
+      "learning_rate": 8.711478270137892e-06,
+      "loss": 0.7673,
+      "step": 2470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7622527529578897,
+      "learning_rate": 8.710350048408885e-06,
+      "loss": 0.6916,
+      "step": 2471
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9881603469858304,
+      "learning_rate": 8.709221406090552e-06,
+      "loss": 0.6934,
+      "step": 2472
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0033670173359495,
+      "learning_rate": 8.70809234331083e-06,
+      "loss": 0.7154,
+      "step": 2473
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.777575663522896,
+      "learning_rate": 8.706962860197707e-06,
+      "loss": 0.6564,
+      "step": 2474
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.967548242810914,
+      "learning_rate": 8.705832956879214e-06,
+      "loss": 0.6439,
+      "step": 2475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.94761621671732,
+      "learning_rate": 8.704702633483431e-06,
+      "loss": 0.6275,
+      "step": 2476
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.017817753341322,
+      "learning_rate": 8.703571890138491e-06,
+      "loss": 0.737,
+      "step": 2477
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6563836459176866,
+      "learning_rate": 8.702440726972565e-06,
+      "loss": 0.5686,
+      "step": 2478
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0567605849561037,
+      "learning_rate": 8.701309144113881e-06,
+      "loss": 0.6831,
+      "step": 2479
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8114888761264734,
+      "learning_rate": 8.700177141690708e-06,
+      "loss": 0.6569,
+      "step": 2480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1350886758417156,
+      "learning_rate": 8.699044719831368e-06,
+      "loss": 0.712,
+      "step": 2481
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8611495208476372,
+      "learning_rate": 8.697911878664222e-06,
+      "loss": 0.6709,
+      "step": 2482
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.940942812297296,
+      "learning_rate": 8.69677861831769e-06,
+      "loss": 0.656,
+      "step": 2483
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8476105475176288,
+      "learning_rate": 8.695644938920229e-06,
+      "loss": 0.6289,
+      "step": 2484
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0744387596970384,
+      "learning_rate": 8.69451084060035e-06,
+      "loss": 0.7183,
+      "step": 2485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9263953191832632,
+      "learning_rate": 8.693376323486609e-06,
+      "loss": 0.5893,
+      "step": 2486
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6553134124256639,
+      "learning_rate": 8.69224138770761e-06,
+      "loss": 0.5539,
+      "step": 2487
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7912036322986,
+      "learning_rate": 8.691106033392004e-06,
+      "loss": 0.6439,
+      "step": 2488
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9551293930565374,
+      "learning_rate": 8.689970260668494e-06,
+      "loss": 0.6473,
+      "step": 2489
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4836561713045784,
+      "learning_rate": 8.688834069665819e-06,
+      "loss": 0.7666,
+      "step": 2490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9702181315374108,
+      "learning_rate": 8.687697460512779e-06,
+      "loss": 0.5778,
+      "step": 2491
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8888601359095911,
+      "learning_rate": 8.686560433338212e-06,
+      "loss": 0.6791,
+      "step": 2492
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8130047281565391,
+      "learning_rate": 8.685422988271005e-06,
+      "loss": 0.7071,
+      "step": 2493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8530436031260942,
+      "learning_rate": 8.684285125440099e-06,
+      "loss": 0.6628,
+      "step": 2494
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8370355003345429,
+      "learning_rate": 8.683146844974473e-06,
+      "loss": 0.6431,
+      "step": 2495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.872017642208934,
+      "learning_rate": 8.682008147003159e-06,
+      "loss": 0.6496,
+      "step": 2496
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.194564775682107,
+      "learning_rate": 8.680869031655234e-06,
+      "loss": 0.7402,
+      "step": 2497
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7391627982124838,
+      "learning_rate": 8.679729499059826e-06,
+      "loss": 0.6092,
+      "step": 2498
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8041114226784951,
+      "learning_rate": 8.678589549346103e-06,
+      "loss": 0.6696,
+      "step": 2499
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9286629979286871,
+      "learning_rate": 8.67744918264329e-06,
+      "loss": 0.6261,
+      "step": 2500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.034026955174544,
+      "learning_rate": 8.67630839908065e-06,
+      "loss": 0.6834,
+      "step": 2501
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7611055542338694,
+      "learning_rate": 8.675167198787497e-06,
+      "loss": 0.6267,
+      "step": 2502
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0451242410055697,
+      "learning_rate": 8.674025581893197e-06,
+      "loss": 0.6848,
+      "step": 2503
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.939191563492884,
+      "learning_rate": 8.672883548527156e-06,
+      "loss": 0.6619,
+      "step": 2504
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5780790969065914,
+      "learning_rate": 8.671741098818829e-06,
+      "loss": 0.7253,
+      "step": 2505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9719588150251077,
+      "learning_rate": 8.67059823289772e-06,
+      "loss": 0.667,
+      "step": 2506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9648668702584733,
+      "learning_rate": 8.669454950893381e-06,
+      "loss": 0.7402,
+      "step": 2507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9258196939936445,
+      "learning_rate": 8.668311252935407e-06,
+      "loss": 0.6772,
+      "step": 2508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0081060724360444,
+      "learning_rate": 8.667167139153443e-06,
+      "loss": 0.6405,
+      "step": 2509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6742198835544373,
+      "learning_rate": 8.666022609677183e-06,
+      "loss": 0.5199,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8564714159417233,
+      "learning_rate": 8.664877664636365e-06,
+      "loss": 0.7075,
+      "step": 2511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7875909865802426,
+      "learning_rate": 8.663732304160772e-06,
+      "loss": 0.6749,
+      "step": 2512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8538473774971693,
+      "learning_rate": 8.66258652838024e-06,
+      "loss": 0.6399,
+      "step": 2513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8806935770428046,
+      "learning_rate": 8.661440337424652e-06,
+      "loss": 0.6695,
+      "step": 2514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8938221894953242,
+      "learning_rate": 8.660293731423929e-06,
+      "loss": 0.5308,
+      "step": 2515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7300456189211295,
+      "learning_rate": 8.659146710508052e-06,
+      "loss": 0.6649,
+      "step": 2516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9670066452324917,
+      "learning_rate": 8.657999274807036e-06,
+      "loss": 0.6295,
+      "step": 2517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.421249790517567,
+      "learning_rate": 8.656851424450954e-06,
+      "loss": 0.5711,
+      "step": 2518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7941611225137957,
+      "learning_rate": 8.655703159569919e-06,
+      "loss": 0.6625,
+      "step": 2519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8645532699032437,
+      "learning_rate": 8.654554480294094e-06,
+      "loss": 0.6557,
+      "step": 2520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7914760505530758,
+      "learning_rate": 8.653405386753688e-06,
+      "loss": 0.6373,
+      "step": 2521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9904369150091747,
+      "learning_rate": 8.652255879078959e-06,
+      "loss": 0.6712,
+      "step": 2522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0797350090765514,
+      "learning_rate": 8.651105957400208e-06,
+      "loss": 0.75,
+      "step": 2523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0592593153415586,
+      "learning_rate": 8.649955621847787e-06,
+      "loss": 0.6536,
+      "step": 2524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9125424066671115,
+      "learning_rate": 8.648804872552092e-06,
+      "loss": 0.7074,
+      "step": 2525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9638190738119656,
+      "learning_rate": 8.647653709643566e-06,
+      "loss": 0.7199,
+      "step": 2526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7916417184920295,
+      "learning_rate": 8.646502133252702e-06,
+      "loss": 0.6874,
+      "step": 2527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8071976978774118,
+      "learning_rate": 8.645350143510036e-06,
+      "loss": 0.8068,
+      "step": 2528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9139163770847392,
+      "learning_rate": 8.644197740546153e-06,
+      "loss": 0.7273,
+      "step": 2529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.902494211289062,
+      "learning_rate": 8.643044924491688e-06,
+      "loss": 0.7555,
+      "step": 2530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7887237449586197,
+      "learning_rate": 8.641891695477314e-06,
+      "loss": 0.7074,
+      "step": 2531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6889392543519213,
+      "learning_rate": 8.640738053633758e-06,
+      "loss": 0.6728,
+      "step": 2532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.045472851926104,
+      "learning_rate": 8.639583999091792e-06,
+      "loss": 0.7177,
+      "step": 2533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8990351279183315,
+      "learning_rate": 8.638429531982235e-06,
+      "loss": 0.7403,
+      "step": 2534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7419245372014982,
+      "learning_rate": 8.637274652435954e-06,
+      "loss": 0.5934,
+      "step": 2535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0825259365129316,
+      "learning_rate": 8.636119360583857e-06,
+      "loss": 0.6051,
+      "step": 2536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8906745088867434,
+      "learning_rate": 8.634963656556904e-06,
+      "loss": 0.6227,
+      "step": 2537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8357251628237146,
+      "learning_rate": 8.633807540486105e-06,
+      "loss": 0.6398,
+      "step": 2538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8989039062807476,
+      "learning_rate": 8.632651012502508e-06,
+      "loss": 0.6443,
+      "step": 2539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7261862083314314,
+      "learning_rate": 8.631494072737215e-06,
+      "loss": 0.6061,
+      "step": 2540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.76712139420288,
+      "learning_rate": 8.630336721321368e-06,
+      "loss": 0.564,
+      "step": 2541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8432886950858667,
+      "learning_rate": 8.629178958386162e-06,
+      "loss": 0.6403,
+      "step": 2542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9592036465215612,
+      "learning_rate": 8.628020784062837e-06,
+      "loss": 0.5919,
+      "step": 2543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8828642264460216,
+      "learning_rate": 8.626862198482676e-06,
+      "loss": 0.6855,
+      "step": 2544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8946806088170294,
+      "learning_rate": 8.62570320177701e-06,
+      "loss": 0.527,
+      "step": 2545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9271775419029487,
+      "learning_rate": 8.624543794077223e-06,
+      "loss": 0.7033,
+      "step": 2546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.079607663238331,
+      "learning_rate": 8.623383975514736e-06,
+      "loss": 0.6912,
+      "step": 2547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.923055418753926,
+      "learning_rate": 8.622223746221024e-06,
+      "loss": 0.7492,
+      "step": 2548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8699947533876318,
+      "learning_rate": 8.621063106327604e-06,
+      "loss": 0.6444,
+      "step": 2549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0911040969561068,
+      "learning_rate": 8.619902055966043e-06,
+      "loss": 0.7458,
+      "step": 2550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0972275911219986,
+      "learning_rate": 8.618740595267949e-06,
+      "loss": 0.7584,
+      "step": 2551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8754911401151375,
+      "learning_rate": 8.617578724364984e-06,
+      "loss": 0.636,
+      "step": 2552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0750752889395683,
+      "learning_rate": 8.616416443388849e-06,
+      "loss": 0.643,
+      "step": 2553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8972081219498664,
+      "learning_rate": 8.615253752471297e-06,
+      "loss": 0.6895,
+      "step": 2554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8417165581548547,
+      "learning_rate": 8.61409065174413e-06,
+      "loss": 0.6493,
+      "step": 2555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7181649813698556,
+      "learning_rate": 8.612927141339184e-06,
+      "loss": 0.5672,
+      "step": 2556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8312017035141213,
+      "learning_rate": 8.611763221388356e-06,
+      "loss": 0.6054,
+      "step": 2557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.618425682813372,
+      "learning_rate": 8.610598892023578e-06,
+      "loss": 0.7074,
+      "step": 2558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9623646243149804,
+      "learning_rate": 8.609434153376839e-06,
+      "loss": 0.6801,
+      "step": 2559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7973271647036762,
+      "learning_rate": 8.608269005580164e-06,
+      "loss": 0.5592,
+      "step": 2560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8497187638997796,
+      "learning_rate": 8.60710344876563e-06,
+      "loss": 0.6174,
+      "step": 2561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6527277765494606,
+      "learning_rate": 8.605937483065361e-06,
+      "loss": 0.6983,
+      "step": 2562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8421684748336877,
+      "learning_rate": 8.604771108611525e-06,
+      "loss": 0.6781,
+      "step": 2563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9856957053064686,
+      "learning_rate": 8.603604325536338e-06,
+      "loss": 0.6826,
+      "step": 2564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9390978240487184,
+      "learning_rate": 8.60243713397206e-06,
+      "loss": 0.6792,
+      "step": 2565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8237849953951095,
+      "learning_rate": 8.601269534051e-06,
+      "loss": 0.7281,
+      "step": 2566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0146077881578974,
+      "learning_rate": 8.600101525905512e-06,
+      "loss": 0.6844,
+      "step": 2567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8280236844474076,
+      "learning_rate": 8.598933109667995e-06,
+      "loss": 0.6021,
+      "step": 2568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7861961197050547,
+      "learning_rate": 8.597764285470897e-06,
+      "loss": 0.5682,
+      "step": 2569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9280191246157654,
+      "learning_rate": 8.596595053446713e-06,
+      "loss": 0.5181,
+      "step": 2570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.822416375671748,
+      "learning_rate": 8.595425413727979e-06,
+      "loss": 0.6518,
+      "step": 2571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2655542490958416,
+      "learning_rate": 8.59425536644728e-06,
+      "loss": 0.6768,
+      "step": 2572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8253287565260945,
+      "learning_rate": 8.593084911737249e-06,
+      "loss": 0.5769,
+      "step": 2573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1113111922024586,
+      "learning_rate": 8.591914049730561e-06,
+      "loss": 0.6449,
+      "step": 2574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0885463074387056,
+      "learning_rate": 8.590742780559945e-06,
+      "loss": 0.6036,
+      "step": 2575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0324217465623797,
+      "learning_rate": 8.589571104358168e-06,
+      "loss": 0.6981,
+      "step": 2576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1051060069157144,
+      "learning_rate": 8.588399021258046e-06,
+      "loss": 0.6511,
+      "step": 2577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.849243572140936,
+      "learning_rate": 8.587226531392443e-06,
+      "loss": 0.5977,
+      "step": 2578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9913608116550334,
+      "learning_rate": 8.586053634894264e-06,
+      "loss": 0.6965,
+      "step": 2579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.015092471598591,
+      "learning_rate": 8.584880331896467e-06,
+      "loss": 0.7351,
+      "step": 2580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9372041452614617,
+      "learning_rate": 8.58370662253205e-06,
+      "loss": 0.7634,
+      "step": 2581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.832886288665305,
+      "learning_rate": 8.582532506934063e-06,
+      "loss": 0.6648,
+      "step": 2582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9608951852109446,
+      "learning_rate": 8.581357985235595e-06,
+      "loss": 0.6949,
+      "step": 2583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.957191923854249,
+      "learning_rate": 8.580183057569788e-06,
+      "loss": 0.665,
+      "step": 2584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0127234717259053,
+      "learning_rate": 8.579007724069823e-06,
+      "loss": 0.6466,
+      "step": 2585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8341028012931,
+      "learning_rate": 8.577831984868934e-06,
+      "loss": 0.6804,
+      "step": 2586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8705519139581988,
+      "learning_rate": 8.576655840100397e-06,
+      "loss": 0.6218,
+      "step": 2587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.776464815901934,
+      "learning_rate": 8.575479289897533e-06,
+      "loss": 0.6891,
+      "step": 2588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7028339642895287,
+      "learning_rate": 8.574302334393712e-06,
+      "loss": 0.5559,
+      "step": 2589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9135910153508136,
+      "learning_rate": 8.573124973722349e-06,
+      "loss": 0.6412,
+      "step": 2590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7931655177631687,
+      "learning_rate": 8.571947208016904e-06,
+      "loss": 0.6621,
+      "step": 2591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8046680636422858,
+      "learning_rate": 8.570769037410885e-06,
+      "loss": 0.6493,
+      "step": 2592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7590315838924429,
+      "learning_rate": 8.56959046203784e-06,
+      "loss": 0.6755,
+      "step": 2593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8703325066678136,
+      "learning_rate": 8.568411482031372e-06,
+      "loss": 0.644,
+      "step": 2594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9303932644741164,
+      "learning_rate": 8.567232097525123e-06,
+      "loss": 0.6309,
+      "step": 2595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8518576249903604,
+      "learning_rate": 8.566052308652783e-06,
+      "loss": 0.6986,
+      "step": 2596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8176810029507842,
+      "learning_rate": 8.56487211554809e-06,
+      "loss": 0.6124,
+      "step": 2597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.806871269882299,
+      "learning_rate": 8.563691518344822e-06,
+      "loss": 0.5968,
+      "step": 2598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9774455484279878,
+      "learning_rate": 8.562510517176807e-06,
+      "loss": 0.6801,
+      "step": 2599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8888905845005723,
+      "learning_rate": 8.561329112177918e-06,
+      "loss": 0.6651,
+      "step": 2600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9363855520254039,
+      "learning_rate": 8.560147303482078e-06,
+      "loss": 0.7647,
+      "step": 2601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1291448409247398,
+      "learning_rate": 8.558965091223248e-06,
+      "loss": 0.7111,
+      "step": 2602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0391902138624625,
+      "learning_rate": 8.55778247553544e-06,
+      "loss": 0.6569,
+      "step": 2603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0624650689714334,
+      "learning_rate": 8.55659945655271e-06,
+      "loss": 0.6995,
+      "step": 2604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6672557776522174,
+      "learning_rate": 8.555416034409158e-06,
+      "loss": 0.5768,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8330406340117817,
+      "learning_rate": 8.554232209238935e-06,
+      "loss": 0.7001,
+      "step": 2606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9879512597907347,
+      "learning_rate": 8.553047981176232e-06,
+      "loss": 0.6188,
+      "step": 2607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.945608350625391,
+      "learning_rate": 8.55186335035529e-06,
+      "loss": 0.6776,
+      "step": 2608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0410863507175847,
+      "learning_rate": 8.55067831691039e-06,
+      "loss": 0.8017,
+      "step": 2609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9040127392524986,
+      "learning_rate": 8.549492880975866e-06,
+      "loss": 0.7686,
+      "step": 2610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7321283064625452,
+      "learning_rate": 8.548307042686093e-06,
+      "loss": 0.5713,
+      "step": 2611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.016949276627831,
+      "learning_rate": 8.54712080217549e-06,
+      "loss": 0.6542,
+      "step": 2612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6897248893935806,
+      "learning_rate": 8.545934159578527e-06,
+      "loss": 0.6149,
+      "step": 2613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7965717499797786,
+      "learning_rate": 8.544747115029717e-06,
+      "loss": 0.6785,
+      "step": 2614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.869943587447054,
+      "learning_rate": 8.543559668663616e-06,
+      "loss": 0.6481,
+      "step": 2615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9399110552269427,
+      "learning_rate": 8.54237182061483e-06,
+      "loss": 0.5535,
+      "step": 2616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8931184533667633,
+      "learning_rate": 8.541183571018006e-06,
+      "loss": 0.6471,
+      "step": 2617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9115419127837032,
+      "learning_rate": 8.53999492000784e-06,
+      "loss": 0.6745,
+      "step": 2618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8487042882352491,
+      "learning_rate": 8.538805867719073e-06,
+      "loss": 0.7526,
+      "step": 2619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9130883130049305,
+      "learning_rate": 8.537616414286491e-06,
+      "loss": 0.6541,
+      "step": 2620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.804592790152373,
+      "learning_rate": 8.536426559844923e-06,
+      "loss": 0.6664,
+      "step": 2621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9861938106078298,
+      "learning_rate": 8.53523630452925e-06,
+      "loss": 0.7571,
+      "step": 2622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6363374746611283,
+      "learning_rate": 8.53404564847439e-06,
+      "loss": 0.5683,
+      "step": 2623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.77836846394158,
+      "learning_rate": 8.532854591815313e-06,
+      "loss": 0.6264,
+      "step": 2624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8115689056920217,
+      "learning_rate": 8.531663134687031e-06,
+      "loss": 0.649,
+      "step": 2625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.734131206157076,
+      "learning_rate": 8.530471277224603e-06,
+      "loss": 0.5828,
+      "step": 2626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1028348732686246,
+      "learning_rate": 8.529279019563133e-06,
+      "loss": 0.6539,
+      "step": 2627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1826963739866594,
+      "learning_rate": 8.528086361837771e-06,
+      "loss": 0.7534,
+      "step": 2628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.852410404097433,
+      "learning_rate": 8.526893304183708e-06,
+      "loss": 0.591,
+      "step": 2629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6674908894896996,
+      "learning_rate": 8.525699846736189e-06,
+      "loss": 0.5761,
+      "step": 2630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1536689365949493,
+      "learning_rate": 8.524505989630493e-06,
+      "loss": 0.6215,
+      "step": 2631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7369192872406605,
+      "learning_rate": 8.523311733001957e-06,
+      "loss": 0.6372,
+      "step": 2632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.491539215247505,
+      "learning_rate": 8.522117076985955e-06,
+      "loss": 0.6566,
+      "step": 2633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8660934135316272,
+      "learning_rate": 8.520922021717903e-06,
+      "loss": 0.6222,
+      "step": 2634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.058563343023167,
+      "learning_rate": 8.519726567333273e-06,
+      "loss": 0.6289,
+      "step": 2635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0092068256350415,
+      "learning_rate": 8.518530713967575e-06,
+      "loss": 0.6672,
+      "step": 2636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8080677008085388,
+      "learning_rate": 8.517334461756366e-06,
+      "loss": 0.6223,
+      "step": 2637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9543683749556946,
+      "learning_rate": 8.516137810835248e-06,
+      "loss": 0.5789,
+      "step": 2638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2093630526599357,
+      "learning_rate": 8.514940761339867e-06,
+      "loss": 0.7268,
+      "step": 2639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2424846594153105,
+      "learning_rate": 8.513743313405916e-06,
+      "loss": 0.6877,
+      "step": 2640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.710291182584347,
+      "learning_rate": 8.512545467169133e-06,
+      "loss": 0.6036,
+      "step": 2641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6276412226705352,
+      "learning_rate": 8.5113472227653e-06,
+      "loss": 0.517,
+      "step": 2642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.802571808921564,
+      "learning_rate": 8.510148580330246e-06,
+      "loss": 0.5983,
+      "step": 2643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.067739256881607,
+      "learning_rate": 8.508949539999845e-06,
+      "loss": 0.6387,
+      "step": 2644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7614511548342808,
+      "learning_rate": 8.50775010191001e-06,
+      "loss": 0.6034,
+      "step": 2645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.852977153736663,
+      "learning_rate": 8.50655026619671e-06,
+      "loss": 0.6811,
+      "step": 2646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7666746273554579,
+      "learning_rate": 8.505350032995946e-06,
+      "loss": 0.662,
+      "step": 2647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9788840184781433,
+      "learning_rate": 8.504149402443782e-06,
+      "loss": 0.6714,
+      "step": 2648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.763046798035941,
+      "learning_rate": 8.502948374676307e-06,
+      "loss": 0.6534,
+      "step": 2649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6872062703419626,
+      "learning_rate": 8.501746949829668e-06,
+      "loss": 0.6629,
+      "step": 2650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9880445367487545,
+      "learning_rate": 8.500545128040052e-06,
+      "loss": 0.6643,
+      "step": 2651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.65755205744229,
+      "learning_rate": 8.499342909443697e-06,
+      "loss": 0.6829,
+      "step": 2652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9105964610284794,
+      "learning_rate": 8.498140294176874e-06,
+      "loss": 0.5986,
+      "step": 2653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7850157986482749,
+      "learning_rate": 8.496937282375912e-06,
+      "loss": 0.598,
+      "step": 2654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.830316459639485,
+      "learning_rate": 8.495733874177176e-06,
+      "loss": 0.6988,
+      "step": 2655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8863090070314195,
+      "learning_rate": 8.49453006971708e-06,
+      "loss": 0.6376,
+      "step": 2656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9381452812775855,
+      "learning_rate": 8.493325869132083e-06,
+      "loss": 0.5406,
+      "step": 2657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9249092805322674,
+      "learning_rate": 8.492121272558687e-06,
+      "loss": 0.6696,
+      "step": 2658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.110242213687224,
+      "learning_rate": 8.49091628013344e-06,
+      "loss": 0.7029,
+      "step": 2659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.797032481255268,
+      "learning_rate": 8.489710891992938e-06,
+      "loss": 0.8027,
+      "step": 2660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.070668589376598,
+      "learning_rate": 8.488505108273813e-06,
+      "loss": 0.7835,
+      "step": 2661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8527254181639246,
+      "learning_rate": 8.487298929112751e-06,
+      "loss": 0.5424,
+      "step": 2662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9530291879277955,
+      "learning_rate": 8.486092354646478e-06,
+      "loss": 0.6221,
+      "step": 2663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0275784280967564,
+      "learning_rate": 8.484885385011765e-06,
+      "loss": 0.6157,
+      "step": 2664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7405853747938145,
+      "learning_rate": 8.483678020345433e-06,
+      "loss": 0.6301,
+      "step": 2665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.249414015385447,
+      "learning_rate": 8.48247026078434e-06,
+      "loss": 0.7239,
+      "step": 2666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8456342897721276,
+      "learning_rate": 8.481262106465395e-06,
+      "loss": 0.6632,
+      "step": 2667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9602153469722,
+      "learning_rate": 8.480053557525544e-06,
+      "loss": 0.795,
+      "step": 2668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.969295867379851,
+      "learning_rate": 8.478844614101792e-06,
+      "loss": 0.6968,
+      "step": 2669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.966399979210272,
+      "learning_rate": 8.47763527633117e-06,
+      "loss": 0.7567,
+      "step": 2670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9079604001646893,
+      "learning_rate": 8.476425544350768e-06,
+      "loss": 0.6965,
+      "step": 2671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8146309399032956,
+      "learning_rate": 8.475215418297718e-06,
+      "loss": 0.6707,
+      "step": 2672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6768376922133157,
+      "learning_rate": 8.47400489830919e-06,
+      "loss": 0.6126,
+      "step": 2673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9499300869990017,
+      "learning_rate": 8.472793984522406e-06,
+      "loss": 0.5496,
+      "step": 2674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1684600034276653,
+      "learning_rate": 8.47158267707463e-06,
+      "loss": 0.6461,
+      "step": 2675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.817694567431662,
+      "learning_rate": 8.470370976103171e-06,
+      "loss": 0.5364,
+      "step": 2676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.217641057731939,
+      "learning_rate": 8.46915888174538e-06,
+      "loss": 0.689,
+      "step": 2677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8021492176632457,
+      "learning_rate": 8.467946394138657e-06,
+      "loss": 0.6299,
+      "step": 2678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9107035431154975,
+      "learning_rate": 8.466733513420442e-06,
+      "loss": 0.5893,
+      "step": 2679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8742357813979735,
+      "learning_rate": 8.465520239728225e-06,
+      "loss": 0.6131,
+      "step": 2680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9727390371432398,
+      "learning_rate": 8.464306573199536e-06,
+      "loss": 0.6842,
+      "step": 2681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7240250249963895,
+      "learning_rate": 8.46309251397195e-06,
+      "loss": 0.6465,
+      "step": 2682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9189298287066967,
+      "learning_rate": 8.461878062183092e-06,
+      "loss": 0.6417,
+      "step": 2683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.800900644472255,
+      "learning_rate": 8.46066321797062e-06,
+      "loss": 0.7715,
+      "step": 2684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8212217697731836,
+      "learning_rate": 8.459447981472249e-06,
+      "loss": 0.7056,
+      "step": 2685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9752112975069571,
+      "learning_rate": 8.45823235282573e-06,
+      "loss": 0.6689,
+      "step": 2686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9746070435343586,
+      "learning_rate": 8.457016332168862e-06,
+      "loss": 0.7456,
+      "step": 2687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.432989926354537,
+      "learning_rate": 8.455799919639489e-06,
+      "loss": 0.593,
+      "step": 2688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.982266556212586,
+      "learning_rate": 8.454583115375498e-06,
+      "loss": 0.6421,
+      "step": 2689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7985726638165644,
+      "learning_rate": 8.45336591951482e-06,
+      "loss": 0.5681,
+      "step": 2690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1677613588933418,
+      "learning_rate": 8.452148332195434e-06,
+      "loss": 0.7061,
+      "step": 2691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8779569941032115,
+      "learning_rate": 8.450930353555355e-06,
+      "loss": 0.6452,
+      "step": 2692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8980254003942776,
+      "learning_rate": 8.449711983732652e-06,
+      "loss": 0.7254,
+      "step": 2693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7453475241560552,
+      "learning_rate": 8.448493222865432e-06,
+      "loss": 0.6741,
+      "step": 2694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8262450441941662,
+      "learning_rate": 8.447274071091848e-06,
+      "loss": 0.7417,
+      "step": 2695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8776699685839715,
+      "learning_rate": 8.446054528550104e-06,
+      "loss": 0.6622,
+      "step": 2696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0960061725496963,
+      "learning_rate": 8.444834595378434e-06,
+      "loss": 0.6906,
+      "step": 2697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9118518648617624,
+      "learning_rate": 8.443614271715128e-06,
+      "loss": 0.6912,
+      "step": 2698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9194363297045363,
+      "learning_rate": 8.442393557698517e-06,
+      "loss": 0.5579,
+      "step": 2699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9081413101253448,
+      "learning_rate": 8.441172453466974e-06,
+      "loss": 0.6608,
+      "step": 2700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9249403721722926,
+      "learning_rate": 8.43995095915892e-06,
+      "loss": 0.7314,
+      "step": 2701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9096647611073485,
+      "learning_rate": 8.438729074912819e-06,
+      "loss": 0.5881,
+      "step": 2702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1034642501676415,
+      "learning_rate": 8.437506800867176e-06,
+      "loss": 0.7437,
+      "step": 2703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0906568978132567,
+      "learning_rate": 8.436284137160544e-06,
+      "loss": 0.685,
+      "step": 2704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7856893381417736,
+      "learning_rate": 8.435061083931519e-06,
+      "loss": 0.6323,
+      "step": 2705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9666249344130147,
+      "learning_rate": 8.433837641318741e-06,
+      "loss": 0.6732,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.122621535668428,
+      "learning_rate": 8.432613809460895e-06,
+      "loss": 0.6447,
+      "step": 2707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0054957981285253,
+      "learning_rate": 8.431389588496708e-06,
+      "loss": 0.6584,
+      "step": 2708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9393041610845445,
+      "learning_rate": 8.430164978564952e-06,
+      "loss": 0.6929,
+      "step": 2709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8108812484917312,
+      "learning_rate": 8.428939979804445e-06,
+      "loss": 0.6626,
+      "step": 2710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.021521385977441,
+      "learning_rate": 8.427714592354046e-06,
+      "loss": 0.7671,
+      "step": 2711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0810704085129377,
+      "learning_rate": 8.426488816352662e-06,
+      "loss": 0.7076,
+      "step": 2712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9862619828862162,
+      "learning_rate": 8.425262651939238e-06,
+      "loss": 0.669,
+      "step": 2713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.951010662132229,
+      "learning_rate": 8.424036099252772e-06,
+      "loss": 0.6356,
+      "step": 2714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9863665255721001,
+      "learning_rate": 8.422809158432296e-06,
+      "loss": 0.6026,
+      "step": 2715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6639586057276734,
+      "learning_rate": 8.421581829616893e-06,
+      "loss": 0.6527,
+      "step": 2716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.103415737649651,
+      "learning_rate": 8.42035411294569e-06,
+      "loss": 0.5964,
+      "step": 2717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5707603548039,
+      "learning_rate": 8.41912600855785e-06,
+      "loss": 0.6068,
+      "step": 2718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8106080608902415,
+      "learning_rate": 8.417897516592589e-06,
+      "loss": 0.6864,
+      "step": 2719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1259873253643327,
+      "learning_rate": 8.416668637189162e-06,
+      "loss": 0.7358,
+      "step": 2720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.083747722616653,
+      "learning_rate": 8.415439370486872e-06,
+      "loss": 0.6944,
+      "step": 2721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.063196434811401,
+      "learning_rate": 8.414209716625062e-06,
+      "loss": 0.741,
+      "step": 2722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7907869386100392,
+      "learning_rate": 8.41297967574312e-06,
+      "loss": 0.7097,
+      "step": 2723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.74621161890205,
+      "learning_rate": 8.411749247980478e-06,
+      "loss": 0.6878,
+      "step": 2724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7515260020420864,
+      "learning_rate": 8.410518433476613e-06,
+      "loss": 0.7515,
+      "step": 2725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9016574392369556,
+      "learning_rate": 8.409287232371043e-06,
+      "loss": 0.5963,
+      "step": 2726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8126661652045508,
+      "learning_rate": 8.408055644803335e-06,
+      "loss": 0.6511,
+      "step": 2727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8555841277858098,
+      "learning_rate": 8.406823670913093e-06,
+      "loss": 0.5926,
+      "step": 2728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0801913638478475,
+      "learning_rate": 8.405591310839972e-06,
+      "loss": 0.7424,
+      "step": 2729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8777432597286765,
+      "learning_rate": 8.404358564723663e-06,
+      "loss": 0.7011,
+      "step": 2730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8320426461813335,
+      "learning_rate": 8.403125432703904e-06,
+      "loss": 0.6469,
+      "step": 2731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8676234472685271,
+      "learning_rate": 8.401891914920483e-06,
+      "loss": 0.7347,
+      "step": 2732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1091699541707456,
+      "learning_rate": 8.400658011513223e-06,
+      "loss": 0.6774,
+      "step": 2733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.778219564489563,
+      "learning_rate": 8.399423722621994e-06,
+      "loss": 0.708,
+      "step": 2734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.791116263128344,
+      "learning_rate": 8.398189048386708e-06,
+      "loss": 0.5954,
+      "step": 2735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9654234959498464,
+      "learning_rate": 8.396953988947327e-06,
+      "loss": 0.7262,
+      "step": 2736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.923570947146556,
+      "learning_rate": 8.39571854444385e-06,
+      "loss": 0.6263,
+      "step": 2737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9762691203393106,
+      "learning_rate": 8.394482715016318e-06,
+      "loss": 0.7634,
+      "step": 2738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.785056907642776,
+      "learning_rate": 8.393246500804825e-06,
+      "loss": 0.6711,
+      "step": 2739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.92259639664194,
+      "learning_rate": 8.3920099019495e-06,
+      "loss": 0.6447,
+      "step": 2740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9325806794278027,
+      "learning_rate": 8.390772918590517e-06,
+      "loss": 0.6193,
+      "step": 2741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.796897789122905,
+      "learning_rate": 8.389535550868098e-06,
+      "loss": 0.5864,
+      "step": 2742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5789258620706366,
+      "learning_rate": 8.388297798922505e-06,
+      "loss": 0.6275,
+      "step": 2743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9466379670977967,
+      "learning_rate": 8.387059662894043e-06,
+      "loss": 0.6843,
+      "step": 2744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7072700706186492,
+      "learning_rate": 8.385821142923064e-06,
+      "loss": 0.669,
+      "step": 2745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.848227815607721,
+      "learning_rate": 8.38458223914996e-06,
+      "loss": 0.6242,
+      "step": 2746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8308049578888184,
+      "learning_rate": 8.383342951715165e-06,
+      "loss": 0.6314,
+      "step": 2747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1493743941726224,
+      "learning_rate": 8.382103280759164e-06,
+      "loss": 0.6465,
+      "step": 2748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9101945267736256,
+      "learning_rate": 8.380863226422478e-06,
+      "loss": 0.6816,
+      "step": 2749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9718539473467542,
+      "learning_rate": 8.379622788845675e-06,
+      "loss": 0.7033,
+      "step": 2750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.038835831098079,
+      "learning_rate": 8.378381968169368e-06,
+      "loss": 0.6988,
+      "step": 2751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7900118328139933,
+      "learning_rate": 8.377140764534206e-06,
+      "loss": 0.5064,
+      "step": 2752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7774886330392579,
+      "learning_rate": 8.37589917808089e-06,
+      "loss": 0.6096,
+      "step": 2753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.809054839530392,
+      "learning_rate": 8.37465720895016e-06,
+      "loss": 0.6773,
+      "step": 2754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5992093185602265,
+      "learning_rate": 8.373414857282802e-06,
+      "loss": 0.5207,
+      "step": 2755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.010983378957128,
+      "learning_rate": 8.372172123219639e-06,
+      "loss": 0.7494,
+      "step": 2756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9938803891005408,
+      "learning_rate": 8.370929006901547e-06,
+      "loss": 0.5927,
+      "step": 2757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8945446825004,
+      "learning_rate": 8.36968550846944e-06,
+      "loss": 0.6086,
+      "step": 2758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9169728105853157,
+      "learning_rate": 8.368441628064273e-06,
+      "loss": 0.6921,
+      "step": 2759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9901330548989136,
+      "learning_rate": 8.367197365827047e-06,
+      "loss": 0.6753,
+      "step": 2760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0824344592889297,
+      "learning_rate": 8.365952721898806e-06,
+      "loss": 0.6964,
+      "step": 2761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1360108572284773,
+      "learning_rate": 8.364707696420642e-06,
+      "loss": 0.7696,
+      "step": 2762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6831167248161611,
+      "learning_rate": 8.363462289533681e-06,
+      "loss": 0.5605,
+      "step": 2763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.989356336067944,
+      "learning_rate": 8.362216501379096e-06,
+      "loss": 0.7506,
+      "step": 2764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5598840637317084,
+      "learning_rate": 8.360970332098111e-06,
+      "loss": 0.6175,
+      "step": 2765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1096417150952536,
+      "learning_rate": 8.359723781831978e-06,
+      "loss": 0.7308,
+      "step": 2766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6011126729548892,
+      "learning_rate": 8.358476850722007e-06,
+      "loss": 0.6811,
+      "step": 2767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9464596799550342,
+      "learning_rate": 8.357229538909542e-06,
+      "loss": 0.663,
+      "step": 2768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9137397893266275,
+      "learning_rate": 8.355981846535972e-06,
+      "loss": 0.6147,
+      "step": 2769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.764067948839555,
+      "learning_rate": 8.354733773742734e-06,
+      "loss": 0.6074,
+      "step": 2770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9148936860052028,
+      "learning_rate": 8.353485320671298e-06,
+      "loss": 0.7877,
+      "step": 2771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7112798128547464,
+      "learning_rate": 8.352236487463188e-06,
+      "loss": 0.5686,
+      "step": 2772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.992101185683906,
+      "learning_rate": 8.350987274259966e-06,
+      "loss": 0.7142,
+      "step": 2773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8753999534127264,
+      "learning_rate": 8.349737681203234e-06,
+      "loss": 0.6127,
+      "step": 2774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.142936199132452,
+      "learning_rate": 8.348487708434644e-06,
+      "loss": 0.6528,
+      "step": 2775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1341310375948317,
+      "learning_rate": 8.347237356095888e-06,
+      "loss": 0.6646,
+      "step": 2776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9282855617561891,
+      "learning_rate": 8.3459866243287e-06,
+      "loss": 0.7142,
+      "step": 2777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8331210754757201,
+      "learning_rate": 8.344735513274853e-06,
+      "loss": 0.7037,
+      "step": 2778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9595823974640136,
+      "learning_rate": 8.343484023076175e-06,
+      "loss": 0.6657,
+      "step": 2779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8681476948952542,
+      "learning_rate": 8.342232153874521e-06,
+      "loss": 0.6335,
+      "step": 2780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7368048649716803,
+      "learning_rate": 8.340979905811805e-06,
+      "loss": 0.6599,
+      "step": 2781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.688572412587195,
+      "learning_rate": 8.339727279029974e-06,
+      "loss": 0.6573,
+      "step": 2782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7254670078388807,
+      "learning_rate": 8.33847427367102e-06,
+      "loss": 0.7033,
+      "step": 2783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7445016985559958,
+      "learning_rate": 8.337220889876978e-06,
+      "loss": 0.6888,
+      "step": 2784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8956447209405711,
+      "learning_rate": 8.335967127789929e-06,
+      "loss": 0.6261,
+      "step": 2785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8284382496111118,
+      "learning_rate": 8.334712987551989e-06,
+      "loss": 0.6161,
+      "step": 2786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.828969267565181,
+      "learning_rate": 8.333458469305324e-06,
+      "loss": 0.7489,
+      "step": 2787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7603821492238025,
+      "learning_rate": 8.332203573192143e-06,
+      "loss": 0.5797,
+      "step": 2788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0787481786361366,
+      "learning_rate": 8.330948299354694e-06,
+      "loss": 0.6936,
+      "step": 2789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0677329033718657,
+      "learning_rate": 8.329692647935269e-06,
+      "loss": 0.6972,
+      "step": 2790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.775462102298857,
+      "learning_rate": 8.328436619076203e-06,
+      "loss": 0.6864,
+      "step": 2791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0303699301163443,
+      "learning_rate": 8.327180212919877e-06,
+      "loss": 0.6823,
+      "step": 2792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.210558036257578,
+      "learning_rate": 8.32592342960871e-06,
+      "loss": 0.7242,
+      "step": 2793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1828645450156468,
+      "learning_rate": 8.324666269285161e-06,
+      "loss": 0.6827,
+      "step": 2794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6089239490667053,
+      "learning_rate": 8.323408732091743e-06,
+      "loss": 0.5493,
+      "step": 2795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7779698846657657,
+      "learning_rate": 8.322150818171002e-06,
+      "loss": 0.6296,
+      "step": 2796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0643568991443635,
+      "learning_rate": 8.32089252766553e-06,
+      "loss": 0.5536,
+      "step": 2797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.86510472363077,
+      "learning_rate": 8.319633860717963e-06,
+      "loss": 0.6896,
+      "step": 2798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9510095493605686,
+      "learning_rate": 8.318374817470976e-06,
+      "loss": 0.6448,
+      "step": 2799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.885516169850448,
+      "learning_rate": 8.317115398067289e-06,
+      "loss": 0.669,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2572097683864305,
+      "learning_rate": 8.315855602649662e-06,
+      "loss": 0.5818,
+      "step": 2801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.759056075445115,
+      "learning_rate": 8.314595431360906e-06,
+      "loss": 0.6457,
+      "step": 2802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6425939825571452,
+      "learning_rate": 8.313334884343866e-06,
+      "loss": 0.5231,
+      "step": 2803
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7130113252654398,
+      "learning_rate": 8.31207396174143e-06,
+      "loss": 0.7554,
+      "step": 2804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.955914565838696,
+      "learning_rate": 8.310812663696531e-06,
+      "loss": 0.627,
+      "step": 2805
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6720147103849268,
+      "learning_rate": 8.309550990352146e-06,
+      "loss": 0.6079,
+      "step": 2806
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.237069935203191,
+      "learning_rate": 8.308288941851295e-06,
+      "loss": 0.7222,
+      "step": 2807
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8290522376346103,
+      "learning_rate": 8.307026518337033e-06,
+      "loss": 0.5718,
+      "step": 2808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8328742008169372,
+      "learning_rate": 8.305763719952467e-06,
+      "loss": 0.6836,
+      "step": 2809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0006170517993533,
+      "learning_rate": 8.304500546840742e-06,
+      "loss": 0.6471,
+      "step": 2810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.863878297149142,
+      "learning_rate": 8.303236999145044e-06,
+      "loss": 0.7006,
+      "step": 2811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9753606075612662,
+      "learning_rate": 8.301973077008604e-06,
+      "loss": 0.6209,
+      "step": 2812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.85924478525618,
+      "learning_rate": 8.300708780574695e-06,
+      "loss": 0.58,
+      "step": 2813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9617682288103584,
+      "learning_rate": 8.299444109986631e-06,
+      "loss": 0.6333,
+      "step": 2814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9932942836246554,
+      "learning_rate": 8.298179065387774e-06,
+      "loss": 0.6311,
+      "step": 2815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9582877206850164,
+      "learning_rate": 8.29691364692152e-06,
+      "loss": 0.6961,
+      "step": 2816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.011029682430342,
+      "learning_rate": 8.295647854731312e-06,
+      "loss": 0.6541,
+      "step": 2817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9245441043938138,
+      "learning_rate": 8.294381688960634e-06,
+      "loss": 0.6018,
+      "step": 2818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1310980060323508,
+      "learning_rate": 8.293115149753016e-06,
+      "loss": 0.7579,
+      "step": 2819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.775629952527209,
+      "learning_rate": 8.291848237252025e-06,
+      "loss": 0.6678,
+      "step": 2820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8880663986965478,
+      "learning_rate": 8.290580951601272e-06,
+      "loss": 0.5957,
+      "step": 2821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.269045732221966,
+      "learning_rate": 8.289313292944415e-06,
+      "loss": 0.6567,
+      "step": 2822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.157095157644829,
+      "learning_rate": 8.288045261425146e-06,
+      "loss": 0.7711,
+      "step": 2823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7717605727416954,
+      "learning_rate": 8.286776857187205e-06,
+      "loss": 0.5993,
+      "step": 2824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2377882664385167,
+      "learning_rate": 8.285508080374376e-06,
+      "loss": 0.6189,
+      "step": 2825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9421413495128026,
+      "learning_rate": 8.284238931130476e-06,
+      "loss": 0.6647,
+      "step": 2826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9791339729087056,
+      "learning_rate": 8.282969409599375e-06,
+      "loss": 0.7541,
+      "step": 2827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2188976160973986,
+      "learning_rate": 8.28169951592498e-06,
+      "loss": 0.6117,
+      "step": 2828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.018124174549173,
+      "learning_rate": 8.280429250251238e-06,
+      "loss": 0.6828,
+      "step": 2829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9341218682789678,
+      "learning_rate": 8.279158612722145e-06,
+      "loss": 0.628,
+      "step": 2830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.861973629599436,
+      "learning_rate": 8.27788760348173e-06,
+      "loss": 0.6198,
+      "step": 2831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7662295415002316,
+      "learning_rate": 8.276616222674072e-06,
+      "loss": 0.6158,
+      "step": 2832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7746613606468469,
+      "learning_rate": 8.275344470443292e-06,
+      "loss": 0.6234,
+      "step": 2833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.77084820465646,
+      "learning_rate": 8.274072346933544e-06,
+      "loss": 0.6325,
+      "step": 2834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9053142262270337,
+      "learning_rate": 8.272799852289036e-06,
+      "loss": 0.5998,
+      "step": 2835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9878612490236216,
+      "learning_rate": 8.27152698665401e-06,
+      "loss": 0.8256,
+      "step": 2836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7506593303708355,
+      "learning_rate": 8.270253750172754e-06,
+      "loss": 0.6899,
+      "step": 2837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9372791674384715,
+      "learning_rate": 8.268980142989594e-06,
+      "loss": 0.6556,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9580485272446944,
+      "learning_rate": 8.267706165248901e-06,
+      "loss": 0.6014,
+      "step": 2839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.94607284887436,
+      "learning_rate": 8.266431817095094e-06,
+      "loss": 0.7343,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0233162109195444,
+      "learning_rate": 8.265157098672617e-06,
+      "loss": 0.6424,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7612601997749828,
+      "learning_rate": 8.263882010125974e-06,
+      "loss": 0.6468,
+      "step": 2842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8166003406613074,
+      "learning_rate": 8.262606551599701e-06,
+      "loss": 0.7198,
+      "step": 2843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1777514169067005,
+      "learning_rate": 8.261330723238381e-06,
+      "loss": 0.6666,
+      "step": 2844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8431539782484165,
+      "learning_rate": 8.260054525186634e-06,
+      "loss": 0.6959,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9351049160946372,
+      "learning_rate": 8.258777957589124e-06,
+      "loss": 0.6634,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.785975155097984,
+      "learning_rate": 8.257501020590557e-06,
+      "loss": 0.7822,
+      "step": 2847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7172233034670437,
+      "learning_rate": 8.256223714335685e-06,
+      "loss": 0.593,
+      "step": 2848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8331082125147111,
+      "learning_rate": 8.254946038969294e-06,
+      "loss": 0.5986,
+      "step": 2849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5989269880200359,
+      "learning_rate": 8.253667994636216e-06,
+      "loss": 0.6494,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8126460703093885,
+      "learning_rate": 8.252389581481328e-06,
+      "loss": 0.6088,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7710013320024187,
+      "learning_rate": 8.25111079964954e-06,
+      "loss": 0.692,
+      "step": 2852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7261220654564688,
+      "learning_rate": 8.249831649285813e-06,
+      "loss": 0.6587,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8275573012523363,
+      "learning_rate": 8.248552130535146e-06,
+      "loss": 0.6926,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8146408797397333,
+      "learning_rate": 8.247272243542579e-06,
+      "loss": 0.603,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9636780696037397,
+      "learning_rate": 8.245991988453193e-06,
+      "loss": 0.7721,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7874387137159384,
+      "learning_rate": 8.244711365412113e-06,
+      "loss": 0.5659,
+      "step": 2857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.836389067622902,
+      "learning_rate": 8.243430374564507e-06,
+      "loss": 0.6837,
+      "step": 2858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7888108908675027,
+      "learning_rate": 8.242149016055582e-06,
+      "loss": 0.5791,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8335701181284372,
+      "learning_rate": 8.240867290030585e-06,
+      "loss": 0.6948,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0901484654745426,
+      "learning_rate": 8.239585196634808e-06,
+      "loss": 0.748,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7432176517498563,
+      "learning_rate": 8.238302736013587e-06,
+      "loss": 0.6321,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8447676911357693,
+      "learning_rate": 8.237019908312289e-06,
+      "loss": 0.6296,
+      "step": 2863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6818923204848246,
+      "learning_rate": 8.235736713676336e-06,
+      "loss": 0.5814,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8930984108194275,
+      "learning_rate": 8.234453152251183e-06,
+      "loss": 0.5862,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7856506372753727,
+      "learning_rate": 8.23316922418233e-06,
+      "loss": 0.6485,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7134171960426727,
+      "learning_rate": 8.231884929615315e-06,
+      "loss": 0.7412,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.034932618216676,
+      "learning_rate": 8.230600268695724e-06,
+      "loss": 0.8004,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8280220472034374,
+      "learning_rate": 8.229315241569177e-06,
+      "loss": 0.6732,
+      "step": 2869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2761201014870887,
+      "learning_rate": 8.228029848381343e-06,
+      "loss": 0.6222,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9435641548208933,
+      "learning_rate": 8.226744089277927e-06,
+      "loss": 0.706,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7705164240793172,
+      "learning_rate": 8.225457964404675e-06,
+      "loss": 0.6072,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7796138793949614,
+      "learning_rate": 8.224171473907379e-06,
+      "loss": 0.7511,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.294358735763917,
+      "learning_rate": 8.222884617931868e-06,
+      "loss": 0.7495,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9779657166432165,
+      "learning_rate": 8.221597396624017e-06,
+      "loss": 0.6661,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.821430815356854,
+      "learning_rate": 8.220309810129739e-06,
+      "loss": 0.6018,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9649085361169925,
+      "learning_rate": 8.219021858594989e-06,
+      "loss": 0.6216,
+      "step": 2877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9075400329966081,
+      "learning_rate": 8.217733542165762e-06,
+      "loss": 0.5262,
+      "step": 2878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.787468247986847,
+      "learning_rate": 8.216444860988098e-06,
+      "loss": 0.7078,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9202828017719724,
+      "learning_rate": 8.215155815208075e-06,
+      "loss": 0.6898,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.074321753545804,
+      "learning_rate": 8.213866404971817e-06,
+      "loss": 0.6434,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9239815198105046,
+      "learning_rate": 8.212576630425482e-06,
+      "loss": 0.6708,
+      "step": 2882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0109810934562273,
+      "learning_rate": 8.211286491715274e-06,
+      "loss": 0.6282,
+      "step": 2883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.877253484772073,
+      "learning_rate": 8.20999598898744e-06,
+      "loss": 0.5615,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9652383152117894,
+      "learning_rate": 8.208705122388263e-06,
+      "loss": 0.6393,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9491407973172676,
+      "learning_rate": 8.207413892064073e-06,
+      "loss": 0.6992,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7003145780895765,
+      "learning_rate": 8.206122298161236e-06,
+      "loss": 0.5622,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8707928859913363,
+      "learning_rate": 8.204830340826161e-06,
+      "loss": 0.751,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9977017222580762,
+      "learning_rate": 8.203538020205301e-06,
+      "loss": 0.6936,
+      "step": 2889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7954445118532534,
+      "learning_rate": 8.202245336445146e-06,
+      "loss": 0.6332,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1212523482878796,
+      "learning_rate": 8.200952289692233e-06,
+      "loss": 0.7295,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9203803961045498,
+      "learning_rate": 8.199658880093132e-06,
+      "loss": 0.6838,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9804119787254033,
+      "learning_rate": 8.198365107794457e-06,
+      "loss": 0.6464,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7118907526474592,
+      "learning_rate": 8.19707097294287e-06,
+      "loss": 0.6199,
+      "step": 2894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8173439975730064,
+      "learning_rate": 8.195776475685061e-06,
+      "loss": 0.6636,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9947503806681284,
+      "learning_rate": 8.194481616167777e-06,
+      "loss": 0.7348,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0059450385174444,
+      "learning_rate": 8.193186394537792e-06,
+      "loss": 0.671,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8995286372082676,
+      "learning_rate": 8.191890810941932e-06,
+      "loss": 0.646,
+      "step": 2898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6680630849647606,
+      "learning_rate": 8.190594865527052e-06,
+      "loss": 0.6015,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0300903756694364,
+      "learning_rate": 8.18929855844006e-06,
+      "loss": 0.6532,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.03213398770465,
+      "learning_rate": 8.188001889827897e-06,
+      "loss": 0.745,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9409976976712575,
+      "learning_rate": 8.18670485983755e-06,
+      "loss": 0.6487,
+      "step": 2902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0263585046810744,
+      "learning_rate": 8.185407468616042e-06,
+      "loss": 0.5663,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.003893682630391,
+      "learning_rate": 8.18410971631044e-06,
+      "loss": 0.7368,
+      "step": 2904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8199743964798658,
+      "learning_rate": 8.182811603067855e-06,
+      "loss": 0.6108,
+      "step": 2905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9114442676440553,
+      "learning_rate": 8.181513129035434e-06,
+      "loss": 0.7262,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.138825809398571,
+      "learning_rate": 8.180214294360365e-06,
+      "loss": 0.6708,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7635191853249457,
+      "learning_rate": 8.178915099189877e-06,
+      "loss": 0.5041,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7070240123779603,
+      "learning_rate": 8.177615543671247e-06,
+      "loss": 0.5458,
+      "step": 2909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9043026323384222,
+      "learning_rate": 8.176315627951781e-06,
+      "loss": 0.6708,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8512788197178098,
+      "learning_rate": 8.175015352178835e-06,
+      "loss": 0.6638,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.69736575656168,
+      "learning_rate": 8.173714716499801e-06,
+      "loss": 0.6792,
+      "step": 2912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7299590678396837,
+      "learning_rate": 8.172413721062115e-06,
+      "loss": 0.576,
+      "step": 2913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8580558642183693,
+      "learning_rate": 8.171112366013252e-06,
+      "loss": 0.6543,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8468439712054643,
+      "learning_rate": 8.169810651500728e-06,
+      "loss": 0.6622,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8010109531550258,
+      "learning_rate": 8.168508577672096e-06,
+      "loss": 0.6466,
+      "step": 2916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9057592922152613,
+      "learning_rate": 8.167206144674959e-06,
+      "loss": 0.5711,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.050683545427632,
+      "learning_rate": 8.165903352656954e-06,
+      "loss": 0.6602,
+      "step": 2918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8054973854791319,
+      "learning_rate": 8.164600201765758e-06,
+      "loss": 0.6585,
+      "step": 2919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9761637090582729,
+      "learning_rate": 8.163296692149093e-06,
+      "loss": 0.6079,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.88880213967494,
+      "learning_rate": 8.161992823954715e-06,
+      "loss": 0.6007,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8416518885541622,
+      "learning_rate": 8.160688597330428e-06,
+      "loss": 0.6713,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.873829102606367,
+      "learning_rate": 8.159384012424074e-06,
+      "loss": 0.6165,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8160124127691164,
+      "learning_rate": 8.158079069383535e-06,
+      "loss": 0.7594,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.863086365003482,
+      "learning_rate": 8.156773768356733e-06,
+      "loss": 0.6637,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6903193748478216,
+      "learning_rate": 8.155468109491632e-06,
+      "loss": 0.5883,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9542532517773603,
+      "learning_rate": 8.154162092936233e-06,
+      "loss": 0.7481,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.772285860147228,
+      "learning_rate": 8.152855718838583e-06,
+      "loss": 0.5753,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9175001165132863,
+      "learning_rate": 8.151548987346768e-06,
+      "loss": 0.6404,
+      "step": 2929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1361197261230918,
+      "learning_rate": 8.150241898608911e-06,
+      "loss": 0.7385,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6543334650805261,
+      "learning_rate": 8.14893445277318e-06,
+      "loss": 0.5488,
+      "step": 2931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2819717034684923,
+      "learning_rate": 8.147626649987779e-06,
+      "loss": 0.5632,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9251150848421499,
+      "learning_rate": 8.146318490400958e-06,
+      "loss": 0.73,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9096484516775554,
+      "learning_rate": 8.145009974161002e-06,
+      "loss": 0.6355,
+      "step": 2934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.014743746968432,
+      "learning_rate": 8.14370110141624e-06,
+      "loss": 0.604,
+      "step": 2935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7574927081153062,
+      "learning_rate": 8.142391872315038e-06,
+      "loss": 0.5701,
+      "step": 2936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7754012395866758,
+      "learning_rate": 8.141082287005808e-06,
+      "loss": 0.7174,
+      "step": 2937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7225274879562285,
+      "learning_rate": 8.139772345636996e-06,
+      "loss": 0.5452,
+      "step": 2938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.88076198186382,
+      "learning_rate": 8.138462048357093e-06,
+      "loss": 0.6864,
+      "step": 2939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.099299123705203,
+      "learning_rate": 8.137151395314628e-06,
+      "loss": 0.6782,
+      "step": 2940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8633877364101679,
+      "learning_rate": 8.13584038665817e-06,
+      "loss": 0.6702,
+      "step": 2941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7629071756590513,
+      "learning_rate": 8.134529022536332e-06,
+      "loss": 0.6396,
+      "step": 2942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8434231548356848,
+      "learning_rate": 8.133217303097764e-06,
+      "loss": 0.7665,
+      "step": 2943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0000437586736637,
+      "learning_rate": 8.131905228491155e-06,
+      "loss": 0.655,
+      "step": 2944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0562680118666377,
+      "learning_rate": 8.130592798865237e-06,
+      "loss": 0.623,
+      "step": 2945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8727560280886,
+      "learning_rate": 8.129280014368781e-06,
+      "loss": 0.5804,
+      "step": 2946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7842157138360704,
+      "learning_rate": 8.1279668751506e-06,
+      "loss": 0.7247,
+      "step": 2947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8652448850439833,
+      "learning_rate": 8.126653381359543e-06,
+      "loss": 0.6968,
+      "step": 2948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2497973252323638,
+      "learning_rate": 8.125339533144507e-06,
+      "loss": 0.6329,
+      "step": 2949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7867873852205007,
+      "learning_rate": 8.12402533065442e-06,
+      "loss": 0.5789,
+      "step": 2950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.077399777170234,
+      "learning_rate": 8.122710774038253e-06,
+      "loss": 0.6199,
+      "step": 2951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9012648354483244,
+      "learning_rate": 8.121395863445023e-06,
+      "loss": 0.7289,
+      "step": 2952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.824018411625984,
+      "learning_rate": 8.120080599023781e-06,
+      "loss": 0.6642,
+      "step": 2953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9300181390182036,
+      "learning_rate": 8.118764980923619e-06,
+      "loss": 0.6938,
+      "step": 2954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8854008336004575,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.6819,
+      "step": 2955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.966959364350495,
+      "learning_rate": 8.116132684283104e-06,
+      "loss": 0.6794,
+      "step": 2956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.905465999690345,
+      "learning_rate": 8.11481600604114e-06,
+      "loss": 0.6242,
+      "step": 2957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.945562288125105,
+      "learning_rate": 8.113498974717027e-06,
+      "loss": 0.5482,
+      "step": 2958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.086970708973281,
+      "learning_rate": 8.11218159046006e-06,
+      "loss": 0.7742,
+      "step": 2959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8443539974736436,
+      "learning_rate": 8.110863853419568e-06,
+      "loss": 0.7127,
+      "step": 2960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.890393352231358,
+      "learning_rate": 8.10954576374493e-06,
+      "loss": 0.6969,
+      "step": 2961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.033669898074698,
+      "learning_rate": 8.108227321585554e-06,
+      "loss": 0.6983,
+      "step": 2962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9019326562773995,
+      "learning_rate": 8.106908527090895e-06,
+      "loss": 0.6637,
+      "step": 2963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9436938800459622,
+      "learning_rate": 8.105589380410448e-06,
+      "loss": 0.6138,
+      "step": 2964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.931517898910024,
+      "learning_rate": 8.10426988169374e-06,
+      "loss": 0.6288,
+      "step": 2965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.01533596468131,
+      "learning_rate": 8.10295003109035e-06,
+      "loss": 0.6901,
+      "step": 2966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9307107562996824,
+      "learning_rate": 8.101629828749887e-06,
+      "loss": 0.6474,
+      "step": 2967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7937486278251964,
+      "learning_rate": 8.100309274822002e-06,
+      "loss": 0.6022,
+      "step": 2968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8036146364537236,
+      "learning_rate": 8.098988369456392e-06,
+      "loss": 0.6685,
+      "step": 2969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8502562122591037,
+      "learning_rate": 8.097667112802784e-06,
+      "loss": 0.7181,
+      "step": 2970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8619088903053767,
+      "learning_rate": 8.096345505010956e-06,
+      "loss": 0.6382,
+      "step": 2971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0072848292508727,
+      "learning_rate": 8.095023546230715e-06,
+      "loss": 0.6269,
+      "step": 2972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0408515277535035,
+      "learning_rate": 8.093701236611914e-06,
+      "loss": 0.7009,
+      "step": 2973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8022765607025482,
+      "learning_rate": 8.092378576304443e-06,
+      "loss": 0.6222,
+      "step": 2974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1356198828409583,
+      "learning_rate": 8.091055565458236e-06,
+      "loss": 0.8086,
+      "step": 2975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7900732105124129,
+      "learning_rate": 8.089732204223263e-06,
+      "loss": 0.6219,
+      "step": 2976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9052543682026897,
+      "learning_rate": 8.088408492749534e-06,
+      "loss": 0.6403,
+      "step": 2977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.831209876851521,
+      "learning_rate": 8.087084431187096e-06,
+      "loss": 0.6012,
+      "step": 2978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.015018382951914,
+      "learning_rate": 8.085760019686044e-06,
+      "loss": 0.7143,
+      "step": 2979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.990275012744501,
+      "learning_rate": 8.084435258396504e-06,
+      "loss": 0.6116,
+      "step": 2980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7622362259074167,
+      "learning_rate": 8.08311014746865e-06,
+      "loss": 0.62,
+      "step": 2981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.102214569940529,
+      "learning_rate": 8.081784687052683e-06,
+      "loss": 0.5395,
+      "step": 2982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8920824650503145,
+      "learning_rate": 8.080458877298861e-06,
+      "loss": 0.6173,
+      "step": 2983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.825716574244372,
+      "learning_rate": 8.079132718357465e-06,
+      "loss": 0.715,
+      "step": 2984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8137038198264506,
+      "learning_rate": 8.077806210378824e-06,
+      "loss": 0.6763,
+      "step": 2985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9100852008920197,
+      "learning_rate": 8.076479353513308e-06,
+      "loss": 0.7405,
+      "step": 2986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.008643615507193,
+      "learning_rate": 8.07515214791132e-06,
+      "loss": 0.6724,
+      "step": 2987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9322056861892911,
+      "learning_rate": 8.073824593723309e-06,
+      "loss": 0.7223,
+      "step": 2988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9330171400421932,
+      "learning_rate": 8.07249669109976e-06,
+      "loss": 0.6259,
+      "step": 2989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.132240605171131,
+      "learning_rate": 8.071168440191199e-06,
+      "loss": 0.6683,
+      "step": 2990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.71224736007814,
+      "learning_rate": 8.06983984114819e-06,
+      "loss": 0.6645,
+      "step": 2991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1475200177604536,
+      "learning_rate": 8.068510894121338e-06,
+      "loss": 0.7044,
+      "step": 2992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8431375374949153,
+      "learning_rate": 8.067181599261285e-06,
+      "loss": 0.5972,
+      "step": 2993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.823307176827159,
+      "learning_rate": 8.065851956718716e-06,
+      "loss": 0.6707,
+      "step": 2994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9937691220905867,
+      "learning_rate": 8.064521966644351e-06,
+      "loss": 0.6577,
+      "step": 2995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7340263444241693,
+      "learning_rate": 8.063191629188958e-06,
+      "loss": 0.5489,
+      "step": 2996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9549885348959344,
+      "learning_rate": 8.06186094450333e-06,
+      "loss": 0.7293,
+      "step": 2997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.759727563771705,
+      "learning_rate": 8.060529912738316e-06,
+      "loss": 0.7183,
+      "step": 2998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8394811728272813,
+      "learning_rate": 8.05919853404479e-06,
+      "loss": 0.6787,
+      "step": 2999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8405697026850263,
+      "learning_rate": 8.057866808573672e-06,
+      "loss": 0.6668,
+      "step": 3000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8378753932405179,
+      "learning_rate": 8.056534736475923e-06,
+      "loss": 0.6546,
+      "step": 3001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7713096886810131,
+      "learning_rate": 8.05520231790254e-06,
+      "loss": 0.7496,
+      "step": 3002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8134475570918778,
+      "learning_rate": 8.053869553004561e-06,
+      "loss": 0.6435,
+      "step": 3003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.847138795743933,
+      "learning_rate": 8.052536441933062e-06,
+      "loss": 0.626,
+      "step": 3004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8868783461313385,
+      "learning_rate": 8.051202984839157e-06,
+      "loss": 0.6443,
+      "step": 3005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6274506637183248,
+      "learning_rate": 8.049869181874002e-06,
+      "loss": 0.6456,
+      "step": 3006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7934457191553317,
+      "learning_rate": 8.048535033188794e-06,
+      "loss": 0.5856,
+      "step": 3007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.925470936240325,
+      "learning_rate": 8.04720053893476e-06,
+      "loss": 0.7103,
+      "step": 3008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0746079297930913,
+      "learning_rate": 8.04586569926318e-06,
+      "loss": 0.8232,
+      "step": 3009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8229354357781145,
+      "learning_rate": 8.04453051432536e-06,
+      "loss": 0.5973,
+      "step": 3010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.17337743919196,
+      "learning_rate": 8.043194984272656e-06,
+      "loss": 0.7325,
+      "step": 3011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8513812475733578,
+      "learning_rate": 8.041859109256452e-06,
+      "loss": 0.6306,
+      "step": 3012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.117685022304302,
+      "learning_rate": 8.04052288942818e-06,
+      "loss": 0.7018,
+      "step": 3013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8467774888427202,
+      "learning_rate": 8.03918632493931e-06,
+      "loss": 0.6535,
+      "step": 3014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.841308511882857,
+      "learning_rate": 8.037849415941346e-06,
+      "loss": 0.5757,
+      "step": 3015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.971472468230098,
+      "learning_rate": 8.036512162585834e-06,
+      "loss": 0.641,
+      "step": 3016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0178646419829653,
+      "learning_rate": 8.035174565024362e-06,
+      "loss": 0.671,
+      "step": 3017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6146581959501902,
+      "learning_rate": 8.033836623408556e-06,
+      "loss": 0.6094,
+      "step": 3018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.873134384365632,
+      "learning_rate": 8.032498337890073e-06,
+      "loss": 0.7449,
+      "step": 3019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0365935522496765,
+      "learning_rate": 8.03115970862062e-06,
+      "loss": 0.6673,
+      "step": 3020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8359556335408065,
+      "learning_rate": 8.029820735751936e-06,
+      "loss": 0.5654,
+      "step": 3021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7270952911249733,
+      "learning_rate": 8.028481419435803e-06,
+      "loss": 0.6331,
+      "step": 3022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2156910867047936,
+      "learning_rate": 8.02714175982404e-06,
+      "loss": 0.7084,
+      "step": 3023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.995649963769197,
+      "learning_rate": 8.025801757068504e-06,
+      "loss": 0.6732,
+      "step": 3024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9708209577525355,
+      "learning_rate": 8.024461411321092e-06,
+      "loss": 0.6283,
+      "step": 3025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7503337006314423,
+      "learning_rate": 8.02312072273374e-06,
+      "loss": 0.6433,
+      "step": 3026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9351669006196723,
+      "learning_rate": 8.021779691458422e-06,
+      "loss": 0.6875,
+      "step": 3027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.654807381732136,
+      "learning_rate": 8.020438317647155e-06,
+      "loss": 0.6628,
+      "step": 3028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8618456773903507,
+      "learning_rate": 8.019096601451987e-06,
+      "loss": 0.6953,
+      "step": 3029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9551349205252637,
+      "learning_rate": 8.017754543025012e-06,
+      "loss": 0.6496,
+      "step": 3030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.780040566722473,
+      "learning_rate": 8.016412142518356e-06,
+      "loss": 0.5778,
+      "step": 3031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8404946597473568,
+      "learning_rate": 8.015069400084194e-06,
+      "loss": 0.7011,
+      "step": 3032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0243422271373612,
+      "learning_rate": 8.013726315874729e-06,
+      "loss": 0.579,
+      "step": 3033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7886309414150088,
+      "learning_rate": 8.012382890042208e-06,
+      "loss": 0.6459,
+      "step": 3034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8805496619276878,
+      "learning_rate": 8.011039122738918e-06,
+      "loss": 0.6622,
+      "step": 3035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.839648694046673,
+      "learning_rate": 8.00969501411718e-06,
+      "loss": 0.6199,
+      "step": 3036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.397442942159792,
+      "learning_rate": 8.008350564329356e-06,
+      "loss": 0.6051,
+      "step": 3037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.031673857387813,
+      "learning_rate": 8.00700577352785e-06,
+      "loss": 0.6209,
+      "step": 3038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6669416983842271,
+      "learning_rate": 8.0056606418651e-06,
+      "loss": 0.6872,
+      "step": 3039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8867515977043507,
+      "learning_rate": 8.004315169493586e-06,
+      "loss": 0.5904,
+      "step": 3040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.755108581811021,
+      "learning_rate": 8.002969356565822e-06,
+      "loss": 0.7178,
+      "step": 3041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.955755740139808,
+      "learning_rate": 8.001623203234366e-06,
+      "loss": 0.6451,
+      "step": 3042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7756454101534862,
+      "learning_rate": 8.00027670965181e-06,
+      "loss": 0.6337,
+      "step": 3043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8908982993281924,
+      "learning_rate": 7.998929875970788e-06,
+      "loss": 0.7074,
+      "step": 3044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8663129056512406,
+      "learning_rate": 7.99758270234397e-06,
+      "loss": 0.5291,
+      "step": 3045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.701969315380072,
+      "learning_rate": 7.99623518892407e-06,
+      "loss": 0.6038,
+      "step": 3046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.104064850307174,
+      "learning_rate": 7.994887335863832e-06,
+      "loss": 0.7409,
+      "step": 3047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8828855956551462,
+      "learning_rate": 7.993539143316044e-06,
+      "loss": 0.6762,
+      "step": 3048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7104446736903443,
+      "learning_rate": 7.992190611433532e-06,
+      "loss": 0.5661,
+      "step": 3049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.07219151701992,
+      "learning_rate": 7.99084174036916e-06,
+      "loss": 0.69,
+      "step": 3050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9078394195572277,
+      "learning_rate": 7.989492530275829e-06,
+      "loss": 0.767,
+      "step": 3051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8675014184687138,
+      "learning_rate": 7.988142981306479e-06,
+      "loss": 0.615,
+      "step": 3052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.713066568608695,
+      "learning_rate": 7.98679309361409e-06,
+      "loss": 0.6929,
+      "step": 3053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9366046584274337,
+      "learning_rate": 7.985442867351682e-06,
+      "loss": 0.7056,
+      "step": 3054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7526108241072664,
+      "learning_rate": 7.984092302672306e-06,
+      "loss": 0.618,
+      "step": 3055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8741774360912886,
+      "learning_rate": 7.98274139972906e-06,
+      "loss": 0.7573,
+      "step": 3056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8106219976633589,
+      "learning_rate": 7.981390158675076e-06,
+      "loss": 0.6292,
+      "step": 3057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8965265207010533,
+      "learning_rate": 7.980038579663523e-06,
+      "loss": 0.7095,
+      "step": 3058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7954077915009716,
+      "learning_rate": 7.978686662847612e-06,
+      "loss": 0.6326,
+      "step": 3059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0120709348450947,
+      "learning_rate": 7.977334408380588e-06,
+      "loss": 0.6684,
+      "step": 3060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7127762611593402,
+      "learning_rate": 7.975981816415741e-06,
+      "loss": 0.5592,
+      "step": 3061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9838766263136216,
+      "learning_rate": 7.974628887106391e-06,
+      "loss": 0.5739,
+      "step": 3062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8024132426453374,
+      "learning_rate": 7.973275620605903e-06,
+      "loss": 0.6057,
+      "step": 3063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.662452177175643,
+      "learning_rate": 7.971922017067674e-06,
+      "loss": 0.5933,
+      "step": 3064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9625977664980205,
+      "learning_rate": 7.970568076645149e-06,
+      "loss": 0.5994,
+      "step": 3065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.006758324902627,
+      "learning_rate": 7.969213799491799e-06,
+      "loss": 0.6422,
+      "step": 3066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.883116643768088,
+      "learning_rate": 7.96785918576114e-06,
+      "loss": 0.6657,
+      "step": 3067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.162848986767221,
+      "learning_rate": 7.966504235606726e-06,
+      "loss": 0.7306,
+      "step": 3068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1148968110067687,
+      "learning_rate": 7.965148949182148e-06,
+      "loss": 0.6631,
+      "step": 3069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7761884516910715,
+      "learning_rate": 7.963793326641038e-06,
+      "loss": 0.6629,
+      "step": 3070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9040220164550217,
+      "learning_rate": 7.96243736813706e-06,
+      "loss": 0.6321,
+      "step": 3071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.993692247613143,
+      "learning_rate": 7.961081073823921e-06,
+      "loss": 0.6548,
+      "step": 3072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7461663526967959,
+      "learning_rate": 7.959724443855366e-06,
+      "loss": 0.6201,
+      "step": 3073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9928755049849687,
+      "learning_rate": 7.958367478385172e-06,
+      "loss": 0.6746,
+      "step": 3074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7754560830370942,
+      "learning_rate": 7.957010177567167e-06,
+      "loss": 0.6824,
+      "step": 3075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.019669666508399,
+      "learning_rate": 7.955652541555198e-06,
+      "loss": 0.6759,
+      "step": 3076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.868951468579273,
+      "learning_rate": 7.954294570503171e-06,
+      "loss": 0.7548,
+      "step": 3077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.845669333759357,
+      "learning_rate": 7.952936264565015e-06,
+      "loss": 0.6105,
+      "step": 3078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.821381002259593,
+      "learning_rate": 7.951577623894701e-06,
+      "loss": 0.6323,
+      "step": 3079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9570419308484384,
+      "learning_rate": 7.950218648646242e-06,
+      "loss": 0.708,
+      "step": 3080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8858894304328615,
+      "learning_rate": 7.948859338973682e-06,
+      "loss": 0.6006,
+      "step": 3081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7831579814605896,
+      "learning_rate": 7.947499695031108e-06,
+      "loss": 0.7099,
+      "step": 3082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8783313094793848,
+      "learning_rate": 7.946139716972644e-06,
+      "loss": 0.732,
+      "step": 3083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0736000767971667,
+      "learning_rate": 7.94477940495245e-06,
+      "loss": 0.7732,
+      "step": 3084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.077886034109067,
+      "learning_rate": 7.943418759124727e-06,
+      "loss": 0.6822,
+      "step": 3085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7930850277354924,
+      "learning_rate": 7.94205777964371e-06,
+      "loss": 0.6518,
+      "step": 3086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8907545370407564,
+      "learning_rate": 7.940696466663674e-06,
+      "loss": 0.6399,
+      "step": 3087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.838018871743233,
+      "learning_rate": 7.939334820338933e-06,
+      "loss": 0.5909,
+      "step": 3088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9243946073025104,
+      "learning_rate": 7.937972840823836e-06,
+      "loss": 0.6856,
+      "step": 3089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0112204626320285,
+      "learning_rate": 7.93661052827277e-06,
+      "loss": 0.7325,
+      "step": 3090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9773351694370116,
+      "learning_rate": 7.935247882840164e-06,
+      "loss": 0.6951,
+      "step": 3091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8575930787334212,
+      "learning_rate": 7.933884904680482e-06,
+      "loss": 0.6292,
+      "step": 3092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8549941118391462,
+      "learning_rate": 7.93252159394822e-06,
+      "loss": 0.7249,
+      "step": 3093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8390883570912002,
+      "learning_rate": 7.931157950797923e-06,
+      "loss": 0.5553,
+      "step": 3094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.979121650127033,
+      "learning_rate": 7.929793975384164e-06,
+      "loss": 0.6871,
+      "step": 3095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8303793323181003,
+      "learning_rate": 7.92842966786156e-06,
+      "loss": 0.6484,
+      "step": 3096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8775457173794086,
+      "learning_rate": 7.92706502838476e-06,
+      "loss": 0.7069,
+      "step": 3097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7512552354294855,
+      "learning_rate": 7.925700057108455e-06,
+      "loss": 0.6866,
+      "step": 3098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.036055977573231,
+      "learning_rate": 7.924334754187373e-06,
+      "loss": 0.6938,
+      "step": 3099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9444583395675479,
+      "learning_rate": 7.92296911977628e-06,
+      "loss": 0.6017,
+      "step": 3100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7671077634825694,
+      "learning_rate": 7.921603154029976e-06,
+      "loss": 0.6438,
+      "step": 3101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7985327525448345,
+      "learning_rate": 7.920236857103301e-06,
+      "loss": 0.5896,
+      "step": 3102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9693985426170422,
+      "learning_rate": 7.918870229151134e-06,
+      "loss": 0.6559,
+      "step": 3103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.756742744849359,
+      "learning_rate": 7.91750327032839e-06,
+      "loss": 0.6919,
+      "step": 3104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7404813813771645,
+      "learning_rate": 7.91613598079002e-06,
+      "loss": 0.6627,
+      "step": 3105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7682475752042812,
+      "learning_rate": 7.914768360691017e-06,
+      "loss": 0.6542,
+      "step": 3106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9117815109314706,
+      "learning_rate": 7.913400410186406e-06,
+      "loss": 0.6916,
+      "step": 3107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0622229043342157,
+      "learning_rate": 7.912032129431251e-06,
+      "loss": 0.7196,
+      "step": 3108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.767446056151579,
+      "learning_rate": 7.91066351858066e-06,
+      "loss": 0.6045,
+      "step": 3109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8688982836066614,
+      "learning_rate": 7.909294577789765e-06,
+      "loss": 0.6622,
+      "step": 3110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.005588406771868,
+      "learning_rate": 7.907925307213748e-06,
+      "loss": 0.7336,
+      "step": 3111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8008940358030987,
+      "learning_rate": 7.906555707007823e-06,
+      "loss": 0.6675,
+      "step": 3112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.912571767612056,
+      "learning_rate": 7.905185777327242e-06,
+      "loss": 0.6646,
+      "step": 3113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6446609630621263,
+      "learning_rate": 7.903815518327295e-06,
+      "loss": 0.6515,
+      "step": 3114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.982274144676952,
+      "learning_rate": 7.902444930163308e-06,
+      "loss": 0.6789,
+      "step": 3115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8551387879880945,
+      "learning_rate": 7.901074012990645e-06,
+      "loss": 0.5955,
+      "step": 3116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.943758584895994,
+      "learning_rate": 7.899702766964705e-06,
+      "loss": 0.7111,
+      "step": 3117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8648274627942667,
+      "learning_rate": 7.898331192240929e-06,
+      "loss": 0.7082,
+      "step": 3118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9477847308322247,
+      "learning_rate": 7.896959288974792e-06,
+      "loss": 0.7108,
+      "step": 3119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.885086662138018,
+      "learning_rate": 7.89558705732181e-06,
+      "loss": 0.6246,
+      "step": 3120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9888061668612529,
+      "learning_rate": 7.894214497437528e-06,
+      "loss": 0.7098,
+      "step": 3121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9341899975077337,
+      "learning_rate": 7.892841609477538e-06,
+      "loss": 0.5994,
+      "step": 3122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9079692239038495,
+      "learning_rate": 7.891468393597464e-06,
+      "loss": 0.6772,
+      "step": 3123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.038691877984083,
+      "learning_rate": 7.890094849952964e-06,
+      "loss": 0.6783,
+      "step": 3124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7539539211322792,
+      "learning_rate": 7.888720978699742e-06,
+      "loss": 0.6765,
+      "step": 3125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8011122274916558,
+      "learning_rate": 7.88734677999353e-06,
+      "loss": 0.6645,
+      "step": 3126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9309782177573798,
+      "learning_rate": 7.885972253990104e-06,
+      "loss": 0.711,
+      "step": 3127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8744956494093434,
+      "learning_rate": 7.884597400845273e-06,
+      "loss": 0.6176,
+      "step": 3128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8638622827816949,
+      "learning_rate": 7.883222220714886e-06,
+      "loss": 0.7032,
+      "step": 3129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.776577837690079,
+      "learning_rate": 7.881846713754826e-06,
+      "loss": 0.6416,
+      "step": 3130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9796058522752333,
+      "learning_rate": 7.880470880121015e-06,
+      "loss": 0.628,
+      "step": 3131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9181012593383686,
+      "learning_rate": 7.879094719969412e-06,
+      "loss": 0.6968,
+      "step": 3132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7820935673678577,
+      "learning_rate": 7.87771823345601e-06,
+      "loss": 0.6633,
+      "step": 3133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7651158173417194,
+      "learning_rate": 7.876341420736847e-06,
+      "loss": 0.6512,
+      "step": 3134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8206169688036722,
+      "learning_rate": 7.874964281967988e-06,
+      "loss": 0.6045,
+      "step": 3135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0786032378811115,
+      "learning_rate": 7.87358681730554e-06,
+      "loss": 0.7924,
+      "step": 3136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1478641904169677,
+      "learning_rate": 7.872209026905648e-06,
+      "loss": 0.8035,
+      "step": 3137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8482328317441383,
+      "learning_rate": 7.870830910924491e-06,
+      "loss": 0.6109,
+      "step": 3138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9746459109210937,
+      "learning_rate": 7.869452469518291e-06,
+      "loss": 0.6593,
+      "step": 3139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3263447551499503,
+      "learning_rate": 7.868073702843294e-06,
+      "loss": 0.7088,
+      "step": 3140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.790599477651934,
+      "learning_rate": 7.866694611055796e-06,
+      "loss": 0.6343,
+      "step": 3141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.031239757144938,
+      "learning_rate": 7.865315194312125e-06,
+      "loss": 0.6643,
+      "step": 3142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0555372887136754,
+      "learning_rate": 7.863935452768645e-06,
+      "loss": 0.6539,
+      "step": 3143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9539749091015681,
+      "learning_rate": 7.862555386581758e-06,
+      "loss": 0.6238,
+      "step": 3144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7643599459994184,
+      "learning_rate": 7.861174995907901e-06,
+      "loss": 0.6501,
+      "step": 3145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7358705962081895,
+      "learning_rate": 7.85979428090355e-06,
+      "loss": 0.6352,
+      "step": 3146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8716834585310764,
+      "learning_rate": 7.858413241725219e-06,
+      "loss": 0.6743,
+      "step": 3147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9704127723150717,
+      "learning_rate": 7.857031878529452e-06,
+      "loss": 0.7033,
+      "step": 3148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9876812359612106,
+      "learning_rate": 7.855650191472836e-06,
+      "loss": 0.6894,
+      "step": 3149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8676202472340744,
+      "learning_rate": 7.854268180711997e-06,
+      "loss": 0.6776,
+      "step": 3150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.016838800702593,
+      "learning_rate": 7.852885846403591e-06,
+      "loss": 0.5783,
+      "step": 3151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.059826072634291,
+      "learning_rate": 7.851503188704312e-06,
+      "loss": 0.6996,
+      "step": 3152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9390691952182482,
+      "learning_rate": 7.850120207770893e-06,
+      "loss": 0.7219,
+      "step": 3153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8477942788865895,
+      "learning_rate": 7.848736903760106e-06,
+      "loss": 0.6796,
+      "step": 3154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7632667358112966,
+      "learning_rate": 7.847353276828751e-06,
+      "loss": 0.719,
+      "step": 3155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7943354805743594,
+      "learning_rate": 7.845969327133673e-06,
+      "loss": 0.6816,
+      "step": 3156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.114565393931848,
+      "learning_rate": 7.84458505483175e-06,
+      "loss": 0.7039,
+      "step": 3157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0487194410229215,
+      "learning_rate": 7.8432004600799e-06,
+      "loss": 0.6646,
+      "step": 3158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.196991395817911,
+      "learning_rate": 7.84181554303507e-06,
+      "loss": 0.7162,
+      "step": 3159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.805330609982835,
+      "learning_rate": 7.840430303854251e-06,
+      "loss": 0.6998,
+      "step": 3160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.945875095821067,
+      "learning_rate": 7.839044742694466e-06,
+      "loss": 0.6201,
+      "step": 3161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8994297100188413,
+      "learning_rate": 7.83765885971278e-06,
+      "loss": 0.6215,
+      "step": 3162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7735390094819343,
+      "learning_rate": 7.836272655066286e-06,
+      "loss": 0.6738,
+      "step": 3163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0506083726297994,
+      "learning_rate": 7.83488612891212e-06,
+      "loss": 0.6146,
+      "step": 3164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0616518744706984,
+      "learning_rate": 7.833499281407455e-06,
+      "loss": 0.6962,
+      "step": 3165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9640484511919805,
+      "learning_rate": 7.832112112709496e-06,
+      "loss": 0.6537,
+      "step": 3166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8916743371522864,
+      "learning_rate": 7.830724622975485e-06,
+      "loss": 0.6021,
+      "step": 3167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.068753121157626,
+      "learning_rate": 7.829336812362703e-06,
+      "loss": 0.6653,
+      "step": 3168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8414392119014225,
+      "learning_rate": 7.827948681028467e-06,
+      "loss": 0.6749,
+      "step": 3169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8041772814742751,
+      "learning_rate": 7.826560229130132e-06,
+      "loss": 0.6561,
+      "step": 3170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.277001270558421,
+      "learning_rate": 7.82517145682508e-06,
+      "loss": 0.7026,
+      "step": 3171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9932702774934739,
+      "learning_rate": 7.823782364270743e-06,
+      "loss": 0.6764,
+      "step": 3172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9152540765115291,
+      "learning_rate": 7.82239295162458e-06,
+      "loss": 0.5991,
+      "step": 3173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0171036692649142,
+      "learning_rate": 7.821003219044087e-06,
+      "loss": 0.6495,
+      "step": 3174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9017373728063542,
+      "learning_rate": 7.819613166686802e-06,
+      "loss": 0.6444,
+      "step": 3175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9178976145922144,
+      "learning_rate": 7.818222794710293e-06,
+      "loss": 0.61,
+      "step": 3176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3536736709136723,
+      "learning_rate": 7.816832103272165e-06,
+      "loss": 0.5589,
+      "step": 3177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5473526112248663,
+      "learning_rate": 7.815441092530064e-06,
+      "loss": 0.7009,
+      "step": 3178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9025338748284415,
+      "learning_rate": 7.814049762641668e-06,
+      "loss": 0.665,
+      "step": 3179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9662648376077074,
+      "learning_rate": 7.812658113764691e-06,
+      "loss": 0.8229,
+      "step": 3180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8632716806392016,
+      "learning_rate": 7.811266146056886e-06,
+      "loss": 0.7367,
+      "step": 3181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.082465852054203,
+      "learning_rate": 7.80987385967604e-06,
+      "loss": 0.696,
+      "step": 3182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9024013485348124,
+      "learning_rate": 7.808481254779975e-06,
+      "loss": 0.7804,
+      "step": 3183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.785436285512577,
+      "learning_rate": 7.807088331526553e-06,
+      "loss": 0.4849,
+      "step": 3184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.933460760859134,
+      "learning_rate": 7.805695090073668e-06,
+      "loss": 0.6479,
+      "step": 3185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8728996996369847,
+      "learning_rate": 7.804301530579253e-06,
+      "loss": 0.6551,
+      "step": 3186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9368897835831256,
+      "learning_rate": 7.802907653201275e-06,
+      "loss": 0.7015,
+      "step": 3187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8425480322102283,
+      "learning_rate": 7.801513458097741e-06,
+      "loss": 0.6353,
+      "step": 3188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9522627071910101,
+      "learning_rate": 7.800118945426684e-06,
+      "loss": 0.6558,
+      "step": 3189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9459414208532624,
+      "learning_rate": 7.798724115346188e-06,
+      "loss": 0.6414,
+      "step": 3190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8198890780658505,
+      "learning_rate": 7.797328968014359e-06,
+      "loss": 0.6503,
+      "step": 3191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.989239342204068,
+      "learning_rate": 7.795933503589349e-06,
+      "loss": 0.6533,
+      "step": 3192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9017341494606674,
+      "learning_rate": 7.79453772222934e-06,
+      "loss": 0.6688,
+      "step": 3193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.790821468643048,
+      "learning_rate": 7.793141624092551e-06,
+      "loss": 0.6868,
+      "step": 3194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1081525065527806,
+      "learning_rate": 7.791745209337239e-06,
+      "loss": 0.7204,
+      "step": 3195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9121572482641518,
+      "learning_rate": 7.790348478121695e-06,
+      "loss": 0.7305,
+      "step": 3196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.197598958597094,
+      "learning_rate": 7.788951430604246e-06,
+      "loss": 0.7218,
+      "step": 3197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7643539813858016,
+      "learning_rate": 7.787554066943256e-06,
+      "loss": 0.5945,
+      "step": 3198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9199918036679502,
+      "learning_rate": 7.786156387297126e-06,
+      "loss": 0.6765,
+      "step": 3199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9184094585478977,
+      "learning_rate": 7.784758391824286e-06,
+      "loss": 0.649,
+      "step": 3200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9359112716415736,
+      "learning_rate": 7.783360080683212e-06,
+      "loss": 0.6215,
+      "step": 3201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6650658879147164,
+      "learning_rate": 7.781961454032407e-06,
+      "loss": 0.6007,
+      "step": 3202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8712142665583873,
+      "learning_rate": 7.780562512030414e-06,
+      "loss": 0.6509,
+      "step": 3203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9979865163071713,
+      "learning_rate": 7.77916325483581e-06,
+      "loss": 0.7924,
+      "step": 3204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9307099915203076,
+      "learning_rate": 7.777763682607214e-06,
+      "loss": 0.6831,
+      "step": 3205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9719246908216845,
+      "learning_rate": 7.77636379550327e-06,
+      "loss": 0.6781,
+      "step": 3206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.281909905417469,
+      "learning_rate": 7.774963593682667e-06,
+      "loss": 0.7001,
+      "step": 3207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9134686519705317,
+      "learning_rate": 7.773563077304123e-06,
+      "loss": 0.6533,
+      "step": 3208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7375846606061787,
+      "learning_rate": 7.772162246526394e-06,
+      "loss": 0.6,
+      "step": 3209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9210680843386594,
+      "learning_rate": 7.770761101508274e-06,
+      "loss": 0.6591,
+      "step": 3210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.858930054263607,
+      "learning_rate": 7.769359642408591e-06,
+      "loss": 0.7036,
+      "step": 3211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.024890925973368,
+      "learning_rate": 7.767957869386208e-06,
+      "loss": 0.6565,
+      "step": 3212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8619952533612665,
+      "learning_rate": 7.766555782600023e-06,
+      "loss": 0.5979,
+      "step": 3213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6727164069719362,
+      "learning_rate": 7.765153382208972e-06,
+      "loss": 0.6987,
+      "step": 3214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8507203779819241,
+      "learning_rate": 7.763750668372023e-06,
+      "loss": 0.628,
+      "step": 3215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9341511347273956,
+      "learning_rate": 7.762347641248182e-06,
+      "loss": 0.607,
+      "step": 3216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8954280950807452,
+      "learning_rate": 7.760944300996494e-06,
+      "loss": 0.6493,
+      "step": 3217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.103132213288938,
+      "learning_rate": 7.759540647776031e-06,
+      "loss": 0.656,
+      "step": 3218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.964201049747272,
+      "learning_rate": 7.758136681745907e-06,
+      "loss": 0.6294,
+      "step": 3219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.731784592727429,
+      "learning_rate": 7.756732403065269e-06,
+      "loss": 0.633,
+      "step": 3220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8723486158466771,
+      "learning_rate": 7.7553278118933e-06,
+      "loss": 0.5642,
+      "step": 3221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.912846816349759,
+      "learning_rate": 7.75392290838922e-06,
+      "loss": 0.6209,
+      "step": 3222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0355518598418763,
+      "learning_rate": 7.75251769271228e-06,
+      "loss": 0.6866,
+      "step": 3223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8024911163155974,
+      "learning_rate": 7.751112165021771e-06,
+      "loss": 0.6965,
+      "step": 3224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.992035602097195,
+      "learning_rate": 7.749706325477017e-06,
+      "loss": 0.727,
+      "step": 3225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.012344638600577,
+      "learning_rate": 7.74830017423738e-06,
+      "loss": 0.6522,
+      "step": 3226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7334097631987118,
+      "learning_rate": 7.74689371146225e-06,
+      "loss": 0.7054,
+      "step": 3227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.910822938912061,
+      "learning_rate": 7.745486937311065e-06,
+      "loss": 0.699,
+      "step": 3228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.189272683241941,
+      "learning_rate": 7.744079851943286e-06,
+      "loss": 0.6648,
+      "step": 3229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.777075786244854,
+      "learning_rate": 7.742672455518413e-06,
+      "loss": 0.7164,
+      "step": 3230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7858757362545894,
+      "learning_rate": 7.741264748195984e-06,
+      "loss": 0.6449,
+      "step": 3231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9753861500283088,
+      "learning_rate": 7.739856730135575e-06,
+      "loss": 0.6902,
+      "step": 3232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6698498743407861,
+      "learning_rate": 7.738448401496785e-06,
+      "loss": 0.6012,
+      "step": 3233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0095261492551115,
+      "learning_rate": 7.737039762439263e-06,
+      "loss": 0.7093,
+      "step": 3234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7863240773519469,
+      "learning_rate": 7.735630813122683e-06,
+      "loss": 0.6411,
+      "step": 3235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7699717846306777,
+      "learning_rate": 7.734221553706756e-06,
+      "loss": 0.6196,
+      "step": 3236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8298275662451435,
+      "learning_rate": 7.732811984351232e-06,
+      "loss": 0.6401,
+      "step": 3237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8757903882230937,
+      "learning_rate": 7.731402105215892e-06,
+      "loss": 0.6505,
+      "step": 3238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8771166804066328,
+      "learning_rate": 7.729991916460558e-06,
+      "loss": 0.6448,
+      "step": 3239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0422800495620725,
+      "learning_rate": 7.728581418245078e-06,
+      "loss": 0.7242,
+      "step": 3240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9500585202348317,
+      "learning_rate": 7.727170610729341e-06,
+      "loss": 0.6406,
+      "step": 3241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9179487116731881,
+      "learning_rate": 7.725759494073272e-06,
+      "loss": 0.709,
+      "step": 3242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8442830499995486,
+      "learning_rate": 7.72434806843683e-06,
+      "loss": 0.5964,
+      "step": 3243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9596900219378959,
+      "learning_rate": 7.722936333980002e-06,
+      "loss": 0.6349,
+      "step": 3244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8473944971850758,
+      "learning_rate": 7.721524290862821e-06,
+      "loss": 0.6101,
+      "step": 3245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.903216419149971,
+      "learning_rate": 7.720111939245351e-06,
+      "loss": 0.6553,
+      "step": 3246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1706479120378255,
+      "learning_rate": 7.71869927928769e-06,
+      "loss": 0.5984,
+      "step": 3247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9082270614604948,
+      "learning_rate": 7.717286311149967e-06,
+      "loss": 0.5559,
+      "step": 3248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8074082902284876,
+      "learning_rate": 7.715873034992354e-06,
+      "loss": 0.6667,
+      "step": 3249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4590764946541355,
+      "learning_rate": 7.714459450975052e-06,
+      "loss": 0.6183,
+      "step": 3250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7625406627291564,
+      "learning_rate": 7.7130455592583e-06,
+      "loss": 0.7122,
+      "step": 3251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.104638475133327,
+      "learning_rate": 7.71163136000237e-06,
+      "loss": 0.6527,
+      "step": 3252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8891265825485541,
+      "learning_rate": 7.710216853367568e-06,
+      "loss": 0.6901,
+      "step": 3253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2146816583793743,
+      "learning_rate": 7.708802039514238e-06,
+      "loss": 0.7759,
+      "step": 3254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9382427214432836,
+      "learning_rate": 7.707386918602759e-06,
+      "loss": 0.6797,
+      "step": 3255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.070382342712708,
+      "learning_rate": 7.70597149079354e-06,
+      "loss": 0.6483,
+      "step": 3256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.767128407067787,
+      "learning_rate": 7.70455575624703e-06,
+      "loss": 0.5626,
+      "step": 3257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7058603097337257,
+      "learning_rate": 7.70313971512371e-06,
+      "loss": 0.4997,
+      "step": 3258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7513131467756609,
+      "learning_rate": 7.701723367584094e-06,
+      "loss": 0.6709,
+      "step": 3259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8236846567523217,
+      "learning_rate": 7.700306713788735e-06,
+      "loss": 0.6769,
+      "step": 3260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1815094050601966,
+      "learning_rate": 7.69888975389822e-06,
+      "loss": 0.7312,
+      "step": 3261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8049118711336998,
+      "learning_rate": 7.697472488073168e-06,
+      "loss": 0.6029,
+      "step": 3262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6981106891148934,
+      "learning_rate": 7.696054916474235e-06,
+      "loss": 0.6433,
+      "step": 3263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.891118335430464,
+      "learning_rate": 7.694637039262109e-06,
+      "loss": 0.6485,
+      "step": 3264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7889227423611733,
+      "learning_rate": 7.693218856597515e-06,
+      "loss": 0.6024,
+      "step": 3265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9373484978474884,
+      "learning_rate": 7.691800368641214e-06,
+      "loss": 0.6609,
+      "step": 3266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7546189984312635,
+      "learning_rate": 7.690381575553998e-06,
+      "loss": 0.6001,
+      "step": 3267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.742031273653571,
+      "learning_rate": 7.688962477496696e-06,
+      "loss": 0.6134,
+      "step": 3268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8461145661001945,
+      "learning_rate": 7.68754307463017e-06,
+      "loss": 0.6021,
+      "step": 3269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.852835261703502,
+      "learning_rate": 7.68612336711532e-06,
+      "loss": 0.6454,
+      "step": 3270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2319112287773493,
+      "learning_rate": 7.684703355113074e-06,
+      "loss": 0.799,
+      "step": 3271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8107744027642172,
+      "learning_rate": 7.683283038784402e-06,
+      "loss": 0.6029,
+      "step": 3272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0367129090031293,
+      "learning_rate": 7.681862418290302e-06,
+      "loss": 0.7111,
+      "step": 3273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8995914346640856,
+      "learning_rate": 7.68044149379181e-06,
+      "loss": 0.6072,
+      "step": 3274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8071434072693784,
+      "learning_rate": 7.679020265449999e-06,
+      "loss": 0.6942,
+      "step": 3275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8946083630450248,
+      "learning_rate": 7.67759873342597e-06,
+      "loss": 0.666,
+      "step": 3276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8185369692074362,
+      "learning_rate": 7.676176897880862e-06,
+      "loss": 0.6498,
+      "step": 3277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8510420885532568,
+      "learning_rate": 7.67475475897585e-06,
+      "loss": 0.5591,
+      "step": 3278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8125241947077075,
+      "learning_rate": 7.67333231687214e-06,
+      "loss": 0.6652,
+      "step": 3279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6649027457440944,
+      "learning_rate": 7.671909571730974e-06,
+      "loss": 0.6202,
+      "step": 3280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6610788823267293,
+      "learning_rate": 7.67048652371363e-06,
+      "loss": 0.5144,
+      "step": 3281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.051859824910609,
+      "learning_rate": 7.669063172981415e-06,
+      "loss": 0.8101,
+      "step": 3282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8135258363900928,
+      "learning_rate": 7.667639519695678e-06,
+      "loss": 0.6548,
+      "step": 3283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0789065749765734,
+      "learning_rate": 7.666215564017797e-06,
+      "loss": 0.7535,
+      "step": 3284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9976097148077065,
+      "learning_rate": 7.664791306109183e-06,
+      "loss": 0.6333,
+      "step": 3285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.792724714904917,
+      "learning_rate": 7.663366746131286e-06,
+      "loss": 0.6677,
+      "step": 3286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.01046056944132,
+      "learning_rate": 7.661941884245589e-06,
+      "loss": 0.5846,
+      "step": 3287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8378314305059873,
+      "learning_rate": 7.660516720613606e-06,
+      "loss": 0.7189,
+      "step": 3288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.116097571217621,
+      "learning_rate": 7.659091255396888e-06,
+      "loss": 0.6876,
+      "step": 3289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.452763925603077,
+      "learning_rate": 7.65766548875702e-06,
+      "loss": 0.6209,
+      "step": 3290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7086733403610908,
+      "learning_rate": 7.656239420855621e-06,
+      "loss": 0.6561,
+      "step": 3291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1055982116655105,
+      "learning_rate": 7.654813051854345e-06,
+      "loss": 0.6497,
+      "step": 3292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.945103058631345,
+      "learning_rate": 7.653386381914874e-06,
+      "loss": 0.7458,
+      "step": 3293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8639467359374053,
+      "learning_rate": 7.651959411198934e-06,
+      "loss": 0.6667,
+      "step": 3294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9314142195588189,
+      "learning_rate": 7.65053213986828e-06,
+      "loss": 0.6434,
+      "step": 3295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.968743977461504,
+      "learning_rate": 7.649104568084701e-06,
+      "loss": 0.6727,
+      "step": 3296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6173702518058752,
+      "learning_rate": 7.64767669601002e-06,
+      "loss": 0.6738,
+      "step": 3297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.819260191517557,
+      "learning_rate": 7.646248523806092e-06,
+      "loss": 0.7187,
+      "step": 3298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.711390241856095,
+      "learning_rate": 7.644820051634813e-06,
+      "loss": 0.5777,
+      "step": 3299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.371999947619344,
+      "learning_rate": 7.643391279658106e-06,
+      "loss": 0.5607,
+      "step": 3300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.080268194932253,
+      "learning_rate": 7.64196220803793e-06,
+      "loss": 0.6454,
+      "step": 3301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8161890897658788,
+      "learning_rate": 7.640532836936279e-06,
+      "loss": 0.6145,
+      "step": 3302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8867414994265745,
+      "learning_rate": 7.639103166515179e-06,
+      "loss": 0.7367,
+      "step": 3303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0829266047249013,
+      "learning_rate": 7.637673196936694e-06,
+      "loss": 0.7419,
+      "step": 3304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.021524026140808,
+      "learning_rate": 7.636242928362918e-06,
+      "loss": 0.7149,
+      "step": 3305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9487343164474362,
+      "learning_rate": 7.634812360955982e-06,
+      "loss": 0.7065,
+      "step": 3306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0147112197597092,
+      "learning_rate": 7.633381494878042e-06,
+      "loss": 0.6809,
+      "step": 3307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.033897246264916,
+      "learning_rate": 7.631950330291305e-06,
+      "loss": 0.6792,
+      "step": 3308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7911815368426338,
+      "learning_rate": 7.630518867357994e-06,
+      "loss": 0.6244,
+      "step": 3309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8465211407886768,
+      "learning_rate": 7.629087106240376e-06,
+      "loss": 0.6385,
+      "step": 3310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8768778769605858,
+      "learning_rate": 7.6276550471007486e-06,
+      "loss": 0.6325,
+      "step": 3311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8761662832688346,
+      "learning_rate": 7.626222690101445e-06,
+      "loss": 0.6617,
+      "step": 3312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8842109296938618,
+      "learning_rate": 7.624790035404831e-06,
+      "loss": 0.6187,
+      "step": 3313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.998556004167101,
+      "learning_rate": 7.623357083173306e-06,
+      "loss": 0.6287,
+      "step": 3314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9369889183368985,
+      "learning_rate": 7.621923833569301e-06,
+      "loss": 0.6092,
+      "step": 3315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8729477963392878,
+      "learning_rate": 7.620490286755286e-06,
+      "loss": 0.6911,
+      "step": 3316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9822928384022405,
+      "learning_rate": 7.619056442893762e-06,
+      "loss": 0.662,
+      "step": 3317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7248626081500673,
+      "learning_rate": 7.61762230214726e-06,
+      "loss": 0.6479,
+      "step": 3318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0789146795929487,
+      "learning_rate": 7.616187864678352e-06,
+      "loss": 0.6182,
+      "step": 3319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.838839059416301,
+      "learning_rate": 7.614753130649638e-06,
+      "loss": 0.656,
+      "step": 3320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7187828108507008,
+      "learning_rate": 7.613318100223752e-06,
+      "loss": 0.6863,
+      "step": 3321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.695488574423953,
+      "learning_rate": 7.611882773563364e-06,
+      "loss": 0.6258,
+      "step": 3322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.758042093943283,
+      "learning_rate": 7.610447150831175e-06,
+      "loss": 0.6084,
+      "step": 3323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9600377009168812,
+      "learning_rate": 7.609011232189925e-06,
+      "loss": 0.6382,
+      "step": 3324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.004029691758048,
+      "learning_rate": 7.60757501780238e-06,
+      "loss": 0.6046,
+      "step": 3325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6927785653074294,
+      "learning_rate": 7.6061385078313424e-06,
+      "loss": 0.6248,
+      "step": 3326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7976204301469354,
+      "learning_rate": 7.604701702439652e-06,
+      "loss": 0.6413,
+      "step": 3327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8544041865004253,
+      "learning_rate": 7.603264601790178e-06,
+      "loss": 0.6129,
+      "step": 3328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.708847319417589,
+      "learning_rate": 7.601827206045822e-06,
+      "loss": 0.5682,
+      "step": 3329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8900069868675693,
+      "learning_rate": 7.600389515369522e-06,
+      "loss": 0.6713,
+      "step": 3330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.667577105224544,
+      "learning_rate": 7.59895152992425e-06,
+      "loss": 0.5176,
+      "step": 3331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.823563725373987,
+      "learning_rate": 7.597513249873008e-06,
+      "loss": 0.6509,
+      "step": 3332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1014119569769187,
+      "learning_rate": 7.5960746753788335e-06,
+      "loss": 0.7477,
+      "step": 3333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9398729490943276,
+      "learning_rate": 7.594635806604797e-06,
+      "loss": 0.6432,
+      "step": 3334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.689141864675774,
+      "learning_rate": 7.593196643714005e-06,
+      "loss": 0.6297,
+      "step": 3335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8997762411676566,
+      "learning_rate": 7.5917571868695905e-06,
+      "loss": 0.6331,
+      "step": 3336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0613676536332335,
+      "learning_rate": 7.5903174362347265e-06,
+      "loss": 0.6842,
+      "step": 3337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9957895885750523,
+      "learning_rate": 7.5888773919726176e-06,
+      "loss": 0.7165,
+      "step": 3338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.018970903898421,
+      "learning_rate": 7.5874370542465005e-06,
+      "loss": 0.6857,
+      "step": 3339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.961825625490136,
+      "learning_rate": 7.585996423219643e-06,
+      "loss": 0.6977,
+      "step": 3340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1271983391631766,
+      "learning_rate": 7.584555499055355e-06,
+      "loss": 0.7688,
+      "step": 3341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7660688433101426,
+      "learning_rate": 7.5831142819169664e-06,
+      "loss": 0.5719,
+      "step": 3342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9487869971090308,
+      "learning_rate": 7.581672771967854e-06,
+      "loss": 0.7091,
+      "step": 3343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9584035184171076,
+      "learning_rate": 7.5802309693714145e-06,
+      "loss": 0.6265,
+      "step": 3344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8973868826432616,
+      "learning_rate": 7.57878887429109e-06,
+      "loss": 0.5412,
+      "step": 3345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8292437136670205,
+      "learning_rate": 7.5773464868903465e-06,
+      "loss": 0.5505,
+      "step": 3346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.110059781420011,
+      "learning_rate": 7.57590380733269e-06,
+      "loss": 0.6323,
+      "step": 3347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8430200609730625,
+      "learning_rate": 7.574460835781654e-06,
+      "loss": 0.5947,
+      "step": 3348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8841432114527707,
+      "learning_rate": 7.573017572400807e-06,
+      "loss": 0.6783,
+      "step": 3349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.013283679598883,
+      "learning_rate": 7.571574017353755e-06,
+      "loss": 0.6922,
+      "step": 3350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.773668795659689,
+      "learning_rate": 7.570130170804129e-06,
+      "loss": 0.6641,
+      "step": 3351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.746076214972598,
+      "learning_rate": 7.5686860329156e-06,
+      "loss": 0.5992,
+      "step": 3352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8380004264405159,
+      "learning_rate": 7.567241603851866e-06,
+      "loss": 0.6498,
+      "step": 3353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.908526988253742,
+      "learning_rate": 7.565796883776666e-06,
+      "loss": 0.679,
+      "step": 3354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7663503430585459,
+      "learning_rate": 7.564351872853763e-06,
+      "loss": 0.6488,
+      "step": 3355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8698403441208586,
+      "learning_rate": 7.5629065712469595e-06,
+      "loss": 0.6297,
+      "step": 3356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7693710236384135,
+      "learning_rate": 7.561460979120088e-06,
+      "loss": 0.6881,
+      "step": 3357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8736827010560169,
+      "learning_rate": 7.560015096637015e-06,
+      "loss": 0.643,
+      "step": 3358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9932305255398057,
+      "learning_rate": 7.558568923961638e-06,
+      "loss": 0.6615,
+      "step": 3359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7329221135497108,
+      "learning_rate": 7.557122461257891e-06,
+      "loss": 0.5559,
+      "step": 3360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7334989136140522,
+      "learning_rate": 7.555675708689738e-06,
+      "loss": 0.7119,
+      "step": 3361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.941030820066443,
+      "learning_rate": 7.554228666421176e-06,
+      "loss": 0.7062,
+      "step": 3362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6938476484173985,
+      "learning_rate": 7.552781334616237e-06,
+      "loss": 0.7022,
+      "step": 3363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.086372882874223,
+      "learning_rate": 7.551333713438982e-06,
+      "loss": 0.5912,
+      "step": 3364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0681430639202976,
+      "learning_rate": 7.549885803053509e-06,
+      "loss": 0.6965,
+      "step": 3365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8155626765570805,
+      "learning_rate": 7.548437603623947e-06,
+      "loss": 0.6455,
+      "step": 3366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0742329496020178,
+      "learning_rate": 7.546989115314456e-06,
+      "loss": 0.6947,
+      "step": 3367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7556510523856663,
+      "learning_rate": 7.5455403382892325e-06,
+      "loss": 0.6135,
+      "step": 3368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9078576489775798,
+      "learning_rate": 7.544091272712501e-06,
+      "loss": 0.6017,
+      "step": 3369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6812110139609036,
+      "learning_rate": 7.542641918748526e-06,
+      "loss": 0.584,
+      "step": 3370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1753283760243174,
+      "learning_rate": 7.541192276561594e-06,
+      "loss": 0.6506,
+      "step": 3371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9246927540441021,
+      "learning_rate": 7.539742346316035e-06,
+      "loss": 0.6418,
+      "step": 3372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0764921092291715,
+      "learning_rate": 7.538292128176204e-06,
+      "loss": 0.7311,
+      "step": 3373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0016266904008444,
+      "learning_rate": 7.536841622306491e-06,
+      "loss": 0.6698,
+      "step": 3374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7476993025192442,
+      "learning_rate": 7.535390828871322e-06,
+      "loss": 0.5843,
+      "step": 3375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1443050656742795,
+      "learning_rate": 7.5339397480351525e-06,
+      "loss": 0.6661,
+      "step": 3376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9640498566143108,
+      "learning_rate": 7.532488379962468e-06,
+      "loss": 0.6737,
+      "step": 3377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.894660279182222,
+      "learning_rate": 7.531036724817791e-06,
+      "loss": 0.6363,
+      "step": 3378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.865129793714187,
+      "learning_rate": 7.529584782765675e-06,
+      "loss": 0.6154,
+      "step": 3379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0631351014194634,
+      "learning_rate": 7.528132553970706e-06,
+      "loss": 0.6748,
+      "step": 3380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6704861005463274,
+      "learning_rate": 7.526680038597502e-06,
+      "loss": 0.6883,
+      "step": 3381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.093149404086043,
+      "learning_rate": 7.525227236810715e-06,
+      "loss": 0.5749,
+      "step": 3382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0045663651778387,
+      "learning_rate": 7.523774148775027e-06,
+      "loss": 0.7112,
+      "step": 3383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.879588638216167,
+      "learning_rate": 7.522320774655154e-06,
+      "loss": 0.632,
+      "step": 3384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.796450452129187,
+      "learning_rate": 7.520867114615844e-06,
+      "loss": 0.6143,
+      "step": 3385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9221269954863394,
+      "learning_rate": 7.519413168821878e-06,
+      "loss": 0.6371,
+      "step": 3386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.245278008970985,
+      "learning_rate": 7.5179589374380705e-06,
+      "loss": 0.6296,
+      "step": 3387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.928866713539612,
+      "learning_rate": 7.516504420629264e-06,
+      "loss": 0.6113,
+      "step": 3388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.295181510678045,
+      "learning_rate": 7.515049618560337e-06,
+      "loss": 0.6031,
+      "step": 3389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.195352983128163,
+      "learning_rate": 7.513594531396202e-06,
+      "loss": 0.6993,
+      "step": 3390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7327908331334299,
+      "learning_rate": 7.512139159301801e-06,
+      "loss": 0.5751,
+      "step": 3391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7569641738070485,
+      "learning_rate": 7.510683502442105e-06,
+      "loss": 0.6413,
+      "step": 3392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6720116832024288,
+      "learning_rate": 7.5092275609821254e-06,
+      "loss": 0.6304,
+      "step": 3393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.164668470616303,
+      "learning_rate": 7.5077713350869e-06,
+      "loss": 0.7014,
+      "step": 3394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0418743773840204,
+      "learning_rate": 7.506314824921498e-06,
+      "loss": 0.7441,
+      "step": 3395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9352876217904793,
+      "learning_rate": 7.504858030651026e-06,
+      "loss": 0.7133,
+      "step": 3396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9893676571328642,
+      "learning_rate": 7.503400952440618e-06,
+      "loss": 0.679,
+      "step": 3397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0088136972631725,
+      "learning_rate": 7.501943590455445e-06,
+      "loss": 0.622,
+      "step": 3398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.635984512656272,
+      "learning_rate": 7.500485944860705e-06,
+      "loss": 0.5652,
+      "step": 3399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.870219344342775,
+      "learning_rate": 7.49902801582163e-06,
+      "loss": 0.676,
+      "step": 3400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.19177716835321,
+      "learning_rate": 7.497569803503486e-06,
+      "loss": 0.7368,
+      "step": 3401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9242849370350925,
+      "learning_rate": 7.49611130807157e-06,
+      "loss": 0.7252,
+      "step": 3402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9182235720120104,
+      "learning_rate": 7.494652529691209e-06,
+      "loss": 0.713,
+      "step": 3403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7468572198863737,
+      "learning_rate": 7.493193468527764e-06,
+      "loss": 0.643,
+      "step": 3404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7308546976807355,
+      "learning_rate": 7.491734124746628e-06,
+      "loss": 0.6003,
+      "step": 3405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8679022278259034,
+      "learning_rate": 7.490274498513228e-06,
+      "loss": 0.7414,
+      "step": 3406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0451720769063906,
+      "learning_rate": 7.488814589993019e-06,
+      "loss": 0.6454,
+      "step": 3407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.954452155359003,
+      "learning_rate": 7.487354399351491e-06,
+      "loss": 0.7069,
+      "step": 3408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0412457459452837,
+      "learning_rate": 7.485893926754164e-06,
+      "loss": 0.706,
+      "step": 3409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.756504654826153,
+      "learning_rate": 7.484433172366592e-06,
+      "loss": 0.5949,
+      "step": 3410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0975397217258207,
+      "learning_rate": 7.482972136354359e-06,
+      "loss": 0.827,
+      "step": 3411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9019894290221375,
+      "learning_rate": 7.48151081888308e-06,
+      "loss": 0.6112,
+      "step": 3412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.961151827754993,
+      "learning_rate": 7.480049220118407e-06,
+      "loss": 0.6559,
+      "step": 3413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1285287007697966,
+      "learning_rate": 7.478587340226019e-06,
+      "loss": 0.6444,
+      "step": 3414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.986160997932849,
+      "learning_rate": 7.477125179371628e-06,
+      "loss": 0.6846,
+      "step": 3415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8237130387040967,
+      "learning_rate": 7.475662737720981e-06,
+      "loss": 0.6,
+      "step": 3416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.851416521972734,
+      "learning_rate": 7.47420001543985e-06,
+      "loss": 0.7632,
+      "step": 3417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9104748163467893,
+      "learning_rate": 7.472737012694045e-06,
+      "loss": 0.5842,
+      "step": 3418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9035555954382448,
+      "learning_rate": 7.471273729649404e-06,
+      "loss": 0.7967,
+      "step": 3419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0514769058465228,
+      "learning_rate": 7.469810166471802e-06,
+      "loss": 0.6309,
+      "step": 3420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.831229869638447,
+      "learning_rate": 7.46834632332714e-06,
+      "loss": 0.6609,
+      "step": 3421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0840899836154985,
+      "learning_rate": 7.466882200381352e-06,
+      "loss": 0.6543,
+      "step": 3422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7908059902561366,
+      "learning_rate": 7.465417797800406e-06,
+      "loss": 0.6881,
+      "step": 3423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9837969625415905,
+      "learning_rate": 7.463953115750302e-06,
+      "loss": 0.633,
+      "step": 3424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6172551301392455,
+      "learning_rate": 7.462488154397067e-06,
+      "loss": 0.6234,
+      "step": 3425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.927272898423346,
+      "learning_rate": 7.461022913906764e-06,
+      "loss": 0.6417,
+      "step": 3426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0551613282894583,
+      "learning_rate": 7.459557394445486e-06,
+      "loss": 0.6011,
+      "step": 3427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0067846400605243,
+      "learning_rate": 7.458091596179359e-06,
+      "loss": 0.6854,
+      "step": 3428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.242882214853698,
+      "learning_rate": 7.4566255192745384e-06,
+      "loss": 0.6827,
+      "step": 3429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9060789364117876,
+      "learning_rate": 7.455159163897213e-06,
+      "loss": 0.6727,
+      "step": 3430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9612708236391978,
+      "learning_rate": 7.453692530213603e-06,
+      "loss": 0.6663,
+      "step": 3431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3743911219968425,
+      "learning_rate": 7.452225618389959e-06,
+      "loss": 0.6794,
+      "step": 3432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8840472801485948,
+      "learning_rate": 7.4507584285925625e-06,
+      "loss": 0.6563,
+      "step": 3433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7866444653145974,
+      "learning_rate": 7.4492909609877304e-06,
+      "loss": 0.7027,
+      "step": 3434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8941736336086783,
+      "learning_rate": 7.447823215741807e-06,
+      "loss": 0.6318,
+      "step": 3435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.782248833519918,
+      "learning_rate": 7.446355193021171e-06,
+      "loss": 0.7005,
+      "step": 3436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9330931951718446,
+      "learning_rate": 7.444886892992229e-06,
+      "loss": 0.7213,
+      "step": 3437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8453021176715279,
+      "learning_rate": 7.443418315821422e-06,
+      "loss": 0.6963,
+      "step": 3438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7615865281146448,
+      "learning_rate": 7.441949461675223e-06,
+      "loss": 0.6233,
+      "step": 3439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9187017752193514,
+      "learning_rate": 7.4404803307201345e-06,
+      "loss": 0.6402,
+      "step": 3440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8966505444836648,
+      "learning_rate": 7.4390109231226895e-06,
+      "loss": 0.6392,
+      "step": 3441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6365254365980388,
+      "learning_rate": 7.437541239049453e-06,
+      "loss": 0.6298,
+      "step": 3442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7596906938070285,
+      "learning_rate": 7.436071278667024e-06,
+      "loss": 0.599,
+      "step": 3443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.787527162481923,
+      "learning_rate": 7.4346010421420314e-06,
+      "loss": 0.6659,
+      "step": 3444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1953328947218864,
+      "learning_rate": 7.433130529641133e-06,
+      "loss": 0.7025,
+      "step": 3445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8839951723071167,
+      "learning_rate": 7.431659741331022e-06,
+      "loss": 0.6574,
+      "step": 3446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7025871231466374,
+      "learning_rate": 7.430188677378418e-06,
+      "loss": 0.6093,
+      "step": 3447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8718629399629925,
+      "learning_rate": 7.4287173379500764e-06,
+      "loss": 0.6797,
+      "step": 3448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9082997423204202,
+      "learning_rate": 7.427245723212781e-06,
+      "loss": 0.5676,
+      "step": 3449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4745611002431454,
+      "learning_rate": 7.425773833333349e-06,
+      "loss": 0.6926,
+      "step": 3450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7970418489642042,
+      "learning_rate": 7.424301668478626e-06,
+      "loss": 0.7225,
+      "step": 3451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0818049027970567,
+      "learning_rate": 7.422829228815491e-06,
+      "loss": 0.6821,
+      "step": 3452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9152217689013808,
+      "learning_rate": 7.421356514510853e-06,
+      "loss": 0.674,
+      "step": 3453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.948628354944815,
+      "learning_rate": 7.419883525731653e-06,
+      "loss": 0.6712,
+      "step": 3454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7921525560109597,
+      "learning_rate": 7.418410262644862e-06,
+      "loss": 0.7164,
+      "step": 3455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8241838649494904,
+      "learning_rate": 7.416936725417483e-06,
+      "loss": 0.6467,
+      "step": 3456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9392896030044404,
+      "learning_rate": 7.415462914216551e-06,
+      "loss": 0.647,
+      "step": 3457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9242618966552065,
+      "learning_rate": 7.41398882920913e-06,
+      "loss": 0.7629,
+      "step": 3458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8239619089818286,
+      "learning_rate": 7.4125144705623155e-06,
+      "loss": 0.5769,
+      "step": 3459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9414170395727977,
+      "learning_rate": 7.411039838443234e-06,
+      "loss": 0.7273,
+      "step": 3460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.796824341854242,
+      "learning_rate": 7.409564933019046e-06,
+      "loss": 0.6165,
+      "step": 3461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.834416215786427,
+      "learning_rate": 7.408089754456939e-06,
+      "loss": 0.5974,
+      "step": 3462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8172898388225946,
+      "learning_rate": 7.406614302924131e-06,
+      "loss": 0.5895,
+      "step": 3463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9176599567813366,
+      "learning_rate": 7.405138578587876e-06,
+      "loss": 0.6056,
+      "step": 3464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.892048274582568,
+      "learning_rate": 7.403662581615454e-06,
+      "loss": 0.7066,
+      "step": 3465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8893025379528445,
+      "learning_rate": 7.402186312174177e-06,
+      "loss": 0.6479,
+      "step": 3466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8654581009570594,
+      "learning_rate": 7.4007097704313894e-06,
+      "loss": 0.5893,
+      "step": 3467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8272464271313358,
+      "learning_rate": 7.399232956554468e-06,
+      "loss": 0.7083,
+      "step": 3468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.810292265366264,
+      "learning_rate": 7.397755870710813e-06,
+      "loss": 0.6151,
+      "step": 3469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8509332314834515,
+      "learning_rate": 7.396278513067865e-06,
+      "loss": 0.616,
+      "step": 3470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6915564968768657,
+      "learning_rate": 7.394800883793087e-06,
+      "loss": 0.6099,
+      "step": 3471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.735592555743763,
+      "learning_rate": 7.393322983053982e-06,
+      "loss": 0.5545,
+      "step": 3472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2703853679664783,
+      "learning_rate": 7.391844811018074e-06,
+      "loss": 0.6984,
+      "step": 3473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0112032870742116,
+      "learning_rate": 7.390366367852923e-06,
+      "loss": 0.6332,
+      "step": 3474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8616175779663644,
+      "learning_rate": 7.38888765372612e-06,
+      "loss": 0.5883,
+      "step": 3475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.712242784645238,
+      "learning_rate": 7.387408668805285e-06,
+      "loss": 0.7109,
+      "step": 3476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6442545413702605,
+      "learning_rate": 7.38592941325807e-06,
+      "loss": 0.6177,
+      "step": 3477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9352886207798174,
+      "learning_rate": 7.384449887252156e-06,
+      "loss": 0.7331,
+      "step": 3478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.918089359140376,
+      "learning_rate": 7.382970090955258e-06,
+      "loss": 0.607,
+      "step": 3479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.653803914611329,
+      "learning_rate": 7.381490024535117e-06,
+      "loss": 0.5688,
+      "step": 3480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8646016700584298,
+      "learning_rate": 7.380009688159507e-06,
+      "loss": 0.7111,
+      "step": 3481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9421559203462602,
+      "learning_rate": 7.378529081996233e-06,
+      "loss": 0.5779,
+      "step": 3482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7791816060422943,
+      "learning_rate": 7.377048206213132e-06,
+      "loss": 0.5849,
+      "step": 3483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9856777896158313,
+      "learning_rate": 7.375567060978067e-06,
+      "loss": 0.6638,
+      "step": 3484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6344717716116572,
+      "learning_rate": 7.374085646458935e-06,
+      "loss": 0.6223,
+      "step": 3485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7495251385174526,
+      "learning_rate": 7.372603962823664e-06,
+      "loss": 0.7742,
+      "step": 3486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.020767726437643,
+      "learning_rate": 7.3711220102402105e-06,
+      "loss": 0.5837,
+      "step": 3487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0961729815094214,
+      "learning_rate": 7.369639788876561e-06,
+      "loss": 0.6203,
+      "step": 3488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9022781368389061,
+      "learning_rate": 7.3681572989007365e-06,
+      "loss": 0.7023,
+      "step": 3489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0866961157120456,
+      "learning_rate": 7.366674540480784e-06,
+      "loss": 0.7635,
+      "step": 3490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.507173647122414,
+      "learning_rate": 7.365191513784782e-06,
+      "loss": 0.5723,
+      "step": 3491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0748742922266414,
+      "learning_rate": 7.363708218980841e-06,
+      "loss": 0.7909,
+      "step": 3492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8333987770622944,
+      "learning_rate": 7.3622246562371e-06,
+      "loss": 0.6826,
+      "step": 3493
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.70355206649613,
+      "learning_rate": 7.360740825721732e-06,
+      "loss": 0.5902,
+      "step": 3494
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9576190398806632,
+      "learning_rate": 7.3592567276029336e-06,
+      "loss": 0.6535,
+      "step": 3495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8723793599589569,
+      "learning_rate": 7.357772362048939e-06,
+      "loss": 0.6424,
+      "step": 3496
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8059004571656276,
+      "learning_rate": 7.356287729228007e-06,
+      "loss": 0.6448,
+      "step": 3497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7440998752846177,
+      "learning_rate": 7.354802829308432e-06,
+      "loss": 0.68,
+      "step": 3498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9870844742740057,
+      "learning_rate": 7.353317662458532e-06,
+      "loss": 0.7523,
+      "step": 3499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6846891263344534,
+      "learning_rate": 7.351832228846664e-06,
+      "loss": 0.5949,
+      "step": 3500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.146763984783295,
+      "learning_rate": 7.3503465286412064e-06,
+      "loss": 0.732,
+      "step": 3501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9106542801719677,
+      "learning_rate": 7.348860562010574e-06,
+      "loss": 0.571,
+      "step": 3502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7867908684838822,
+      "learning_rate": 7.347374329123209e-06,
+      "loss": 0.6187,
+      "step": 3503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8933431553615583,
+      "learning_rate": 7.345887830147583e-06,
+      "loss": 0.5439,
+      "step": 3504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1452897179479513,
+      "learning_rate": 7.344401065252203e-06,
+      "loss": 0.6209,
+      "step": 3505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8240488714231755,
+      "learning_rate": 7.3429140346055975e-06,
+      "loss": 0.6676,
+      "step": 3506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1529547922669914,
+      "learning_rate": 7.341426738376332e-06,
+      "loss": 0.5676,
+      "step": 3507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0283041565520574,
+      "learning_rate": 7.339939176733e-06,
+      "loss": 0.7162,
+      "step": 3508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1571072349133633,
+      "learning_rate": 7.338451349844225e-06,
+      "loss": 0.8388,
+      "step": 3509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8719672067212767,
+      "learning_rate": 7.336963257878662e-06,
+      "loss": 0.7124,
+      "step": 3510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.877417040669118,
+      "learning_rate": 7.335474901004992e-06,
+      "loss": 0.7178,
+      "step": 3511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8934927176612808,
+      "learning_rate": 7.33398627939193e-06,
+      "loss": 0.5567,
+      "step": 3512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8796050148787842,
+      "learning_rate": 7.332497393208221e-06,
+      "loss": 0.6233,
+      "step": 3513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.83490148528102,
+      "learning_rate": 7.331008242622637e-06,
+      "loss": 0.6343,
+      "step": 3514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8143441207248123,
+      "learning_rate": 7.329518827803983e-06,
+      "loss": 0.6808,
+      "step": 3515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1087748223352754,
+      "learning_rate": 7.328029148921093e-06,
+      "loss": 0.5901,
+      "step": 3516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7491095171764681,
+      "learning_rate": 7.326539206142829e-06,
+      "loss": 0.566,
+      "step": 3517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8990983602988085,
+      "learning_rate": 7.325048999638083e-06,
+      "loss": 0.647,
+      "step": 3518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7817325075047465,
+      "learning_rate": 7.323558529575783e-06,
+      "loss": 0.6294,
+      "step": 3519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.822167484873732,
+      "learning_rate": 7.32206779612488e-06,
+      "loss": 0.599,
+      "step": 3520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.910179885511139,
+      "learning_rate": 7.320576799454355e-06,
+      "loss": 0.6137,
+      "step": 3521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9373630796040067,
+      "learning_rate": 7.319085539733225e-06,
+      "loss": 0.5921,
+      "step": 3522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8558805706601034,
+      "learning_rate": 7.317594017130529e-06,
+      "loss": 0.6571,
+      "step": 3523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7509386776872393,
+      "learning_rate": 7.316102231815343e-06,
+      "loss": 0.65,
+      "step": 3524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.87626594632814,
+      "learning_rate": 7.3146101839567665e-06,
+      "loss": 0.6148,
+      "step": 3525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9780604464642324,
+      "learning_rate": 7.313117873723933e-06,
+      "loss": 0.5868,
+      "step": 3526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7614477042809984,
+      "learning_rate": 7.311625301286005e-06,
+      "loss": 0.6357,
+      "step": 3527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8327860152596689,
+      "learning_rate": 7.310132466812172e-06,
+      "loss": 0.5737,
+      "step": 3528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8844138267432626,
+      "learning_rate": 7.308639370471658e-06,
+      "loss": 0.6322,
+      "step": 3529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8674359027203937,
+      "learning_rate": 7.30714601243371e-06,
+      "loss": 0.5701,
+      "step": 3530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8123042050551081,
+      "learning_rate": 7.3056523928676145e-06,
+      "loss": 0.6558,
+      "step": 3531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7883251931061648,
+      "learning_rate": 7.304158511942676e-06,
+      "loss": 0.6277,
+      "step": 3532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0045012224678205,
+      "learning_rate": 7.302664369828238e-06,
+      "loss": 0.7663,
+      "step": 3533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9524543531778042,
+      "learning_rate": 7.3011699666936685e-06,
+      "loss": 0.7106,
+      "step": 3534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9413034212258677,
+      "learning_rate": 7.299675302708368e-06,
+      "loss": 0.6501,
+      "step": 3535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6892597011874646,
+      "learning_rate": 7.298180378041763e-06,
+      "loss": 0.5715,
+      "step": 3536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9208949618622175,
+      "learning_rate": 7.296685192863313e-06,
+      "loss": 0.6828,
+      "step": 3537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1550436825626966,
+      "learning_rate": 7.295189747342507e-06,
+      "loss": 0.6368,
+      "step": 3538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6240302138516085,
+      "learning_rate": 7.29369404164886e-06,
+      "loss": 0.5729,
+      "step": 3539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9753475779628318,
+      "learning_rate": 7.2921980759519195e-06,
+      "loss": 0.6921,
+      "step": 3540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.871059765624354,
+      "learning_rate": 7.290701850421263e-06,
+      "loss": 0.7101,
+      "step": 3541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8488824199834615,
+      "learning_rate": 7.289205365226495e-06,
+      "loss": 0.7121,
+      "step": 3542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.979776721083964,
+      "learning_rate": 7.28770862053725e-06,
+      "loss": 0.6629,
+      "step": 3543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7690159529643779,
+      "learning_rate": 7.286211616523193e-06,
+      "loss": 0.5671,
+      "step": 3544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0002953479570516,
+      "learning_rate": 7.28471435335402e-06,
+      "loss": 0.5838,
+      "step": 3545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8849684513390108,
+      "learning_rate": 7.2832168311994514e-06,
+      "loss": 0.6947,
+      "step": 3546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8271816423627594,
+      "learning_rate": 7.281719050229241e-06,
+      "loss": 0.6798,
+      "step": 3547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9154738481542926,
+      "learning_rate": 7.280221010613171e-06,
+      "loss": 0.5558,
+      "step": 3548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7236896270573931,
+      "learning_rate": 7.278722712521054e-06,
+      "loss": 0.6129,
+      "step": 3549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.858416081191063,
+      "learning_rate": 7.277224156122728e-06,
+      "loss": 0.7925,
+      "step": 3550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9349250790205446,
+      "learning_rate": 7.275725341588064e-06,
+      "loss": 0.7287,
+      "step": 3551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0965449424230016,
+      "learning_rate": 7.2742262690869615e-06,
+      "loss": 0.7357,
+      "step": 3552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9572073961053111,
+      "learning_rate": 7.272726938789348e-06,
+      "loss": 0.6857,
+      "step": 3553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6355516242540316,
+      "learning_rate": 7.2712273508651834e-06,
+      "loss": 0.578,
+      "step": 3554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8240038253760955,
+      "learning_rate": 7.269727505484452e-06,
+      "loss": 0.7063,
+      "step": 3555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9522137436656415,
+      "learning_rate": 7.268227402817171e-06,
+      "loss": 0.6384,
+      "step": 3556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8803868241469999,
+      "learning_rate": 7.266727043033386e-06,
+      "loss": 0.6292,
+      "step": 3557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7416415007506427,
+      "learning_rate": 7.26522642630317e-06,
+      "loss": 0.5993,
+      "step": 3558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7285779931370737,
+      "learning_rate": 7.263725552796628e-06,
+      "loss": 0.6292,
+      "step": 3559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8865673304095096,
+      "learning_rate": 7.262224422683891e-06,
+      "loss": 0.6817,
+      "step": 3560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8541330421002968,
+      "learning_rate": 7.260723036135122e-06,
+      "loss": 0.5544,
+      "step": 3561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5749650639128956,
+      "learning_rate": 7.259221393320511e-06,
+      "loss": 0.637,
+      "step": 3562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.821273194460373,
+      "learning_rate": 7.257719494410278e-06,
+      "loss": 0.5897,
+      "step": 3563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8149563401389326,
+      "learning_rate": 7.2562173395746725e-06,
+      "loss": 0.6168,
+      "step": 3564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0211465063160863,
+      "learning_rate": 7.25471492898397e-06,
+      "loss": 0.6834,
+      "step": 3565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6758100886152418,
+      "learning_rate": 7.25321226280848e-06,
+      "loss": 0.5116,
+      "step": 3566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9332635703531122,
+      "learning_rate": 7.251709341218536e-06,
+      "loss": 0.7101,
+      "step": 3567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9021202447821917,
+      "learning_rate": 7.250206164384506e-06,
+      "loss": 0.7104,
+      "step": 3568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1677945215060594,
+      "learning_rate": 7.24870273247678e-06,
+      "loss": 0.7328,
+      "step": 3569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8317538366672943,
+      "learning_rate": 7.247199045665781e-06,
+      "loss": 0.6999,
+      "step": 3570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9648854126394588,
+      "learning_rate": 7.245695104121963e-06,
+      "loss": 0.6013,
+      "step": 3571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7893874702530328,
+      "learning_rate": 7.244190908015805e-06,
+      "loss": 0.6356,
+      "step": 3572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4006419796596727,
+      "learning_rate": 7.242686457517815e-06,
+      "loss": 0.7903,
+      "step": 3573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8682533212867887,
+      "learning_rate": 7.241181752798534e-06,
+      "loss": 0.7252,
+      "step": 3574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.968275479562997,
+      "learning_rate": 7.239676794028526e-06,
+      "loss": 0.667,
+      "step": 3575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5473586071098482,
+      "learning_rate": 7.238171581378388e-06,
+      "loss": 0.5816,
+      "step": 3576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.773785035800144,
+      "learning_rate": 7.236666115018744e-06,
+      "loss": 0.6212,
+      "step": 3577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8332500658652024,
+      "learning_rate": 7.235160395120247e-06,
+      "loss": 0.6962,
+      "step": 3578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5568737353280147,
+      "learning_rate": 7.2336544218535776e-06,
+      "loss": 0.559,
+      "step": 3579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8913861420333955,
+      "learning_rate": 7.23214819538945e-06,
+      "loss": 0.553,
+      "step": 3580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9256262438864884,
+      "learning_rate": 7.230641715898602e-06,
+      "loss": 0.7225,
+      "step": 3581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9428757885250234,
+      "learning_rate": 7.2291349835518e-06,
+      "loss": 0.6232,
+      "step": 3582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.847661038967175,
+      "learning_rate": 7.227627998519843e-06,
+      "loss": 0.6888,
+      "step": 3583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7968971064453072,
+      "learning_rate": 7.226120760973554e-06,
+      "loss": 0.6808,
+      "step": 3584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8316340872317762,
+      "learning_rate": 7.224613271083789e-06,
+      "loss": 0.6496,
+      "step": 3585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7751193674739747,
+      "learning_rate": 7.22310552902143e-06,
+      "loss": 0.6898,
+      "step": 3586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0056699940699625,
+      "learning_rate": 7.221597534957389e-06,
+      "loss": 0.6359,
+      "step": 3587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2114079963540183,
+      "learning_rate": 7.220089289062603e-06,
+      "loss": 0.7437,
+      "step": 3588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8708371960322898,
+      "learning_rate": 7.218580791508043e-06,
+      "loss": 0.6704,
+      "step": 3589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8610230872176512,
+      "learning_rate": 7.217072042464706e-06,
+      "loss": 0.5883,
+      "step": 3590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7328612845093743,
+      "learning_rate": 7.215563042103614e-06,
+      "loss": 0.6694,
+      "step": 3591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7591041086632069,
+      "learning_rate": 7.214053790595823e-06,
+      "loss": 0.7389,
+      "step": 3592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9763016947888363,
+      "learning_rate": 7.212544288112415e-06,
+      "loss": 0.7703,
+      "step": 3593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.739904384320081,
+      "learning_rate": 7.211034534824503e-06,
+      "loss": 0.6461,
+      "step": 3594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.725500951731821,
+      "learning_rate": 7.209524530903223e-06,
+      "loss": 0.6184,
+      "step": 3595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9718658533362245,
+      "learning_rate": 7.208014276519741e-06,
+      "loss": 0.6249,
+      "step": 3596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7782536173137125,
+      "learning_rate": 7.206503771845259e-06,
+      "loss": 0.7426,
+      "step": 3597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0669465358505135,
+      "learning_rate": 7.2049930170509965e-06,
+      "loss": 0.6984,
+      "step": 3598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9149032831444073,
+      "learning_rate": 7.2034820123082075e-06,
+      "loss": 0.5994,
+      "step": 3599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8824750319453147,
+      "learning_rate": 7.201970757788172e-06,
+      "loss": 0.5991,
+      "step": 3600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.897325064986783,
+      "learning_rate": 7.200459253662202e-06,
+      "loss": 0.6076,
+      "step": 3601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9589204398556992,
+      "learning_rate": 7.198947500101632e-06,
+      "loss": 0.6531,
+      "step": 3602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6587850379948488,
+      "learning_rate": 7.19743549727783e-06,
+      "loss": 0.6824,
+      "step": 3603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9014993866573917,
+      "learning_rate": 7.195923245362188e-06,
+      "loss": 0.6682,
+      "step": 3604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.872178215470094,
+      "learning_rate": 7.194410744526132e-06,
+      "loss": 0.5496,
+      "step": 3605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.971682951834879,
+      "learning_rate": 7.192897994941111e-06,
+      "loss": 0.6729,
+      "step": 3606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.097197443870445,
+      "learning_rate": 7.191384996778601e-06,
+      "loss": 0.7267,
+      "step": 3607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.104883778259818,
+      "learning_rate": 7.189871750210111e-06,
+      "loss": 0.5755,
+      "step": 3608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.038992765805888,
+      "learning_rate": 7.1883582554071776e-06,
+      "loss": 0.596,
+      "step": 3609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8255786200273927,
+      "learning_rate": 7.1868445125413625e-06,
+      "loss": 0.5536,
+      "step": 3610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.649347481136232,
+      "learning_rate": 7.1853305217842565e-06,
+      "loss": 0.6403,
+      "step": 3611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7709514913740092,
+      "learning_rate": 7.183816283307481e-06,
+      "loss": 0.6644,
+      "step": 3612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8054415691274552,
+      "learning_rate": 7.1823017972826815e-06,
+      "loss": 0.5759,
+      "step": 3613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.908794948479207,
+      "learning_rate": 7.180787063881534e-06,
+      "loss": 0.7358,
+      "step": 3614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8067734916298452,
+      "learning_rate": 7.179272083275744e-06,
+      "loss": 0.6164,
+      "step": 3615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1927195824585604,
+      "learning_rate": 7.177756855637042e-06,
+      "loss": 0.7013,
+      "step": 3616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.853731488800181,
+      "learning_rate": 7.1762413811371855e-06,
+      "loss": 0.6514,
+      "step": 3617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6905411710563716,
+      "learning_rate": 7.174725659947966e-06,
+      "loss": 0.5478,
+      "step": 3618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8778815442471106,
+      "learning_rate": 7.173209692241199e-06,
+      "loss": 0.6467,
+      "step": 3619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.066809489639149,
+      "learning_rate": 7.171693478188724e-06,
+      "loss": 0.7792,
+      "step": 3620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.182844071235007,
+      "learning_rate": 7.170177017962415e-06,
+      "loss": 0.6966,
+      "step": 3621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9173244135871332,
+      "learning_rate": 7.168660311734173e-06,
+      "loss": 0.6599,
+      "step": 3622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8071165425130356,
+      "learning_rate": 7.167143359675924e-06,
+      "loss": 0.5384,
+      "step": 3623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.818643370303639,
+      "learning_rate": 7.1656261619596205e-06,
+      "loss": 0.5815,
+      "step": 3624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9981940575117898,
+      "learning_rate": 7.1641087187572485e-06,
+      "loss": 0.6087,
+      "step": 3625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7155423758343529,
+      "learning_rate": 7.16259103024082e-06,
+      "loss": 0.5151,
+      "step": 3626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8508271654042858,
+      "learning_rate": 7.161073096582371e-06,
+      "loss": 0.6903,
+      "step": 3627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9064304886131862,
+      "learning_rate": 7.159554917953968e-06,
+      "loss": 0.6077,
+      "step": 3628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9992795993108614,
+      "learning_rate": 7.158036494527707e-06,
+      "loss": 0.8044,
+      "step": 3629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8289307895653182,
+      "learning_rate": 7.156517826475708e-06,
+      "loss": 0.6744,
+      "step": 3630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9640255147580412,
+      "learning_rate": 7.154998913970124e-06,
+      "loss": 0.6093,
+      "step": 3631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8704100255026406,
+      "learning_rate": 7.153479757183127e-06,
+      "loss": 0.5818,
+      "step": 3632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8530421196673053,
+      "learning_rate": 7.1519603562869265e-06,
+      "loss": 0.5494,
+      "step": 3633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.859849961294324,
+      "learning_rate": 7.150440711453754e-06,
+      "loss": 0.6875,
+      "step": 3634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9767165128139512,
+      "learning_rate": 7.148920822855869e-06,
+      "loss": 0.6612,
+      "step": 3635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0058716132606684,
+      "learning_rate": 7.1474006906655605e-06,
+      "loss": 0.7287,
+      "step": 3636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7254558229061596,
+      "learning_rate": 7.145880315055145e-06,
+      "loss": 0.6355,
+      "step": 3637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5572686981014636,
+      "learning_rate": 7.144359696196964e-06,
+      "loss": 0.6584,
+      "step": 3638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9052815448096254,
+      "learning_rate": 7.142838834263388e-06,
+      "loss": 0.6688,
+      "step": 3639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6860169875157103,
+      "learning_rate": 7.141317729426817e-06,
+      "loss": 0.6196,
+      "step": 3640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7721664305732985,
+      "learning_rate": 7.139796381859676e-06,
+      "loss": 0.6633,
+      "step": 3641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7826626559988565,
+      "learning_rate": 7.138274791734421e-06,
+      "loss": 0.6367,
+      "step": 3642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.990311955470244,
+      "learning_rate": 7.136752959223527e-06,
+      "loss": 0.6497,
+      "step": 3643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.074898158189017,
+      "learning_rate": 7.1352308844995086e-06,
+      "loss": 0.6088,
+      "step": 3644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9277929482740352,
+      "learning_rate": 7.133708567734898e-06,
+      "loss": 0.5938,
+      "step": 3645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7516961666712483,
+      "learning_rate": 7.13218600910226e-06,
+      "loss": 0.7127,
+      "step": 3646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.889977062266395,
+      "learning_rate": 7.1306632087741844e-06,
+      "loss": 0.6515,
+      "step": 3647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0117401818077862,
+      "learning_rate": 7.12914016692329e-06,
+      "loss": 0.708,
+      "step": 3648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8988679344144204,
+      "learning_rate": 7.1276168837222215e-06,
+      "loss": 0.707,
+      "step": 3649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.992469048910512,
+      "learning_rate": 7.1260933593436535e-06,
+      "loss": 0.6132,
+      "step": 3650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8079967436056992,
+      "learning_rate": 7.1245695939602834e-06,
+      "loss": 0.6561,
+      "step": 3651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6363093178305284,
+      "learning_rate": 7.12304558774484e-06,
+      "loss": 0.5939,
+      "step": 3652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7910381053908828,
+      "learning_rate": 7.121521340870079e-06,
+      "loss": 0.673,
+      "step": 3653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9706296825523928,
+      "learning_rate": 7.119996853508781e-06,
+      "loss": 0.6262,
+      "step": 3654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.869840494562361,
+      "learning_rate": 7.1184721258337575e-06,
+      "loss": 0.6058,
+      "step": 3655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.926379615057285,
+      "learning_rate": 7.116947158017842e-06,
+      "loss": 0.7076,
+      "step": 3656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7795167202195803,
+      "learning_rate": 7.115421950233902e-06,
+      "loss": 0.6051,
+      "step": 3657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9578043275804096,
+      "learning_rate": 7.113896502654824e-06,
+      "loss": 0.605,
+      "step": 3658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7369262214424697,
+      "learning_rate": 7.112370815453531e-06,
+      "loss": 0.5664,
+      "step": 3659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0109937094073906,
+      "learning_rate": 7.110844888802966e-06,
+      "loss": 0.6973,
+      "step": 3660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9212899018541152,
+      "learning_rate": 7.1093187228760995e-06,
+      "loss": 0.7831,
+      "step": 3661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9675443765140932,
+      "learning_rate": 7.107792317845934e-06,
+      "loss": 0.7564,
+      "step": 3662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5186964557046534,
+      "learning_rate": 7.106265673885494e-06,
+      "loss": 0.8091,
+      "step": 3663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.172007104728858,
+      "learning_rate": 7.104738791167837e-06,
+      "loss": 0.6913,
+      "step": 3664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.767126716740684,
+      "learning_rate": 7.103211669866039e-06,
+      "loss": 0.6973,
+      "step": 3665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7754637022542625,
+      "learning_rate": 7.1016843101532115e-06,
+      "loss": 0.6218,
+      "step": 3666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7831012329808487,
+      "learning_rate": 7.100156712202488e-06,
+      "loss": 0.685,
+      "step": 3667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8985315383249024,
+      "learning_rate": 7.098628876187031e-06,
+      "loss": 0.6087,
+      "step": 3668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9257692672338116,
+      "learning_rate": 7.0971008022800295e-06,
+      "loss": 0.6588,
+      "step": 3669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.048096408415564,
+      "learning_rate": 7.095572490654698e-06,
+      "loss": 0.7406,
+      "step": 3670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.80679778795583,
+      "learning_rate": 7.094043941484282e-06,
+      "loss": 0.6449,
+      "step": 3671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.924580745209938,
+      "learning_rate": 7.092515154942048e-06,
+      "loss": 0.7301,
+      "step": 3672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7817683854261057,
+      "learning_rate": 7.090986131201294e-06,
+      "loss": 0.6471,
+      "step": 3673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7928879478359547,
+      "learning_rate": 7.089456870435344e-06,
+      "loss": 0.6514,
+      "step": 3674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0696609601855616,
+      "learning_rate": 7.087927372817549e-06,
+      "loss": 0.6171,
+      "step": 3675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6884810491961613,
+      "learning_rate": 7.086397638521285e-06,
+      "loss": 0.6483,
+      "step": 3676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7331759057428129,
+      "learning_rate": 7.084867667719957e-06,
+      "loss": 0.5589,
+      "step": 3677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9600799820017218,
+      "learning_rate": 7.083337460586995e-06,
+      "loss": 0.7712,
+      "step": 3678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9698872325167316,
+      "learning_rate": 7.0818070172958585e-06,
+      "loss": 0.6482,
+      "step": 3679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.772990633413805,
+      "learning_rate": 7.080276338020029e-06,
+      "loss": 0.6664,
+      "step": 3680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9925905028374946,
+      "learning_rate": 7.07874542293302e-06,
+      "loss": 0.7052,
+      "step": 3681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9775724860452595,
+      "learning_rate": 7.077214272208369e-06,
+      "loss": 0.6453,
+      "step": 3682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.101909580042845,
+      "learning_rate": 7.07568288601964e-06,
+      "loss": 0.7348,
+      "step": 3683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7865633356521249,
+      "learning_rate": 7.074151264540425e-06,
+      "loss": 0.5937,
+      "step": 3684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.912028189475162,
+      "learning_rate": 7.072619407944343e-06,
+      "loss": 0.6418,
+      "step": 3685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7555446608686078,
+      "learning_rate": 7.071087316405037e-06,
+      "loss": 0.7161,
+      "step": 3686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7423318666131733,
+      "learning_rate": 7.069554990096178e-06,
+      "loss": 0.6073,
+      "step": 3687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.727743961687725,
+      "learning_rate": 7.068022429191465e-06,
+      "loss": 0.6346,
+      "step": 3688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.020998879626979,
+      "learning_rate": 7.066489633864624e-06,
+      "loss": 0.6432,
+      "step": 3689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7904324410819448,
+      "learning_rate": 7.064956604289402e-06,
+      "loss": 0.5901,
+      "step": 3690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6681103595815747,
+      "learning_rate": 7.0634233406395806e-06,
+      "loss": 0.6236,
+      "step": 3691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.019485959225672,
+      "learning_rate": 7.061889843088961e-06,
+      "loss": 0.6031,
+      "step": 3692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6974985157799625,
+      "learning_rate": 7.060356111811376e-06,
+      "loss": 0.5825,
+      "step": 3693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.007966294692052,
+      "learning_rate": 7.058822146980684e-06,
+      "loss": 0.6565,
+      "step": 3694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8677225337817733,
+      "learning_rate": 7.0572879487707645e-06,
+      "loss": 0.6306,
+      "step": 3695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9268857672131279,
+      "learning_rate": 7.05575351735553e-06,
+      "loss": 0.6039,
+      "step": 3696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8004826415621484,
+      "learning_rate": 7.054218852908918e-06,
+      "loss": 0.7229,
+      "step": 3697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.86814066122037,
+      "learning_rate": 7.05268395560489e-06,
+      "loss": 0.6712,
+      "step": 3698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8861582510455044,
+      "learning_rate": 7.051148825617435e-06,
+      "loss": 0.6199,
+      "step": 3699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0271375461938907,
+      "learning_rate": 7.0496134631205705e-06,
+      "loss": 0.6532,
+      "step": 3700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7322689461414547,
+      "learning_rate": 7.048077868288338e-06,
+      "loss": 0.6175,
+      "step": 3701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7427810559006416,
+      "learning_rate": 7.046542041294804e-06,
+      "loss": 0.624,
+      "step": 3702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.814216437196377,
+      "learning_rate": 7.045005982314065e-06,
+      "loss": 0.5836,
+      "step": 3703
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.008142687023798,
+      "learning_rate": 7.0434696915202415e-06,
+      "loss": 0.7704,
+      "step": 3704
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7855230650520368,
+      "learning_rate": 7.041933169087482e-06,
+      "loss": 0.6452,
+      "step": 3705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.985668972035989,
+      "learning_rate": 7.040396415189959e-06,
+      "loss": 0.7503,
+      "step": 3706
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8844199217844504,
+      "learning_rate": 7.038859430001872e-06,
+      "loss": 0.6682,
+      "step": 3707
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6502541814904381,
+      "learning_rate": 7.037322213697448e-06,
+      "loss": 0.5502,
+      "step": 3708
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.323083971904915,
+      "learning_rate": 7.035784766450938e-06,
+      "loss": 0.5438,
+      "step": 3709
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9612136420018857,
+      "learning_rate": 7.034247088436621e-06,
+      "loss": 0.7139,
+      "step": 3710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8172381656336853,
+      "learning_rate": 7.032709179828803e-06,
+      "loss": 0.6152,
+      "step": 3711
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8903375116755927,
+      "learning_rate": 7.031171040801813e-06,
+      "loss": 0.6615,
+      "step": 3712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0513450644693183,
+      "learning_rate": 7.029632671530008e-06,
+      "loss": 0.7287,
+      "step": 3713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.364069746083148,
+      "learning_rate": 7.02809407218777e-06,
+      "loss": 0.7201,
+      "step": 3714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.789762597807389,
+      "learning_rate": 7.026555242949511e-06,
+      "loss": 0.6513,
+      "step": 3715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7415445101122051,
+      "learning_rate": 7.0250161839896636e-06,
+      "loss": 0.5658,
+      "step": 3716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9838077381226866,
+      "learning_rate": 7.02347689548269e-06,
+      "loss": 0.643,
+      "step": 3717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0244134936944933,
+      "learning_rate": 7.021937377603076e-06,
+      "loss": 0.7017,
+      "step": 3718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9745168936105955,
+      "learning_rate": 7.020397630525336e-06,
+      "loss": 0.6745,
+      "step": 3719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7289255604430116,
+      "learning_rate": 7.018857654424008e-06,
+      "loss": 0.6488,
+      "step": 3720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9155400352019034,
+      "learning_rate": 7.017317449473658e-06,
+      "loss": 0.6324,
+      "step": 3721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5674009208190292,
+      "learning_rate": 7.015777015848877e-06,
+      "loss": 0.6442,
+      "step": 3722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.876596679522037,
+      "learning_rate": 7.0142363537242815e-06,
+      "loss": 0.6725,
+      "step": 3723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8965716533361012,
+      "learning_rate": 7.012695463274515e-06,
+      "loss": 0.6449,
+      "step": 3724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8872516694937251,
+      "learning_rate": 7.0111543446742444e-06,
+      "loss": 0.6431,
+      "step": 3725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1515953153188776,
+      "learning_rate": 7.0096129980981674e-06,
+      "loss": 0.6978,
+      "step": 3726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0270766691639523,
+      "learning_rate": 7.008071423721004e-06,
+      "loss": 0.6044,
+      "step": 3727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0184644859313225,
+      "learning_rate": 7.006529621717496e-06,
+      "loss": 0.696,
+      "step": 3728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0626010069202447,
+      "learning_rate": 7.00498759226242e-06,
+      "loss": 0.6551,
+      "step": 3729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.016147799155721,
+      "learning_rate": 7.003445335530572e-06,
+      "loss": 0.641,
+      "step": 3730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8731480687364606,
+      "learning_rate": 7.001902851696775e-06,
+      "loss": 0.5508,
+      "step": 3731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7525210729280494,
+      "learning_rate": 7.000360140935881e-06,
+      "loss": 0.5715,
+      "step": 3732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7791094074416964,
+      "learning_rate": 6.998817203422763e-06,
+      "loss": 0.6188,
+      "step": 3733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7107178217967776,
+      "learning_rate": 6.997274039332323e-06,
+      "loss": 0.6057,
+      "step": 3734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1689931360984835,
+      "learning_rate": 6.995730648839485e-06,
+      "loss": 0.6081,
+      "step": 3735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9944082514802253,
+      "learning_rate": 6.9941870321192015e-06,
+      "loss": 0.6827,
+      "step": 3736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9902720881288807,
+      "learning_rate": 6.992643189346453e-06,
+      "loss": 0.6835,
+      "step": 3737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9676878976709407,
+      "learning_rate": 6.991099120696243e-06,
+      "loss": 0.6268,
+      "step": 3738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9265199475254238,
+      "learning_rate": 6.989554826343597e-06,
+      "loss": 0.6599,
+      "step": 3739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6498327098254237,
+      "learning_rate": 6.988010306463571e-06,
+      "loss": 0.5858,
+      "step": 3740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8263087431164569,
+      "learning_rate": 6.986465561231246e-06,
+      "loss": 0.5895,
+      "step": 3741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9555452689076078,
+      "learning_rate": 6.984920590821726e-06,
+      "loss": 0.6573,
+      "step": 3742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0437770842120933,
+      "learning_rate": 6.983375395410146e-06,
+      "loss": 0.6526,
+      "step": 3743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.869433885945104,
+      "learning_rate": 6.981829975171658e-06,
+      "loss": 0.6379,
+      "step": 3744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8994878040677412,
+      "learning_rate": 6.9802843302814475e-06,
+      "loss": 0.673,
+      "step": 3745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8642847780082168,
+      "learning_rate": 6.97873846091472e-06,
+      "loss": 0.6918,
+      "step": 3746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.940150418676658,
+      "learning_rate": 6.977192367246709e-06,
+      "loss": 0.6778,
+      "step": 3747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.952659328164777,
+      "learning_rate": 6.975646049452673e-06,
+      "loss": 0.6567,
+      "step": 3748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.897858212581815,
+      "learning_rate": 6.9740995077079e-06,
+      "loss": 0.6589,
+      "step": 3749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9798414395768777,
+      "learning_rate": 6.972552742187693e-06,
+      "loss": 0.6554,
+      "step": 3750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.947474102381345,
+      "learning_rate": 6.971005753067391e-06,
+      "loss": 0.6939,
+      "step": 3751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8593020322548108,
+      "learning_rate": 6.96945854052235e-06,
+      "loss": 0.6908,
+      "step": 3752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8790691411742364,
+      "learning_rate": 6.96791110472796e-06,
+      "loss": 0.6918,
+      "step": 3753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.009560441823809,
+      "learning_rate": 6.966363445859629e-06,
+      "loss": 0.701,
+      "step": 3754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7247879670505806,
+      "learning_rate": 6.964815564092792e-06,
+      "loss": 0.5531,
+      "step": 3755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9237024626907298,
+      "learning_rate": 6.9632674596029135e-06,
+      "loss": 0.7353,
+      "step": 3756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8423072923305135,
+      "learning_rate": 6.9617191325654785e-06,
+      "loss": 0.6292,
+      "step": 3757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.095217792173662,
+      "learning_rate": 6.9601705831559985e-06,
+      "loss": 0.7027,
+      "step": 3758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9172392464043688,
+      "learning_rate": 6.95862181155001e-06,
+      "loss": 0.5968,
+      "step": 3759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7885688923040335,
+      "learning_rate": 6.957072817923074e-06,
+      "loss": 0.6571,
+      "step": 3760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0189100338935813,
+      "learning_rate": 6.95552360245078e-06,
+      "loss": 0.6102,
+      "step": 3761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.176085342804834,
+      "learning_rate": 6.95397416530874e-06,
+      "loss": 0.7103,
+      "step": 3762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8281330590058102,
+      "learning_rate": 6.95242450667259e-06,
+      "loss": 0.5678,
+      "step": 3763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.780282107248268,
+      "learning_rate": 6.950874626717996e-06,
+      "loss": 0.5708,
+      "step": 3764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.696269074894569,
+      "learning_rate": 6.949324525620642e-06,
+      "loss": 0.627,
+      "step": 3765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.028234243684498,
+      "learning_rate": 6.947774203556241e-06,
+      "loss": 0.6769,
+      "step": 3766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8100157337051184,
+      "learning_rate": 6.946223660700535e-06,
+      "loss": 0.6096,
+      "step": 3767
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0304897261390935,
+      "learning_rate": 6.944672897229282e-06,
+      "loss": 0.7408,
+      "step": 3768
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8186687992645447,
+      "learning_rate": 6.943121913318272e-06,
+      "loss": 0.6727,
+      "step": 3769
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.014696798801752,
+      "learning_rate": 6.941570709143317e-06,
+      "loss": 0.6679,
+      "step": 3770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9573005847526657,
+      "learning_rate": 6.9400192848802575e-06,
+      "loss": 0.7203,
+      "step": 3771
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.870931516660422,
+      "learning_rate": 6.938467640704953e-06,
+      "loss": 0.6407,
+      "step": 3772
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.924348926934089,
+      "learning_rate": 6.936915776793293e-06,
+      "loss": 0.7739,
+      "step": 3773
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8219644545452585,
+      "learning_rate": 6.935363693321189e-06,
+      "loss": 0.7424,
+      "step": 3774
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8909820062964133,
+      "learning_rate": 6.93381139046458e-06,
+      "loss": 0.5874,
+      "step": 3775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.851095708702634,
+      "learning_rate": 6.932258868399426e-06,
+      "loss": 0.5811,
+      "step": 3776
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9626956710437433,
+      "learning_rate": 6.930706127301718e-06,
+      "loss": 0.6269,
+      "step": 3777
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8588515617089818,
+      "learning_rate": 6.9291531673474645e-06,
+      "loss": 0.6689,
+      "step": 3778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0129296892502437,
+      "learning_rate": 6.9275999887127045e-06,
+      "loss": 0.692,
+      "step": 3779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.967573552691157,
+      "learning_rate": 6.926046591573498e-06,
+      "loss": 0.6498,
+      "step": 3780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9177145711520847,
+      "learning_rate": 6.924492976105932e-06,
+      "loss": 0.6621,
+      "step": 3781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8662266389846618,
+      "learning_rate": 6.922939142486118e-06,
+      "loss": 0.7561,
+      "step": 3782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0519301872983853,
+      "learning_rate": 6.921385090890193e-06,
+      "loss": 0.613,
+      "step": 3783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9506463097256144,
+      "learning_rate": 6.919830821494314e-06,
+      "loss": 0.771,
+      "step": 3784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.035175531222123,
+      "learning_rate": 6.918276334474671e-06,
+      "loss": 0.7823,
+      "step": 3785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7749634341037155,
+      "learning_rate": 6.916721630007471e-06,
+      "loss": 0.6171,
+      "step": 3786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.135332931852134,
+      "learning_rate": 6.91516670826895e-06,
+      "loss": 0.5635,
+      "step": 3787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7881778397885084,
+      "learning_rate": 6.913611569435366e-06,
+      "loss": 0.5461,
+      "step": 3788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7450638856633924,
+      "learning_rate": 6.912056213683001e-06,
+      "loss": 0.6384,
+      "step": 3789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9663931312471843,
+      "learning_rate": 6.9105006411881695e-06,
+      "loss": 0.6423,
+      "step": 3790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.012129630221285,
+      "learning_rate": 6.9089448521271995e-06,
+      "loss": 0.6986,
+      "step": 3791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.933878118917115,
+      "learning_rate": 6.9073888466764495e-06,
+      "loss": 0.6312,
+      "step": 3792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8464628698639192,
+      "learning_rate": 6.905832625012301e-06,
+      "loss": 0.6321,
+      "step": 3793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.062510373103962,
+      "learning_rate": 6.904276187311163e-06,
+      "loss": 0.6971,
+      "step": 3794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.93374853425991,
+      "learning_rate": 6.9027195337494645e-06,
+      "loss": 0.5843,
+      "step": 3795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1647078071719457,
+      "learning_rate": 6.901162664503662e-06,
+      "loss": 0.5992,
+      "step": 3796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9599458350519412,
+      "learning_rate": 6.899605579750236e-06,
+      "loss": 0.6586,
+      "step": 3797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2163360975899313,
+      "learning_rate": 6.898048279665689e-06,
+      "loss": 0.6925,
+      "step": 3798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7824221315080275,
+      "learning_rate": 6.896490764426551e-06,
+      "loss": 0.6436,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9842651710271497,
+      "learning_rate": 6.8949330342093756e-06,
+      "loss": 0.7,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8505223212020494,
+      "learning_rate": 6.893375089190741e-06,
+      "loss": 0.6553,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8419752520577877,
+      "learning_rate": 6.891816929547247e-06,
+      "loss": 0.5906,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6868313861269602,
+      "learning_rate": 6.890258555455521e-06,
+      "loss": 0.6291,
+      "step": 3803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.906131986603101,
+      "learning_rate": 6.888699967092215e-06,
+      "loss": 0.6008,
+      "step": 3804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.756561257424716,
+      "learning_rate": 6.887141164634001e-06,
+      "loss": 0.5851,
+      "step": 3805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8379054874503786,
+      "learning_rate": 6.885582148257579e-06,
+      "loss": 0.6615,
+      "step": 3806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8589992016721577,
+      "learning_rate": 6.884022918139675e-06,
+      "loss": 0.5656,
+      "step": 3807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7750855346733305,
+      "learning_rate": 6.882463474457034e-06,
+      "loss": 0.684,
+      "step": 3808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7331464972792372,
+      "learning_rate": 6.8809038173864285e-06,
+      "loss": 0.5875,
+      "step": 3809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8145592971627584,
+      "learning_rate": 6.879343947104653e-06,
+      "loss": 0.6518,
+      "step": 3810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7186609313479941,
+      "learning_rate": 6.877783863788531e-06,
+      "loss": 0.6398,
+      "step": 3811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0593901439717217,
+      "learning_rate": 6.876223567614904e-06,
+      "loss": 0.6199,
+      "step": 3812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.051921874821278,
+      "learning_rate": 6.874663058760642e-06,
+      "loss": 0.6652,
+      "step": 3813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.983268125178121,
+      "learning_rate": 6.873102337402637e-06,
+      "loss": 0.6742,
+      "step": 3814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9250957085532272,
+      "learning_rate": 6.871541403717808e-06,
+      "loss": 0.5963,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8923565653388212,
+      "learning_rate": 6.86998025788309e-06,
+      "loss": 0.6559,
+      "step": 3816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7717289543577657,
+      "learning_rate": 6.868418900075452e-06,
+      "loss": 0.6317,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8393903443302437,
+      "learning_rate": 6.866857330471882e-06,
+      "loss": 0.6645,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.907056188925765,
+      "learning_rate": 6.8652955492493944e-06,
+      "loss": 0.6674,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7792952155871635,
+      "learning_rate": 6.863733556585023e-06,
+      "loss": 0.6211,
+      "step": 3820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0070176175353485,
+      "learning_rate": 6.862171352655831e-06,
+      "loss": 0.5849,
+      "step": 3821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.927108955804964,
+      "learning_rate": 6.8606089376389006e-06,
+      "loss": 0.6479,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8876654284510865,
+      "learning_rate": 6.859046311711344e-06,
+      "loss": 0.6571,
+      "step": 3823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9645805494724002,
+      "learning_rate": 6.85748347505029e-06,
+      "loss": 0.6443,
+      "step": 3824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.033078807453262,
+      "learning_rate": 6.855920427832898e-06,
+      "loss": 0.7693,
+      "step": 3825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.741023709894303,
+      "learning_rate": 6.854357170236346e-06,
+      "loss": 0.5633,
+      "step": 3826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8327831597190622,
+      "learning_rate": 6.85279370243784e-06,
+      "loss": 0.6629,
+      "step": 3827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.138936128449846,
+      "learning_rate": 6.851230024614608e-06,
+      "loss": 0.6216,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9998323937030877,
+      "learning_rate": 6.8496661369439e-06,
+      "loss": 0.6371,
+      "step": 3829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.956674866863184,
+      "learning_rate": 6.848102039602993e-06,
+      "loss": 0.6175,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1315466856909713,
+      "learning_rate": 6.846537732769185e-06,
+      "loss": 0.7057,
+      "step": 3831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8023054657266382,
+      "learning_rate": 6.844973216619801e-06,
+      "loss": 0.7285,
+      "step": 3832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.331237965000702,
+      "learning_rate": 6.843408491332186e-06,
+      "loss": 0.6631,
+      "step": 3833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8747243292576776,
+      "learning_rate": 6.841843557083714e-06,
+      "loss": 0.623,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.081104098621447,
+      "learning_rate": 6.840278414051774e-06,
+      "loss": 0.6736,
+      "step": 3835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.027824141424956,
+      "learning_rate": 6.838713062413788e-06,
+      "loss": 0.7356,
+      "step": 3836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8448960739370412,
+      "learning_rate": 6.8371475023471945e-06,
+      "loss": 0.689,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9094274178849335,
+      "learning_rate": 6.835581734029462e-06,
+      "loss": 0.64,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8412279763358383,
+      "learning_rate": 6.834015757638076e-06,
+      "loss": 0.6305,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8963150898529104,
+      "learning_rate": 6.8324495733505515e-06,
+      "loss": 0.5916,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.006814141934046,
+      "learning_rate": 6.830883181344423e-06,
+      "loss": 0.6003,
+      "step": 3841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.761995258515656,
+      "learning_rate": 6.829316581797249e-06,
+      "loss": 0.5449,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8751428173287201,
+      "learning_rate": 6.827749774886616e-06,
+      "loss": 0.6368,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8238649095624873,
+      "learning_rate": 6.826182760790127e-06,
+      "loss": 0.5485,
+      "step": 3844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8050371465050767,
+      "learning_rate": 6.824615539685413e-06,
+      "loss": 0.6776,
+      "step": 3845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8185148092018413,
+      "learning_rate": 6.823048111750128e-06,
+      "loss": 0.5611,
+      "step": 3846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9137674036077919,
+      "learning_rate": 6.821480477161948e-06,
+      "loss": 0.7279,
+      "step": 3847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.278371530895591,
+      "learning_rate": 6.819912636098574e-06,
+      "loss": 0.6727,
+      "step": 3848
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9336374035262587,
+      "learning_rate": 6.81834458873773e-06,
+      "loss": 0.5828,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.608424844887298,
+      "learning_rate": 6.816776335257162e-06,
+      "loss": 0.6486,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9733951460532413,
+      "learning_rate": 6.815207875834641e-06,
+      "loss": 0.7597,
+      "step": 3851
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8337771137234302,
+      "learning_rate": 6.8136392106479624e-06,
+      "loss": 0.6421,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.88536398864315,
+      "learning_rate": 6.81207033987494e-06,
+      "loss": 0.5933,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8852332420912083,
+      "learning_rate": 6.810501263693416e-06,
+      "loss": 0.6265,
+      "step": 3854
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6270870662218326,
+      "learning_rate": 6.808931982281255e-06,
+      "loss": 0.5371,
+      "step": 3855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9198351870779027,
+      "learning_rate": 6.807362495816344e-06,
+      "loss": 0.6244,
+      "step": 3856
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9626894008745388,
+      "learning_rate": 6.805792804476592e-06,
+      "loss": 0.6162,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.847790506596514,
+      "learning_rate": 6.8042229084399325e-06,
+      "loss": 0.656,
+      "step": 3858
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9491524882255473,
+      "learning_rate": 6.802652807884322e-06,
+      "loss": 0.6687,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8982335002271724,
+      "learning_rate": 6.801082502987742e-06,
+      "loss": 0.6663,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0630552583208925,
+      "learning_rate": 6.799511993928195e-06,
+      "loss": 0.5751,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9407115444354701,
+      "learning_rate": 6.797941280883706e-06,
+      "loss": 0.6156,
+      "step": 3862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.855317202450553,
+      "learning_rate": 6.796370364032324e-06,
+      "loss": 0.6032,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8158488358917717,
+      "learning_rate": 6.794799243552123e-06,
+      "loss": 0.5607,
+      "step": 3864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7489231484293017,
+      "learning_rate": 6.793227919621197e-06,
+      "loss": 0.7611,
+      "step": 3865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0671408277409413,
+      "learning_rate": 6.791656392417666e-06,
+      "loss": 0.7821,
+      "step": 3866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6697050142330605,
+      "learning_rate": 6.790084662119671e-06,
+      "loss": 0.6472,
+      "step": 3867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9443126027690214,
+      "learning_rate": 6.7885127289053765e-06,
+      "loss": 0.6885,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7689567655002967,
+      "learning_rate": 6.78694059295297e-06,
+      "loss": 0.6111,
+      "step": 3869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9659917977999166,
+      "learning_rate": 6.785368254440661e-06,
+      "loss": 0.5973,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9024907067927386,
+      "learning_rate": 6.783795713546686e-06,
+      "loss": 0.6616,
+      "step": 3871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8879653406046633,
+      "learning_rate": 6.782222970449298e-06,
+      "loss": 0.6182,
+      "step": 3872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8947196759807639,
+      "learning_rate": 6.780650025326778e-06,
+      "loss": 0.6899,
+      "step": 3873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8450879931056883,
+      "learning_rate": 6.779076878357429e-06,
+      "loss": 0.5877,
+      "step": 3874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8933404415566144,
+      "learning_rate": 6.777503529719576e-06,
+      "loss": 0.7101,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7558373826056832,
+      "learning_rate": 6.775929979591565e-06,
+      "loss": 0.6396,
+      "step": 3876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0003562552284433,
+      "learning_rate": 6.774356228151768e-06,
+      "loss": 0.6069,
+      "step": 3877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9255259994750547,
+      "learning_rate": 6.772782275578582e-06,
+      "loss": 0.5854,
+      "step": 3878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8642977093526387,
+      "learning_rate": 6.771208122050418e-06,
+      "loss": 0.6715,
+      "step": 3879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8022100159125873,
+      "learning_rate": 6.769633767745718e-06,
+      "loss": 0.6321,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8278453367976983,
+      "learning_rate": 6.768059212842944e-06,
+      "loss": 0.592,
+      "step": 3881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8746751891331823,
+      "learning_rate": 6.7664844575205816e-06,
+      "loss": 0.6962,
+      "step": 3882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.999998865342274,
+      "learning_rate": 6.764909501957136e-06,
+      "loss": 0.665,
+      "step": 3883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8866066346814458,
+      "learning_rate": 6.76333434633114e-06,
+      "loss": 0.7158,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.854320769553393,
+      "learning_rate": 6.761758990821143e-06,
+      "loss": 0.6665,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0265802747703274,
+      "learning_rate": 6.760183435605725e-06,
+      "loss": 0.6098,
+      "step": 3886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8563571894286828,
+      "learning_rate": 6.758607680863481e-06,
+      "loss": 0.6027,
+      "step": 3887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8388299982466285,
+      "learning_rate": 6.757031726773033e-06,
+      "loss": 0.6205,
+      "step": 3888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9306457408739781,
+      "learning_rate": 6.755455573513025e-06,
+      "loss": 0.6471,
+      "step": 3889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8282397443953233,
+      "learning_rate": 6.75387922126212e-06,
+      "loss": 0.6413,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8013519821627202,
+      "learning_rate": 6.752302670199009e-06,
+      "loss": 0.5619,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.279002945506568,
+      "learning_rate": 6.750725920502402e-06,
+      "loss": 0.7127,
+      "step": 3892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0640115448453646,
+      "learning_rate": 6.749148972351034e-06,
+      "loss": 0.6407,
+      "step": 3893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0687523081357693,
+      "learning_rate": 6.74757182592366e-06,
+      "loss": 0.6546,
+      "step": 3894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.116509052754373,
+      "learning_rate": 6.7459944813990585e-06,
+      "loss": 0.737,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0452835675623184,
+      "learning_rate": 6.744416938956031e-06,
+      "loss": 0.7485,
+      "step": 3896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8635310962131926,
+      "learning_rate": 6.7428391987734e-06,
+      "loss": 0.639,
+      "step": 3897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.00514277899685,
+      "learning_rate": 6.741261261030013e-06,
+      "loss": 0.7099,
+      "step": 3898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7909492085460657,
+      "learning_rate": 6.739683125904737e-06,
+      "loss": 0.7444,
+      "step": 3899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9311294006580535,
+      "learning_rate": 6.7381047935764625e-06,
+      "loss": 0.6704,
+      "step": 3900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.066959770065368,
+      "learning_rate": 6.736526264224101e-06,
+      "loss": 0.6768,
+      "step": 3901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7730917255734966,
+      "learning_rate": 6.7349475380265926e-06,
+      "loss": 0.6414,
+      "step": 3902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7538215933569536,
+      "learning_rate": 6.7333686151628895e-06,
+      "loss": 0.7574,
+      "step": 3903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8843523002908729,
+      "learning_rate": 6.731789495811975e-06,
+      "loss": 0.6446,
+      "step": 3904
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9711212147816433,
+      "learning_rate": 6.730210180152852e-06,
+      "loss": 0.6436,
+      "step": 3905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7667975906547377,
+      "learning_rate": 6.728630668364541e-06,
+      "loss": 0.5743,
+      "step": 3906
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.747514957971742,
+      "learning_rate": 6.7270509606260915e-06,
+      "loss": 0.5884,
+      "step": 3907
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8617280757448222,
+      "learning_rate": 6.725471057116573e-06,
+      "loss": 0.7538,
+      "step": 3908
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.103545907282549,
+      "learning_rate": 6.7238909580150735e-06,
+      "loss": 0.6881,
+      "step": 3909
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8513840301905578,
+      "learning_rate": 6.7223106635007085e-06,
+      "loss": 0.671,
+      "step": 3910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.049540873747753,
+      "learning_rate": 6.720730173752613e-06,
+      "loss": 0.5919,
+      "step": 3911
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8190429314370908,
+      "learning_rate": 6.719149488949945e-06,
+      "loss": 0.5832,
+      "step": 3912
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8679025089120667,
+      "learning_rate": 6.717568609271883e-06,
+      "loss": 0.7133,
+      "step": 3913
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0178764743414006,
+      "learning_rate": 6.715987534897629e-06,
+      "loss": 0.7242,
+      "step": 3914
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9092744073317511,
+      "learning_rate": 6.714406266006408e-06,
+      "loss": 0.6669,
+      "step": 3915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7023005026783422,
+      "learning_rate": 6.712824802777465e-06,
+      "loss": 0.6614,
+      "step": 3916
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.077414840829166,
+      "learning_rate": 6.711243145390066e-06,
+      "loss": 0.7272,
+      "step": 3917
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5880707527072129,
+      "learning_rate": 6.709661294023504e-06,
+      "loss": 0.5592,
+      "step": 3918
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0116835193591376,
+      "learning_rate": 6.708079248857091e-06,
+      "loss": 0.7295,
+      "step": 3919
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9509358664824894,
+      "learning_rate": 6.706497010070157e-06,
+      "loss": 0.6443,
+      "step": 3920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9452163843629293,
+      "learning_rate": 6.704914577842062e-06,
+      "loss": 0.6771,
+      "step": 3921
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7279799757953567,
+      "learning_rate": 6.703331952352181e-06,
+      "loss": 0.6676,
+      "step": 3922
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9451199834666655,
+      "learning_rate": 6.701749133779916e-06,
+      "loss": 0.5865,
+      "step": 3923
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.813795054616006,
+      "learning_rate": 6.700166122304686e-06,
+      "loss": 0.6689,
+      "step": 3924
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8965308229743518,
+      "learning_rate": 6.698582918105934e-06,
+      "loss": 0.6256,
+      "step": 3925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7884201565417062,
+      "learning_rate": 6.696999521363128e-06,
+      "loss": 0.6363,
+      "step": 3926
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9635557773440546,
+      "learning_rate": 6.695415932255753e-06,
+      "loss": 0.653,
+      "step": 3927
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0890326348091453,
+      "learning_rate": 6.69383215096332e-06,
+      "loss": 0.7069,
+      "step": 3928
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6538146640381228,
+      "learning_rate": 6.692248177665357e-06,
+      "loss": 0.5993,
+      "step": 3929
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9705437684955363,
+      "learning_rate": 6.690664012541418e-06,
+      "loss": 0.7516,
+      "step": 3930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7444055802195977,
+      "learning_rate": 6.689079655771076e-06,
+      "loss": 0.6767,
+      "step": 3931
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8819144065805602,
+      "learning_rate": 6.687495107533928e-06,
+      "loss": 0.7314,
+      "step": 3932
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8479615838505195,
+      "learning_rate": 6.685910368009592e-06,
+      "loss": 0.6272,
+      "step": 3933
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0909081795401017,
+      "learning_rate": 6.684325437377704e-06,
+      "loss": 0.681,
+      "step": 3934
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9637392727671243,
+      "learning_rate": 6.682740315817929e-06,
+      "loss": 0.6803,
+      "step": 3935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7789446449482587,
+      "learning_rate": 6.681155003509949e-06,
+      "loss": 0.6188,
+      "step": 3936
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8522898997130828,
+      "learning_rate": 6.679569500633466e-06,
+      "loss": 0.6477,
+      "step": 3937
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5571151829309526,
+      "learning_rate": 6.6779838073682066e-06,
+      "loss": 0.5387,
+      "step": 3938
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7670313948733032,
+      "learning_rate": 6.676397923893918e-06,
+      "loss": 0.6602,
+      "step": 3939
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.027370857469276,
+      "learning_rate": 6.67481185039037e-06,
+      "loss": 0.6638,
+      "step": 3940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9916936077066447,
+      "learning_rate": 6.673225587037354e-06,
+      "loss": 0.6213,
+      "step": 3941
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9133600244144617,
+      "learning_rate": 6.671639134014679e-06,
+      "loss": 0.5868,
+      "step": 3942
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8755522182462092,
+      "learning_rate": 6.670052491502182e-06,
+      "loss": 0.6124,
+      "step": 3943
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3078342997813386,
+      "learning_rate": 6.668465659679714e-06,
+      "loss": 0.7289,
+      "step": 3944
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9514380326839884,
+      "learning_rate": 6.666878638727154e-06,
+      "loss": 0.6594,
+      "step": 3945
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.133201334464456,
+      "learning_rate": 6.6652914288243996e-06,
+      "loss": 0.5717,
+      "step": 3946
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9012313927156967,
+      "learning_rate": 6.66370403015137e-06,
+      "loss": 0.5791,
+      "step": 3947
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9722937856370126,
+      "learning_rate": 6.662116442888007e-06,
+      "loss": 0.6285,
+      "step": 3948
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.903817678915892,
+      "learning_rate": 6.66052866721427e-06,
+      "loss": 0.6406,
+      "step": 3949
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9979888015586342,
+      "learning_rate": 6.6589407033101435e-06,
+      "loss": 0.7432,
+      "step": 3950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6259230128595419,
+      "learning_rate": 6.657352551355634e-06,
+      "loss": 0.6007,
+      "step": 3951
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9194025746637722,
+      "learning_rate": 6.655764211530767e-06,
+      "loss": 0.7022,
+      "step": 3952
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.735014946320013,
+      "learning_rate": 6.654175684015587e-06,
+      "loss": 0.6331,
+      "step": 3953
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9823415478074962,
+      "learning_rate": 6.652586968990164e-06,
+      "loss": 0.7171,
+      "step": 3954
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7583769540998349,
+      "learning_rate": 6.650998066634589e-06,
+      "loss": 0.6572,
+      "step": 3955
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8600867587835666,
+      "learning_rate": 6.649408977128975e-06,
+      "loss": 0.6734,
+      "step": 3956
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.862698659637831,
+      "learning_rate": 6.64781970065345e-06,
+      "loss": 0.615,
+      "step": 3957
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9564006592363403,
+      "learning_rate": 6.646230237388172e-06,
+      "loss": 0.6748,
+      "step": 3958
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8968835571667682,
+      "learning_rate": 6.644640587513313e-06,
+      "loss": 0.6716,
+      "step": 3959
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.869462572063707,
+      "learning_rate": 6.643050751209067e-06,
+      "loss": 0.7545,
+      "step": 3960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8778880002152252,
+      "learning_rate": 6.641460728655654e-06,
+      "loss": 0.6778,
+      "step": 3961
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7399480890625565,
+      "learning_rate": 6.6398705200333125e-06,
+      "loss": 0.5642,
+      "step": 3962
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5578811285592717,
+      "learning_rate": 6.6382801255223e-06,
+      "loss": 0.4749,
+      "step": 3963
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6948492903335945,
+      "learning_rate": 6.636689545302898e-06,
+      "loss": 0.6076,
+      "step": 3964
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5766728927525278,
+      "learning_rate": 6.6350987795554056e-06,
+      "loss": 0.5974,
+      "step": 3965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9685332103178712,
+      "learning_rate": 6.633507828460148e-06,
+      "loss": 0.712,
+      "step": 3966
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.748252872817549,
+      "learning_rate": 6.631916692197466e-06,
+      "loss": 0.5779,
+      "step": 3967
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2011682557023478,
+      "learning_rate": 6.6303253709477276e-06,
+      "loss": 0.756,
+      "step": 3968
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.870775828998848,
+      "learning_rate": 6.628733864891315e-06,
+      "loss": 0.6796,
+      "step": 3969
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9934558724921358,
+      "learning_rate": 6.627142174208634e-06,
+      "loss": 0.6594,
+      "step": 3970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.762238827723651,
+      "learning_rate": 6.625550299080115e-06,
+      "loss": 0.673,
+      "step": 3971
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7996931164327568,
+      "learning_rate": 6.623958239686204e-06,
+      "loss": 0.599,
+      "step": 3972
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7062603543348756,
+      "learning_rate": 6.622365996207368e-06,
+      "loss": 0.6358,
+      "step": 3973
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8352990023889209,
+      "learning_rate": 6.620773568824101e-06,
+      "loss": 0.5925,
+      "step": 3974
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.777374114911824,
+      "learning_rate": 6.619180957716913e-06,
+      "loss": 0.5977,
+      "step": 3975
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8645416702797122,
+      "learning_rate": 6.617588163066333e-06,
+      "loss": 0.6726,
+      "step": 3976
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.986094087122366,
+      "learning_rate": 6.615995185052915e-06,
+      "loss": 0.6859,
+      "step": 3977
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9671637070485966,
+      "learning_rate": 6.614402023857231e-06,
+      "loss": 0.6191,
+      "step": 3978
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.992735241698695,
+      "learning_rate": 6.612808679659878e-06,
+      "loss": 0.7177,
+      "step": 3979
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8594009010630315,
+      "learning_rate": 6.611215152641466e-06,
+      "loss": 0.6817,
+      "step": 3980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8366393363968638,
+      "learning_rate": 6.609621442982634e-06,
+      "loss": 0.7283,
+      "step": 3981
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8880200232029714,
+      "learning_rate": 6.608027550864038e-06,
+      "loss": 0.6786,
+      "step": 3982
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7713852962235712,
+      "learning_rate": 6.606433476466352e-06,
+      "loss": 0.6204,
+      "step": 3983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.951459700873205,
+      "learning_rate": 6.604839219970276e-06,
+      "loss": 0.5822,
+      "step": 3984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.941984778501092,
+      "learning_rate": 6.603244781556527e-06,
+      "loss": 0.816,
+      "step": 3985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6869770005515161,
+      "learning_rate": 6.601650161405844e-06,
+      "loss": 0.6415,
+      "step": 3986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7618809871558578,
+      "learning_rate": 6.600055359698984e-06,
+      "loss": 0.718,
+      "step": 3987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8940876608861175,
+      "learning_rate": 6.598460376616731e-06,
+      "loss": 0.695,
+      "step": 3988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.582975947302647,
+      "learning_rate": 6.596865212339885e-06,
+      "loss": 0.5889,
+      "step": 3989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0959689544385873,
+      "learning_rate": 6.595269867049262e-06,
+      "loss": 0.7246,
+      "step": 3990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.875274315991869,
+      "learning_rate": 6.5936743409257085e-06,
+      "loss": 0.6586,
+      "step": 3991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.952777633657918,
+      "learning_rate": 6.592078634150084e-06,
+      "loss": 0.6749,
+      "step": 3992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0330590435053755,
+      "learning_rate": 6.590482746903273e-06,
+      "loss": 0.694,
+      "step": 3993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.040171030726033,
+      "learning_rate": 6.588886679366177e-06,
+      "loss": 0.8347,
+      "step": 3994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9779967255959818,
+      "learning_rate": 6.587290431719718e-06,
+      "loss": 0.5998,
+      "step": 3995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9489545122956449,
+      "learning_rate": 6.585694004144844e-06,
+      "loss": 0.622,
+      "step": 3996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7578029357926648,
+      "learning_rate": 6.584097396822514e-06,
+      "loss": 0.6302,
+      "step": 3997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.851928626062996,
+      "learning_rate": 6.582500609933715e-06,
+      "loss": 0.6829,
+      "step": 3998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7174238910726383,
+      "learning_rate": 6.580903643659453e-06,
+      "loss": 0.6014,
+      "step": 3999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8121816554070762,
+      "learning_rate": 6.579306498180753e-06,
+      "loss": 0.6708,
+      "step": 4000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8179688321936576,
+      "learning_rate": 6.577709173678658e-06,
+      "loss": 0.6709,
+      "step": 4001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0852265758278596,
+      "learning_rate": 6.5761116703342365e-06,
+      "loss": 0.7801,
+      "step": 4002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.725498003504001,
+      "learning_rate": 6.574513988328572e-06,
+      "loss": 0.6907,
+      "step": 4003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9877426312997266,
+      "learning_rate": 6.572916127842775e-06,
+      "loss": 0.5943,
+      "step": 4004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7800607172380565,
+      "learning_rate": 6.5713180890579675e-06,
+      "loss": 0.5852,
+      "step": 4005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0937490532286516,
+      "learning_rate": 6.569719872155299e-06,
+      "loss": 0.7188,
+      "step": 4006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7241389489460623,
+      "learning_rate": 6.568121477315936e-06,
+      "loss": 0.4979,
+      "step": 4007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.007324352220353,
+      "learning_rate": 6.566522904721066e-06,
+      "loss": 0.6385,
+      "step": 4008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2119954176168126,
+      "learning_rate": 6.564924154551895e-06,
+      "loss": 0.6259,
+      "step": 4009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1210351984877147,
+      "learning_rate": 6.563325226989652e-06,
+      "loss": 0.707,
+      "step": 4010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8997902827936242,
+      "learning_rate": 6.561726122215585e-06,
+      "loss": 0.7243,
+      "step": 4011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6957497867748266,
+      "learning_rate": 6.560126840410958e-06,
+      "loss": 0.6136,
+      "step": 4012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7441208168504188,
+      "learning_rate": 6.558527381757063e-06,
+      "loss": 0.6152,
+      "step": 4013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.790485685409849,
+      "learning_rate": 6.556927746435204e-06,
+      "loss": 0.6593,
+      "step": 4014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0974652382452303,
+      "learning_rate": 6.555327934626714e-06,
+      "loss": 0.6737,
+      "step": 4015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8269551144468505,
+      "learning_rate": 6.553727946512935e-06,
+      "loss": 0.6143,
+      "step": 4016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7153314087472655,
+      "learning_rate": 6.552127782275238e-06,
+      "loss": 0.6397,
+      "step": 4017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.090028098890312,
+      "learning_rate": 6.55052744209501e-06,
+      "loss": 0.6303,
+      "step": 4018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.009734384358688,
+      "learning_rate": 6.548926926153659e-06,
+      "loss": 0.6369,
+      "step": 4019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.770723520702836,
+      "learning_rate": 6.5473262346326125e-06,
+      "loss": 0.5863,
+      "step": 4020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0075361710627084,
+      "learning_rate": 6.545725367713317e-06,
+      "loss": 0.5489,
+      "step": 4021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7245698235105078,
+      "learning_rate": 6.5441243255772426e-06,
+      "loss": 0.6108,
+      "step": 4022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8035379353837957,
+      "learning_rate": 6.542523108405873e-06,
+      "loss": 0.8067,
+      "step": 4023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8715746504239854,
+      "learning_rate": 6.540921716380717e-06,
+      "loss": 0.6967,
+      "step": 4024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.096205344372492,
+      "learning_rate": 6.539320149683302e-06,
+      "loss": 0.645,
+      "step": 4025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7646255765648868,
+      "learning_rate": 6.537718408495174e-06,
+      "loss": 0.7169,
+      "step": 4026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8536865089842989,
+      "learning_rate": 6.536116492997899e-06,
+      "loss": 0.5697,
+      "step": 4027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0152237179046915,
+      "learning_rate": 6.534514403373064e-06,
+      "loss": 0.6279,
+      "step": 4028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8954231608810324,
+      "learning_rate": 6.5329121398022756e-06,
+      "loss": 0.665,
+      "step": 4029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8970903222431375,
+      "learning_rate": 6.531309702467159e-06,
+      "loss": 0.7078,
+      "step": 4030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.135112232684426,
+      "learning_rate": 6.52970709154936e-06,
+      "loss": 0.7513,
+      "step": 4031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9549109974607481,
+      "learning_rate": 6.528104307230542e-06,
+      "loss": 0.6653,
+      "step": 4032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7139953691811927,
+      "learning_rate": 6.526501349692392e-06,
+      "loss": 0.591,
+      "step": 4033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6627286877312124,
+      "learning_rate": 6.524898219116612e-06,
+      "loss": 0.6157,
+      "step": 4034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0560769425457464,
+      "learning_rate": 6.523294915684928e-06,
+      "loss": 0.634,
+      "step": 4035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8113394568340626,
+      "learning_rate": 6.5216914395790806e-06,
+      "loss": 0.6815,
+      "step": 4036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9270712852849294,
+      "learning_rate": 6.520087790980838e-06,
+      "loss": 0.6843,
+      "step": 4037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9527943178109233,
+      "learning_rate": 6.51848397007198e-06,
+      "loss": 0.709,
+      "step": 4038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.023544759218534,
+      "learning_rate": 6.516879977034307e-06,
+      "loss": 0.7007,
+      "step": 4039
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0956614909800018,
+      "learning_rate": 6.515275812049644e-06,
+      "loss": 0.6164,
+      "step": 4040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0323171139807004,
+      "learning_rate": 6.51367147529983e-06,
+      "loss": 0.706,
+      "step": 4041
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8683385638918015,
+      "learning_rate": 6.512066966966728e-06,
+      "loss": 0.6089,
+      "step": 4042
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.02288354013809,
+      "learning_rate": 6.510462287232216e-06,
+      "loss": 0.7069,
+      "step": 4043
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8013988532696286,
+      "learning_rate": 6.508857436278195e-06,
+      "loss": 0.6228,
+      "step": 4044
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0158224191382668,
+      "learning_rate": 6.5072524142865825e-06,
+      "loss": 0.7294,
+      "step": 4045
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6978751958811078,
+      "learning_rate": 6.505647221439317e-06,
+      "loss": 0.527,
+      "step": 4046
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7626559524080982,
+      "learning_rate": 6.504041857918359e-06,
+      "loss": 0.5494,
+      "step": 4047
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.052676030295689,
+      "learning_rate": 6.502436323905683e-06,
+      "loss": 0.7799,
+      "step": 4048
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9461214027379186,
+      "learning_rate": 6.500830619583286e-06,
+      "loss": 0.6377,
+      "step": 4049
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8652319862322915,
+      "learning_rate": 6.499224745133184e-06,
+      "loss": 0.6572,
+      "step": 4050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.077328808264135,
+      "learning_rate": 6.4976187007374116e-06,
+      "loss": 0.7387,
+      "step": 4051
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8648108669301284,
+      "learning_rate": 6.496012486578024e-06,
+      "loss": 0.6635,
+      "step": 4052
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.535449841127006,
+      "learning_rate": 6.494406102837093e-06,
+      "loss": 0.573,
+      "step": 4053
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7169141635665444,
+      "learning_rate": 6.492799549696712e-06,
+      "loss": 0.5881,
+      "step": 4054
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8801509275305481,
+      "learning_rate": 6.4911928273389946e-06,
+      "loss": 0.7089,
+      "step": 4055
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6912175530324325,
+      "learning_rate": 6.4895859359460714e-06,
+      "loss": 0.6018,
+      "step": 4056
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8383688047454536,
+      "learning_rate": 6.487978875700091e-06,
+      "loss": 0.6198,
+      "step": 4057
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8638435936380096,
+      "learning_rate": 6.486371646783223e-06,
+      "loss": 0.6198,
+      "step": 4058
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7715201966443364,
+      "learning_rate": 6.4847642493776585e-06,
+      "loss": 0.6964,
+      "step": 4059
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7922800486547616,
+      "learning_rate": 6.4831566836656024e-06,
+      "loss": 0.5816,
+      "step": 4060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7727987030469636,
+      "learning_rate": 6.481548949829282e-06,
+      "loss": 0.6143,
+      "step": 4061
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.024752836474955,
+      "learning_rate": 6.479941048050944e-06,
+      "loss": 0.6124,
+      "step": 4062
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8250783973226659,
+      "learning_rate": 6.478332978512853e-06,
+      "loss": 0.6522,
+      "step": 4063
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9493846686603185,
+      "learning_rate": 6.476724741397293e-06,
+      "loss": 0.6507,
+      "step": 4064
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.755309105876831,
+      "learning_rate": 6.4751163368865665e-06,
+      "loss": 0.6361,
+      "step": 4065
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.810159349170945,
+      "learning_rate": 6.473507765162994e-06,
+      "loss": 0.6304,
+      "step": 4066
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8676885311500284,
+      "learning_rate": 6.47189902640892e-06,
+      "loss": 0.7234,
+      "step": 4067
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6933054623850488,
+      "learning_rate": 6.4702901208067e-06,
+      "loss": 0.6732,
+      "step": 4068
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7455580715939196,
+      "learning_rate": 6.468681048538715e-06,
+      "loss": 0.6288,
+      "step": 4069
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8551247654424352,
+      "learning_rate": 6.467071809787363e-06,
+      "loss": 0.6904,
+      "step": 4070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2777286315308367,
+      "learning_rate": 6.4654624047350575e-06,
+      "loss": 0.7492,
+      "step": 4071
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6700799417004726,
+      "learning_rate": 6.463852833564236e-06,
+      "loss": 0.5802,
+      "step": 4072
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.829777392822749,
+      "learning_rate": 6.462243096457352e-06,
+      "loss": 0.6539,
+      "step": 4073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8592653629194054,
+      "learning_rate": 6.460633193596879e-06,
+      "loss": 0.5977,
+      "step": 4074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9504686835427847,
+      "learning_rate": 6.459023125165308e-06,
+      "loss": 0.7809,
+      "step": 4075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0166498080138373,
+      "learning_rate": 6.4574128913451495e-06,
+      "loss": 0.6218,
+      "step": 4076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9365597515134503,
+      "learning_rate": 6.4558024923189336e-06,
+      "loss": 0.678,
+      "step": 4077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7631181530060356,
+      "learning_rate": 6.454191928269207e-06,
+      "loss": 0.6319,
+      "step": 4078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.861501700664707,
+      "learning_rate": 6.452581199378536e-06,
+      "loss": 0.6315,
+      "step": 4079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1103702750302067,
+      "learning_rate": 6.450970305829507e-06,
+      "loss": 0.7092,
+      "step": 4080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.740618758516033,
+      "learning_rate": 6.449359247804724e-06,
+      "loss": 0.6223,
+      "step": 4081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8901679485332747,
+      "learning_rate": 6.447748025486809e-06,
+      "loss": 0.5887,
+      "step": 4082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9819488844550182,
+      "learning_rate": 6.4461366390584025e-06,
+      "loss": 0.7952,
+      "step": 4083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9219125754878177,
+      "learning_rate": 6.444525088702166e-06,
+      "loss": 0.8131,
+      "step": 4084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7999306822008587,
+      "learning_rate": 6.442913374600778e-06,
+      "loss": 0.6919,
+      "step": 4085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9952306260295138,
+      "learning_rate": 6.441301496936934e-06,
+      "loss": 0.6228,
+      "step": 4086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6397261652327657,
+      "learning_rate": 6.4396894558933495e-06,
+      "loss": 0.5822,
+      "step": 4087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0119623346471935,
+      "learning_rate": 6.438077251652759e-06,
+      "loss": 0.6775,
+      "step": 4088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.883160572533884,
+      "learning_rate": 6.436464884397917e-06,
+      "loss": 0.5655,
+      "step": 4089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9595553513484605,
+      "learning_rate": 6.434852354311592e-06,
+      "loss": 0.6677,
+      "step": 4090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8563543420640487,
+      "learning_rate": 6.433239661576574e-06,
+      "loss": 0.6959,
+      "step": 4091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.979965895251359,
+      "learning_rate": 6.431626806375671e-06,
+      "loss": 0.6135,
+      "step": 4092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.895092434982727,
+      "learning_rate": 6.4300137888917104e-06,
+      "loss": 0.5857,
+      "step": 4093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8786371189307325,
+      "learning_rate": 6.428400609307535e-06,
+      "loss": 0.6763,
+      "step": 4094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9249867638405949,
+      "learning_rate": 6.426787267806009e-06,
+      "loss": 0.5852,
+      "step": 4095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.076984042439822,
+      "learning_rate": 6.425173764570014e-06,
+      "loss": 0.5923,
+      "step": 4096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9017582414587852,
+      "learning_rate": 6.42356009978245e-06,
+      "loss": 0.6385,
+      "step": 4097
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.98427043016857,
+      "learning_rate": 6.421946273626234e-06,
+      "loss": 0.6316,
+      "step": 4098
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.085670532511142,
+      "learning_rate": 6.420332286284303e-06,
+      "loss": 0.6466,
+      "step": 4099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9356842581183769,
+      "learning_rate": 6.418718137939614e-06,
+      "loss": 0.6055,
+      "step": 4100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.099063254616092,
+      "learning_rate": 6.417103828775135e-06,
+      "loss": 0.6641,
+      "step": 4101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2228334066963855,
+      "learning_rate": 6.41548935897386e-06,
+      "loss": 0.784,
+      "step": 4102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7704647109970402,
+      "learning_rate": 6.4138747287187984e-06,
+      "loss": 0.6703,
+      "step": 4103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9802068841142038,
+      "learning_rate": 6.412259938192978e-06,
+      "loss": 0.6845,
+      "step": 4104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4689051914936173,
+      "learning_rate": 6.410644987579444e-06,
+      "loss": 0.594,
+      "step": 4105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.212128219379263,
+      "learning_rate": 6.409029877061259e-06,
+      "loss": 0.7405,
+      "step": 4106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.875314160139368,
+      "learning_rate": 6.407414606821507e-06,
+      "loss": 0.6231,
+      "step": 4107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8218343066015268,
+      "learning_rate": 6.405799177043289e-06,
+      "loss": 0.6286,
+      "step": 4108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0688539285382137,
+      "learning_rate": 6.4041835879097205e-06,
+      "loss": 0.7444,
+      "step": 4109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7472797608134105,
+      "learning_rate": 6.402567839603937e-06,
+      "loss": 0.6215,
+      "step": 4110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9131271318603962,
+      "learning_rate": 6.400951932309097e-06,
+      "loss": 0.6368,
+      "step": 4111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.028012533221227,
+      "learning_rate": 6.399335866208367e-06,
+      "loss": 0.6568,
+      "step": 4112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8051750494724248,
+      "learning_rate": 6.397719641484943e-06,
+      "loss": 0.6733,
+      "step": 4113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9604187629140273,
+      "learning_rate": 6.39610325832203e-06,
+      "loss": 0.6956,
+      "step": 4114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7146717528468935,
+      "learning_rate": 6.394486716902857e-06,
+      "loss": 0.567,
+      "step": 4115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8587330203323895,
+      "learning_rate": 6.392870017410665e-06,
+      "loss": 0.7145,
+      "step": 4116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.886070724526401,
+      "learning_rate": 6.3912531600287166e-06,
+      "loss": 0.6478,
+      "step": 4117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8676316071674293,
+      "learning_rate": 6.389636144940294e-06,
+      "loss": 0.6376,
+      "step": 4118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.152777218544474,
+      "learning_rate": 6.388018972328693e-06,
+      "loss": 0.6583,
+      "step": 4119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7736923671339342,
+      "learning_rate": 6.386401642377231e-06,
+      "loss": 0.6683,
+      "step": 4120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7318840943278404,
+      "learning_rate": 6.384784155269239e-06,
+      "loss": 0.6245,
+      "step": 4121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1659556826960933,
+      "learning_rate": 6.383166511188072e-06,
+      "loss": 0.7848,
+      "step": 4122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8961148222331055,
+      "learning_rate": 6.381548710317096e-06,
+      "loss": 0.7207,
+      "step": 4123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8419789827418658,
+      "learning_rate": 6.3799307528397e-06,
+      "loss": 0.7183,
+      "step": 4124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8897939224972424,
+      "learning_rate": 6.378312638939286e-06,
+      "loss": 0.5796,
+      "step": 4125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9689256228329488,
+      "learning_rate": 6.37669436879928e-06,
+      "loss": 0.7358,
+      "step": 4126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8892196667301127,
+      "learning_rate": 6.375075942603119e-06,
+      "loss": 0.7024,
+      "step": 4127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9259521778924813,
+      "learning_rate": 6.373457360534263e-06,
+      "loss": 0.6811,
+      "step": 4128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8004120338996794,
+      "learning_rate": 6.371838622776187e-06,
+      "loss": 0.6131,
+      "step": 4129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.935184387430914,
+      "learning_rate": 6.370219729512383e-06,
+      "loss": 0.6478,
+      "step": 4130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.909323332437752,
+      "learning_rate": 6.368600680926364e-06,
+      "loss": 0.7664,
+      "step": 4131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1700474542918915,
+      "learning_rate": 6.3669814772016555e-06,
+      "loss": 0.7379,
+      "step": 4132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8763611045110336,
+      "learning_rate": 6.365362118521807e-06,
+      "loss": 0.6204,
+      "step": 4133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7139413076250838,
+      "learning_rate": 6.363742605070379e-06,
+      "loss": 0.7147,
+      "step": 4134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9044300533734067,
+      "learning_rate": 6.362122937030952e-06,
+      "loss": 0.6914,
+      "step": 4135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9154252110669694,
+      "learning_rate": 6.360503114587129e-06,
+      "loss": 0.615,
+      "step": 4136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.739297540790688,
+      "learning_rate": 6.3588831379225226e-06,
+      "loss": 0.6519,
+      "step": 4137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7433831357411251,
+      "learning_rate": 6.357263007220767e-06,
+      "loss": 0.6645,
+      "step": 4138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9514132458981708,
+      "learning_rate": 6.355642722665512e-06,
+      "loss": 0.7732,
+      "step": 4139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8361076741980893,
+      "learning_rate": 6.354022284440429e-06,
+      "loss": 0.6559,
+      "step": 4140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9369982517217328,
+      "learning_rate": 6.352401692729202e-06,
+      "loss": 0.6168,
+      "step": 4141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5500501779631404,
+      "learning_rate": 6.3507809477155335e-06,
+      "loss": 0.6486,
+      "step": 4142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.762058079092894,
+      "learning_rate": 6.349160049583146e-06,
+      "loss": 0.6379,
+      "step": 4143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6547077639256238,
+      "learning_rate": 6.347538998515778e-06,
+      "loss": 0.6427,
+      "step": 4144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.863644426003244,
+      "learning_rate": 6.345917794697183e-06,
+      "loss": 0.682,
+      "step": 4145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0289147358075525,
+      "learning_rate": 6.344296438311134e-06,
+      "loss": 0.6736,
+      "step": 4146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0453181169725765,
+      "learning_rate": 6.342674929541424e-06,
+      "loss": 0.6646,
+      "step": 4147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.814620528092348,
+      "learning_rate": 6.341053268571855e-06,
+      "loss": 0.6626,
+      "step": 4148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9840600275863907,
+      "learning_rate": 6.3394314555862545e-06,
+      "loss": 0.6618,
+      "step": 4149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8379449832426429,
+      "learning_rate": 6.337809490768465e-06,
+      "loss": 0.6177,
+      "step": 4150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8309354258877755,
+      "learning_rate": 6.336187374302344e-06,
+      "loss": 0.7409,
+      "step": 4151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7755918403707267,
+      "learning_rate": 6.334565106371768e-06,
+      "loss": 0.6348,
+      "step": 4152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8224093319923627,
+      "learning_rate": 6.332942687160632e-06,
+      "loss": 0.6293,
+      "step": 4153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0566967543569863,
+      "learning_rate": 6.331320116852842e-06,
+      "loss": 0.6957,
+      "step": 4154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8244220543225507,
+      "learning_rate": 6.329697395632332e-06,
+      "loss": 0.6038,
+      "step": 4155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.623371055393367,
+      "learning_rate": 6.328074523683041e-06,
+      "loss": 0.6673,
+      "step": 4156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2657238349497706,
+      "learning_rate": 6.326451501188933e-06,
+      "loss": 0.6658,
+      "step": 4157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9014605197412302,
+      "learning_rate": 6.324828328333986e-06,
+      "loss": 0.6146,
+      "step": 4158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7894161936132815,
+      "learning_rate": 6.323205005302199e-06,
+      "loss": 0.6886,
+      "step": 4159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8682449812414834,
+      "learning_rate": 6.321581532277581e-06,
+      "loss": 0.6771,
+      "step": 4160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7046598099665882,
+      "learning_rate": 6.319957909444163e-06,
+      "loss": 0.6186,
+      "step": 4161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.963434024206442,
+      "learning_rate": 6.318334136985993e-06,
+      "loss": 0.6581,
+      "step": 4162
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9731526290449264,
+      "learning_rate": 6.316710215087136e-06,
+      "loss": 0.6845,
+      "step": 4163
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8947817321763945,
+      "learning_rate": 6.31508614393167e-06,
+      "loss": 0.71,
+      "step": 4164
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6776835190313186,
+      "learning_rate": 6.313461923703693e-06,
+      "loss": 0.5618,
+      "step": 4165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0755723255709397,
+      "learning_rate": 6.311837554587322e-06,
+      "loss": 0.6423,
+      "step": 4166
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8678892690364675,
+      "learning_rate": 6.3102130367666855e-06,
+      "loss": 0.6015,
+      "step": 4167
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9535946234388366,
+      "learning_rate": 6.308588370425934e-06,
+      "loss": 0.6249,
+      "step": 4168
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9754689426313052,
+      "learning_rate": 6.306963555749231e-06,
+      "loss": 0.6902,
+      "step": 4169
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7827061123763464,
+      "learning_rate": 6.305338592920762e-06,
+      "loss": 0.6728,
+      "step": 4170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.737874266961729,
+      "learning_rate": 6.303713482124721e-06,
+      "loss": 0.6458,
+      "step": 4171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8841336594214666,
+      "learning_rate": 6.302088223545327e-06,
+      "loss": 0.6342,
+      "step": 4172
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8336447518739336,
+      "learning_rate": 6.30046281736681e-06,
+      "loss": 0.5722,
+      "step": 4173
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9792552077265053,
+      "learning_rate": 6.298837263773423e-06,
+      "loss": 0.619,
+      "step": 4174
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8990343746392675,
+      "learning_rate": 6.297211562949427e-06,
+      "loss": 0.6661,
+      "step": 4175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.012978566826468,
+      "learning_rate": 6.2955857150791055e-06,
+      "loss": 0.6569,
+      "step": 4176
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.988969857473836,
+      "learning_rate": 6.29395972034676e-06,
+      "loss": 0.6604,
+      "step": 4177
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7627911436862862,
+      "learning_rate": 6.2923335789367044e-06,
+      "loss": 0.5787,
+      "step": 4178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9374367211852295,
+      "learning_rate": 6.290707291033272e-06,
+      "loss": 0.6177,
+      "step": 4179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0083371897370674,
+      "learning_rate": 6.289080856820811e-06,
+      "loss": 0.6268,
+      "step": 4180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0569173930723967,
+      "learning_rate": 6.287454276483687e-06,
+      "loss": 0.6635,
+      "step": 4181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7807436006960025,
+      "learning_rate": 6.285827550206282e-06,
+      "loss": 0.535,
+      "step": 4182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.101638627215469,
+      "learning_rate": 6.284200678172997e-06,
+      "loss": 0.7426,
+      "step": 4183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7072537972165742,
+      "learning_rate": 6.282573660568245e-06,
+      "loss": 0.7568,
+      "step": 4184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9695668809384006,
+      "learning_rate": 6.2809464975764575e-06,
+      "loss": 0.5676,
+      "step": 4185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.767265477943679,
+      "learning_rate": 6.279319189382084e-06,
+      "loss": 0.601,
+      "step": 4186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7094500604340592,
+      "learning_rate": 6.2776917361695876e-06,
+      "loss": 0.6147,
+      "step": 4187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8402827980509162,
+      "learning_rate": 6.276064138123453e-06,
+      "loss": 0.732,
+      "step": 4188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9402004321921997,
+      "learning_rate": 6.274436395428171e-06,
+      "loss": 0.6128,
+      "step": 4189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9031106270161509,
+      "learning_rate": 6.272808508268262e-06,
+      "loss": 0.6311,
+      "step": 4190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.033607806685722,
+      "learning_rate": 6.2711804768282535e-06,
+      "loss": 0.5969,
+      "step": 4191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.800733372504521,
+      "learning_rate": 6.269552301292693e-06,
+      "loss": 0.5638,
+      "step": 4192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.639474658639473,
+      "learning_rate": 6.267923981846141e-06,
+      "loss": 0.5257,
+      "step": 4193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.023559643318534,
+      "learning_rate": 6.26629551867318e-06,
+      "loss": 0.5674,
+      "step": 4194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.874292710750846,
+      "learning_rate": 6.264666911958404e-06,
+      "loss": 0.6333,
+      "step": 4195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8563287999328968,
+      "learning_rate": 6.263038161886426e-06,
+      "loss": 0.7358,
+      "step": 4196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7453380961988425,
+      "learning_rate": 6.261409268641872e-06,
+      "loss": 0.6227,
+      "step": 4197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.751372462067731,
+      "learning_rate": 6.259780232409389e-06,
+      "loss": 0.6074,
+      "step": 4198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.880312499572035,
+      "learning_rate": 6.2581510533736346e-06,
+      "loss": 0.6836,
+      "step": 4199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.866018226892596,
+      "learning_rate": 6.25652173171929e-06,
+      "loss": 0.6002,
+      "step": 4200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8337619090346275,
+      "learning_rate": 6.254892267631042e-06,
+      "loss": 0.6665,
+      "step": 4201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.821660317307866,
+      "learning_rate": 6.2532626612936035e-06,
+      "loss": 0.6837,
+      "step": 4202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.852300675093971,
+      "learning_rate": 6.2516329128917e-06,
+      "loss": 0.6855,
+      "step": 4203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.067378628400363,
+      "learning_rate": 6.250003022610071e-06,
+      "loss": 0.684,
+      "step": 4204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.766568461268017,
+      "learning_rate": 6.248372990633475e-06,
+      "loss": 0.6828,
+      "step": 4205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.121014004372851,
+      "learning_rate": 6.246742817146684e-06,
+      "loss": 0.6887,
+      "step": 4206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.916239882668666,
+      "learning_rate": 6.2451125023344895e-06,
+      "loss": 0.7135,
+      "step": 4207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0778830065986345,
+      "learning_rate": 6.243482046381696e-06,
+      "loss": 0.6779,
+      "step": 4208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.869012297808762,
+      "learning_rate": 6.2418514494731245e-06,
+      "loss": 0.6217,
+      "step": 4209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.650570523552766,
+      "learning_rate": 6.240220711793612e-06,
+      "loss": 0.6305,
+      "step": 4210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.775718554562567,
+      "learning_rate": 6.238589833528015e-06,
+      "loss": 0.6848,
+      "step": 4211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.010000193787045,
+      "learning_rate": 6.236958814861199e-06,
+      "loss": 0.6825,
+      "step": 4212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7651378794347117,
+      "learning_rate": 6.2353276559780515e-06,
+      "loss": 0.5645,
+      "step": 4213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.722675513011501,
+      "learning_rate": 6.233696357063472e-06,
+      "loss": 0.668,
+      "step": 4214
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1181437150465214,
+      "learning_rate": 6.23206491830238e-06,
+      "loss": 0.7165,
+      "step": 4215
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9627449801841879,
+      "learning_rate": 6.230433339879706e-06,
+      "loss": 0.6373,
+      "step": 4216
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.747089518779566,
+      "learning_rate": 6.2288016219804e-06,
+      "loss": 0.5995,
+      "step": 4217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6960776760193497,
+      "learning_rate": 6.2271697647894265e-06,
+      "loss": 0.6389,
+      "step": 4218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9381104103996185,
+      "learning_rate": 6.225537768491766e-06,
+      "loss": 0.6122,
+      "step": 4219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9867789174214119,
+      "learning_rate": 6.223905633272414e-06,
+      "loss": 0.6503,
+      "step": 4220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2355853167261124,
+      "learning_rate": 6.2222733593163805e-06,
+      "loss": 0.7641,
+      "step": 4221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.796152824199846,
+      "learning_rate": 6.220640946808697e-06,
+      "loss": 0.5726,
+      "step": 4222
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8490238088076443,
+      "learning_rate": 6.219008395934405e-06,
+      "loss": 0.5832,
+      "step": 4223
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.734880478046965,
+      "learning_rate": 6.217375706878561e-06,
+      "loss": 0.6244,
+      "step": 4224
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8760334967705816,
+      "learning_rate": 6.215742879826244e-06,
+      "loss": 0.6718,
+      "step": 4225
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8641079257282516,
+      "learning_rate": 6.214109914962542e-06,
+      "loss": 0.6378,
+      "step": 4226
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9736071636844352,
+      "learning_rate": 6.21247681247256e-06,
+      "loss": 0.7905,
+      "step": 4227
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.657253663363964,
+      "learning_rate": 6.210843572541421e-06,
+      "loss": 0.6086,
+      "step": 4228
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9013689343551632,
+      "learning_rate": 6.20921019535426e-06,
+      "loss": 0.6301,
+      "step": 4229
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9593538236517356,
+      "learning_rate": 6.207576681096233e-06,
+      "loss": 0.7359,
+      "step": 4230
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9607098259709497,
+      "learning_rate": 6.205943029952505e-06,
+      "loss": 0.6411,
+      "step": 4231
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.917909444411163,
+      "learning_rate": 6.204309242108262e-06,
+      "loss": 0.6877,
+      "step": 4232
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9283331737073521,
+      "learning_rate": 6.202675317748702e-06,
+      "loss": 0.6473,
+      "step": 4233
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.820140476133725,
+      "learning_rate": 6.201041257059039e-06,
+      "loss": 0.623,
+      "step": 4234
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8303703060759442,
+      "learning_rate": 6.199407060224503e-06,
+      "loss": 0.6681,
+      "step": 4235
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1784835352215315,
+      "learning_rate": 6.197772727430341e-06,
+      "loss": 0.6553,
+      "step": 4236
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.057402259459153,
+      "learning_rate": 6.196138258861815e-06,
+      "loss": 0.6533,
+      "step": 4237
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9346080271003194,
+      "learning_rate": 6.194503654704198e-06,
+      "loss": 0.6091,
+      "step": 4238
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7055354507122449,
+      "learning_rate": 6.192868915142782e-06,
+      "loss": 0.6281,
+      "step": 4239
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9143105967875074,
+      "learning_rate": 6.191234040362879e-06,
+      "loss": 0.5833,
+      "step": 4240
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7670799000079251,
+      "learning_rate": 6.189599030549804e-06,
+      "loss": 0.6947,
+      "step": 4241
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8540914584065187,
+      "learning_rate": 6.1879638858889e-06,
+      "loss": 0.7129,
+      "step": 4242
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7533761314777891,
+      "learning_rate": 6.186328606565518e-06,
+      "loss": 0.5931,
+      "step": 4243
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8074134710317675,
+      "learning_rate": 6.184693192765028e-06,
+      "loss": 0.7204,
+      "step": 4244
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0892333656674364,
+      "learning_rate": 6.18305764467281e-06,
+      "loss": 0.7689,
+      "step": 4245
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.903938283307967,
+      "learning_rate": 6.181421962474267e-06,
+      "loss": 0.6832,
+      "step": 4246
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9806689141914189,
+      "learning_rate": 6.179786146354808e-06,
+      "loss": 0.6931,
+      "step": 4247
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.977935900264373,
+      "learning_rate": 6.178150196499868e-06,
+      "loss": 0.6945,
+      "step": 4248
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8313727726369613,
+      "learning_rate": 6.176514113094885e-06,
+      "loss": 0.5794,
+      "step": 4249
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.081898696849267,
+      "learning_rate": 6.174877896325323e-06,
+      "loss": 0.6605,
+      "step": 4250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9063195075574626,
+      "learning_rate": 6.173241546376654e-06,
+      "loss": 0.6784,
+      "step": 4251
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8626522818570608,
+      "learning_rate": 6.171605063434368e-06,
+      "loss": 0.6452,
+      "step": 4252
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8826580292136437,
+      "learning_rate": 6.169968447683971e-06,
+      "loss": 0.6847,
+      "step": 4253
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7157387012544394,
+      "learning_rate": 6.168331699310982e-06,
+      "loss": 0.5853,
+      "step": 4254
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9401064441075455,
+      "learning_rate": 6.1666948185009355e-06,
+      "loss": 0.6954,
+      "step": 4255
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0024822246974736,
+      "learning_rate": 6.165057805439382e-06,
+      "loss": 0.6232,
+      "step": 4256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.171084207492589,
+      "learning_rate": 6.1634206603118844e-06,
+      "loss": 0.672,
+      "step": 4257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8833760760024472,
+      "learning_rate": 6.161783383304024e-06,
+      "loss": 0.6084,
+      "step": 4258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8417561591740161,
+      "learning_rate": 6.160145974601397e-06,
+      "loss": 0.6039,
+      "step": 4259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0709052955398577,
+      "learning_rate": 6.158508434389608e-06,
+      "loss": 0.6919,
+      "step": 4260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8776488381311174,
+      "learning_rate": 6.156870762854287e-06,
+      "loss": 0.7309,
+      "step": 4261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.873816021513749,
+      "learning_rate": 6.155232960181071e-06,
+      "loss": 0.6691,
+      "step": 4262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7901395019323316,
+      "learning_rate": 6.153595026555613e-06,
+      "loss": 0.6377,
+      "step": 4263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.75387062711662,
+      "learning_rate": 6.151956962163584e-06,
+      "loss": 0.6408,
+      "step": 4264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.770309730929508,
+      "learning_rate": 6.150318767190668e-06,
+      "loss": 0.6725,
+      "step": 4265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.919943910487086,
+      "learning_rate": 6.148680441822563e-06,
+      "loss": 0.5916,
+      "step": 4266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8486164691741107,
+      "learning_rate": 6.1470419862449825e-06,
+      "loss": 0.5883,
+      "step": 4267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9032761553068864,
+      "learning_rate": 6.1454034006436545e-06,
+      "loss": 0.5558,
+      "step": 4268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9283442009335932,
+      "learning_rate": 6.143764685204323e-06,
+      "loss": 0.6445,
+      "step": 4269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7842124977670573,
+      "learning_rate": 6.142125840112746e-06,
+      "loss": 0.6604,
+      "step": 4270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6599561789243407,
+      "learning_rate": 6.140486865554693e-06,
+      "loss": 0.6057,
+      "step": 4271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7994386291779092,
+      "learning_rate": 6.138847761715955e-06,
+      "loss": 0.5795,
+      "step": 4272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1473723111235663,
+      "learning_rate": 6.137208528782331e-06,
+      "loss": 0.6959,
+      "step": 4273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.725204692644165,
+      "learning_rate": 6.1355691669396386e-06,
+      "loss": 0.5396,
+      "step": 4274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0090637781380067,
+      "learning_rate": 6.133929676373709e-06,
+      "loss": 0.6082,
+      "step": 4275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9772578144450519,
+      "learning_rate": 6.132290057270387e-06,
+      "loss": 0.6552,
+      "step": 4276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0108173014531148,
+      "learning_rate": 6.130650309815535e-06,
+      "loss": 0.6897,
+      "step": 4277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.823894686140687,
+      "learning_rate": 6.129010434195023e-06,
+      "loss": 0.6536,
+      "step": 4278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6203568412684168,
+      "learning_rate": 6.127370430594745e-06,
+      "loss": 0.5311,
+      "step": 4279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.014024329591531,
+      "learning_rate": 6.125730299200601e-06,
+      "loss": 0.6031,
+      "step": 4280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.944148231455446,
+      "learning_rate": 6.124090040198514e-06,
+      "loss": 0.667,
+      "step": 4281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.960939325536016,
+      "learning_rate": 6.122449653774411e-06,
+      "loss": 0.7211,
+      "step": 4282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9195307639129888,
+      "learning_rate": 6.120809140114243e-06,
+      "loss": 0.6355,
+      "step": 4283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7758071657350973,
+      "learning_rate": 6.119168499403971e-06,
+      "loss": 0.6558,
+      "step": 4284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9774042812835415,
+      "learning_rate": 6.11752773182957e-06,
+      "loss": 0.6619,
+      "step": 4285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8617259628540834,
+      "learning_rate": 6.115886837577031e-06,
+      "loss": 0.5504,
+      "step": 4286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8015457930515442,
+      "learning_rate": 6.114245816832359e-06,
+      "loss": 0.5942,
+      "step": 4287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8809551738481627,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.6016,
+      "step": 4288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8090247840030873,
+      "learning_rate": 6.110963396610705e-06,
+      "loss": 0.583,
+      "step": 4289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.000768037728457,
+      "learning_rate": 6.109321997505804e-06,
+      "loss": 0.6731,
+      "step": 4290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9743242871047149,
+      "learning_rate": 6.107680472652931e-06,
+      "loss": 0.715,
+      "step": 4291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9282462151001922,
+      "learning_rate": 6.106038822238165e-06,
+      "loss": 0.6739,
+      "step": 4292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1015185797276303,
+      "learning_rate": 6.104397046447593e-06,
+      "loss": 0.708,
+      "step": 4293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.748887075305813,
+      "learning_rate": 6.1027551454673204e-06,
+      "loss": 0.7053,
+      "step": 4294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8395064781208474,
+      "learning_rate": 6.1011131194834675e-06,
+      "loss": 0.6444,
+      "step": 4295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7803548664428446,
+      "learning_rate": 6.099470968682168e-06,
+      "loss": 0.5716,
+      "step": 4296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6009153734353425,
+      "learning_rate": 6.097828693249565e-06,
+      "loss": 0.6457,
+      "step": 4297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.647971075805321,
+      "learning_rate": 6.0961862933718215e-06,
+      "loss": 0.5746,
+      "step": 4298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8482386722821733,
+      "learning_rate": 6.0945437692351166e-06,
+      "loss": 0.6898,
+      "step": 4299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8232321333867145,
+      "learning_rate": 6.092901121025634e-06,
+      "loss": 0.6632,
+      "step": 4300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.881039170540099,
+      "learning_rate": 6.091258348929581e-06,
+      "loss": 0.6537,
+      "step": 4301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0396835214722606,
+      "learning_rate": 6.089615453133173e-06,
+      "loss": 0.6914,
+      "step": 4302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8316939185491101,
+      "learning_rate": 6.0879724338226454e-06,
+      "loss": 0.5984,
+      "step": 4303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7744519568370833,
+      "learning_rate": 6.086329291184238e-06,
+      "loss": 0.5976,
+      "step": 4304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.945342822305884,
+      "learning_rate": 6.084686025404216e-06,
+      "loss": 0.5689,
+      "step": 4305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.660347128085138,
+      "learning_rate": 6.08304263666885e-06,
+      "loss": 0.5238,
+      "step": 4306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9805416996449299,
+      "learning_rate": 6.081399125164429e-06,
+      "loss": 0.7111,
+      "step": 4307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8201579374736838,
+      "learning_rate": 6.079755491077251e-06,
+      "loss": 0.62,
+      "step": 4308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.671229966988917,
+      "learning_rate": 6.0781117345936345e-06,
+      "loss": 0.5775,
+      "step": 4309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8676293114126867,
+      "learning_rate": 6.07646785589991e-06,
+      "loss": 0.7568,
+      "step": 4310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8390499419653883,
+      "learning_rate": 6.074823855182416e-06,
+      "loss": 0.6798,
+      "step": 4311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7786191474157553,
+      "learning_rate": 6.073179732627512e-06,
+      "loss": 0.5254,
+      "step": 4312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7017606975225539,
+      "learning_rate": 6.0715354884215685e-06,
+      "loss": 0.5916,
+      "step": 4313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8331971488804444,
+      "learning_rate": 6.069891122750971e-06,
+      "loss": 0.6283,
+      "step": 4314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0707708625988377,
+      "learning_rate": 6.068246635802115e-06,
+      "loss": 0.6521,
+      "step": 4315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1080846184243596,
+      "learning_rate": 6.066602027761414e-06,
+      "loss": 0.6613,
+      "step": 4316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8541610045993289,
+      "learning_rate": 6.064957298815295e-06,
+      "loss": 0.7219,
+      "step": 4317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7233308318810572,
+      "learning_rate": 6.063312449150196e-06,
+      "loss": 0.7256,
+      "step": 4318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7786958252668963,
+      "learning_rate": 6.06166747895257e-06,
+      "loss": 0.6196,
+      "step": 4319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8992649487983346,
+      "learning_rate": 6.060022388408883e-06,
+      "loss": 0.6449,
+      "step": 4320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8783947566207382,
+      "learning_rate": 6.0583771777056166e-06,
+      "loss": 0.6849,
+      "step": 4321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.947749299526801,
+      "learning_rate": 6.056731847029265e-06,
+      "loss": 0.7143,
+      "step": 4322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9019855284634992,
+      "learning_rate": 6.055086396566334e-06,
+      "loss": 0.5855,
+      "step": 4323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.831529691156291,
+      "learning_rate": 6.0534408265033485e-06,
+      "loss": 0.6443,
+      "step": 4324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8942202831397477,
+      "learning_rate": 6.05179513702684e-06,
+      "loss": 0.5609,
+      "step": 4325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9975511737435045,
+      "learning_rate": 6.050149328323358e-06,
+      "loss": 0.6611,
+      "step": 4326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.970690832774122,
+      "learning_rate": 6.048503400579463e-06,
+      "loss": 0.657,
+      "step": 4327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.903442299604656,
+      "learning_rate": 6.046857353981732e-06,
+      "loss": 0.6878,
+      "step": 4328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9033342278105545,
+      "learning_rate": 6.045211188716753e-06,
+      "loss": 0.6339,
+      "step": 4329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9358946001912258,
+      "learning_rate": 6.043564904971129e-06,
+      "loss": 0.7204,
+      "step": 4330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.562197508869273,
+      "learning_rate": 6.041918502931473e-06,
+      "loss": 0.5725,
+      "step": 4331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7487153609771282,
+      "learning_rate": 6.040271982784417e-06,
+      "loss": 0.6848,
+      "step": 4332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0537934311353845,
+      "learning_rate": 6.038625344716603e-06,
+      "loss": 0.7616,
+      "step": 4333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.880860377128153,
+      "learning_rate": 6.036978588914684e-06,
+      "loss": 0.6721,
+      "step": 4334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7029662258523208,
+      "learning_rate": 6.035331715565333e-06,
+      "loss": 0.6612,
+      "step": 4335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0649190844874443,
+      "learning_rate": 6.0336847248552335e-06,
+      "loss": 0.7124,
+      "step": 4336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8863676505857916,
+      "learning_rate": 6.032037616971075e-06,
+      "loss": 0.6383,
+      "step": 4337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6913686311949407,
+      "learning_rate": 6.030390392099571e-06,
+      "loss": 0.5815,
+      "step": 4338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6854186885544427,
+      "learning_rate": 6.028743050427442e-06,
+      "loss": 0.5173,
+      "step": 4339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0603592121426995,
+      "learning_rate": 6.027095592141428e-06,
+      "loss": 0.65,
+      "step": 4340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8627823283033291,
+      "learning_rate": 6.025448017428272e-06,
+      "loss": 0.5187,
+      "step": 4341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8550463706984868,
+      "learning_rate": 6.023800326474738e-06,
+      "loss": 0.6554,
+      "step": 4342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.01413399369854,
+      "learning_rate": 6.022152519467601e-06,
+      "loss": 0.5523,
+      "step": 4343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.871011023204209,
+      "learning_rate": 6.020504596593652e-06,
+      "loss": 0.5976,
+      "step": 4344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9291226273099116,
+      "learning_rate": 6.018856558039689e-06,
+      "loss": 0.5815,
+      "step": 4345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1033708436285283,
+      "learning_rate": 6.017208403992527e-06,
+      "loss": 0.6804,
+      "step": 4346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7746131482935668,
+      "learning_rate": 6.015560134638997e-06,
+      "loss": 0.6312,
+      "step": 4347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.736794612224939,
+      "learning_rate": 6.013911750165935e-06,
+      "loss": 0.688,
+      "step": 4348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8071088364104275,
+      "learning_rate": 6.012263250760199e-06,
+      "loss": 0.6993,
+      "step": 4349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.767310249059785,
+      "learning_rate": 6.0106146366086514e-06,
+      "loss": 0.5741,
+      "step": 4350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0526654951756416,
+      "learning_rate": 6.0089659078981765e-06,
+      "loss": 0.6669,
+      "step": 4351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6899837975267995,
+      "learning_rate": 6.007317064815664e-06,
+      "loss": 0.5892,
+      "step": 4352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9256094251339038,
+      "learning_rate": 6.0056681075480206e-06,
+      "loss": 0.6298,
+      "step": 4353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.018125474079877,
+      "learning_rate": 6.004019036282165e-06,
+      "loss": 0.687,
+      "step": 4354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8216632857004333,
+      "learning_rate": 6.002369851205029e-06,
+      "loss": 0.6037,
+      "step": 4355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.934654363397389,
+      "learning_rate": 6.000720552503557e-06,
+      "loss": 0.6146,
+      "step": 4356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.816258512462237,
+      "learning_rate": 5.999071140364708e-06,
+      "loss": 0.6523,
+      "step": 4357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8239092670075012,
+      "learning_rate": 5.997421614975449e-06,
+      "loss": 0.6458,
+      "step": 4358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9149484450077814,
+      "learning_rate": 5.995771976522765e-06,
+      "loss": 0.7412,
+      "step": 4359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9121047542571723,
+      "learning_rate": 5.9941222251936525e-06,
+      "loss": 0.6096,
+      "step": 4360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8481439735718255,
+      "learning_rate": 5.992472361175118e-06,
+      "loss": 0.6237,
+      "step": 4361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0655990944889124,
+      "learning_rate": 5.990822384654187e-06,
+      "loss": 0.6851,
+      "step": 4362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8831730913470377,
+      "learning_rate": 5.989172295817889e-06,
+      "loss": 0.618,
+      "step": 4363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8724322214686773,
+      "learning_rate": 5.9875220948532745e-06,
+      "loss": 0.6682,
+      "step": 4364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.905634382989326,
+      "learning_rate": 5.9858717819474e-06,
+      "loss": 0.647,
+      "step": 4365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0386048289605534,
+      "learning_rate": 5.984221357287342e-06,
+      "loss": 0.6118,
+      "step": 4366
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8630627415405536,
+      "learning_rate": 5.982570821060182e-06,
+      "loss": 0.6513,
+      "step": 4367
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.146377572209457,
+      "learning_rate": 5.980920173453019e-06,
+      "loss": 0.6575,
+      "step": 4368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8648330548708527,
+      "learning_rate": 5.979269414652964e-06,
+      "loss": 0.5981,
+      "step": 4369
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7479863045414308,
+      "learning_rate": 5.977618544847139e-06,
+      "loss": 0.6314,
+      "step": 4370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9122892283559052,
+      "learning_rate": 5.975967564222679e-06,
+      "loss": 0.6404,
+      "step": 4371
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7702998645102086,
+      "learning_rate": 5.974316472966732e-06,
+      "loss": 0.6091,
+      "step": 4372
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9419264167413117,
+      "learning_rate": 5.9726652712664625e-06,
+      "loss": 0.6424,
+      "step": 4373
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.924232841922094,
+      "learning_rate": 5.971013959309038e-06,
+      "loss": 0.6804,
+      "step": 4374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0329707067511555,
+      "learning_rate": 5.969362537281647e-06,
+      "loss": 0.7667,
+      "step": 4375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8286541545437696,
+      "learning_rate": 5.967711005371487e-06,
+      "loss": 0.6649,
+      "step": 4376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.036373560381382,
+      "learning_rate": 5.966059363765771e-06,
+      "loss": 0.6073,
+      "step": 4377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8738880734455083,
+      "learning_rate": 5.9644076126517166e-06,
+      "loss": 0.6915,
+      "step": 4378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.049257494525163,
+      "learning_rate": 5.962755752216564e-06,
+      "loss": 0.6387,
+      "step": 4379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.693094678717967,
+      "learning_rate": 5.961103782647558e-06,
+      "loss": 0.4983,
+      "step": 4380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9267764438897257,
+      "learning_rate": 5.959451704131962e-06,
+      "loss": 0.594,
+      "step": 4381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.966784263635863,
+      "learning_rate": 5.957799516857046e-06,
+      "loss": 0.6696,
+      "step": 4382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.917751357367465,
+      "learning_rate": 5.9561472210100955e-06,
+      "loss": 0.616,
+      "step": 4383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9782860809483898,
+      "learning_rate": 5.954494816778408e-06,
+      "loss": 0.719,
+      "step": 4384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8408410543098181,
+      "learning_rate": 5.952842304349291e-06,
+      "loss": 0.6837,
+      "step": 4385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0510629750021216,
+      "learning_rate": 5.951189683910069e-06,
+      "loss": 0.6411,
+      "step": 4386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7191965103067213,
+      "learning_rate": 5.949536955648074e-06,
+      "loss": 0.5532,
+      "step": 4387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7432450197461355,
+      "learning_rate": 5.947884119750656e-06,
+      "loss": 0.652,
+      "step": 4388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7602986319167306,
+      "learning_rate": 5.946231176405166e-06,
+      "loss": 0.5931,
+      "step": 4389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9640168883565872,
+      "learning_rate": 5.944578125798981e-06,
+      "loss": 0.6641,
+      "step": 4390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2045820879343583,
+      "learning_rate": 5.94292496811948e-06,
+      "loss": 0.6991,
+      "step": 4391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6992537195980884,
+      "learning_rate": 5.94127170355406e-06,
+      "loss": 0.5314,
+      "step": 4392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7760267943580061,
+      "learning_rate": 5.939618332290128e-06,
+      "loss": 0.6334,
+      "step": 4393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0270949968085743,
+      "learning_rate": 5.937964854515101e-06,
+      "loss": 0.7455,
+      "step": 4394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.853300870630623,
+      "learning_rate": 5.936311270416415e-06,
+      "loss": 0.5298,
+      "step": 4395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.094047846726863,
+      "learning_rate": 5.9346575801815064e-06,
+      "loss": 0.6803,
+      "step": 4396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5348679190305892,
+      "learning_rate": 5.933003783997835e-06,
+      "loss": 0.5962,
+      "step": 4397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7849133821203453,
+      "learning_rate": 5.931349882052866e-06,
+      "loss": 0.6919,
+      "step": 4398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7708118222363871,
+      "learning_rate": 5.929695874534081e-06,
+      "loss": 0.701,
+      "step": 4399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9364917331436853,
+      "learning_rate": 5.928041761628968e-06,
+      "loss": 0.614,
+      "step": 4400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.81362935636992,
+      "learning_rate": 5.926387543525031e-06,
+      "loss": 0.7095,
+      "step": 4401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9855548797846845,
+      "learning_rate": 5.924733220409786e-06,
+      "loss": 0.6378,
+      "step": 4402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9860147094040368,
+      "learning_rate": 5.9230787924707625e-06,
+      "loss": 0.717,
+      "step": 4403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.799813809623052,
+      "learning_rate": 5.921424259895493e-06,
+      "loss": 0.6095,
+      "step": 4404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.854974605347049,
+      "learning_rate": 5.919769622871533e-06,
+      "loss": 0.614,
+      "step": 4405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8929793362499432,
+      "learning_rate": 5.918114881586444e-06,
+      "loss": 0.6377,
+      "step": 4406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8735159541726778,
+      "learning_rate": 5.9164600362278005e-06,
+      "loss": 0.5564,
+      "step": 4407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9266281340327402,
+      "learning_rate": 5.914805086983187e-06,
+      "loss": 0.6363,
+      "step": 4408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.91652283894417,
+      "learning_rate": 5.913150034040203e-06,
+      "loss": 0.5966,
+      "step": 4409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8967692186120793,
+      "learning_rate": 5.9114948775864585e-06,
+      "loss": 0.5763,
+      "step": 4410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.868068960782959,
+      "learning_rate": 5.909839617809574e-06,
+      "loss": 0.5677,
+      "step": 4411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.949627880291475,
+      "learning_rate": 5.908184254897183e-06,
+      "loss": 0.6456,
+      "step": 4412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8952854730038935,
+      "learning_rate": 5.906528789036929e-06,
+      "loss": 0.7244,
+      "step": 4413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.88180194047267,
+      "learning_rate": 5.904873220416472e-06,
+      "loss": 0.6549,
+      "step": 4414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9535874023387245,
+      "learning_rate": 5.903217549223477e-06,
+      "loss": 0.6613,
+      "step": 4415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8438707598961703,
+      "learning_rate": 5.901561775645623e-06,
+      "loss": 0.5778,
+      "step": 4416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0575987277313716,
+      "learning_rate": 5.8999058998706046e-06,
+      "loss": 0.656,
+      "step": 4417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8539834505786594,
+      "learning_rate": 5.898249922086123e-06,
+      "loss": 0.7293,
+      "step": 4418
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8979032405655898,
+      "learning_rate": 5.896593842479893e-06,
+      "loss": 0.7049,
+      "step": 4419
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0946306162111976,
+      "learning_rate": 5.89493766123964e-06,
+      "loss": 0.6983,
+      "step": 4420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1908352500873205,
+      "learning_rate": 5.893281378553104e-06,
+      "loss": 0.7129,
+      "step": 4421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0150748552490465,
+      "learning_rate": 5.891624994608029e-06,
+      "loss": 0.6083,
+      "step": 4422
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7215755278309917,
+      "learning_rate": 5.8899685095921814e-06,
+      "loss": 0.656,
+      "step": 4423
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9343024267813007,
+      "learning_rate": 5.888311923693328e-06,
+      "loss": 0.5971,
+      "step": 4424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8744245976889982,
+      "learning_rate": 5.886655237099257e-06,
+      "loss": 0.6771,
+      "step": 4425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9031240731966292,
+      "learning_rate": 5.88499844999776e-06,
+      "loss": 0.6218,
+      "step": 4426
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9729786683428097,
+      "learning_rate": 5.8833415625766455e-06,
+      "loss": 0.5538,
+      "step": 4427
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.834573070869529,
+      "learning_rate": 5.881684575023729e-06,
+      "loss": 0.6215,
+      "step": 4428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.760291997270579,
+      "learning_rate": 5.880027487526842e-06,
+      "loss": 0.5801,
+      "step": 4429
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9670809366698832,
+      "learning_rate": 5.878370300273821e-06,
+      "loss": 0.6537,
+      "step": 4430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.011201336258629,
+      "learning_rate": 5.876713013452521e-06,
+      "loss": 0.7965,
+      "step": 4431
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0312721214511242,
+      "learning_rate": 5.875055627250804e-06,
+      "loss": 0.6555,
+      "step": 4432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7256318681000062,
+      "learning_rate": 5.873398141856545e-06,
+      "loss": 0.7285,
+      "step": 4433
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7330021830336968,
+      "learning_rate": 5.871740557457626e-06,
+      "loss": 0.5818,
+      "step": 4434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7225502241433897,
+      "learning_rate": 5.870082874241947e-06,
+      "loss": 0.5089,
+      "step": 4435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9286229634377934,
+      "learning_rate": 5.868425092397416e-06,
+      "loss": 0.7606,
+      "step": 4436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9414863379591762,
+      "learning_rate": 5.86676721211195e-06,
+      "loss": 0.6719,
+      "step": 4437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8080730220213328,
+      "learning_rate": 5.86510923357348e-06,
+      "loss": 0.6629,
+      "step": 4438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7384087585471595,
+      "learning_rate": 5.8634511569699486e-06,
+      "loss": 0.728,
+      "step": 4439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8723704862477661,
+      "learning_rate": 5.861792982489306e-06,
+      "loss": 0.6252,
+      "step": 4440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7642246326221103,
+      "learning_rate": 5.860134710319517e-06,
+      "loss": 0.6327,
+      "step": 4441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8926388005449055,
+      "learning_rate": 5.858476340648555e-06,
+      "loss": 0.6858,
+      "step": 4442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8708903747298244,
+      "learning_rate": 5.856817873664409e-06,
+      "loss": 0.7199,
+      "step": 4443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6894652296859378,
+      "learning_rate": 5.855159309555072e-06,
+      "loss": 0.5358,
+      "step": 4444
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8180878241005396,
+      "learning_rate": 5.853500648508552e-06,
+      "loss": 0.6665,
+      "step": 4445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9056202180080377,
+      "learning_rate": 5.8518418907128694e-06,
+      "loss": 0.6349,
+      "step": 4446
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0279056733244483,
+      "learning_rate": 5.850183036356054e-06,
+      "loss": 0.6681,
+      "step": 4447
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7326930647206062,
+      "learning_rate": 5.8485240856261446e-06,
+      "loss": 0.5012,
+      "step": 4448
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7860636370758964,
+      "learning_rate": 5.846865038711194e-06,
+      "loss": 0.7176,
+      "step": 4449
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9843512953094944,
+      "learning_rate": 5.845205895799264e-06,
+      "loss": 0.7526,
+      "step": 4450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.010199635494966,
+      "learning_rate": 5.8435466570784295e-06,
+      "loss": 0.6164,
+      "step": 4451
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.892725992106848,
+      "learning_rate": 5.8418873227367724e-06,
+      "loss": 0.6222,
+      "step": 4452
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0294869818788763,
+      "learning_rate": 5.840227892962388e-06,
+      "loss": 0.6154,
+      "step": 4453
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0869333452841192,
+      "learning_rate": 5.838568367943383e-06,
+      "loss": 0.6628,
+      "step": 4454
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6602804318639692,
+      "learning_rate": 5.8369087478678755e-06,
+      "loss": 0.6215,
+      "step": 4455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8406790377489561,
+      "learning_rate": 5.835249032923989e-06,
+      "loss": 0.6173,
+      "step": 4456
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6276517300331053,
+      "learning_rate": 5.833589223299865e-06,
+      "loss": 0.6091,
+      "step": 4457
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.820277411526944,
+      "learning_rate": 5.831929319183651e-06,
+      "loss": 0.6295,
+      "step": 4458
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9676714418497274,
+      "learning_rate": 5.830269320763507e-06,
+      "loss": 0.7288,
+      "step": 4459
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8329203395615346,
+      "learning_rate": 5.828609228227603e-06,
+      "loss": 0.7023,
+      "step": 4460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8305837504641471,
+      "learning_rate": 5.82694904176412e-06,
+      "loss": 0.6172,
+      "step": 4461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1586280747536115,
+      "learning_rate": 5.825288761561248e-06,
+      "loss": 0.7142,
+      "step": 4462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8557818338452512,
+      "learning_rate": 5.823628387807193e-06,
+      "loss": 0.5918,
+      "step": 4463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0083257509743015,
+      "learning_rate": 5.821967920690165e-06,
+      "loss": 0.6542,
+      "step": 4464
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7232403898211963,
+      "learning_rate": 5.82030736039839e-06,
+      "loss": 0.5959,
+      "step": 4465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.690743499181069,
+      "learning_rate": 5.818646707120098e-06,
+      "loss": 0.6244,
+      "step": 4466
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.946693473563528,
+      "learning_rate": 5.8169859610435355e-06,
+      "loss": 0.5974,
+      "step": 4467
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8758207732410697,
+      "learning_rate": 5.815325122356959e-06,
+      "loss": 0.6168,
+      "step": 4468
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.847333041662102,
+      "learning_rate": 5.813664191248631e-06,
+      "loss": 0.702,
+      "step": 4469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8080621215350756,
+      "learning_rate": 5.8120031679068315e-06,
+      "loss": 0.5091,
+      "step": 4470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7877506559716392,
+      "learning_rate": 5.810342052519842e-06,
+      "loss": 0.6393,
+      "step": 4471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8901858594534553,
+      "learning_rate": 5.808680845275963e-06,
+      "loss": 0.5721,
+      "step": 4472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.873495548260194,
+      "learning_rate": 5.8070195463635025e-06,
+      "loss": 0.6208,
+      "step": 4473
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9721690840833237,
+      "learning_rate": 5.8053581559707754e-06,
+      "loss": 0.6252,
+      "step": 4474
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8923298290697836,
+      "learning_rate": 5.80369667428611e-06,
+      "loss": 0.6648,
+      "step": 4475
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0497196950258973,
+      "learning_rate": 5.802035101497846e-06,
+      "loss": 0.6719,
+      "step": 4476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7156531826874144,
+      "learning_rate": 5.800373437794334e-06,
+      "loss": 0.6103,
+      "step": 4477
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8561943349164527,
+      "learning_rate": 5.798711683363929e-06,
+      "loss": 0.6015,
+      "step": 4478
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.722452492222899,
+      "learning_rate": 5.797049838395001e-06,
+      "loss": 0.5597,
+      "step": 4479
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8749034479522824,
+      "learning_rate": 5.795387903075933e-06,
+      "loss": 0.6504,
+      "step": 4480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7635470279845804,
+      "learning_rate": 5.7937258775951125e-06,
+      "loss": 0.5879,
+      "step": 4481
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8571087504016002,
+      "learning_rate": 5.792063762140938e-06,
+      "loss": 0.6682,
+      "step": 4482
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8506275346708279,
+      "learning_rate": 5.790401556901822e-06,
+      "loss": 0.7062,
+      "step": 4483
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9054042143788954,
+      "learning_rate": 5.788739262066185e-06,
+      "loss": 0.5929,
+      "step": 4484
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.745527884049558,
+      "learning_rate": 5.787076877822457e-06,
+      "loss": 0.6085,
+      "step": 4485
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8825036097416747,
+      "learning_rate": 5.7854144043590775e-06,
+      "loss": 0.6847,
+      "step": 4486
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.023138211383369,
+      "learning_rate": 5.7837518418645e-06,
+      "loss": 0.6308,
+      "step": 4487
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9547099231146705,
+      "learning_rate": 5.782089190527185e-06,
+      "loss": 0.6312,
+      "step": 4488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.643185636278992,
+      "learning_rate": 5.7804264505356e-06,
+      "loss": 0.6085,
+      "step": 4489
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0414113888224117,
+      "learning_rate": 5.7787636220782294e-06,
+      "loss": 0.6321,
+      "step": 4490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2067355496136902,
+      "learning_rate": 5.777100705343565e-06,
+      "loss": 0.7242,
+      "step": 4491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9603021200117243,
+      "learning_rate": 5.775437700520103e-06,
+      "loss": 0.59,
+      "step": 4492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9097724088822532,
+      "learning_rate": 5.7737746077963605e-06,
+      "loss": 0.714,
+      "step": 4493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6876366699927718,
+      "learning_rate": 5.772111427360855e-06,
+      "loss": 0.5481,
+      "step": 4494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8767910335072366,
+      "learning_rate": 5.770448159402118e-06,
+      "loss": 0.6646,
+      "step": 4495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0294806073429137,
+      "learning_rate": 5.7687848041086905e-06,
+      "loss": 0.6843,
+      "step": 4496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9588128637096367,
+      "learning_rate": 5.767121361669125e-06,
+      "loss": 0.7105,
+      "step": 4497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9716151486499722,
+      "learning_rate": 5.765457832271979e-06,
+      "loss": 0.6373,
+      "step": 4498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.080359918608944,
+      "learning_rate": 5.763794216105826e-06,
+      "loss": 0.7501,
+      "step": 4499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8964238274940508,
+      "learning_rate": 5.762130513359244e-06,
+      "loss": 0.5701,
+      "step": 4500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8669777706044983,
+      "learning_rate": 5.760466724220824e-06,
+      "loss": 0.5573,
+      "step": 4501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9046577228447057,
+      "learning_rate": 5.758802848879169e-06,
+      "loss": 0.6396,
+      "step": 4502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2300958145019742,
+      "learning_rate": 5.757138887522884e-06,
+      "loss": 0.7542,
+      "step": 4503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8800972518796621,
+      "learning_rate": 5.75547484034059e-06,
+      "loss": 0.5727,
+      "step": 4504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2108595105155944,
+      "learning_rate": 5.753810707520918e-06,
+      "loss": 0.5708,
+      "step": 4505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9557531259523118,
+      "learning_rate": 5.7521464892525055e-06,
+      "loss": 0.668,
+      "step": 4506
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6753004989326905,
+      "learning_rate": 5.750482185724001e-06,
+      "loss": 0.5802,
+      "step": 4507
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8277555372897838,
+      "learning_rate": 5.748817797124063e-06,
+      "loss": 0.5803,
+      "step": 4508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9809336995417062,
+      "learning_rate": 5.74715332364136e-06,
+      "loss": 0.6485,
+      "step": 4509
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.944164497840986,
+      "learning_rate": 5.7454887654645706e-06,
+      "loss": 0.554,
+      "step": 4510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9367604540968901,
+      "learning_rate": 5.743824122782379e-06,
+      "loss": 0.7028,
+      "step": 4511
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0476021643490068,
+      "learning_rate": 5.7421593957834835e-06,
+      "loss": 0.5873,
+      "step": 4512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8303359306005593,
+      "learning_rate": 5.74049458465659e-06,
+      "loss": 0.5952,
+      "step": 4513
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6641282597478058,
+      "learning_rate": 5.738829689590415e-06,
+      "loss": 0.6916,
+      "step": 4514
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9218716826790592,
+      "learning_rate": 5.7371647107736824e-06,
+      "loss": 0.5552,
+      "step": 4515
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8021875591091798,
+      "learning_rate": 5.73549964839513e-06,
+      "loss": 0.6477,
+      "step": 4516
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.874696675514852,
+      "learning_rate": 5.7338345026434995e-06,
+      "loss": 0.6552,
+      "step": 4517
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7124112887912268,
+      "learning_rate": 5.732169273707545e-06,
+      "loss": 0.6077,
+      "step": 4518
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.036393377024413,
+      "learning_rate": 5.73050396177603e-06,
+      "loss": 0.7247,
+      "step": 4519
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7701392918915468,
+      "learning_rate": 5.728838567037728e-06,
+      "loss": 0.7127,
+      "step": 4520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9231725547933194,
+      "learning_rate": 5.72717308968142e-06,
+      "loss": 0.6524,
+      "step": 4521
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.080539493901293,
+      "learning_rate": 5.725507529895898e-06,
+      "loss": 0.6405,
+      "step": 4522
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8921086679701191,
+      "learning_rate": 5.723841887869961e-06,
+      "loss": 0.6963,
+      "step": 4523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.982435709140776,
+      "learning_rate": 5.72217616379242e-06,
+      "loss": 0.553,
+      "step": 4524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.869479487988279,
+      "learning_rate": 5.7205103578520956e-06,
+      "loss": 0.5267,
+      "step": 4525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8967605276922301,
+      "learning_rate": 5.7188444702378155e-06,
+      "loss": 0.6433,
+      "step": 4526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1663730345064023,
+      "learning_rate": 5.717178501138416e-06,
+      "loss": 0.7024,
+      "step": 4527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6711389674607338,
+      "learning_rate": 5.715512450742749e-06,
+      "loss": 0.6199,
+      "step": 4528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.037140945002469,
+      "learning_rate": 5.713846319239664e-06,
+      "loss": 0.6665,
+      "step": 4529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.022157294672067,
+      "learning_rate": 5.71218010681803e-06,
+      "loss": 0.7589,
+      "step": 4530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7878692251576156,
+      "learning_rate": 5.710513813666722e-06,
+      "loss": 0.6059,
+      "step": 4531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9980205127983506,
+      "learning_rate": 5.708847439974625e-06,
+      "loss": 0.6536,
+      "step": 4532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.109109182112321,
+      "learning_rate": 5.707180985930629e-06,
+      "loss": 0.6149,
+      "step": 4533
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.035240334490125,
+      "learning_rate": 5.705514451723637e-06,
+      "loss": 0.6236,
+      "step": 4534
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.991215297512859,
+      "learning_rate": 5.703847837542562e-06,
+      "loss": 0.6123,
+      "step": 4535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8754198782296225,
+      "learning_rate": 5.702181143576323e-06,
+      "loss": 0.5805,
+      "step": 4536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8767609705655672,
+      "learning_rate": 5.7005143700138474e-06,
+      "loss": 0.5803,
+      "step": 4537
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8788940690815126,
+      "learning_rate": 5.698847517044076e-06,
+      "loss": 0.6342,
+      "step": 4538
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.046203066451224,
+      "learning_rate": 5.697180584855957e-06,
+      "loss": 0.7154,
+      "step": 4539
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.147823782012038,
+      "learning_rate": 5.6955135736384425e-06,
+      "loss": 0.6689,
+      "step": 4540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9211938256886534,
+      "learning_rate": 5.693846483580501e-06,
+      "loss": 0.7068,
+      "step": 4541
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7172972276987455,
+      "learning_rate": 5.692179314871104e-06,
+      "loss": 0.5695,
+      "step": 4542
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9061594156772637,
+      "learning_rate": 5.69051206769924e-06,
+      "loss": 0.6323,
+      "step": 4543
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9402592924801827,
+      "learning_rate": 5.688844742253895e-06,
+      "loss": 0.7288,
+      "step": 4544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9548934582721385,
+      "learning_rate": 5.687177338724073e-06,
+      "loss": 0.6359,
+      "step": 4545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9058935634603749,
+      "learning_rate": 5.685509857298781e-06,
+      "loss": 0.6532,
+      "step": 4546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8694790048885155,
+      "learning_rate": 5.683842298167041e-06,
+      "loss": 0.6986,
+      "step": 4547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0055822862463293,
+      "learning_rate": 5.68217466151788e-06,
+      "loss": 0.5503,
+      "step": 4548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8668045735488943,
+      "learning_rate": 5.680506947540331e-06,
+      "loss": 0.7325,
+      "step": 4549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6947983432539857,
+      "learning_rate": 5.67883915642344e-06,
+      "loss": 0.692,
+      "step": 4550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8895950172582314,
+      "learning_rate": 5.677171288356263e-06,
+      "loss": 0.7109,
+      "step": 4551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.971379110045142,
+      "learning_rate": 5.675503343527861e-06,
+      "loss": 0.623,
+      "step": 4552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.030884339212526,
+      "learning_rate": 5.673835322127304e-06,
+      "loss": 0.6353,
+      "step": 4553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.928163277407566,
+      "learning_rate": 5.672167224343673e-06,
+      "loss": 0.6159,
+      "step": 4554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.908799846546338,
+      "learning_rate": 5.670499050366055e-06,
+      "loss": 0.5873,
+      "step": 4555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.141916280219293,
+      "learning_rate": 5.668830800383548e-06,
+      "loss": 0.8722,
+      "step": 4556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9235665075383108,
+      "learning_rate": 5.667162474585258e-06,
+      "loss": 0.6594,
+      "step": 4557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9022698171629637,
+      "learning_rate": 5.6654940731602995e-06,
+      "loss": 0.674,
+      "step": 4558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0388621559832467,
+      "learning_rate": 5.663825596297794e-06,
+      "loss": 0.6344,
+      "step": 4559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.689495512348284,
+      "learning_rate": 5.662157044186873e-06,
+      "loss": 0.5544,
+      "step": 4560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0045306909764857,
+      "learning_rate": 5.6604884170166765e-06,
+      "loss": 0.673,
+      "step": 4561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9111890412837886,
+      "learning_rate": 5.658819714976355e-06,
+      "loss": 0.6942,
+      "step": 4562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9998776939328162,
+      "learning_rate": 5.657150938255062e-06,
+      "loss": 0.701,
+      "step": 4563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0427375075903735,
+      "learning_rate": 5.655482087041965e-06,
+      "loss": 0.6718,
+      "step": 4564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0273043034155704,
+      "learning_rate": 5.653813161526237e-06,
+      "loss": 0.6799,
+      "step": 4565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9518930981995333,
+      "learning_rate": 5.6521441618970605e-06,
+      "loss": 0.5647,
+      "step": 4566
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7441388879377546,
+      "learning_rate": 5.6504750883436275e-06,
+      "loss": 0.5699,
+      "step": 4567
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.890032316812201,
+      "learning_rate": 5.648805941055135e-06,
+      "loss": 0.6521,
+      "step": 4568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.880305597341674,
+      "learning_rate": 5.647136720220791e-06,
+      "loss": 0.6335,
+      "step": 4569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8075253023263849,
+      "learning_rate": 5.64546742602981e-06,
+      "loss": 0.6353,
+      "step": 4570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7588936557722914,
+      "learning_rate": 5.643798058671418e-06,
+      "loss": 0.6952,
+      "step": 4571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6995840290285382,
+      "learning_rate": 5.6421286183348465e-06,
+      "loss": 0.6305,
+      "step": 4572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0948395121790315,
+      "learning_rate": 5.640459105209337e-06,
+      "loss": 0.6411,
+      "step": 4573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9736848664004365,
+      "learning_rate": 5.638789519484137e-06,
+      "loss": 0.6777,
+      "step": 4574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0435556431950803,
+      "learning_rate": 5.637119861348504e-06,
+      "loss": 0.7337,
+      "step": 4575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0847525177688038,
+      "learning_rate": 5.6354501309917034e-06,
+      "loss": 0.5759,
+      "step": 4576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.659038326232906,
+      "learning_rate": 5.633780328603008e-06,
+      "loss": 0.5909,
+      "step": 4577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9345623802471068,
+      "learning_rate": 5.6321104543717e-06,
+      "loss": 0.6019,
+      "step": 4578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.963218170933064,
+      "learning_rate": 5.630440508487068e-06,
+      "loss": 0.6093,
+      "step": 4579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7722605156571707,
+      "learning_rate": 5.628770491138414e-06,
+      "loss": 0.5577,
+      "step": 4580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9424797388437025,
+      "learning_rate": 5.627100402515038e-06,
+      "loss": 0.6364,
+      "step": 4581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9240402046319094,
+      "learning_rate": 5.625430242806258e-06,
+      "loss": 0.7009,
+      "step": 4582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.721892694944862,
+      "learning_rate": 5.623760012201394e-06,
+      "loss": 0.6482,
+      "step": 4583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.977992478992338,
+      "learning_rate": 5.62208971088978e-06,
+      "loss": 0.7306,
+      "step": 4584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.847701541976097,
+      "learning_rate": 5.62041933906075e-06,
+      "loss": 0.6842,
+      "step": 4585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8640134483922797,
+      "learning_rate": 5.618748896903652e-06,
+      "loss": 0.6163,
+      "step": 4586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8814485760535749,
+      "learning_rate": 5.617078384607839e-06,
+      "loss": 0.7321,
+      "step": 4587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6377353046992302,
+      "learning_rate": 5.615407802362675e-06,
+      "loss": 0.6036,
+      "step": 4588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8481119867884892,
+      "learning_rate": 5.613737150357528e-06,
+      "loss": 0.629,
+      "step": 4589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.892284220082062,
+      "learning_rate": 5.6120664287817765e-06,
+      "loss": 0.6487,
+      "step": 4590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.021796552026886,
+      "learning_rate": 5.610395637824808e-06,
+      "loss": 0.7287,
+      "step": 4591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8874399763646057,
+      "learning_rate": 5.608724777676013e-06,
+      "loss": 0.6372,
+      "step": 4592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7794562570973433,
+      "learning_rate": 5.607053848524796e-06,
+      "loss": 0.6554,
+      "step": 4593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.741221722198036,
+      "learning_rate": 5.605382850560565e-06,
+      "loss": 0.6421,
+      "step": 4594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7314815048933372,
+      "learning_rate": 5.603711783972738e-06,
+      "loss": 0.5944,
+      "step": 4595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8385199882160486,
+      "learning_rate": 5.6020406489507385e-06,
+      "loss": 0.6879,
+      "step": 4596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9110860400813556,
+      "learning_rate": 5.600369445683999e-06,
+      "loss": 0.555,
+      "step": 4597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9249828682309444,
+      "learning_rate": 5.5986981743619615e-06,
+      "loss": 0.6981,
+      "step": 4598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.949901847401595,
+      "learning_rate": 5.597026835174075e-06,
+      "loss": 0.6773,
+      "step": 4599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8957329701386099,
+      "learning_rate": 5.5953554283097925e-06,
+      "loss": 0.7059,
+      "step": 4600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9662603076387533,
+      "learning_rate": 5.593683953958579e-06,
+      "loss": 0.5705,
+      "step": 4601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9760343575599348,
+      "learning_rate": 5.592012412309905e-06,
+      "loss": 0.5734,
+      "step": 4602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9314288640416741,
+      "learning_rate": 5.590340803553249e-06,
+      "loss": 0.5705,
+      "step": 4603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7749475103591477,
+      "learning_rate": 5.5886691278780995e-06,
+      "loss": 0.5766,
+      "step": 4604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8301652296483235,
+      "learning_rate": 5.586997385473949e-06,
+      "loss": 0.5995,
+      "step": 4605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0224492387711708,
+      "learning_rate": 5.5853255765302995e-06,
+      "loss": 0.7307,
+      "step": 4606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7800813580039285,
+      "learning_rate": 5.583653701236658e-06,
+      "loss": 0.5758,
+      "step": 4607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7387567730906435,
+      "learning_rate": 5.581981759782543e-06,
+      "loss": 0.6738,
+      "step": 4608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8039219026028164,
+      "learning_rate": 5.580309752357479e-06,
+      "loss": 0.582,
+      "step": 4609
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9569486267401757,
+      "learning_rate": 5.578637679150997e-06,
+      "loss": 0.6614,
+      "step": 4610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7888601680101468,
+      "learning_rate": 5.576965540352637e-06,
+      "loss": 0.5691,
+      "step": 4611
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7861531175452023,
+      "learning_rate": 5.575293336151943e-06,
+      "loss": 0.5806,
+      "step": 4612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9083526025670454,
+      "learning_rate": 5.573621066738471e-06,
+      "loss": 0.6261,
+      "step": 4613
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.737846510421099,
+      "learning_rate": 5.571948732301781e-06,
+      "loss": 0.6108,
+      "step": 4614
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0423564349394105,
+      "learning_rate": 5.570276333031441e-06,
+      "loss": 0.6206,
+      "step": 4615
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.125783143784693,
+      "learning_rate": 5.568603869117029e-06,
+      "loss": 0.7213,
+      "step": 4616
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8570347144204054,
+      "learning_rate": 5.56693134074813e-06,
+      "loss": 0.6892,
+      "step": 4617
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7588279865564456,
+      "learning_rate": 5.56525874811433e-06,
+      "loss": 0.618,
+      "step": 4618
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7432706073652564,
+      "learning_rate": 5.563586091405229e-06,
+      "loss": 0.5623,
+      "step": 4619
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.847040133653717,
+      "learning_rate": 5.561913370810432e-06,
+      "loss": 0.6304,
+      "step": 4620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0289255167643967,
+      "learning_rate": 5.560240586519553e-06,
+      "loss": 0.5503,
+      "step": 4621
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.781866236186172,
+      "learning_rate": 5.558567738722208e-06,
+      "loss": 0.5261,
+      "step": 4622
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.270523192504574,
+      "learning_rate": 5.556894827608027e-06,
+      "loss": 0.7636,
+      "step": 4623
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.925337493063739,
+      "learning_rate": 5.555221853366644e-06,
+      "loss": 0.503,
+      "step": 4624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7839496985845051,
+      "learning_rate": 5.5535488161876994e-06,
+      "loss": 0.612,
+      "step": 4625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6814669083880103,
+      "learning_rate": 5.55187571626084e-06,
+      "loss": 0.6839,
+      "step": 4626
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6368160551509423,
+      "learning_rate": 5.550202553775723e-06,
+      "loss": 0.6327,
+      "step": 4627
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7793440986251825,
+      "learning_rate": 5.548529328922012e-06,
+      "loss": 0.5759,
+      "step": 4628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9763629323931438,
+      "learning_rate": 5.546856041889374e-06,
+      "loss": 0.6881,
+      "step": 4629
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9111816163433664,
+      "learning_rate": 5.545182692867486e-06,
+      "loss": 0.5799,
+      "step": 4630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.825419245447919,
+      "learning_rate": 5.543509282046031e-06,
+      "loss": 0.6236,
+      "step": 4631
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8821953354276433,
+      "learning_rate": 5.541835809614704e-06,
+      "loss": 0.6645,
+      "step": 4632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.800286495436237,
+      "learning_rate": 5.540162275763198e-06,
+      "loss": 0.5669,
+      "step": 4633
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9310568135270056,
+      "learning_rate": 5.53848868068122e-06,
+      "loss": 0.6105,
+      "step": 4634
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.894980337439149,
+      "learning_rate": 5.53681502455848e-06,
+      "loss": 0.6786,
+      "step": 4635
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.966074480215953,
+      "learning_rate": 5.535141307584697e-06,
+      "loss": 0.6413,
+      "step": 4636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8636160131590067,
+      "learning_rate": 5.5334675299495975e-06,
+      "loss": 0.5534,
+      "step": 4637
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9035811073752469,
+      "learning_rate": 5.531793691842912e-06,
+      "loss": 0.7617,
+      "step": 4638
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9211832192257259,
+      "learning_rate": 5.530119793454381e-06,
+      "loss": 0.6682,
+      "step": 4639
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.990268702208091,
+      "learning_rate": 5.52844583497375e-06,
+      "loss": 0.6658,
+      "step": 4640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.048537657034261,
+      "learning_rate": 5.52677181659077e-06,
+      "loss": 0.7124,
+      "step": 4641
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.241227644645234,
+      "learning_rate": 5.525097738495204e-06,
+      "loss": 0.6851,
+      "step": 4642
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7727776823270593,
+      "learning_rate": 5.523423600876816e-06,
+      "loss": 0.5747,
+      "step": 4643
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9624937546211163,
+      "learning_rate": 5.521749403925379e-06,
+      "loss": 0.6018,
+      "step": 4644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8723540282901674,
+      "learning_rate": 5.520075147830674e-06,
+      "loss": 0.6188,
+      "step": 4645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8822938715140052,
+      "learning_rate": 5.518400832782485e-06,
+      "loss": 0.5811,
+      "step": 4646
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9055839171728313,
+      "learning_rate": 5.516726458970608e-06,
+      "loss": 0.5566,
+      "step": 4647
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.805471104267127,
+      "learning_rate": 5.515052026584842e-06,
+      "loss": 0.6422,
+      "step": 4648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8946969960370972,
+      "learning_rate": 5.513377535814992e-06,
+      "loss": 0.7232,
+      "step": 4649
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.996885687942003,
+      "learning_rate": 5.511702986850873e-06,
+      "loss": 0.5903,
+      "step": 4650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9418199161319338,
+      "learning_rate": 5.510028379882304e-06,
+      "loss": 0.6816,
+      "step": 4651
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9209458855987755,
+      "learning_rate": 5.508353715099111e-06,
+      "loss": 0.7325,
+      "step": 4652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.737261300068071,
+      "learning_rate": 5.506678992691126e-06,
+      "loss": 0.578,
+      "step": 4653
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.887091236693906,
+      "learning_rate": 5.50500421284819e-06,
+      "loss": 0.648,
+      "step": 4654
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9141217370293513,
+      "learning_rate": 5.503329375760148e-06,
+      "loss": 0.5891,
+      "step": 4655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.240905822821042,
+      "learning_rate": 5.5016544816168515e-06,
+      "loss": 0.6495,
+      "step": 4656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.008628318504785,
+      "learning_rate": 5.49997953060816e-06,
+      "loss": 0.6581,
+      "step": 4657
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.848748059234952,
+      "learning_rate": 5.498304522923941e-06,
+      "loss": 0.6991,
+      "step": 4658
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9226179328864823,
+      "learning_rate": 5.4966294587540626e-06,
+      "loss": 0.62,
+      "step": 4659
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.731765557271535,
+      "learning_rate": 5.494954338288404e-06,
+      "loss": 0.5619,
+      "step": 4660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9932290535373656,
+      "learning_rate": 5.493279161716851e-06,
+      "loss": 0.6834,
+      "step": 4661
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7461404202900574,
+      "learning_rate": 5.491603929229293e-06,
+      "loss": 0.6477,
+      "step": 4662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9588893964565277,
+      "learning_rate": 5.4899286410156275e-06,
+      "loss": 0.7275,
+      "step": 4663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.888923777184811,
+      "learning_rate": 5.488253297265757e-06,
+      "loss": 0.6459,
+      "step": 4664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.990188644636464,
+      "learning_rate": 5.486577898169595e-06,
+      "loss": 0.6675,
+      "step": 4665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8803794705216603,
+      "learning_rate": 5.484902443917053e-06,
+      "loss": 0.6864,
+      "step": 4666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0001158625188173,
+      "learning_rate": 5.483226934698055e-06,
+      "loss": 0.6569,
+      "step": 4667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8830150639111904,
+      "learning_rate": 5.4815513707025306e-06,
+      "loss": 0.6279,
+      "step": 4668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.023732196650425,
+      "learning_rate": 5.479875752120414e-06,
+      "loss": 0.6609,
+      "step": 4669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0625613008373773,
+      "learning_rate": 5.478200079141644e-06,
+      "loss": 0.643,
+      "step": 4670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9591030994917111,
+      "learning_rate": 5.47652435195617e-06,
+      "loss": 0.6557,
+      "step": 4671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7978853519603617,
+      "learning_rate": 5.4748485707539435e-06,
+      "loss": 0.5881,
+      "step": 4672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9473749824242983,
+      "learning_rate": 5.473172735724927e-06,
+      "loss": 0.6415,
+      "step": 4673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.058316827182397,
+      "learning_rate": 5.471496847059082e-06,
+      "loss": 0.7541,
+      "step": 4674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0498700867540416,
+      "learning_rate": 5.469820904946383e-06,
+      "loss": 0.6762,
+      "step": 4675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9141503530490744,
+      "learning_rate": 5.468144909576807e-06,
+      "loss": 0.5577,
+      "step": 4676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.797991597019518,
+      "learning_rate": 5.466468861140337e-06,
+      "loss": 0.5654,
+      "step": 4677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8613963899363541,
+      "learning_rate": 5.464792759826962e-06,
+      "loss": 0.6946,
+      "step": 4678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8335011635462246,
+      "learning_rate": 5.463116605826678e-06,
+      "loss": 0.678,
+      "step": 4679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0050714504137965,
+      "learning_rate": 5.4614403993294895e-06,
+      "loss": 0.6334,
+      "step": 4680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.032874636356343,
+      "learning_rate": 5.4597641405254e-06,
+      "loss": 0.6813,
+      "step": 4681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.790986980802673,
+      "learning_rate": 5.458087829604423e-06,
+      "loss": 0.6778,
+      "step": 4682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.751145689736165,
+      "learning_rate": 5.456411466756584e-06,
+      "loss": 0.5752,
+      "step": 4683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.697096974282548,
+      "learning_rate": 5.4547350521719e-06,
+      "loss": 0.567,
+      "step": 4684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1498813486938393,
+      "learning_rate": 5.453058586040406e-06,
+      "loss": 0.6605,
+      "step": 4685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9907803657150827,
+      "learning_rate": 5.45138206855214e-06,
+      "loss": 0.6133,
+      "step": 4686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1637345099530534,
+      "learning_rate": 5.4497054998971445e-06,
+      "loss": 0.6557,
+      "step": 4687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8252761461235034,
+      "learning_rate": 5.448028880265467e-06,
+      "loss": 0.6198,
+      "step": 4688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.82343633197036,
+      "learning_rate": 5.446352209847161e-06,
+      "loss": 0.7194,
+      "step": 4689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0614527120208,
+      "learning_rate": 5.444675488832288e-06,
+      "loss": 0.6723,
+      "step": 4690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1820987619351175,
+      "learning_rate": 5.442998717410916e-06,
+      "loss": 0.778,
+      "step": 4691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0025828276893134,
+      "learning_rate": 5.441321895773112e-06,
+      "loss": 0.6472,
+      "step": 4692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9803099147219958,
+      "learning_rate": 5.439645024108956e-06,
+      "loss": 0.6677,
+      "step": 4693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8356847480376939,
+      "learning_rate": 5.4379681026085305e-06,
+      "loss": 0.5864,
+      "step": 4694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9702510121981012,
+      "learning_rate": 5.436291131461926e-06,
+      "loss": 0.643,
+      "step": 4695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.259676272993004,
+      "learning_rate": 5.434614110859233e-06,
+      "loss": 0.7192,
+      "step": 4696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8027623450713939,
+      "learning_rate": 5.432937040990553e-06,
+      "loss": 0.6741,
+      "step": 4697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2863564801237377,
+      "learning_rate": 5.431259922045995e-06,
+      "loss": 0.7341,
+      "step": 4698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8466916983922386,
+      "learning_rate": 5.429582754215664e-06,
+      "loss": 0.5891,
+      "step": 4699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7862194556644164,
+      "learning_rate": 5.427905537689679e-06,
+      "loss": 0.638,
+      "step": 4700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8250276402626973,
+      "learning_rate": 5.426228272658163e-06,
+      "loss": 0.6587,
+      "step": 4701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9226029242682035,
+      "learning_rate": 5.424550959311244e-06,
+      "loss": 0.6189,
+      "step": 4702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.778075048144127,
+      "learning_rate": 5.422873597839052e-06,
+      "loss": 0.5834,
+      "step": 4703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8704703373919398,
+      "learning_rate": 5.4211961884317285e-06,
+      "loss": 0.609,
+      "step": 4704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9872381122558112,
+      "learning_rate": 5.4195187312794165e-06,
+      "loss": 0.6613,
+      "step": 4705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2208206315817356,
+      "learning_rate": 5.417841226572263e-06,
+      "loss": 0.6202,
+      "step": 4706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8030691134210728,
+      "learning_rate": 5.416163674500429e-06,
+      "loss": 0.5852,
+      "step": 4707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9087645300174885,
+      "learning_rate": 5.4144860752540675e-06,
+      "loss": 0.5521,
+      "step": 4708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.003121434948416,
+      "learning_rate": 5.412808429023346e-06,
+      "loss": 0.6437,
+      "step": 4709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.805314904231961,
+      "learning_rate": 5.4111307359984375e-06,
+      "loss": 0.6137,
+      "step": 4710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8556671218808498,
+      "learning_rate": 5.409452996369517e-06,
+      "loss": 0.5666,
+      "step": 4711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.913852383240618,
+      "learning_rate": 5.407775210326765e-06,
+      "loss": 0.593,
+      "step": 4712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8579997869434959,
+      "learning_rate": 5.40609737806037e-06,
+      "loss": 0.6143,
+      "step": 4713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.955230915460718,
+      "learning_rate": 5.404419499760521e-06,
+      "loss": 0.6543,
+      "step": 4714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.918928720060036,
+      "learning_rate": 5.402741575617417e-06,
+      "loss": 0.5767,
+      "step": 4715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9450222578852545,
+      "learning_rate": 5.401063605821259e-06,
+      "loss": 0.599,
+      "step": 4716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.830693102223202,
+      "learning_rate": 5.399385590562257e-06,
+      "loss": 0.6338,
+      "step": 4717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.86583491311485,
+      "learning_rate": 5.397707530030621e-06,
+      "loss": 0.5858,
+      "step": 4718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9159187468657468,
+      "learning_rate": 5.3960294244165705e-06,
+      "loss": 0.66,
+      "step": 4719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9568612561515344,
+      "learning_rate": 5.394351273910327e-06,
+      "loss": 0.6735,
+      "step": 4720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8762483550864508,
+      "learning_rate": 5.392673078702118e-06,
+      "loss": 0.6049,
+      "step": 4721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7436999227726706,
+      "learning_rate": 5.390994838982178e-06,
+      "loss": 0.5753,
+      "step": 4722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.012137021393326,
+      "learning_rate": 5.3893165549407435e-06,
+      "loss": 0.6212,
+      "step": 4723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8836151547134943,
+      "learning_rate": 5.38763822676806e-06,
+      "loss": 0.6909,
+      "step": 4724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.804096332876596,
+      "learning_rate": 5.385959854654374e-06,
+      "loss": 0.5916,
+      "step": 4725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.737330588597606,
+      "learning_rate": 5.384281438789937e-06,
+      "loss": 0.6128,
+      "step": 4726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8569231685201164,
+      "learning_rate": 5.382602979365009e-06,
+      "loss": 0.6701,
+      "step": 4727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7276227842031793,
+      "learning_rate": 5.380924476569854e-06,
+      "loss": 0.5959,
+      "step": 4728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8705463387699754,
+      "learning_rate": 5.379245930594738e-06,
+      "loss": 0.6266,
+      "step": 4729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9820668249263957,
+      "learning_rate": 5.3775673416299325e-06,
+      "loss": 0.7306,
+      "step": 4730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1991728959408166,
+      "learning_rate": 5.375888709865718e-06,
+      "loss": 0.6499,
+      "step": 4731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9049944223482733,
+      "learning_rate": 5.374210035492375e-06,
+      "loss": 0.6072,
+      "step": 4732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7630256763054708,
+      "learning_rate": 5.372531318700192e-06,
+      "loss": 0.5893,
+      "step": 4733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8797816070120186,
+      "learning_rate": 5.370852559679461e-06,
+      "loss": 0.6232,
+      "step": 4734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9335377315595026,
+      "learning_rate": 5.36917375862048e-06,
+      "loss": 0.6255,
+      "step": 4735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8935577231729566,
+      "learning_rate": 5.367494915713547e-06,
+      "loss": 0.6538,
+      "step": 4736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.159238536605417,
+      "learning_rate": 5.365816031148971e-06,
+      "loss": 0.6783,
+      "step": 4737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2691128833223444,
+      "learning_rate": 5.364137105117062e-06,
+      "loss": 0.6657,
+      "step": 4738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9293045676431337,
+      "learning_rate": 5.362458137808139e-06,
+      "loss": 0.6729,
+      "step": 4739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9461395972368574,
+      "learning_rate": 5.360779129412519e-06,
+      "loss": 0.5758,
+      "step": 4740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7469143360040644,
+      "learning_rate": 5.359100080120527e-06,
+      "loss": 0.5882,
+      "step": 4741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.525512152522639,
+      "learning_rate": 5.357420990122495e-06,
+      "loss": 0.5615,
+      "step": 4742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8314068903897658,
+      "learning_rate": 5.355741859608756e-06,
+      "loss": 0.6769,
+      "step": 4743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8789868416667825,
+      "learning_rate": 5.35406268876965e-06,
+      "loss": 0.5629,
+      "step": 4744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9337216447416474,
+      "learning_rate": 5.352383477795522e-06,
+      "loss": 0.6372,
+      "step": 4745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.991741251184157,
+      "learning_rate": 5.3507042268767165e-06,
+      "loss": 0.7707,
+      "step": 4746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7991416686125543,
+      "learning_rate": 5.3490249362035875e-06,
+      "loss": 0.6465,
+      "step": 4747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2012059216577566,
+      "learning_rate": 5.347345605966493e-06,
+      "loss": 0.7182,
+      "step": 4748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9724821199819178,
+      "learning_rate": 5.345666236355794e-06,
+      "loss": 0.6756,
+      "step": 4749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8347424754598582,
+      "learning_rate": 5.343986827561859e-06,
+      "loss": 0.6407,
+      "step": 4750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8192154867135653,
+      "learning_rate": 5.342307379775053e-06,
+      "loss": 0.6398,
+      "step": 4751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9939590161819776,
+      "learning_rate": 5.340627893185757e-06,
+      "loss": 0.5802,
+      "step": 4752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1746888356632277,
+      "learning_rate": 5.338948367984347e-06,
+      "loss": 0.7248,
+      "step": 4753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7532309609679348,
+      "learning_rate": 5.337268804361208e-06,
+      "loss": 0.5655,
+      "step": 4754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.727385613972203,
+      "learning_rate": 5.335589202506727e-06,
+      "loss": 0.5939,
+      "step": 4755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0887581011841823,
+      "learning_rate": 5.3339095626112965e-06,
+      "loss": 0.6019,
+      "step": 4756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8319707154087208,
+      "learning_rate": 5.332229884865316e-06,
+      "loss": 0.6287,
+      "step": 4757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.935043766822851,
+      "learning_rate": 5.3305501694591836e-06,
+      "loss": 0.6483,
+      "step": 4758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8084928100607975,
+      "learning_rate": 5.3288704165833035e-06,
+      "loss": 0.5682,
+      "step": 4759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.884803293181149,
+      "learning_rate": 5.327190626428089e-06,
+      "loss": 0.6029,
+      "step": 4760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.932522766945742,
+      "learning_rate": 5.325510799183953e-06,
+      "loss": 0.6771,
+      "step": 4761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9866449112920452,
+      "learning_rate": 5.32383093504131e-06,
+      "loss": 0.5524,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8167555076320083,
+      "learning_rate": 5.3221510341905855e-06,
+      "loss": 0.6437,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8329283982720728,
+      "learning_rate": 5.320471096822206e-06,
+      "loss": 0.5299,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9138426461164462,
+      "learning_rate": 5.318791123126601e-06,
+      "loss": 0.62,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1010868001824097,
+      "learning_rate": 5.3171111132942045e-06,
+      "loss": 0.658,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9813575064514743,
+      "learning_rate": 5.315431067515456e-06,
+      "loss": 0.7796,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9144527346597267,
+      "learning_rate": 5.313750985980799e-06,
+      "loss": 0.6166,
+      "step": 4768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9394591482862733,
+      "learning_rate": 5.312070868880678e-06,
+      "loss": 0.6915,
+      "step": 4769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7787970809473537,
+      "learning_rate": 5.310390716405546e-06,
+      "loss": 0.5507,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9257353515354803,
+      "learning_rate": 5.308710528745856e-06,
+      "loss": 0.5736,
+      "step": 4771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.899196278729454,
+      "learning_rate": 5.3070303060920706e-06,
+      "loss": 0.6948,
+      "step": 4772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8548166587030892,
+      "learning_rate": 5.305350048634648e-06,
+      "loss": 0.6213,
+      "step": 4773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.72877940583896,
+      "learning_rate": 5.303669756564057e-06,
+      "loss": 0.5737,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.81760933387422,
+      "learning_rate": 5.301989430070767e-06,
+      "loss": 0.5618,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9269910883598707,
+      "learning_rate": 5.300309069345257e-06,
+      "loss": 0.6532,
+      "step": 4776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9463546960698885,
+      "learning_rate": 5.298628674578e-06,
+      "loss": 0.6358,
+      "step": 4777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8442409478011446,
+      "learning_rate": 5.296948245959481e-06,
+      "loss": 0.6798,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6872881787647978,
+      "learning_rate": 5.295267783680186e-06,
+      "loss": 0.615,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.020457824524539,
+      "learning_rate": 5.293587287930605e-06,
+      "loss": 0.7365,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9419792797429891,
+      "learning_rate": 5.291906758901231e-06,
+      "loss": 0.6422,
+      "step": 4781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9961829611233421,
+      "learning_rate": 5.290226196782562e-06,
+      "loss": 0.6491,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8346912090258714,
+      "learning_rate": 5.2885456017651e-06,
+      "loss": 0.6669,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9113822633234845,
+      "learning_rate": 5.286864974039349e-06,
+      "loss": 0.6546,
+      "step": 4784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9088972328489655,
+      "learning_rate": 5.285184313795818e-06,
+      "loss": 0.6967,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9770384150188363,
+      "learning_rate": 5.28350362122502e-06,
+      "loss": 0.6551,
+      "step": 4786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.983698674791684,
+      "learning_rate": 5.281822896517471e-06,
+      "loss": 0.6672,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9723552036785195,
+      "learning_rate": 5.280142139863689e-06,
+      "loss": 0.6123,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7502841190125016,
+      "learning_rate": 5.278461351454199e-06,
+      "loss": 0.6269,
+      "step": 4789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8380073595702189,
+      "learning_rate": 5.276780531479528e-06,
+      "loss": 0.629,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9008197624107583,
+      "learning_rate": 5.275099680130207e-06,
+      "loss": 0.686,
+      "step": 4791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6690452997820755,
+      "learning_rate": 5.273418797596769e-06,
+      "loss": 0.541,
+      "step": 4792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7704221672792175,
+      "learning_rate": 5.271737884069751e-06,
+      "loss": 0.7384,
+      "step": 4793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8289870533935189,
+      "learning_rate": 5.270056939739695e-06,
+      "loss": 0.7028,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9957521734241734,
+      "learning_rate": 5.268375964797147e-06,
+      "loss": 0.8097,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8245717485844726,
+      "learning_rate": 5.266694959432651e-06,
+      "loss": 0.5797,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6513891472143982,
+      "learning_rate": 5.265013923836763e-06,
+      "loss": 0.5582,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8439248589354222,
+      "learning_rate": 5.263332858200037e-06,
+      "loss": 0.5971,
+      "step": 4798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7693969050117528,
+      "learning_rate": 5.261651762713029e-06,
+      "loss": 0.6627,
+      "step": 4799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8799400076854698,
+      "learning_rate": 5.259970637566303e-06,
+      "loss": 0.6186,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.810317566636923,
+      "learning_rate": 5.2582894829504225e-06,
+      "loss": 0.6289,
+      "step": 4801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1064888897228604,
+      "learning_rate": 5.256608299055959e-06,
+      "loss": 0.7368,
+      "step": 4802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7934611876160385,
+      "learning_rate": 5.254927086073481e-06,
+      "loss": 0.5756,
+      "step": 4803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.918347419758823,
+      "learning_rate": 5.253245844193564e-06,
+      "loss": 0.6471,
+      "step": 4804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.921103053022868,
+      "learning_rate": 5.251564573606789e-06,
+      "loss": 0.6237,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.881055921289424,
+      "learning_rate": 5.249883274503734e-06,
+      "loss": 0.655,
+      "step": 4806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1128203412785695,
+      "learning_rate": 5.248201947074986e-06,
+      "loss": 0.6767,
+      "step": 4807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8473802354895168,
+      "learning_rate": 5.246520591511133e-06,
+      "loss": 0.7472,
+      "step": 4808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.939482551753464,
+      "learning_rate": 5.244839208002766e-06,
+      "loss": 0.7308,
+      "step": 4809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.924260276947548,
+      "learning_rate": 5.243157796740478e-06,
+      "loss": 0.6115,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0874541041869157,
+      "learning_rate": 5.241476357914869e-06,
+      "loss": 0.7236,
+      "step": 4811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.837733357562263,
+      "learning_rate": 5.239794891716538e-06,
+      "loss": 0.6035,
+      "step": 4812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8910288208809416,
+      "learning_rate": 5.238113398336089e-06,
+      "loss": 0.6271,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0685585225934435,
+      "learning_rate": 5.236431877964129e-06,
+      "loss": 0.7462,
+      "step": 4814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.901940101007261,
+      "learning_rate": 5.234750330791268e-06,
+      "loss": 0.6331,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7745538373385725,
+      "learning_rate": 5.23306875700812e-06,
+      "loss": 0.5965,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.914938178129142,
+      "learning_rate": 5.231387156805299e-06,
+      "loss": 0.6152,
+      "step": 4817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9602214851083848,
+      "learning_rate": 5.229705530373424e-06,
+      "loss": 0.619,
+      "step": 4818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8208173429792096,
+      "learning_rate": 5.228023877903119e-06,
+      "loss": 0.5953,
+      "step": 4819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8886106395780093,
+      "learning_rate": 5.22634219958501e-06,
+      "loss": 0.6042,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8052470224663513,
+      "learning_rate": 5.224660495609719e-06,
+      "loss": 0.5851,
+      "step": 4821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8864420425653845,
+      "learning_rate": 5.222978766167881e-06,
+      "loss": 0.6513,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8881947704523596,
+      "learning_rate": 5.221297011450129e-06,
+      "loss": 0.6889,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7349782274775427,
+      "learning_rate": 5.219615231647102e-06,
+      "loss": 0.7054,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.876831428970342,
+      "learning_rate": 5.2179334269494345e-06,
+      "loss": 0.698,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7635223691101098,
+      "learning_rate": 5.21625159754777e-06,
+      "loss": 0.6371,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7059481236334688,
+      "learning_rate": 5.214569743632756e-06,
+      "loss": 0.5452,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5553605927294225,
+      "learning_rate": 5.212887865395038e-06,
+      "loss": 0.566,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9616001578735676,
+      "learning_rate": 5.211205963025268e-06,
+      "loss": 0.7019,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8648642120464367,
+      "learning_rate": 5.209524036714096e-06,
+      "loss": 0.67,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0163559085077853,
+      "learning_rate": 5.207842086652183e-06,
+      "loss": 0.6602,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.035789772992233,
+      "learning_rate": 5.206160113030182e-06,
+      "loss": 0.6346,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9600216866470923,
+      "learning_rate": 5.204478116038758e-06,
+      "loss": 0.6541,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7741794956531685,
+      "learning_rate": 5.202796095868574e-06,
+      "loss": 0.5933,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7335196713309575,
+      "learning_rate": 5.201114052710299e-06,
+      "loss": 0.6089,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9376194087329808,
+      "learning_rate": 5.1994319867545974e-06,
+      "loss": 0.7502,
+      "step": 4836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8325179415301216,
+      "learning_rate": 5.197749898192144e-06,
+      "loss": 0.6688,
+      "step": 4837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9692153336399745,
+      "learning_rate": 5.196067787213611e-06,
+      "loss": 0.6632,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.800237625965685,
+      "learning_rate": 5.1943856540096795e-06,
+      "loss": 0.648,
+      "step": 4839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.795377760515686,
+      "learning_rate": 5.1927034987710245e-06,
+      "loss": 0.6032,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.895454260117348,
+      "learning_rate": 5.19102132168833e-06,
+      "loss": 0.6731,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.039687589229877,
+      "learning_rate": 5.189339122952281e-06,
+      "loss": 0.6824,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8781273228678457,
+      "learning_rate": 5.18765690275356e-06,
+      "loss": 0.6945,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8013673164406765,
+      "learning_rate": 5.185974661282862e-06,
+      "loss": 0.6765,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.98407236347405,
+      "learning_rate": 5.184292398730876e-06,
+      "loss": 0.7248,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0038873123117757,
+      "learning_rate": 5.182610115288296e-06,
+      "loss": 0.6836,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.660770204955728,
+      "learning_rate": 5.180927811145818e-06,
+      "loss": 0.56,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.17998760571231,
+      "learning_rate": 5.179245486494141e-06,
+      "loss": 0.7484,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9112108041835019,
+      "learning_rate": 5.177563141523967e-06,
+      "loss": 0.7095,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.160618119718771,
+      "learning_rate": 5.175880776425999e-06,
+      "loss": 0.7162,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.85331993962675,
+      "learning_rate": 5.174198391390942e-06,
+      "loss": 0.6401,
+      "step": 4851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.939094194390486,
+      "learning_rate": 5.172515986609504e-06,
+      "loss": 0.6593,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9149904541209817,
+      "learning_rate": 5.170833562272398e-06,
+      "loss": 0.6141,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9222041139959078,
+      "learning_rate": 5.169151118570332e-06,
+      "loss": 0.6764,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7669182377758037,
+      "learning_rate": 5.167468655694022e-06,
+      "loss": 0.538,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1867857975643434,
+      "learning_rate": 5.165786173834187e-06,
+      "loss": 0.656,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6534862157740111,
+      "learning_rate": 5.164103673181544e-06,
+      "loss": 0.6398,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1437201670617037,
+      "learning_rate": 5.162421153926814e-06,
+      "loss": 0.7164,
+      "step": 4858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6553379755707762,
+      "learning_rate": 5.16073861626072e-06,
+      "loss": 0.6298,
+      "step": 4859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8648222159713144,
+      "learning_rate": 5.1590560603739885e-06,
+      "loss": 0.697,
+      "step": 4860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.961783728362329,
+      "learning_rate": 5.157373486457346e-06,
+      "loss": 0.6445,
+      "step": 4861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7476062727074282,
+      "learning_rate": 5.15569089470152e-06,
+      "loss": 0.4982,
+      "step": 4862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0169770654335872,
+      "learning_rate": 5.1540082852972455e-06,
+      "loss": 0.6628,
+      "step": 4863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7595848868953443,
+      "learning_rate": 5.152325658435254e-06,
+      "loss": 0.6338,
+      "step": 4864
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6235278140692069,
+      "learning_rate": 5.15064301430628e-06,
+      "loss": 0.5293,
+      "step": 4865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8762127911073818,
+      "learning_rate": 5.148960353101063e-06,
+      "loss": 0.5885,
+      "step": 4866
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.155003322145285,
+      "learning_rate": 5.147277675010339e-06,
+      "loss": 0.749,
+      "step": 4867
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9059516874727496,
+      "learning_rate": 5.145594980224853e-06,
+      "loss": 0.6446,
+      "step": 4868
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7276881296721822,
+      "learning_rate": 5.143912268935345e-06,
+      "loss": 0.5343,
+      "step": 4869
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.069942567716001,
+      "learning_rate": 5.14222954133256e-06,
+      "loss": 0.6633,
+      "step": 4870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0174980514678205,
+      "learning_rate": 5.140546797607248e-06,
+      "loss": 0.6722,
+      "step": 4871
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9082251784968078,
+      "learning_rate": 5.138864037950155e-06,
+      "loss": 0.5613,
+      "step": 4872
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8732227894779176,
+      "learning_rate": 5.137181262552031e-06,
+      "loss": 0.6477,
+      "step": 4873
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6832285896137942,
+      "learning_rate": 5.135498471603629e-06,
+      "loss": 0.6062,
+      "step": 4874
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6344271952482194,
+      "learning_rate": 5.133815665295704e-06,
+      "loss": 0.676,
+      "step": 4875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7922029487358937,
+      "learning_rate": 5.13213284381901e-06,
+      "loss": 0.571,
+      "step": 4876
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8548110893994498,
+      "learning_rate": 5.1304500073643045e-06,
+      "loss": 0.5944,
+      "step": 4877
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.780740593676751,
+      "learning_rate": 5.128767156122347e-06,
+      "loss": 0.5777,
+      "step": 4878
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.247295270187268,
+      "learning_rate": 5.1270842902839e-06,
+      "loss": 0.7154,
+      "step": 4879
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9584193373117975,
+      "learning_rate": 5.125401410039723e-06,
+      "loss": 0.6757,
+      "step": 4880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.095672743950007,
+      "learning_rate": 5.123718515580581e-06,
+      "loss": 0.6605,
+      "step": 4881
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8048789293075735,
+      "learning_rate": 5.1220356070972414e-06,
+      "loss": 0.7379,
+      "step": 4882
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9146554643863576,
+      "learning_rate": 5.120352684780469e-06,
+      "loss": 0.5925,
+      "step": 4883
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1568243392901105,
+      "learning_rate": 5.118669748821034e-06,
+      "loss": 0.7402,
+      "step": 4884
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9630846444451566,
+      "learning_rate": 5.116986799409708e-06,
+      "loss": 0.6983,
+      "step": 4885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.939417972921571,
+      "learning_rate": 5.11530383673726e-06,
+      "loss": 0.6694,
+      "step": 4886
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.15676437427674,
+      "learning_rate": 5.1136208609944644e-06,
+      "loss": 0.6987,
+      "step": 4887
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9490400227157634,
+      "learning_rate": 5.111937872372097e-06,
+      "loss": 0.6834,
+      "step": 4888
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.89411260769405,
+      "learning_rate": 5.110254871060933e-06,
+      "loss": 0.6891,
+      "step": 4889
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8278879530072518,
+      "learning_rate": 5.108571857251754e-06,
+      "loss": 0.5864,
+      "step": 4890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.959630964015898,
+      "learning_rate": 5.106888831135334e-06,
+      "loss": 0.5718,
+      "step": 4891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.887053393441401,
+      "learning_rate": 5.105205792902456e-06,
+      "loss": 0.5991,
+      "step": 4892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8419676864917576,
+      "learning_rate": 5.103522742743901e-06,
+      "loss": 0.6203,
+      "step": 4893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9813179222203285,
+      "learning_rate": 5.101839680850454e-06,
+      "loss": 0.7061,
+      "step": 4894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9051838482841017,
+      "learning_rate": 5.100156607412899e-06,
+      "loss": 0.663,
+      "step": 4895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7946750412220063,
+      "learning_rate": 5.09847352262202e-06,
+      "loss": 0.6395,
+      "step": 4896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.692629416537992,
+      "learning_rate": 5.096790426668608e-06,
+      "loss": 0.6797,
+      "step": 4897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6992150985739385,
+      "learning_rate": 5.095107319743449e-06,
+      "loss": 0.6609,
+      "step": 4898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0328131001769885,
+      "learning_rate": 5.093424202037333e-06,
+      "loss": 0.6063,
+      "step": 4899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8965582288309775,
+      "learning_rate": 5.09174107374105e-06,
+      "loss": 0.7277,
+      "step": 4900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9789015178424736,
+      "learning_rate": 5.090057935045395e-06,
+      "loss": 0.7337,
+      "step": 4901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0892636946117795,
+      "learning_rate": 5.088374786141159e-06,
+      "loss": 0.5095,
+      "step": 4902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9916605863411267,
+      "learning_rate": 5.086691627219137e-06,
+      "loss": 0.6281,
+      "step": 4903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8199559774598433,
+      "learning_rate": 5.085008458470126e-06,
+      "loss": 0.6696,
+      "step": 4904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2905345302233284,
+      "learning_rate": 5.0833252800849205e-06,
+      "loss": 0.723,
+      "step": 4905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8128746657582677,
+      "learning_rate": 5.0816420922543195e-06,
+      "loss": 0.6756,
+      "step": 4906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6940473207507076,
+      "learning_rate": 5.079958895169122e-06,
+      "loss": 0.6189,
+      "step": 4907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.099894774247856,
+      "learning_rate": 5.078275689020129e-06,
+      "loss": 0.6214,
+      "step": 4908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.923333567399948,
+      "learning_rate": 5.076592473998141e-06,
+      "loss": 0.5972,
+      "step": 4909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.742185055247871,
+      "learning_rate": 5.0749092502939575e-06,
+      "loss": 0.6876,
+      "step": 4910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7921342848541095,
+      "learning_rate": 5.073226018098385e-06,
+      "loss": 0.5127,
+      "step": 4911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9020236365196561,
+      "learning_rate": 5.071542777602225e-06,
+      "loss": 0.5513,
+      "step": 4912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8385169703695514,
+      "learning_rate": 5.0698595289962845e-06,
+      "loss": 0.6028,
+      "step": 4913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.868450962690883,
+      "learning_rate": 5.068176272471368e-06,
+      "loss": 0.676,
+      "step": 4914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8850082304527058,
+      "learning_rate": 5.066493008218282e-06,
+      "loss": 0.6854,
+      "step": 4915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9577619267509585,
+      "learning_rate": 5.064809736427835e-06,
+      "loss": 0.6569,
+      "step": 4916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9076916716005377,
+      "learning_rate": 5.0631264572908334e-06,
+      "loss": 0.5663,
+      "step": 4917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2592632992616415,
+      "learning_rate": 5.0614431709980895e-06,
+      "loss": 0.669,
+      "step": 4918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6810611347190536,
+      "learning_rate": 5.059759877740411e-06,
+      "loss": 0.5439,
+      "step": 4919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2498791112061545,
+      "learning_rate": 5.058076577708611e-06,
+      "loss": 0.7829,
+      "step": 4920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7291877715984016,
+      "learning_rate": 5.056393271093498e-06,
+      "loss": 0.6395,
+      "step": 4921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6289988487754408,
+      "learning_rate": 5.0547099580858874e-06,
+      "loss": 0.5376,
+      "step": 4922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.213277225739156,
+      "learning_rate": 5.053026638876591e-06,
+      "loss": 0.7141,
+      "step": 4923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.141769948061226,
+      "learning_rate": 5.0513433136564236e-06,
+      "loss": 0.6407,
+      "step": 4924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.804589027364479,
+      "learning_rate": 5.049659982616199e-06,
+      "loss": 0.5956,
+      "step": 4925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9554108740288227,
+      "learning_rate": 5.047976645946732e-06,
+      "loss": 0.6802,
+      "step": 4926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6593553009678443,
+      "learning_rate": 5.046293303838838e-06,
+      "loss": 0.6175,
+      "step": 4927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.25531640266805,
+      "learning_rate": 5.044609956483335e-06,
+      "loss": 0.7248,
+      "step": 4928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8342676675179805,
+      "learning_rate": 5.042926604071039e-06,
+      "loss": 0.567,
+      "step": 4929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.983517953815852,
+      "learning_rate": 5.0412432467927674e-06,
+      "loss": 0.728,
+      "step": 4930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0957696190309636,
+      "learning_rate": 5.039559884839339e-06,
+      "loss": 0.6731,
+      "step": 4931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7798860206636986,
+      "learning_rate": 5.037876518401572e-06,
+      "loss": 0.6234,
+      "step": 4932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8782106374167673,
+      "learning_rate": 5.036193147670286e-06,
+      "loss": 0.5433,
+      "step": 4933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.965876390350256,
+      "learning_rate": 5.0345097728363e-06,
+      "loss": 0.7375,
+      "step": 4934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6555281848960024,
+      "learning_rate": 5.032826394090435e-06,
+      "loss": 0.5369,
+      "step": 4935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0343517692919293,
+      "learning_rate": 5.031143011623511e-06,
+      "loss": 0.6064,
+      "step": 4936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7335127141852276,
+      "learning_rate": 5.02945962562635e-06,
+      "loss": 0.6594,
+      "step": 4937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7992246411216586,
+      "learning_rate": 5.027776236289772e-06,
+      "loss": 0.7669,
+      "step": 4938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7969757038163299,
+      "learning_rate": 5.026092843804599e-06,
+      "loss": 0.5572,
+      "step": 4939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2179435808357195,
+      "learning_rate": 5.024409448361653e-06,
+      "loss": 0.6424,
+      "step": 4940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7330945557046764,
+      "learning_rate": 5.022726050151756e-06,
+      "loss": 0.6363,
+      "step": 4941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6773497431028328,
+      "learning_rate": 5.0210426493657335e-06,
+      "loss": 0.6059,
+      "step": 4942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7029285789828024,
+      "learning_rate": 5.019359246194406e-06,
+      "loss": 0.5937,
+      "step": 4943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7350035183219865,
+      "learning_rate": 5.017675840828597e-06,
+      "loss": 0.5811,
+      "step": 4944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8633238467146478,
+      "learning_rate": 5.0159924334591316e-06,
+      "loss": 0.6168,
+      "step": 4945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0472909051472272,
+      "learning_rate": 5.014309024276833e-06,
+      "loss": 0.6218,
+      "step": 4946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.889431669608345,
+      "learning_rate": 5.012625613472525e-06,
+      "loss": 0.6791,
+      "step": 4947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7828852466791096,
+      "learning_rate": 5.010942201237031e-06,
+      "loss": 0.5815,
+      "step": 4948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7605094733233677,
+      "learning_rate": 5.009258787761178e-06,
+      "loss": 0.6806,
+      "step": 4949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.106865906855424,
+      "learning_rate": 5.007575373235786e-06,
+      "loss": 0.5734,
+      "step": 4950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.069021850878164,
+      "learning_rate": 5.005891957851683e-06,
+      "loss": 0.6051,
+      "step": 4951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.953305074899751,
+      "learning_rate": 5.004208541799693e-06,
+      "loss": 0.6234,
+      "step": 4952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6461163901567595,
+      "learning_rate": 5.002525125270641e-06,
+      "loss": 0.7198,
+      "step": 4953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9211725095625716,
+      "learning_rate": 5.000841708455351e-06,
+      "loss": 0.5863,
+      "step": 4954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8963853637765036,
+      "learning_rate": 4.99915829154465e-06,
+      "loss": 0.6632,
+      "step": 4955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9670876231350454,
+      "learning_rate": 4.997474874729361e-06,
+      "loss": 0.7981,
+      "step": 4956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7891379226397384,
+      "learning_rate": 4.995791458200309e-06,
+      "loss": 0.5285,
+      "step": 4957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.176966656664999,
+      "learning_rate": 4.994108042148318e-06,
+      "loss": 0.702,
+      "step": 4958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9299774614223446,
+      "learning_rate": 4.992424626764216e-06,
+      "loss": 0.6392,
+      "step": 4959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.699908367772429,
+      "learning_rate": 4.990741212238825e-06,
+      "loss": 0.5981,
+      "step": 4960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6255346707577691,
+      "learning_rate": 4.98905779876297e-06,
+      "loss": 0.5514,
+      "step": 4961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8368275633671942,
+      "learning_rate": 4.987374386527478e-06,
+      "loss": 0.6648,
+      "step": 4962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8212502460777635,
+      "learning_rate": 4.985690975723168e-06,
+      "loss": 0.6234,
+      "step": 4963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5672021052986176,
+      "learning_rate": 4.984007566540869e-06,
+      "loss": 0.5612,
+      "step": 4964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8762480041490517,
+      "learning_rate": 4.982324159171404e-06,
+      "loss": 0.6543,
+      "step": 4965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.867185684724656,
+      "learning_rate": 4.980640753805595e-06,
+      "loss": 0.6743,
+      "step": 4966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.979608168978776,
+      "learning_rate": 4.978957350634267e-06,
+      "loss": 0.6139,
+      "step": 4967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0889995333117413,
+      "learning_rate": 4.977273949848244e-06,
+      "loss": 0.6246,
+      "step": 4968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0304842602058533,
+      "learning_rate": 4.975590551638348e-06,
+      "loss": 0.6666,
+      "step": 4969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0355864818291303,
+      "learning_rate": 4.973907156195405e-06,
+      "loss": 0.637,
+      "step": 4970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8969754159542844,
+      "learning_rate": 4.972223763710231e-06,
+      "loss": 0.6201,
+      "step": 4971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2601088567596532,
+      "learning_rate": 4.970540374373653e-06,
+      "loss": 0.6495,
+      "step": 4972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.86528365828505,
+      "learning_rate": 4.96885698837649e-06,
+      "loss": 0.5782,
+      "step": 4973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9862084426505078,
+      "learning_rate": 4.967173605909566e-06,
+      "loss": 0.6409,
+      "step": 4974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7833941900740438,
+      "learning_rate": 4.9654902271637005e-06,
+      "loss": 0.7328,
+      "step": 4975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8010266943749187,
+      "learning_rate": 4.963806852329715e-06,
+      "loss": 0.6425,
+      "step": 4976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9009609692301752,
+      "learning_rate": 4.962123481598431e-06,
+      "loss": 0.6324,
+      "step": 4977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8052576908809401,
+      "learning_rate": 4.9604401151606626e-06,
+      "loss": 0.6518,
+      "step": 4978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.868428200480606,
+      "learning_rate": 4.958756753207234e-06,
+      "loss": 0.5987,
+      "step": 4979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9367859434196406,
+      "learning_rate": 4.957073395928963e-06,
+      "loss": 0.7087,
+      "step": 4980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7339870392195045,
+      "learning_rate": 4.955390043516666e-06,
+      "loss": 0.5802,
+      "step": 4981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0035904262119333,
+      "learning_rate": 4.953706696161163e-06,
+      "loss": 0.7026,
+      "step": 4982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.961898238315474,
+      "learning_rate": 4.952023354053269e-06,
+      "loss": 0.6831,
+      "step": 4983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8107631912130409,
+      "learning_rate": 4.950340017383802e-06,
+      "loss": 0.6986,
+      "step": 4984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7360843176805136,
+      "learning_rate": 4.948656686343577e-06,
+      "loss": 0.7014,
+      "step": 4985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.880010839932374,
+      "learning_rate": 4.946973361123411e-06,
+      "loss": 0.6949,
+      "step": 4986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0900365683376148,
+      "learning_rate": 4.945290041914114e-06,
+      "loss": 0.7737,
+      "step": 4987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7786822013502936,
+      "learning_rate": 4.943606728906503e-06,
+      "loss": 0.5823,
+      "step": 4988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.00799604136966,
+      "learning_rate": 4.941923422291392e-06,
+      "loss": 0.6422,
+      "step": 4989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0586128906008607,
+      "learning_rate": 4.94024012225959e-06,
+      "loss": 0.7,
+      "step": 4990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7841514369236677,
+      "learning_rate": 4.938556829001912e-06,
+      "loss": 0.5756,
+      "step": 4991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.306439783095156,
+      "learning_rate": 4.936873542709168e-06,
+      "loss": 0.5948,
+      "step": 4992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.615959858500367,
+      "learning_rate": 4.935190263572168e-06,
+      "loss": 0.5653,
+      "step": 4993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7733694335394845,
+      "learning_rate": 4.93350699178172e-06,
+      "loss": 0.5998,
+      "step": 4994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7428928640514691,
+      "learning_rate": 4.931823727528634e-06,
+      "loss": 0.6163,
+      "step": 4995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8644342122695163,
+      "learning_rate": 4.930140471003716e-06,
+      "loss": 0.7548,
+      "step": 4996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0175758276054747,
+      "learning_rate": 4.9284572223977755e-06,
+      "loss": 0.6279,
+      "step": 4997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0749950976688205,
+      "learning_rate": 4.926773981901616e-06,
+      "loss": 0.6444,
+      "step": 4998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.124837336083513,
+      "learning_rate": 4.925090749706045e-06,
+      "loss": 0.719,
+      "step": 4999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.993800913838882,
+      "learning_rate": 4.9234075260018615e-06,
+      "loss": 0.6828,
+      "step": 5000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8242495509149763,
+      "learning_rate": 4.921724310979872e-06,
+      "loss": 0.6616,
+      "step": 5001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8393977819228315,
+      "learning_rate": 4.920041104830879e-06,
+      "loss": 0.5354,
+      "step": 5002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.890344553733854,
+      "learning_rate": 4.918357907745681e-06,
+      "loss": 0.5357,
+      "step": 5003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8790617991665943,
+      "learning_rate": 4.91667471991508e-06,
+      "loss": 0.5759,
+      "step": 5004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8795858588418088,
+      "learning_rate": 4.914991541529875e-06,
+      "loss": 0.6199,
+      "step": 5005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7329455799381723,
+      "learning_rate": 4.913308372780863e-06,
+      "loss": 0.6488,
+      "step": 5006
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9222191221821798,
+      "learning_rate": 4.9116252138588435e-06,
+      "loss": 0.6239,
+      "step": 5007
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8479397466575056,
+      "learning_rate": 4.909942064954607e-06,
+      "loss": 0.6731,
+      "step": 5008
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9196852172679921,
+      "learning_rate": 4.908258926258951e-06,
+      "loss": 0.607,
+      "step": 5009
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8166403880883872,
+      "learning_rate": 4.906575797962669e-06,
+      "loss": 0.6431,
+      "step": 5010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7465182013878113,
+      "learning_rate": 4.904892680256553e-06,
+      "loss": 0.6119,
+      "step": 5011
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7704646382888727,
+      "learning_rate": 4.903209573331393e-06,
+      "loss": 0.6068,
+      "step": 5012
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9521348314630844,
+      "learning_rate": 4.90152647737798e-06,
+      "loss": 0.5722,
+      "step": 5013
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8961939935911354,
+      "learning_rate": 4.899843392587104e-06,
+      "loss": 0.6284,
+      "step": 5014
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.06014596993545,
+      "learning_rate": 4.8981603191495484e-06,
+      "loss": 0.6333,
+      "step": 5015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8220721139660931,
+      "learning_rate": 4.8964772572561e-06,
+      "loss": 0.5629,
+      "step": 5016
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9131734154680402,
+      "learning_rate": 4.894794207097546e-06,
+      "loss": 0.6336,
+      "step": 5017
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0006845737302243,
+      "learning_rate": 4.893111168864668e-06,
+      "loss": 0.6024,
+      "step": 5018
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.835189048852703,
+      "learning_rate": 4.891428142748247e-06,
+      "loss": 0.5964,
+      "step": 5019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9859963215195038,
+      "learning_rate": 4.889745128939067e-06,
+      "loss": 0.7307,
+      "step": 5020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0406766351110996,
+      "learning_rate": 4.888062127627904e-06,
+      "loss": 0.6828,
+      "step": 5021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9519571650177585,
+      "learning_rate": 4.886379139005537e-06,
+      "loss": 0.6814,
+      "step": 5022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8257334568373909,
+      "learning_rate": 4.884696163262742e-06,
+      "loss": 0.5058,
+      "step": 5023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7128904695532874,
+      "learning_rate": 4.883013200590294e-06,
+      "loss": 0.5868,
+      "step": 5024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7085841774434567,
+      "learning_rate": 4.881330251178968e-06,
+      "loss": 0.5638,
+      "step": 5025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.908140564370563,
+      "learning_rate": 4.879647315219533e-06,
+      "loss": 0.5806,
+      "step": 5026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8237970784787472,
+      "learning_rate": 4.87796439290276e-06,
+      "loss": 0.6137,
+      "step": 5027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.884695749390379,
+      "learning_rate": 4.87628148441942e-06,
+      "loss": 0.6748,
+      "step": 5028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8186368146388783,
+      "learning_rate": 4.874598589960279e-06,
+      "loss": 0.695,
+      "step": 5029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7967346368568278,
+      "learning_rate": 4.8729157097161025e-06,
+      "loss": 0.6156,
+      "step": 5030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.786633454376535,
+      "learning_rate": 4.871232843877654e-06,
+      "loss": 0.562,
+      "step": 5031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9850275764992054,
+      "learning_rate": 4.869549992635697e-06,
+      "loss": 0.5847,
+      "step": 5032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8984219117890422,
+      "learning_rate": 4.867867156180992e-06,
+      "loss": 0.622,
+      "step": 5033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7075309788258288,
+      "learning_rate": 4.866184334704297e-06,
+      "loss": 0.5809,
+      "step": 5034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8264943957080289,
+      "learning_rate": 4.864501528396371e-06,
+      "loss": 0.6548,
+      "step": 5035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.025846349325443,
+      "learning_rate": 4.862818737447971e-06,
+      "loss": 0.6318,
+      "step": 5036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7219540198512069,
+      "learning_rate": 4.861135962049847e-06,
+      "loss": 0.6695,
+      "step": 5037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.077354289814275,
+      "learning_rate": 4.859453202392753e-06,
+      "loss": 0.6169,
+      "step": 5038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9060446882946842,
+      "learning_rate": 4.8577704586674405e-06,
+      "loss": 0.5871,
+      "step": 5039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2842872317930714,
+      "learning_rate": 4.856087731064656e-06,
+      "loss": 0.6755,
+      "step": 5040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8767079908441326,
+      "learning_rate": 4.854405019775148e-06,
+      "loss": 0.6126,
+      "step": 5041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7312471324735106,
+      "learning_rate": 4.852722324989661e-06,
+      "loss": 0.6224,
+      "step": 5042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9361134166550507,
+      "learning_rate": 4.851039646898938e-06,
+      "loss": 0.6556,
+      "step": 5043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7027530285331263,
+      "learning_rate": 4.8493569856937215e-06,
+      "loss": 0.6058,
+      "step": 5044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1245655764901317,
+      "learning_rate": 4.847674341564748e-06,
+      "loss": 0.7028,
+      "step": 5045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9143911466692012,
+      "learning_rate": 4.845991714702755e-06,
+      "loss": 0.6948,
+      "step": 5046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.941124018073597,
+      "learning_rate": 4.844309105298481e-06,
+      "loss": 0.7231,
+      "step": 5047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9438368169759088,
+      "learning_rate": 4.842626513542656e-06,
+      "loss": 0.6155,
+      "step": 5048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.828675457167432,
+      "learning_rate": 4.840943939626012e-06,
+      "loss": 0.6513,
+      "step": 5049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8271797659739548,
+      "learning_rate": 4.83926138373928e-06,
+      "loss": 0.5988,
+      "step": 5050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9452851427531885,
+      "learning_rate": 4.8375788460731885e-06,
+      "loss": 0.6387,
+      "step": 5051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9841860268134885,
+      "learning_rate": 4.8358963268184585e-06,
+      "loss": 0.6832,
+      "step": 5052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8792424593951718,
+      "learning_rate": 4.8342138261658145e-06,
+      "loss": 0.5336,
+      "step": 5053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9310634981620276,
+      "learning_rate": 4.832531344305979e-06,
+      "loss": 0.6749,
+      "step": 5054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.02272191555057,
+      "learning_rate": 4.8308488814296695e-06,
+      "loss": 0.7115,
+      "step": 5055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7549744426219025,
+      "learning_rate": 4.829166437727603e-06,
+      "loss": 0.6443,
+      "step": 5056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4305012850264225,
+      "learning_rate": 4.827484013390496e-06,
+      "loss": 0.5472,
+      "step": 5057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1868137796135327,
+      "learning_rate": 4.825801608609059e-06,
+      "loss": 0.6223,
+      "step": 5058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.853626000392213,
+      "learning_rate": 4.824119223574002e-06,
+      "loss": 0.6385,
+      "step": 5059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8454461648434337,
+      "learning_rate": 4.8224368584760345e-06,
+      "loss": 0.6323,
+      "step": 5060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.735519333587952,
+      "learning_rate": 4.82075451350586e-06,
+      "loss": 0.6135,
+      "step": 5061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.858344450733772,
+      "learning_rate": 4.819072188854183e-06,
+      "loss": 0.535,
+      "step": 5062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0553850757925702,
+      "learning_rate": 4.817389884711706e-06,
+      "loss": 0.652,
+      "step": 5063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9772353357592571,
+      "learning_rate": 4.815707601269126e-06,
+      "loss": 0.662,
+      "step": 5064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7625156674615081,
+      "learning_rate": 4.814025338717139e-06,
+      "loss": 0.6364,
+      "step": 5065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9316812486984454,
+      "learning_rate": 4.812343097246442e-06,
+      "loss": 0.6193,
+      "step": 5066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4902859699492574,
+      "learning_rate": 4.8106608770477225e-06,
+      "loss": 0.4888,
+      "step": 5067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.191617491974796,
+      "learning_rate": 4.808978678311672e-06,
+      "loss": 0.5908,
+      "step": 5068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9555631768420776,
+      "learning_rate": 4.807296501228977e-06,
+      "loss": 0.6095,
+      "step": 5069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9063575635693246,
+      "learning_rate": 4.805614345990322e-06,
+      "loss": 0.6828,
+      "step": 5070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7436840319728293,
+      "learning_rate": 4.803932212786389e-06,
+      "loss": 0.7226,
+      "step": 5071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2328858534263847,
+      "learning_rate": 4.802250101807857e-06,
+      "loss": 0.6088,
+      "step": 5072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6911436875533297,
+      "learning_rate": 4.800568013245405e-06,
+      "loss": 0.6307,
+      "step": 5073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9214187162207514,
+      "learning_rate": 4.798885947289705e-06,
+      "loss": 0.6821,
+      "step": 5074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7311494622263512,
+      "learning_rate": 4.797203904131427e-06,
+      "loss": 0.6244,
+      "step": 5075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.00546387408505,
+      "learning_rate": 4.795521883961243e-06,
+      "loss": 0.6301,
+      "step": 5076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7067685876166543,
+      "learning_rate": 4.793839886969819e-06,
+      "loss": 0.6525,
+      "step": 5077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9479052680870534,
+      "learning_rate": 4.792157913347819e-06,
+      "loss": 0.6248,
+      "step": 5078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9878951234618571,
+      "learning_rate": 4.790475963285904e-06,
+      "loss": 0.6752,
+      "step": 5079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8307605231277921,
+      "learning_rate": 4.788794036974733e-06,
+      "loss": 0.6728,
+      "step": 5080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.007850691507178,
+      "learning_rate": 4.787112134604964e-06,
+      "loss": 0.6615,
+      "step": 5081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9682209210971686,
+      "learning_rate": 4.785430256367246e-06,
+      "loss": 0.6212,
+      "step": 5082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9539068327837195,
+      "learning_rate": 4.783748402452231e-06,
+      "loss": 0.6382,
+      "step": 5083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8254711972644915,
+      "learning_rate": 4.782066573050567e-06,
+      "loss": 0.5879,
+      "step": 5084
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9498763509941295,
+      "learning_rate": 4.7803847683529e-06,
+      "loss": 0.6181,
+      "step": 5085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8874003778516166,
+      "learning_rate": 4.77870298854987e-06,
+      "loss": 0.6335,
+      "step": 5086
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8312048896987314,
+      "learning_rate": 4.777021233832119e-06,
+      "loss": 0.5927,
+      "step": 5087
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9016075310030311,
+      "learning_rate": 4.775339504390283e-06,
+      "loss": 0.6854,
+      "step": 5088
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0648643513408897,
+      "learning_rate": 4.7736578004149936e-06,
+      "loss": 0.585,
+      "step": 5089
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7637557948369473,
+      "learning_rate": 4.771976122096882e-06,
+      "loss": 0.638,
+      "step": 5090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2250046913316366,
+      "learning_rate": 4.7702944696265766e-06,
+      "loss": 0.6797,
+      "step": 5091
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8031799845090113,
+      "learning_rate": 4.768612843194703e-06,
+      "loss": 0.6511,
+      "step": 5092
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8677023618333652,
+      "learning_rate": 4.766931242991882e-06,
+      "loss": 0.555,
+      "step": 5093
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9985670391581052,
+      "learning_rate": 4.765249669208733e-06,
+      "loss": 0.6718,
+      "step": 5094
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7495206295185197,
+      "learning_rate": 4.763568122035873e-06,
+      "loss": 0.7245,
+      "step": 5095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8561316308119256,
+      "learning_rate": 4.761886601663913e-06,
+      "loss": 0.7281,
+      "step": 5096
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8464447577024359,
+      "learning_rate": 4.760205108283465e-06,
+      "loss": 0.7014,
+      "step": 5097
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.912072520084057,
+      "learning_rate": 4.758523642085133e-06,
+      "loss": 0.683,
+      "step": 5098
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9941678964102907,
+      "learning_rate": 4.756842203259523e-06,
+      "loss": 0.5755,
+      "step": 5099
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.951575529813139,
+      "learning_rate": 4.755160791997235e-06,
+      "loss": 0.7646,
+      "step": 5100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8620977956553646,
+      "learning_rate": 4.753479408488868e-06,
+      "loss": 0.6415,
+      "step": 5101
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9029340613448509,
+      "learning_rate": 4.751798052925015e-06,
+      "loss": 0.6514,
+      "step": 5102
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8930213798910867,
+      "learning_rate": 4.750116725496268e-06,
+      "loss": 0.6322,
+      "step": 5103
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.804024019566647,
+      "learning_rate": 4.7484354263932136e-06,
+      "loss": 0.6197,
+      "step": 5104
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7759620876156021,
+      "learning_rate": 4.746754155806437e-06,
+      "loss": 0.5931,
+      "step": 5105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.728832946061556,
+      "learning_rate": 4.74507291392652e-06,
+      "loss": 0.5501,
+      "step": 5106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.116357919100813,
+      "learning_rate": 4.743391700944042e-06,
+      "loss": 0.6326,
+      "step": 5107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.841309504995033,
+      "learning_rate": 4.7417105170495775e-06,
+      "loss": 0.6341,
+      "step": 5108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9423787604865643,
+      "learning_rate": 4.740029362433698e-06,
+      "loss": 0.6879,
+      "step": 5109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9253807388055482,
+      "learning_rate": 4.7383482372869735e-06,
+      "loss": 0.5657,
+      "step": 5110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9821619032453928,
+      "learning_rate": 4.736667141799965e-06,
+      "loss": 0.7289,
+      "step": 5111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.908684744520375,
+      "learning_rate": 4.734986076163238e-06,
+      "loss": 0.7003,
+      "step": 5112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9596783586093547,
+      "learning_rate": 4.73330504056735e-06,
+      "loss": 0.6458,
+      "step": 5113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9224228062346809,
+      "learning_rate": 4.731624035202856e-06,
+      "loss": 0.7468,
+      "step": 5114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8680844649352413,
+      "learning_rate": 4.729943060260306e-06,
+      "loss": 0.6674,
+      "step": 5115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.024935667135311,
+      "learning_rate": 4.72826211593025e-06,
+      "loss": 0.7308,
+      "step": 5116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8095294446492074,
+      "learning_rate": 4.726581202403232e-06,
+      "loss": 0.7611,
+      "step": 5117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0807128692653345,
+      "learning_rate": 4.7249003198697955e-06,
+      "loss": 0.6249,
+      "step": 5118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0181745603197236,
+      "learning_rate": 4.723219468520474e-06,
+      "loss": 0.6167,
+      "step": 5119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7391558203953414,
+      "learning_rate": 4.721538648545802e-06,
+      "loss": 0.6373,
+      "step": 5120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7577092949947608,
+      "learning_rate": 4.719857860136312e-06,
+      "loss": 0.6041,
+      "step": 5121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9735884394796157,
+      "learning_rate": 4.7181771034825306e-06,
+      "loss": 0.6166,
+      "step": 5122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9220027882098172,
+      "learning_rate": 4.716496378774981e-06,
+      "loss": 0.524,
+      "step": 5123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9092831897320568,
+      "learning_rate": 4.714815686204182e-06,
+      "loss": 0.5647,
+      "step": 5124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9953171291519414,
+      "learning_rate": 4.713135025960652e-06,
+      "loss": 0.7217,
+      "step": 5125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6846329284575117,
+      "learning_rate": 4.711454398234902e-06,
+      "loss": 0.5023,
+      "step": 5126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8337482270315908,
+      "learning_rate": 4.7097738032174394e-06,
+      "loss": 0.6636,
+      "step": 5127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7405246564961698,
+      "learning_rate": 4.708093241098771e-06,
+      "loss": 0.6347,
+      "step": 5128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7954045338519318,
+      "learning_rate": 4.7064127120693965e-06,
+      "loss": 0.6376,
+      "step": 5129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8846517693769833,
+      "learning_rate": 4.704732216319815e-06,
+      "loss": 0.636,
+      "step": 5130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8830831159833605,
+      "learning_rate": 4.7030517540405195e-06,
+      "loss": 0.6288,
+      "step": 5131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8156056135854486,
+      "learning_rate": 4.701371325422002e-06,
+      "loss": 0.6129,
+      "step": 5132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0101280335971166,
+      "learning_rate": 4.6996909306547455e-06,
+      "loss": 0.7395,
+      "step": 5133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.321181377226723,
+      "learning_rate": 4.6980105699292335e-06,
+      "loss": 0.6699,
+      "step": 5134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.226576509137887,
+      "learning_rate": 4.696330243435945e-06,
+      "loss": 0.687,
+      "step": 5135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9204092632555354,
+      "learning_rate": 4.694649951365354e-06,
+      "loss": 0.5934,
+      "step": 5136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0429760454559283,
+      "learning_rate": 4.692969693907931e-06,
+      "loss": 0.6716,
+      "step": 5137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9277314108498504,
+      "learning_rate": 4.691289471254144e-06,
+      "loss": 0.6667,
+      "step": 5138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.988347172280978,
+      "learning_rate": 4.689609283594454e-06,
+      "loss": 0.6952,
+      "step": 5139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.952429859380367,
+      "learning_rate": 4.6879291311193244e-06,
+      "loss": 0.579,
+      "step": 5140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.804754920642748,
+      "learning_rate": 4.686249014019203e-06,
+      "loss": 0.6225,
+      "step": 5141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.989089474261446,
+      "learning_rate": 4.684568932484546e-06,
+      "loss": 0.6693,
+      "step": 5142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.983466565946788,
+      "learning_rate": 4.682888886705797e-06,
+      "loss": 0.6234,
+      "step": 5143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7356926775014248,
+      "learning_rate": 4.6812088768734e-06,
+      "loss": 0.5841,
+      "step": 5144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9370988809301124,
+      "learning_rate": 4.679528903177795e-06,
+      "loss": 0.6462,
+      "step": 5145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9192368651984106,
+      "learning_rate": 4.6778489658094145e-06,
+      "loss": 0.6877,
+      "step": 5146
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.82458653830636,
+      "learning_rate": 4.676169064958692e-06,
+      "loss": 0.6435,
+      "step": 5147
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0128828499828373,
+      "learning_rate": 4.674489200816051e-06,
+      "loss": 0.6844,
+      "step": 5148
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.079968396447985,
+      "learning_rate": 4.672809373571913e-06,
+      "loss": 0.712,
+      "step": 5149
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9981225791183472,
+      "learning_rate": 4.671129583416697e-06,
+      "loss": 0.6063,
+      "step": 5150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.800390445160367,
+      "learning_rate": 4.669449830540819e-06,
+      "loss": 0.6671,
+      "step": 5151
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7464047867362262,
+      "learning_rate": 4.6677701151346856e-06,
+      "loss": 0.6108,
+      "step": 5152
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8788397288269676,
+      "learning_rate": 4.6660904373887035e-06,
+      "loss": 0.6478,
+      "step": 5153
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8506751756270443,
+      "learning_rate": 4.664410797493275e-06,
+      "loss": 0.6116,
+      "step": 5154
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8466483682885082,
+      "learning_rate": 4.662731195638794e-06,
+      "loss": 0.5673,
+      "step": 5155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9893512470431758,
+      "learning_rate": 4.661051632015655e-06,
+      "loss": 0.5976,
+      "step": 5156
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8473962326091087,
+      "learning_rate": 4.659372106814244e-06,
+      "loss": 0.7179,
+      "step": 5157
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8880905203942298,
+      "learning_rate": 4.657692620224948e-06,
+      "loss": 0.6614,
+      "step": 5158
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8210595211398692,
+      "learning_rate": 4.656013172438143e-06,
+      "loss": 0.6492,
+      "step": 5159
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8642394980355337,
+      "learning_rate": 4.654333763644206e-06,
+      "loss": 0.6072,
+      "step": 5160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.113480454514094,
+      "learning_rate": 4.652654394033508e-06,
+      "loss": 0.7173,
+      "step": 5161
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7522034431581206,
+      "learning_rate": 4.650975063796414e-06,
+      "loss": 0.6015,
+      "step": 5162
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2933826515210276,
+      "learning_rate": 4.649295773123285e-06,
+      "loss": 0.5954,
+      "step": 5163
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5870787892330132,
+      "learning_rate": 4.6476165222044795e-06,
+      "loss": 0.6106,
+      "step": 5164
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.897198998438171,
+      "learning_rate": 4.645937311230351e-06,
+      "loss": 0.581,
+      "step": 5165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9011534788596396,
+      "learning_rate": 4.6442581403912444e-06,
+      "loss": 0.7063,
+      "step": 5166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8255651882129653,
+      "learning_rate": 4.6425790098775064e-06,
+      "loss": 0.7078,
+      "step": 5167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.103934341812746,
+      "learning_rate": 4.6408999198794744e-06,
+      "loss": 0.7187,
+      "step": 5168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8022669695183133,
+      "learning_rate": 4.639220870587484e-06,
+      "loss": 0.6677,
+      "step": 5169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0815425124291265,
+      "learning_rate": 4.637541862191864e-06,
+      "loss": 0.7668,
+      "step": 5170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6168920667058806,
+      "learning_rate": 4.635862894882939e-06,
+      "loss": 0.6407,
+      "step": 5171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8448181111407145,
+      "learning_rate": 4.634183968851031e-06,
+      "loss": 0.5558,
+      "step": 5172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8625072882231783,
+      "learning_rate": 4.632505084286455e-06,
+      "loss": 0.5622,
+      "step": 5173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.02673876945062,
+      "learning_rate": 4.630826241379522e-06,
+      "loss": 0.6759,
+      "step": 5174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.001843218175361,
+      "learning_rate": 4.629147440320539e-06,
+      "loss": 0.7657,
+      "step": 5175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9718566782218825,
+      "learning_rate": 4.627468681299808e-06,
+      "loss": 0.6772,
+      "step": 5176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8470501784821054,
+      "learning_rate": 4.625789964507626e-06,
+      "loss": 0.6576,
+      "step": 5177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9660622527248055,
+      "learning_rate": 4.624111290134284e-06,
+      "loss": 0.7013,
+      "step": 5178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8012877386792032,
+      "learning_rate": 4.622432658370069e-06,
+      "loss": 0.7023,
+      "step": 5179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8651094026382105,
+      "learning_rate": 4.620754069405265e-06,
+      "loss": 0.7387,
+      "step": 5180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984823336261961,
+      "learning_rate": 4.619075523430147e-06,
+      "loss": 0.6433,
+      "step": 5181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8084018306588718,
+      "learning_rate": 4.617397020634991e-06,
+      "loss": 0.4955,
+      "step": 5182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7058235501308092,
+      "learning_rate": 4.615718561210063e-06,
+      "loss": 0.4806,
+      "step": 5183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.135606445952955,
+      "learning_rate": 4.614040145345629e-06,
+      "loss": 0.6489,
+      "step": 5184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9003261832974596,
+      "learning_rate": 4.612361773231942e-06,
+      "loss": 0.6782,
+      "step": 5185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.766062587324516,
+      "learning_rate": 4.610683445059257e-06,
+      "loss": 0.6328,
+      "step": 5186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.772434598760773,
+      "learning_rate": 4.609005161017824e-06,
+      "loss": 0.5695,
+      "step": 5187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9399750167465266,
+      "learning_rate": 4.607326921297883e-06,
+      "loss": 0.7608,
+      "step": 5188
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.834185074488284,
+      "learning_rate": 4.605648726089674e-06,
+      "loss": 0.6278,
+      "step": 5189
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9981329160306083,
+      "learning_rate": 4.60397057558343e-06,
+      "loss": 0.6317,
+      "step": 5190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.075228145121904,
+      "learning_rate": 4.602292469969381e-06,
+      "loss": 0.627,
+      "step": 5191
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7500511714057594,
+      "learning_rate": 4.600614409437745e-06,
+      "loss": 0.6107,
+      "step": 5192
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0761176947995206,
+      "learning_rate": 4.5989363941787415e-06,
+      "loss": 0.6552,
+      "step": 5193
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.788070090950307,
+      "learning_rate": 4.597258424382584e-06,
+      "loss": 0.5552,
+      "step": 5194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8921091625857311,
+      "learning_rate": 4.5955805002394804e-06,
+      "loss": 0.6449,
+      "step": 5195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8761192118955454,
+      "learning_rate": 4.593902621939632e-06,
+      "loss": 0.609,
+      "step": 5196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.054485578286654,
+      "learning_rate": 4.592224789673235e-06,
+      "loss": 0.5932,
+      "step": 5197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9350066838823927,
+      "learning_rate": 4.590547003630484e-06,
+      "loss": 0.691,
+      "step": 5198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9608958757814123,
+      "learning_rate": 4.588869264001563e-06,
+      "loss": 0.5844,
+      "step": 5199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9773425293593354,
+      "learning_rate": 4.587191570976655e-06,
+      "loss": 0.669,
+      "step": 5200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9251634580282069,
+      "learning_rate": 4.585513924745935e-06,
+      "loss": 0.6388,
+      "step": 5201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7757867948720212,
+      "learning_rate": 4.583836325499573e-06,
+      "loss": 0.5827,
+      "step": 5202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.085264127591141,
+      "learning_rate": 4.5821587734277374e-06,
+      "loss": 0.532,
+      "step": 5203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.75056064511947,
+      "learning_rate": 4.580481268720585e-06,
+      "loss": 0.538,
+      "step": 5204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8790372630466574,
+      "learning_rate": 4.578803811568272e-06,
+      "loss": 0.5644,
+      "step": 5205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9390235343364985,
+      "learning_rate": 4.5771264021609494e-06,
+      "loss": 0.5906,
+      "step": 5206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.924353396395898,
+      "learning_rate": 4.575449040688758e-06,
+      "loss": 0.5973,
+      "step": 5207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7020159742103511,
+      "learning_rate": 4.573771727341838e-06,
+      "loss": 0.5145,
+      "step": 5208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7708593482880979,
+      "learning_rate": 4.572094462310322e-06,
+      "loss": 0.6289,
+      "step": 5209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.166457630399841,
+      "learning_rate": 4.570417245784337e-06,
+      "loss": 0.6675,
+      "step": 5210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.877828575644666,
+      "learning_rate": 4.568740077954007e-06,
+      "loss": 0.6011,
+      "step": 5211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.911475306416678,
+      "learning_rate": 4.567062959009446e-06,
+      "loss": 0.569,
+      "step": 5212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8913396335881767,
+      "learning_rate": 4.565385889140767e-06,
+      "loss": 0.6125,
+      "step": 5213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.08561786360724,
+      "learning_rate": 4.563708868538077e-06,
+      "loss": 0.7155,
+      "step": 5214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.154811342394879,
+      "learning_rate": 4.56203189739147e-06,
+      "loss": 0.5944,
+      "step": 5215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0035311683983323,
+      "learning_rate": 4.560354975891045e-06,
+      "loss": 0.6358,
+      "step": 5216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9237950319146873,
+      "learning_rate": 4.55867810422689e-06,
+      "loss": 0.612,
+      "step": 5217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.891746378991884,
+      "learning_rate": 4.557001282589086e-06,
+      "loss": 0.6303,
+      "step": 5218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7132159636798743,
+      "learning_rate": 4.5553245111677124e-06,
+      "loss": 0.5618,
+      "step": 5219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9582069007610816,
+      "learning_rate": 4.55364779015284e-06,
+      "loss": 0.7596,
+      "step": 5220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8416862170002277,
+      "learning_rate": 4.5519711197345365e-06,
+      "loss": 0.6514,
+      "step": 5221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7109777078724329,
+      "learning_rate": 4.550294500102858e-06,
+      "loss": 0.6504,
+      "step": 5222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7965413196363942,
+      "learning_rate": 4.548617931447861e-06,
+      "loss": 0.6029,
+      "step": 5223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7730256764319179,
+      "learning_rate": 4.546941413959595e-06,
+      "loss": 0.6525,
+      "step": 5224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.003273238543562,
+      "learning_rate": 4.545264947828101e-06,
+      "loss": 0.6805,
+      "step": 5225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.010674257758387,
+      "learning_rate": 4.5435885332434184e-06,
+      "loss": 0.5269,
+      "step": 5226
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8445652402675543,
+      "learning_rate": 4.541912170395576e-06,
+      "loss": 0.6639,
+      "step": 5227
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.937635194164857,
+      "learning_rate": 4.540235859474603e-06,
+      "loss": 0.5969,
+      "step": 5228
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7232427319669361,
+      "learning_rate": 4.538559600670513e-06,
+      "loss": 0.575,
+      "step": 5229
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9439853705055208,
+      "learning_rate": 4.5368833941733235e-06,
+      "loss": 0.5975,
+      "step": 5230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8694910106701905,
+      "learning_rate": 4.53520724017304e-06,
+      "loss": 0.5624,
+      "step": 5231
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.012545765815874,
+      "learning_rate": 4.533531138859665e-06,
+      "loss": 0.5813,
+      "step": 5232
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8435415628000271,
+      "learning_rate": 4.531855090423195e-06,
+      "loss": 0.606,
+      "step": 5233
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9556583134206118,
+      "learning_rate": 4.5301790950536175e-06,
+      "loss": 0.6116,
+      "step": 5234
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0787950978231047,
+      "learning_rate": 4.5285031529409184e-06,
+      "loss": 0.6888,
+      "step": 5235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.853465607687972,
+      "learning_rate": 4.526827264275076e-06,
+      "loss": 0.5916,
+      "step": 5236
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7968352682356779,
+      "learning_rate": 4.525151429246057e-06,
+      "loss": 0.6473,
+      "step": 5237
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2194788816263666,
+      "learning_rate": 4.523475648043832e-06,
+      "loss": 0.6571,
+      "step": 5238
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7577666502701064,
+      "learning_rate": 4.521799920858358e-06,
+      "loss": 0.6082,
+      "step": 5239
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9041366214064464,
+      "learning_rate": 4.520124247879588e-06,
+      "loss": 0.678,
+      "step": 5240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.865248415982124,
+      "learning_rate": 4.51844862929747e-06,
+      "loss": 0.6431,
+      "step": 5241
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8755663337410842,
+      "learning_rate": 4.516773065301946e-06,
+      "loss": 0.6337,
+      "step": 5242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8953438735645665,
+      "learning_rate": 4.515097556082949e-06,
+      "loss": 0.5888,
+      "step": 5243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.803506322598369,
+      "learning_rate": 4.513422101830408e-06,
+      "loss": 0.5827,
+      "step": 5244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9563878164296573,
+      "learning_rate": 4.5117467027342435e-06,
+      "loss": 0.5976,
+      "step": 5245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7276638684043466,
+      "learning_rate": 4.510071358984373e-06,
+      "loss": 0.6353,
+      "step": 5246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9039495600414362,
+      "learning_rate": 4.508396070770708e-06,
+      "loss": 0.6116,
+      "step": 5247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.930911155262903,
+      "learning_rate": 4.50672083828315e-06,
+      "loss": 0.6507,
+      "step": 5248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7348548611102907,
+      "learning_rate": 4.505045661711596e-06,
+      "loss": 0.633,
+      "step": 5249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0192760027137764,
+      "learning_rate": 4.503370541245937e-06,
+      "loss": 0.6852,
+      "step": 5250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.291914253911129,
+      "learning_rate": 4.501695477076062e-06,
+      "loss": 0.7381,
+      "step": 5251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8818408471589059,
+      "learning_rate": 4.5000204693918405e-06,
+      "loss": 0.6618,
+      "step": 5252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6450990278976414,
+      "learning_rate": 4.49834551838315e-06,
+      "loss": 0.7165,
+      "step": 5253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8655688952363387,
+      "learning_rate": 4.496670624239854e-06,
+      "loss": 0.5881,
+      "step": 5254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5881419810270465,
+      "learning_rate": 4.494995787151811e-06,
+      "loss": 0.5591,
+      "step": 5255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9868107004461213,
+      "learning_rate": 4.493321007308875e-06,
+      "loss": 0.6265,
+      "step": 5256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.748853627472903,
+      "learning_rate": 4.49164628490089e-06,
+      "loss": 0.6645,
+      "step": 5257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6825341367268272,
+      "learning_rate": 4.489971620117699e-06,
+      "loss": 0.5963,
+      "step": 5258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7378224177710997,
+      "learning_rate": 4.4882970131491286e-06,
+      "loss": 0.6074,
+      "step": 5259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7043117434948243,
+      "learning_rate": 4.4866224641850095e-06,
+      "loss": 0.5871,
+      "step": 5260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.873293864202997,
+      "learning_rate": 4.48494797341516e-06,
+      "loss": 0.6175,
+      "step": 5261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8950760665680089,
+      "learning_rate": 4.4832735410293924e-06,
+      "loss": 0.6416,
+      "step": 5262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.825762291742463,
+      "learning_rate": 4.481599167217515e-06,
+      "loss": 0.6834,
+      "step": 5263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7127983027636853,
+      "learning_rate": 4.479924852169327e-06,
+      "loss": 0.6113,
+      "step": 5264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2666902600585157,
+      "learning_rate": 4.478250596074623e-06,
+      "loss": 0.5987,
+      "step": 5265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0911717000348458,
+      "learning_rate": 4.476576399123187e-06,
+      "loss": 0.5852,
+      "step": 5266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0671663566535403,
+      "learning_rate": 4.4749022615047975e-06,
+      "loss": 0.6781,
+      "step": 5267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7386282025860975,
+      "learning_rate": 4.4732281834092305e-06,
+      "loss": 0.6234,
+      "step": 5268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.031240480955804,
+      "learning_rate": 4.471554165026251e-06,
+      "loss": 0.6807,
+      "step": 5269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9289143528508108,
+      "learning_rate": 4.46988020654562e-06,
+      "loss": 0.6729,
+      "step": 5270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0153197310458943,
+      "learning_rate": 4.4682063081570885e-06,
+      "loss": 0.631,
+      "step": 5271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8926835051803474,
+      "learning_rate": 4.466532470050403e-06,
+      "loss": 0.6079,
+      "step": 5272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9793089069024377,
+      "learning_rate": 4.464858692415304e-06,
+      "loss": 0.6527,
+      "step": 5273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9025404029621105,
+      "learning_rate": 4.463184975441522e-06,
+      "loss": 0.6627,
+      "step": 5274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9089159496126418,
+      "learning_rate": 4.461511319318782e-06,
+      "loss": 0.6441,
+      "step": 5275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.904444956313747,
+      "learning_rate": 4.4598377242368026e-06,
+      "loss": 0.6431,
+      "step": 5276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9622376326034894,
+      "learning_rate": 4.458164190385297e-06,
+      "loss": 0.6903,
+      "step": 5277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8203493701420281,
+      "learning_rate": 4.456490717953969e-06,
+      "loss": 0.5699,
+      "step": 5278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7996933125103711,
+      "learning_rate": 4.454817307132515e-06,
+      "loss": 0.5578,
+      "step": 5279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8862630623229697,
+      "learning_rate": 4.4531439581106295e-06,
+      "loss": 0.6777,
+      "step": 5280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9189441415408823,
+      "learning_rate": 4.451470671077991e-06,
+      "loss": 0.6002,
+      "step": 5281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7339129039063892,
+      "learning_rate": 4.449797446224279e-06,
+      "loss": 0.6599,
+      "step": 5282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.141150906787522,
+      "learning_rate": 4.4481242837391615e-06,
+      "loss": 0.6243,
+      "step": 5283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.006503116981974,
+      "learning_rate": 4.446451183812302e-06,
+      "loss": 0.5795,
+      "step": 5284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.045139071894576,
+      "learning_rate": 4.4447781466333565e-06,
+      "loss": 0.5821,
+      "step": 5285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8897413641576013,
+      "learning_rate": 4.4431051723919725e-06,
+      "loss": 0.4936,
+      "step": 5286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8631944101350673,
+      "learning_rate": 4.441432261277794e-06,
+      "loss": 0.5666,
+      "step": 5287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8750818481890121,
+      "learning_rate": 4.43975941348045e-06,
+      "loss": 0.6188,
+      "step": 5288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7830177570624304,
+      "learning_rate": 4.43808662918957e-06,
+      "loss": 0.5248,
+      "step": 5289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9469663022421002,
+      "learning_rate": 4.436413908594773e-06,
+      "loss": 0.699,
+      "step": 5290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9735300322889926,
+      "learning_rate": 4.434741251885671e-06,
+      "loss": 0.6604,
+      "step": 5291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7706573081264123,
+      "learning_rate": 4.433068659251871e-06,
+      "loss": 0.8055,
+      "step": 5292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1555746130188567,
+      "learning_rate": 4.43139613088297e-06,
+      "loss": 0.6699,
+      "step": 5293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9824503792487942,
+      "learning_rate": 4.429723666968559e-06,
+      "loss": 0.7092,
+      "step": 5294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1311700772834556,
+      "learning_rate": 4.4280512676982215e-06,
+      "loss": 0.7339,
+      "step": 5295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7811696362624156,
+      "learning_rate": 4.426378933261532e-06,
+      "loss": 0.6008,
+      "step": 5296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7760256644912769,
+      "learning_rate": 4.424706663848059e-06,
+      "loss": 0.6531,
+      "step": 5297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7122666917208644,
+      "learning_rate": 4.423034459647365e-06,
+      "loss": 0.6768,
+      "step": 5298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1077740469187316,
+      "learning_rate": 4.421362320849003e-06,
+      "loss": 0.7002,
+      "step": 5299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1373646086172955,
+      "learning_rate": 4.419690247642521e-06,
+      "loss": 0.5956,
+      "step": 5300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8105651786972892,
+      "learning_rate": 4.418018240217457e-06,
+      "loss": 0.6231,
+      "step": 5301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7954524837112216,
+      "learning_rate": 4.416346298763343e-06,
+      "loss": 0.5767,
+      "step": 5302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.797535057274918,
+      "learning_rate": 4.414674423469702e-06,
+      "loss": 0.5235,
+      "step": 5303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8580491980634726,
+      "learning_rate": 4.413002614526052e-06,
+      "loss": 0.6158,
+      "step": 5304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1384624398890035,
+      "learning_rate": 4.411330872121901e-06,
+      "loss": 0.638,
+      "step": 5305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.838188572332752,
+      "learning_rate": 4.4096591964467515e-06,
+      "loss": 0.5328,
+      "step": 5306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.00835612910002,
+      "learning_rate": 4.407987587690097e-06,
+      "loss": 0.6725,
+      "step": 5307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9969651486276907,
+      "learning_rate": 4.406316046041423e-06,
+      "loss": 0.7185,
+      "step": 5308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1341862742091475,
+      "learning_rate": 4.404644571690208e-06,
+      "loss": 0.713,
+      "step": 5309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9766606231136952,
+      "learning_rate": 4.402973164825927e-06,
+      "loss": 0.5984,
+      "step": 5310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8695986242532783,
+      "learning_rate": 4.401301825638039e-06,
+      "loss": 0.6824,
+      "step": 5311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.025287723277109,
+      "learning_rate": 4.399630554316002e-06,
+      "loss": 0.5779,
+      "step": 5312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.843453109578057,
+      "learning_rate": 4.397959351049263e-06,
+      "loss": 0.5695,
+      "step": 5313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9383448410386213,
+      "learning_rate": 4.396288216027264e-06,
+      "loss": 0.5474,
+      "step": 5314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8430720816288715,
+      "learning_rate": 4.394617149439435e-06,
+      "loss": 0.6376,
+      "step": 5315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8636058676938005,
+      "learning_rate": 4.392946151475204e-06,
+      "loss": 0.6708,
+      "step": 5316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8075379505938336,
+      "learning_rate": 4.391275222323989e-06,
+      "loss": 0.623,
+      "step": 5317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.282958892491627,
+      "learning_rate": 4.389604362175194e-06,
+      "loss": 0.6761,
+      "step": 5318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9314202760968853,
+      "learning_rate": 4.387933571218224e-06,
+      "loss": 0.6157,
+      "step": 5319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8084471261587793,
+      "learning_rate": 4.386262849642474e-06,
+      "loss": 0.5921,
+      "step": 5320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7878979830871655,
+      "learning_rate": 4.384592197637327e-06,
+      "loss": 0.6585,
+      "step": 5321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7272057096771558,
+      "learning_rate": 4.382921615392162e-06,
+      "loss": 0.6154,
+      "step": 5322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7715100059111006,
+      "learning_rate": 4.381251103096349e-06,
+      "loss": 0.5159,
+      "step": 5323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8160415729278552,
+      "learning_rate": 4.379580660939253e-06,
+      "loss": 0.6188,
+      "step": 5324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0714128617370844,
+      "learning_rate": 4.377910289110222e-06,
+      "loss": 0.6339,
+      "step": 5325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7709774757228993,
+      "learning_rate": 4.376239987798606e-06,
+      "loss": 0.5522,
+      "step": 5326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6953184185859074,
+      "learning_rate": 4.3745697571937434e-06,
+      "loss": 0.6141,
+      "step": 5327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0788554843615246,
+      "learning_rate": 4.372899597484963e-06,
+      "loss": 0.6108,
+      "step": 5328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8961589961491185,
+      "learning_rate": 4.371229508861588e-06,
+      "loss": 0.6662,
+      "step": 5329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7425613651137846,
+      "learning_rate": 4.369559491512932e-06,
+      "loss": 0.5696,
+      "step": 5330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7939534103840276,
+      "learning_rate": 4.367889545628301e-06,
+      "loss": 0.5928,
+      "step": 5331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8404672607265415,
+      "learning_rate": 4.366219671396995e-06,
+      "loss": 0.607,
+      "step": 5332
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9584221320985873,
+      "learning_rate": 4.364549869008299e-06,
+      "loss": 0.6284,
+      "step": 5333
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9320337232975733,
+      "learning_rate": 4.362880138651498e-06,
+      "loss": 0.7023,
+      "step": 5334
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8876641008267565,
+      "learning_rate": 4.361210480515865e-06,
+      "loss": 0.61,
+      "step": 5335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7938844774124776,
+      "learning_rate": 4.3595408947906644e-06,
+      "loss": 0.7137,
+      "step": 5336
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8127042217334888,
+      "learning_rate": 4.3578713816651535e-06,
+      "loss": 0.5189,
+      "step": 5337
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7419189776304915,
+      "learning_rate": 4.356201941328582e-06,
+      "loss": 0.6613,
+      "step": 5338
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8832170874848826,
+      "learning_rate": 4.354532573970191e-06,
+      "loss": 0.6425,
+      "step": 5339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.821648531778137,
+      "learning_rate": 4.352863279779211e-06,
+      "loss": 0.6567,
+      "step": 5340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8233645260795328,
+      "learning_rate": 4.351194058944866e-06,
+      "loss": 0.5911,
+      "step": 5341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9133087513735505,
+      "learning_rate": 4.349524911656373e-06,
+      "loss": 0.6188,
+      "step": 5342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.672726443042582,
+      "learning_rate": 4.34785583810294e-06,
+      "loss": 0.5733,
+      "step": 5343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9222023216966218,
+      "learning_rate": 4.346186838473764e-06,
+      "loss": 0.6429,
+      "step": 5344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.757050865618816,
+      "learning_rate": 4.344517912958036e-06,
+      "loss": 0.6731,
+      "step": 5345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7741802896340304,
+      "learning_rate": 4.342849061744939e-06,
+      "loss": 0.6015,
+      "step": 5346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7665348586924297,
+      "learning_rate": 4.341180285023648e-06,
+      "loss": 0.5387,
+      "step": 5347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9006331306865378,
+      "learning_rate": 4.339511582983325e-06,
+      "loss": 0.6501,
+      "step": 5348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171585780899604,
+      "learning_rate": 4.337842955813129e-06,
+      "loss": 0.6288,
+      "step": 5349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7377998975322286,
+      "learning_rate": 4.336174403702208e-06,
+      "loss": 0.6438,
+      "step": 5350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9683352192208223,
+      "learning_rate": 4.334505926839702e-06,
+      "loss": 0.6363,
+      "step": 5351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8374943205843048,
+      "learning_rate": 4.332837525414743e-06,
+      "loss": 0.5948,
+      "step": 5352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7886524450076886,
+      "learning_rate": 4.331169199616453e-06,
+      "loss": 0.6511,
+      "step": 5353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0533622511419374,
+      "learning_rate": 4.329500949633947e-06,
+      "loss": 0.6977,
+      "step": 5354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0519868425621772,
+      "learning_rate": 4.32783277565633e-06,
+      "loss": 0.7681,
+      "step": 5355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9748725112796992,
+      "learning_rate": 4.326164677872698e-06,
+      "loss": 0.626,
+      "step": 5356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8229150751926968,
+      "learning_rate": 4.324496656472141e-06,
+      "loss": 0.6979,
+      "step": 5357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8616048764016284,
+      "learning_rate": 4.322828711643737e-06,
+      "loss": 0.5955,
+      "step": 5358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7219793227650209,
+      "learning_rate": 4.32116084357656e-06,
+      "loss": 0.6187,
+      "step": 5359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7727210332779848,
+      "learning_rate": 4.31949305245967e-06,
+      "loss": 0.6097,
+      "step": 5360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7854736896677759,
+      "learning_rate": 4.317825338482123e-06,
+      "loss": 0.667,
+      "step": 5361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0495980515372945,
+      "learning_rate": 4.31615770183296e-06,
+      "loss": 0.631,
+      "step": 5362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0096774454763766,
+      "learning_rate": 4.31449014270122e-06,
+      "loss": 0.6482,
+      "step": 5363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2165036324005603,
+      "learning_rate": 4.312822661275929e-06,
+      "loss": 0.7564,
+      "step": 5364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.857956424239191,
+      "learning_rate": 4.3111552577461066e-06,
+      "loss": 0.592,
+      "step": 5365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0795780405956905,
+      "learning_rate": 4.309487932300762e-06,
+      "loss": 0.7031,
+      "step": 5366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9276173353067232,
+      "learning_rate": 4.307820685128896e-06,
+      "loss": 0.6087,
+      "step": 5367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6763284047956102,
+      "learning_rate": 4.3061535164195e-06,
+      "loss": 0.7254,
+      "step": 5368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0186762339298694,
+      "learning_rate": 4.304486426361561e-06,
+      "loss": 0.6859,
+      "step": 5369
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8808991283500813,
+      "learning_rate": 4.302819415144046e-06,
+      "loss": 0.5656,
+      "step": 5370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.894815287220838,
+      "learning_rate": 4.301152482955926e-06,
+      "loss": 0.6015,
+      "step": 5371
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.781652807228387,
+      "learning_rate": 4.299485629986153e-06,
+      "loss": 0.5619,
+      "step": 5372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9419047981480841,
+      "learning_rate": 4.297818856423679e-06,
+      "loss": 0.6505,
+      "step": 5373
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0120179327646777,
+      "learning_rate": 4.296152162457439e-06,
+      "loss": 0.6558,
+      "step": 5374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8836238693944019,
+      "learning_rate": 4.294485548276363e-06,
+      "loss": 0.6515,
+      "step": 5375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8504429409168075,
+      "learning_rate": 4.292819014069372e-06,
+      "loss": 0.7023,
+      "step": 5376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0390665006750743,
+      "learning_rate": 4.291152560025377e-06,
+      "loss": 0.7807,
+      "step": 5377
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7922684370026463,
+      "learning_rate": 4.2894861863332785e-06,
+      "loss": 0.6747,
+      "step": 5378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9449512984578232,
+      "learning_rate": 4.2878198931819705e-06,
+      "loss": 0.6697,
+      "step": 5379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0122903659992213,
+      "learning_rate": 4.286153680760338e-06,
+      "loss": 0.7135,
+      "step": 5380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9004201271206218,
+      "learning_rate": 4.284487549257254e-06,
+      "loss": 0.4542,
+      "step": 5381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.672893473626159,
+      "learning_rate": 4.282821498861585e-06,
+      "loss": 0.5413,
+      "step": 5382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8436366187307252,
+      "learning_rate": 4.281155529762186e-06,
+      "loss": 0.6247,
+      "step": 5383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.676623195970577,
+      "learning_rate": 4.279489642147906e-06,
+      "loss": 0.6366,
+      "step": 5384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8849736907682577,
+      "learning_rate": 4.277823836207581e-06,
+      "loss": 0.6423,
+      "step": 5385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6504346249070367,
+      "learning_rate": 4.276158112130041e-06,
+      "loss": 0.6051,
+      "step": 5386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.900270675572181,
+      "learning_rate": 4.274492470104105e-06,
+      "loss": 0.7088,
+      "step": 5387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0141651357876653,
+      "learning_rate": 4.272826910318581e-06,
+      "loss": 0.6249,
+      "step": 5388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9112805090759801,
+      "learning_rate": 4.271161432962273e-06,
+      "loss": 0.6067,
+      "step": 5389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.034156020188733,
+      "learning_rate": 4.2694960382239705e-06,
+      "loss": 0.644,
+      "step": 5390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9930605463073325,
+      "learning_rate": 4.2678307262924575e-06,
+      "loss": 0.6368,
+      "step": 5391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.906858020595567,
+      "learning_rate": 4.266165497356503e-06,
+      "loss": 0.5987,
+      "step": 5392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8406121682236543,
+      "learning_rate": 4.264500351604872e-06,
+      "loss": 0.6536,
+      "step": 5393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8254428663126403,
+      "learning_rate": 4.262835289226318e-06,
+      "loss": 0.5699,
+      "step": 5394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9270412949134141,
+      "learning_rate": 4.261170310409586e-06,
+      "loss": 0.7475,
+      "step": 5395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8066282234062603,
+      "learning_rate": 4.259505415343411e-06,
+      "loss": 0.6369,
+      "step": 5396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.758284076012403,
+      "learning_rate": 4.257840604216517e-06,
+      "loss": 0.6009,
+      "step": 5397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6749754336124096,
+      "learning_rate": 4.2561758772176244e-06,
+      "loss": 0.6205,
+      "step": 5398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9275539454621302,
+      "learning_rate": 4.254511234535432e-06,
+      "loss": 0.5659,
+      "step": 5399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9011018789471603,
+      "learning_rate": 4.252846676358641e-06,
+      "loss": 0.6691,
+      "step": 5400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0596941216693483,
+      "learning_rate": 4.251182202875938e-06,
+      "loss": 0.6916,
+      "step": 5401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.857656244516316,
+      "learning_rate": 4.249517814276e-06,
+      "loss": 0.6174,
+      "step": 5402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9254466566790163,
+      "learning_rate": 4.247853510747495e-06,
+      "loss": 0.586,
+      "step": 5403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7917004971343924,
+      "learning_rate": 4.2461892924790825e-06,
+      "loss": 0.5538,
+      "step": 5404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9172145911881517,
+      "learning_rate": 4.244525159659409e-06,
+      "loss": 0.6444,
+      "step": 5405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0655398744581377,
+      "learning_rate": 4.2428611124771184e-06,
+      "loss": 0.6404,
+      "step": 5406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8304720049482348,
+      "learning_rate": 4.241197151120834e-06,
+      "loss": 0.6175,
+      "step": 5407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.81762951481684,
+      "learning_rate": 4.239533275779177e-06,
+      "loss": 0.6268,
+      "step": 5408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9408357676082384,
+      "learning_rate": 4.237869486640758e-06,
+      "loss": 0.6111,
+      "step": 5409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.486652228429193,
+      "learning_rate": 4.236205783894176e-06,
+      "loss": 0.6691,
+      "step": 5410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6509217437191293,
+      "learning_rate": 4.2345421677280215e-06,
+      "loss": 0.6332,
+      "step": 5411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8758194011046228,
+      "learning_rate": 4.232878638330877e-06,
+      "loss": 0.6156,
+      "step": 5412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9285754226908345,
+      "learning_rate": 4.231215195891311e-06,
+      "loss": 0.6819,
+      "step": 5413
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.248208337442681,
+      "learning_rate": 4.229551840597884e-06,
+      "loss": 0.6568,
+      "step": 5414
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7753810739067684,
+      "learning_rate": 4.227888572639148e-06,
+      "loss": 0.5761,
+      "step": 5415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0888773685296074,
+      "learning_rate": 4.226225392203641e-06,
+      "loss": 0.7153,
+      "step": 5416
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7249346313721021,
+      "learning_rate": 4.224562299479897e-06,
+      "loss": 0.547,
+      "step": 5417
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.711135526070868,
+      "learning_rate": 4.222899294656437e-06,
+      "loss": 0.6079,
+      "step": 5418
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9324287494379544,
+      "learning_rate": 4.221236377921771e-06,
+      "loss": 0.5704,
+      "step": 5419
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7926007338434826,
+      "learning_rate": 4.219573549464403e-06,
+      "loss": 0.5625,
+      "step": 5420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9062785913816724,
+      "learning_rate": 4.2179108094728185e-06,
+      "loss": 0.6407,
+      "step": 5421
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7912130192025735,
+      "learning_rate": 4.2162481581355015e-06,
+      "loss": 0.5941,
+      "step": 5422
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.919288046883698,
+      "learning_rate": 4.214585595640923e-06,
+      "loss": 0.6401,
+      "step": 5423
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8434935010857791,
+      "learning_rate": 4.212923122177545e-06,
+      "loss": 0.5237,
+      "step": 5424
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0484794450464237,
+      "learning_rate": 4.211260737933816e-06,
+      "loss": 0.6333,
+      "step": 5425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0744597871080273,
+      "learning_rate": 4.209598443098179e-06,
+      "loss": 0.572,
+      "step": 5426
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.912927529947007,
+      "learning_rate": 4.2079362378590625e-06,
+      "loss": 0.6074,
+      "step": 5427
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.815141363610265,
+      "learning_rate": 4.206274122404891e-06,
+      "loss": 0.5915,
+      "step": 5428
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7372331606973053,
+      "learning_rate": 4.204612096924069e-06,
+      "loss": 0.4789,
+      "step": 5429
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8298977661303382,
+      "learning_rate": 4.202950161605e-06,
+      "loss": 0.6594,
+      "step": 5430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9098142487861431,
+      "learning_rate": 4.201288316636073e-06,
+      "loss": 0.6554,
+      "step": 5431
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0474500584686406,
+      "learning_rate": 4.199626562205668e-06,
+      "loss": 0.6762,
+      "step": 5432
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.64859334265512,
+      "learning_rate": 4.197964898502154e-06,
+      "loss": 0.6535,
+      "step": 5433
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0230013593164036,
+      "learning_rate": 4.1963033257138904e-06,
+      "loss": 0.5574,
+      "step": 5434
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7973657989903222,
+      "learning_rate": 4.194641844029227e-06,
+      "loss": 0.6348,
+      "step": 5435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.949821896226258,
+      "learning_rate": 4.1929804536365e-06,
+      "loss": 0.7132,
+      "step": 5436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8046228407634757,
+      "learning_rate": 4.191319154724038e-06,
+      "loss": 0.7059,
+      "step": 5437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8969165232037644,
+      "learning_rate": 4.189657947480159e-06,
+      "loss": 0.5655,
+      "step": 5438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.803066198351578,
+      "learning_rate": 4.187996832093171e-06,
+      "loss": 0.5827,
+      "step": 5439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9186713994335214,
+      "learning_rate": 4.186335808751369e-06,
+      "loss": 0.6451,
+      "step": 5440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9254356727843782,
+      "learning_rate": 4.184674877643042e-06,
+      "loss": 0.6686,
+      "step": 5441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0999174784887145,
+      "learning_rate": 4.183014038956465e-06,
+      "loss": 0.6833,
+      "step": 5442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.037687236980084,
+      "learning_rate": 4.181353292879904e-06,
+      "loss": 0.7428,
+      "step": 5443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7644928803447903,
+      "learning_rate": 4.179692639601612e-06,
+      "loss": 0.7074,
+      "step": 5444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9200170976748951,
+      "learning_rate": 4.178032079309836e-06,
+      "loss": 0.684,
+      "step": 5445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9507059714006074,
+      "learning_rate": 4.176371612192808e-06,
+      "loss": 0.6933,
+      "step": 5446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7315185417192958,
+      "learning_rate": 4.174711238438752e-06,
+      "loss": 0.6032,
+      "step": 5447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9013248224515067,
+      "learning_rate": 4.173050958235882e-06,
+      "loss": 0.6588,
+      "step": 5448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8963272377797982,
+      "learning_rate": 4.171390771772399e-06,
+      "loss": 0.5986,
+      "step": 5449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9082415935624142,
+      "learning_rate": 4.169730679236496e-06,
+      "loss": 0.6572,
+      "step": 5450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.937334868978178,
+      "learning_rate": 4.168070680816351e-06,
+      "loss": 0.6632,
+      "step": 5451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8696062932881234,
+      "learning_rate": 4.166410776700137e-06,
+      "loss": 0.7399,
+      "step": 5452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1429005075050083,
+      "learning_rate": 4.164750967076012e-06,
+      "loss": 0.7346,
+      "step": 5453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.743702093443798,
+      "learning_rate": 4.163091252132126e-06,
+      "loss": 0.6522,
+      "step": 5454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8587073840092287,
+      "learning_rate": 4.1614316320566174e-06,
+      "loss": 0.6181,
+      "step": 5455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8675162521113273,
+      "learning_rate": 4.159772107037613e-06,
+      "loss": 0.6114,
+      "step": 5456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.912758383634301,
+      "learning_rate": 4.158112677263231e-06,
+      "loss": 0.6577,
+      "step": 5457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.836405720086498,
+      "learning_rate": 4.156453342921573e-06,
+      "loss": 0.7016,
+      "step": 5458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9065901889070593,
+      "learning_rate": 4.154794104200737e-06,
+      "loss": 0.6624,
+      "step": 5459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9075230574920552,
+      "learning_rate": 4.153134961288807e-06,
+      "loss": 0.6226,
+      "step": 5460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8711562433698041,
+      "learning_rate": 4.151475914373856e-06,
+      "loss": 0.6264,
+      "step": 5461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8048208435526953,
+      "learning_rate": 4.149816963643947e-06,
+      "loss": 0.593,
+      "step": 5462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6991033271060485,
+      "learning_rate": 4.1481581092871305e-06,
+      "loss": 0.5303,
+      "step": 5463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7456484367991891,
+      "learning_rate": 4.146499351491448e-06,
+      "loss": 0.5549,
+      "step": 5464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7845115021251257,
+      "learning_rate": 4.144840690444931e-06,
+      "loss": 0.6408,
+      "step": 5465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.07574862295293,
+      "learning_rate": 4.143182126335594e-06,
+      "loss": 0.6681,
+      "step": 5466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.843927894099637,
+      "learning_rate": 4.1415236593514454e-06,
+      "loss": 0.5944,
+      "step": 5467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7210896919710381,
+      "learning_rate": 4.139865289680485e-06,
+      "loss": 0.6904,
+      "step": 5468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1143978026959203,
+      "learning_rate": 4.138207017510696e-06,
+      "loss": 0.7319,
+      "step": 5469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9674427272555244,
+      "learning_rate": 4.136548843030052e-06,
+      "loss": 0.6494,
+      "step": 5470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.753820948119458,
+      "learning_rate": 4.134890766426521e-06,
+      "loss": 0.6728,
+      "step": 5471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6638752593393304,
+      "learning_rate": 4.133232787888052e-06,
+      "loss": 0.5565,
+      "step": 5472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1632980049827677,
+      "learning_rate": 4.131574907602586e-06,
+      "loss": 0.6462,
+      "step": 5473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0407862239585697,
+      "learning_rate": 4.129917125758054e-06,
+      "loss": 0.746,
+      "step": 5474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.854988687102971,
+      "learning_rate": 4.128259442542375e-06,
+      "loss": 0.666,
+      "step": 5475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0598051706252316,
+      "learning_rate": 4.126601858143457e-06,
+      "loss": 0.6856,
+      "step": 5476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2416076777807854,
+      "learning_rate": 4.124944372749197e-06,
+      "loss": 0.7149,
+      "step": 5477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9940483790212515,
+      "learning_rate": 4.12328698654748e-06,
+      "loss": 0.678,
+      "step": 5478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9448515255469294,
+      "learning_rate": 4.12162969972618e-06,
+      "loss": 0.6333,
+      "step": 5479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7150709469269039,
+      "learning_rate": 4.119972512473161e-06,
+      "loss": 0.6189,
+      "step": 5480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8341509373262923,
+      "learning_rate": 4.118315424976272e-06,
+      "loss": 0.5073,
+      "step": 5481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9476966860099154,
+      "learning_rate": 4.116658437423355e-06,
+      "loss": 0.5557,
+      "step": 5482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6553886987138464,
+      "learning_rate": 4.115001550002241e-06,
+      "loss": 0.5275,
+      "step": 5483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9821949785533268,
+      "learning_rate": 4.113344762900745e-06,
+      "loss": 0.6504,
+      "step": 5484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7156562932709354,
+      "learning_rate": 4.111688076306673e-06,
+      "loss": 0.5742,
+      "step": 5485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0232269740630366,
+      "learning_rate": 4.11003149040782e-06,
+      "loss": 0.6493,
+      "step": 5486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8519773132537132,
+      "learning_rate": 4.108375005391972e-06,
+      "loss": 0.6184,
+      "step": 5487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.793916358582051,
+      "learning_rate": 4.106718621446899e-06,
+      "loss": 0.5546,
+      "step": 5488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.036030057119493,
+      "learning_rate": 4.105062338760361e-06,
+      "loss": 0.5802,
+      "step": 5489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7927973737970935,
+      "learning_rate": 4.103406157520108e-06,
+      "loss": 0.6194,
+      "step": 5490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6799135980861009,
+      "learning_rate": 4.101750077913878e-06,
+      "loss": 0.5536,
+      "step": 5491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.009000457580272,
+      "learning_rate": 4.100094100129396e-06,
+      "loss": 0.7814,
+      "step": 5492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2549031104223727,
+      "learning_rate": 4.098438224354377e-06,
+      "loss": 0.628,
+      "step": 5493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7853953706906511,
+      "learning_rate": 4.096782450776526e-06,
+      "loss": 0.6338,
+      "step": 5494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9185594643941042,
+      "learning_rate": 4.09512677958353e-06,
+      "loss": 0.5703,
+      "step": 5495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8102601216846244,
+      "learning_rate": 4.093471210963072e-06,
+      "loss": 0.5525,
+      "step": 5496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9988743835333525,
+      "learning_rate": 4.091815745102818e-06,
+      "loss": 0.6161,
+      "step": 5497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9896802093368413,
+      "learning_rate": 4.0901603821904264e-06,
+      "loss": 0.6264,
+      "step": 5498
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7811960844124857,
+      "learning_rate": 4.088505122413542e-06,
+      "loss": 0.6219,
+      "step": 5499
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.642234193822947,
+      "learning_rate": 4.086849965959797e-06,
+      "loss": 0.5674,
+      "step": 5500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1360764455341803,
+      "learning_rate": 4.0851949130168135e-06,
+      "loss": 0.763,
+      "step": 5501
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8119817360590291,
+      "learning_rate": 4.083539963772202e-06,
+      "loss": 0.6736,
+      "step": 5502
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0075224162284337,
+      "learning_rate": 4.081885118413557e-06,
+      "loss": 0.6061,
+      "step": 5503
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9394585405022449,
+      "learning_rate": 4.0802303771284685e-06,
+      "loss": 0.6544,
+      "step": 5504
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9226282044737457,
+      "learning_rate": 4.078575740104508e-06,
+      "loss": 0.6074,
+      "step": 5505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8428113935987287,
+      "learning_rate": 4.07692120752924e-06,
+      "loss": 0.5472,
+      "step": 5506
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.944596305847222,
+      "learning_rate": 4.075266779590214e-06,
+      "loss": 0.7045,
+      "step": 5507
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.164562137124973,
+      "learning_rate": 4.073612456474969e-06,
+      "loss": 0.5906,
+      "step": 5508
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.632993502982212,
+      "learning_rate": 4.071958238371036e-06,
+      "loss": 0.5673,
+      "step": 5509
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9024162945461678,
+      "learning_rate": 4.070304125465922e-06,
+      "loss": 0.5853,
+      "step": 5510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9007911838792717,
+      "learning_rate": 4.068650117947135e-06,
+      "loss": 0.6144,
+      "step": 5511
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9514097246456006,
+      "learning_rate": 4.066996216002167e-06,
+      "loss": 0.6627,
+      "step": 5512
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.018105134211878,
+      "learning_rate": 4.065342419818495e-06,
+      "loss": 0.6744,
+      "step": 5513
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.885304977006811,
+      "learning_rate": 4.063688729583587e-06,
+      "loss": 0.7094,
+      "step": 5514
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0783420987217838,
+      "learning_rate": 4.0620351454848986e-06,
+      "loss": 0.586,
+      "step": 5515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.088674812109934,
+      "learning_rate": 4.0603816677098735e-06,
+      "loss": 0.648,
+      "step": 5516
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0529187492775622,
+      "learning_rate": 4.058728296445941e-06,
+      "loss": 0.688,
+      "step": 5517
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.07021175838815,
+      "learning_rate": 4.057075031880521e-06,
+      "loss": 0.6682,
+      "step": 5518
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9584875967230313,
+      "learning_rate": 4.055421874201021e-06,
+      "loss": 0.6779,
+      "step": 5519
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9478184359951374,
+      "learning_rate": 4.053768823594835e-06,
+      "loss": 0.6618,
+      "step": 5520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9949873298042147,
+      "learning_rate": 4.052115880249347e-06,
+      "loss": 0.5795,
+      "step": 5521
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9743169662323066,
+      "learning_rate": 4.050463044351927e-06,
+      "loss": 0.5705,
+      "step": 5522
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9749688882641023,
+      "learning_rate": 4.048810316089932e-06,
+      "loss": 0.6088,
+      "step": 5523
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8137222978800382,
+      "learning_rate": 4.0471576956507106e-06,
+      "loss": 0.6511,
+      "step": 5524
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0162638592127324,
+      "learning_rate": 4.045505183221594e-06,
+      "loss": 0.6694,
+      "step": 5525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.231644995017,
+      "learning_rate": 4.043852778989906e-06,
+      "loss": 0.7455,
+      "step": 5526
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1407694111031104,
+      "learning_rate": 4.042200483142955e-06,
+      "loss": 0.574,
+      "step": 5527
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9598142070636422,
+      "learning_rate": 4.040548295868039e-06,
+      "loss": 0.5821,
+      "step": 5528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9740464430151743,
+      "learning_rate": 4.038896217352442e-06,
+      "loss": 0.6735,
+      "step": 5529
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6941300405779327,
+      "learning_rate": 4.037244247783437e-06,
+      "loss": 0.6544,
+      "step": 5530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8565441269875411,
+      "learning_rate": 4.035592387348285e-06,
+      "loss": 0.6343,
+      "step": 5531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0685840879545663,
+      "learning_rate": 4.033940636234233e-06,
+      "loss": 0.6154,
+      "step": 5532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8535815659947346,
+      "learning_rate": 4.0322889946285146e-06,
+      "loss": 0.5411,
+      "step": 5533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8589830851032711,
+      "learning_rate": 4.030637462718354e-06,
+      "loss": 0.5412,
+      "step": 5534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7381406224431815,
+      "learning_rate": 4.028986040690963e-06,
+      "loss": 0.5865,
+      "step": 5535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8576683960163518,
+      "learning_rate": 4.027334728733539e-06,
+      "loss": 0.6374,
+      "step": 5536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9290621874870129,
+      "learning_rate": 4.0256835270332676e-06,
+      "loss": 0.6227,
+      "step": 5537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9821450643649243,
+      "learning_rate": 4.024032435777321e-06,
+      "loss": 0.5536,
+      "step": 5538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7135125156331292,
+      "learning_rate": 4.022381455152863e-06,
+      "loss": 0.6002,
+      "step": 5539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9281196949391677,
+      "learning_rate": 4.020730585347038e-06,
+      "loss": 0.7845,
+      "step": 5540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.013273341855397,
+      "learning_rate": 4.019079826546982e-06,
+      "loss": 0.6069,
+      "step": 5541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8903020483922408,
+      "learning_rate": 4.017429178939819e-06,
+      "loss": 0.6474,
+      "step": 5542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.980021206707033,
+      "learning_rate": 4.01577864271266e-06,
+      "loss": 0.6513,
+      "step": 5543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8662321684702983,
+      "learning_rate": 4.0141282180526e-06,
+      "loss": 0.5716,
+      "step": 5544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9032406858177675,
+      "learning_rate": 4.012477905146726e-06,
+      "loss": 0.6737,
+      "step": 5545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.951128633499513,
+      "learning_rate": 4.010827704182113e-06,
+      "loss": 0.663,
+      "step": 5546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8364003831985076,
+      "learning_rate": 4.009177615345816e-06,
+      "loss": 0.6535,
+      "step": 5547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7446521841903204,
+      "learning_rate": 4.007527638824883e-06,
+      "loss": 0.5733,
+      "step": 5548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1231342278478005,
+      "learning_rate": 4.005877774806349e-06,
+      "loss": 0.6509,
+      "step": 5549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8927153794231493,
+      "learning_rate": 4.004228023477236e-06,
+      "loss": 0.6556,
+      "step": 5550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.867765315475591,
+      "learning_rate": 4.002578385024552e-06,
+      "loss": 0.6168,
+      "step": 5551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8285478053225521,
+      "learning_rate": 4.000928859635293e-06,
+      "loss": 0.6157,
+      "step": 5552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7997858856785665,
+      "learning_rate": 3.999279447496444e-06,
+      "loss": 0.6565,
+      "step": 5553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0598783545296144,
+      "learning_rate": 3.997630148794972e-06,
+      "loss": 0.579,
+      "step": 5554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.809587176377569,
+      "learning_rate": 3.995980963717836e-06,
+      "loss": 0.6207,
+      "step": 5555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.921765056383696,
+      "learning_rate": 3.994331892451981e-06,
+      "loss": 0.6272,
+      "step": 5556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9168894037206115,
+      "learning_rate": 3.992682935184338e-06,
+      "loss": 0.552,
+      "step": 5557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9846547269975716,
+      "learning_rate": 3.991034092101825e-06,
+      "loss": 0.6359,
+      "step": 5558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9654747327693463,
+      "learning_rate": 3.9893853633913485e-06,
+      "loss": 0.7966,
+      "step": 5559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8248422292882072,
+      "learning_rate": 3.987736749239804e-06,
+      "loss": 0.694,
+      "step": 5560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7068304249889794,
+      "learning_rate": 3.986088249834067e-06,
+      "loss": 0.5808,
+      "step": 5561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9279785384262589,
+      "learning_rate": 3.984439865361006e-06,
+      "loss": 0.6398,
+      "step": 5562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8561935533701657,
+      "learning_rate": 3.982791596007474e-06,
+      "loss": 0.6948,
+      "step": 5563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.037470040126886,
+      "learning_rate": 3.981143441960312e-06,
+      "loss": 0.6098,
+      "step": 5564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7159903024585526,
+      "learning_rate": 3.979495403406349e-06,
+      "loss": 0.6738,
+      "step": 5565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9450556424814869,
+      "learning_rate": 3.977847480532399e-06,
+      "loss": 0.6524,
+      "step": 5566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8435450586070443,
+      "learning_rate": 3.976199673525263e-06,
+      "loss": 0.6249,
+      "step": 5567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9349741783559788,
+      "learning_rate": 3.974551982571732e-06,
+      "loss": 0.7301,
+      "step": 5568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1214284817615923,
+      "learning_rate": 3.9729044078585756e-06,
+      "loss": 0.6747,
+      "step": 5569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9991903575884222,
+      "learning_rate": 3.971256949572559e-06,
+      "loss": 0.6091,
+      "step": 5570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.962993073595112,
+      "learning_rate": 3.969609607900431e-06,
+      "loss": 0.5745,
+      "step": 5571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0994401626194543,
+      "learning_rate": 3.967962383028927e-06,
+      "loss": 0.5469,
+      "step": 5572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8638963329996499,
+      "learning_rate": 3.966315275144769e-06,
+      "loss": 0.6496,
+      "step": 5573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.040176832504217,
+      "learning_rate": 3.964668284434666e-06,
+      "loss": 0.6288,
+      "step": 5574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.022872445806577,
+      "learning_rate": 3.963021411085315e-06,
+      "loss": 0.6499,
+      "step": 5575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9155250142977756,
+      "learning_rate": 3.9613746552834e-06,
+      "loss": 0.5525,
+      "step": 5576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0230746897832876,
+      "learning_rate": 3.959728017215585e-06,
+      "loss": 0.5354,
+      "step": 5577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1162250972308643,
+      "learning_rate": 3.958081497068528e-06,
+      "loss": 0.6553,
+      "step": 5578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7692582229440683,
+      "learning_rate": 3.9564350950288735e-06,
+      "loss": 0.6553,
+      "step": 5579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8739200142352377,
+      "learning_rate": 3.954788811283249e-06,
+      "loss": 0.6072,
+      "step": 5580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7011178415660841,
+      "learning_rate": 3.953142646018269e-06,
+      "loss": 0.5354,
+      "step": 5581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8492767816456872,
+      "learning_rate": 3.951496599420538e-06,
+      "loss": 0.7356,
+      "step": 5582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0115047324799478,
+      "learning_rate": 3.949850671676644e-06,
+      "loss": 0.6773,
+      "step": 5583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.068672760742573,
+      "learning_rate": 3.948204862973161e-06,
+      "loss": 0.6645,
+      "step": 5584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9316885249397973,
+      "learning_rate": 3.946559173496653e-06,
+      "loss": 0.7602,
+      "step": 5585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.813002548695951,
+      "learning_rate": 3.944913603433667e-06,
+      "loss": 0.6286,
+      "step": 5586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7041134204449735,
+      "learning_rate": 3.943268152970737e-06,
+      "loss": 0.6044,
+      "step": 5587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9429926106788047,
+      "learning_rate": 3.941622822294385e-06,
+      "loss": 0.7283,
+      "step": 5588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9588914791860121,
+      "learning_rate": 3.939977611591119e-06,
+      "loss": 0.5775,
+      "step": 5589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8812887191683487,
+      "learning_rate": 3.938332521047434e-06,
+      "loss": 0.6613,
+      "step": 5590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9635352056032394,
+      "learning_rate": 3.936687550849807e-06,
+      "loss": 0.6232,
+      "step": 5591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0963605977584754,
+      "learning_rate": 3.935042701184707e-06,
+      "loss": 0.661,
+      "step": 5592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8334345997394716,
+      "learning_rate": 3.9333979722385865e-06,
+      "loss": 0.6448,
+      "step": 5593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.802952264613462,
+      "learning_rate": 3.931753364197886e-06,
+      "loss": 0.6125,
+      "step": 5594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0955138413023127,
+      "learning_rate": 3.93010887724903e-06,
+      "loss": 0.6208,
+      "step": 5595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9727861271093967,
+      "learning_rate": 3.928464511578432e-06,
+      "loss": 0.6802,
+      "step": 5596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7483375199702063,
+      "learning_rate": 3.926820267372489e-06,
+      "loss": 0.5174,
+      "step": 5597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.079593531305218,
+      "learning_rate": 3.925176144817587e-06,
+      "loss": 0.6895,
+      "step": 5598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.115779088659057,
+      "learning_rate": 3.9235321441000925e-06,
+      "loss": 0.5738,
+      "step": 5599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0251026439895456,
+      "learning_rate": 3.921888265406366e-06,
+      "loss": 0.6485,
+      "step": 5600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7558793985516596,
+      "learning_rate": 3.92024450892275e-06,
+      "loss": 0.5925,
+      "step": 5601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.860407325614981,
+      "learning_rate": 3.918600874835573e-06,
+      "loss": 0.5922,
+      "step": 5602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.981471706438414,
+      "learning_rate": 3.9169573633311505e-06,
+      "loss": 0.6904,
+      "step": 5603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8167178665462902,
+      "learning_rate": 3.915313974595784e-06,
+      "loss": 0.5259,
+      "step": 5604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.487307591563184,
+      "learning_rate": 3.913670708815763e-06,
+      "loss": 0.5648,
+      "step": 5605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6661233435057807,
+      "learning_rate": 3.912027566177358e-06,
+      "loss": 0.5738,
+      "step": 5606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.049360076678344,
+      "learning_rate": 3.910384546866828e-06,
+      "loss": 0.6835,
+      "step": 5607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0422476557675906,
+      "learning_rate": 3.9087416510704206e-06,
+      "loss": 0.7541,
+      "step": 5608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0434482018310782,
+      "learning_rate": 3.907098878974367e-06,
+      "loss": 0.6639,
+      "step": 5609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1404651279397346,
+      "learning_rate": 3.905456230764886e-06,
+      "loss": 0.6493,
+      "step": 5610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1477378402307594,
+      "learning_rate": 3.903813706628179e-06,
+      "loss": 0.6731,
+      "step": 5611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6555033642042376,
+      "learning_rate": 3.902171306750436e-06,
+      "loss": 0.6208,
+      "step": 5612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9267343051568924,
+      "learning_rate": 3.900529031317836e-06,
+      "loss": 0.5961,
+      "step": 5613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0221313434196153,
+      "learning_rate": 3.898886880516533e-06,
+      "loss": 0.5812,
+      "step": 5614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9915391709592674,
+      "learning_rate": 3.89724485453268e-06,
+      "loss": 0.5604,
+      "step": 5615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9024492756973865,
+      "learning_rate": 3.895602953552408e-06,
+      "loss": 0.6102,
+      "step": 5616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0749917123549704,
+      "learning_rate": 3.893961177761835e-06,
+      "loss": 0.6328,
+      "step": 5617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6907876354852427,
+      "learning_rate": 3.8923195273470686e-06,
+      "loss": 0.6116,
+      "step": 5618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9039400354395462,
+      "learning_rate": 3.890678002494196e-06,
+      "loss": 0.693,
+      "step": 5619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109421413941969,
+      "learning_rate": 3.889036603389297e-06,
+      "loss": 0.6047,
+      "step": 5620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7020168441413386,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.6034,
+      "step": 5621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9585934375817815,
+      "learning_rate": 3.885754183167642e-06,
+      "loss": 0.5768,
+      "step": 5622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21563878724902,
+      "learning_rate": 3.884113162422971e-06,
+      "loss": 0.6703,
+      "step": 5623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8662332887713378,
+      "learning_rate": 3.882472268170432e-06,
+      "loss": 0.5811,
+      "step": 5624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0872931585370282,
+      "learning_rate": 3.8808315005960305e-06,
+      "loss": 0.6711,
+      "step": 5625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8026638556504297,
+      "learning_rate": 3.879190859885758e-06,
+      "loss": 0.6181,
+      "step": 5626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7741256125064448,
+      "learning_rate": 3.877550346225591e-06,
+      "loss": 0.7367,
+      "step": 5627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.012507567318553,
+      "learning_rate": 3.8759099598014895e-06,
+      "loss": 0.6178,
+      "step": 5628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0531286926195667,
+      "learning_rate": 3.8742697007994e-06,
+      "loss": 0.6073,
+      "step": 5629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8202572945880635,
+      "learning_rate": 3.872629569405257e-06,
+      "loss": 0.6469,
+      "step": 5630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.796969026042274,
+      "learning_rate": 3.870989565804979e-06,
+      "loss": 0.6439,
+      "step": 5631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1220990590037574,
+      "learning_rate": 3.869349690184468e-06,
+      "loss": 0.6905,
+      "step": 5632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0824399282601846,
+      "learning_rate": 3.867709942729613e-06,
+      "loss": 0.693,
+      "step": 5633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7845464618082878,
+      "learning_rate": 3.866070323626291e-06,
+      "loss": 0.5927,
+      "step": 5634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9299347986415278,
+      "learning_rate": 3.864430833060363e-06,
+      "loss": 0.6342,
+      "step": 5635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7117444670691249,
+      "learning_rate": 3.86279147121767e-06,
+      "loss": 0.6953,
+      "step": 5636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.828004952003326,
+      "learning_rate": 3.8611522382840476e-06,
+      "loss": 0.6119,
+      "step": 5637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8602018760685013,
+      "learning_rate": 3.859513134445308e-06,
+      "loss": 0.6153,
+      "step": 5638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8653514922373067,
+      "learning_rate": 3.857874159887256e-06,
+      "loss": 0.5868,
+      "step": 5639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.06982897795329,
+      "learning_rate": 3.856235314795678e-06,
+      "loss": 0.6817,
+      "step": 5640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1179240984389147,
+      "learning_rate": 3.8545965993563454e-06,
+      "loss": 0.6013,
+      "step": 5641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.819858665984067,
+      "learning_rate": 3.85295801375502e-06,
+      "loss": 0.5963,
+      "step": 5642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9063952687832015,
+      "learning_rate": 3.851319558177439e-06,
+      "loss": 0.6431,
+      "step": 5643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.191630045938788,
+      "learning_rate": 3.8496812328093335e-06,
+      "loss": 0.6341,
+      "step": 5644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.070202153822423,
+      "learning_rate": 3.848043037836417e-06,
+      "loss": 0.6904,
+      "step": 5645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.223577708768498,
+      "learning_rate": 3.846404973444388e-06,
+      "loss": 0.6196,
+      "step": 5646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8708658636793065,
+      "learning_rate": 3.8447670398189305e-06,
+      "loss": 0.6542,
+      "step": 5647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9332690937542027,
+      "learning_rate": 3.843129237145713e-06,
+      "loss": 0.6483,
+      "step": 5648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8760128233078004,
+      "learning_rate": 3.841491565610393e-06,
+      "loss": 0.6364,
+      "step": 5649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.833179006583722,
+      "learning_rate": 3.839854025398606e-06,
+      "loss": 0.5539,
+      "step": 5650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7178320341789401,
+      "learning_rate": 3.838216616695977e-06,
+      "loss": 0.5955,
+      "step": 5651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.043925966235596,
+      "learning_rate": 3.836579339688116e-06,
+      "loss": 0.5994,
+      "step": 5652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.013863381361781,
+      "learning_rate": 3.83494219456062e-06,
+      "loss": 0.7002,
+      "step": 5653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7479848563782898,
+      "learning_rate": 3.833305181499065e-06,
+      "loss": 0.6117,
+      "step": 5654
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.843085118521347,
+      "learning_rate": 3.831668300689019e-06,
+      "loss": 0.5921,
+      "step": 5655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8414043257902304,
+      "learning_rate": 3.8300315523160295e-06,
+      "loss": 0.6534,
+      "step": 5656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8521076862406898,
+      "learning_rate": 3.828394936565633e-06,
+      "loss": 0.6813,
+      "step": 5657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.964166539127327,
+      "learning_rate": 3.826758453623348e-06,
+      "loss": 0.5947,
+      "step": 5658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7844201494772378,
+      "learning_rate": 3.825122103674679e-06,
+      "loss": 0.6167,
+      "step": 5659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7340840036629404,
+      "learning_rate": 3.823485886905116e-06,
+      "loss": 0.6266,
+      "step": 5660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8099618683785281,
+      "learning_rate": 3.821849803500134e-06,
+      "loss": 0.6138,
+      "step": 5661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.944243244863238,
+      "learning_rate": 3.820213853645193e-06,
+      "loss": 0.6171,
+      "step": 5662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8178877996961225,
+      "learning_rate": 3.8185780375257356e-06,
+      "loss": 0.6275,
+      "step": 5663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9352643805301815,
+      "learning_rate": 3.816942355327191e-06,
+      "loss": 0.5816,
+      "step": 5664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8325507491789237,
+      "learning_rate": 3.815306807234974e-06,
+      "loss": 0.6927,
+      "step": 5665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.837901350432591,
+      "learning_rate": 3.8136713934344825e-06,
+      "loss": 0.6256,
+      "step": 5666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.111254838340173,
+      "learning_rate": 3.812036114111101e-06,
+      "loss": 0.5796,
+      "step": 5667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8187308591668736,
+      "learning_rate": 3.8104009694501963e-06,
+      "loss": 0.6156,
+      "step": 5668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8485495400562826,
+      "learning_rate": 3.808765959637123e-06,
+      "loss": 0.7708,
+      "step": 5669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8249275555676456,
+      "learning_rate": 3.8071310848572173e-06,
+      "loss": 0.5503,
+      "step": 5670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1032973373284682,
+      "learning_rate": 3.8054963452958025e-06,
+      "loss": 0.6755,
+      "step": 5671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9971403602282247,
+      "learning_rate": 3.8038617411381876e-06,
+      "loss": 0.6573,
+      "step": 5672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.795638769861292,
+      "learning_rate": 3.80222727256966e-06,
+      "loss": 0.6491,
+      "step": 5673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.871045292034552,
+      "learning_rate": 3.800592939775498e-06,
+      "loss": 0.5779,
+      "step": 5674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1159675641961306,
+      "learning_rate": 3.798958742940963e-06,
+      "loss": 0.6505,
+      "step": 5675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9957808122960192,
+      "learning_rate": 3.7973246822513e-06,
+      "loss": 0.6975,
+      "step": 5676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8929123275811768,
+      "learning_rate": 3.795690757891739e-06,
+      "loss": 0.6495,
+      "step": 5677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8000004498952757,
+      "learning_rate": 3.794056970047495e-06,
+      "loss": 0.6205,
+      "step": 5678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.016000272727104,
+      "learning_rate": 3.7924233189037697e-06,
+      "loss": 0.6353,
+      "step": 5679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9150551526282293,
+      "learning_rate": 3.7907898046457416e-06,
+      "loss": 0.6361,
+      "step": 5680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.905395762182365,
+      "learning_rate": 3.789156427458581e-06,
+      "loss": 0.6509,
+      "step": 5681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0270735773997557,
+      "learning_rate": 3.787523187527442e-06,
+      "loss": 0.6225,
+      "step": 5682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8523260107443824,
+      "learning_rate": 3.7858900850374596e-06,
+      "loss": 0.5487,
+      "step": 5683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8400704301179405,
+      "learning_rate": 3.7842571201737568e-06,
+      "loss": 0.6822,
+      "step": 5684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.213633936194134,
+      "learning_rate": 3.7826242931214386e-06,
+      "loss": 0.6599,
+      "step": 5685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.847807446734992,
+      "learning_rate": 3.780991604065598e-06,
+      "loss": 0.5762,
+      "step": 5686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7717364388265258,
+      "learning_rate": 3.7793590531913047e-06,
+      "loss": 0.5142,
+      "step": 5687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.06824980539038,
+      "learning_rate": 3.777726640683621e-06,
+      "loss": 0.6043,
+      "step": 5688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0951430699266713,
+      "learning_rate": 3.7760943667275884e-06,
+      "loss": 0.6901,
+      "step": 5689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9928453372788413,
+      "learning_rate": 3.7744622315082358e-06,
+      "loss": 0.5764,
+      "step": 5690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0971652173091364,
+      "learning_rate": 3.7728302352105743e-06,
+      "loss": 0.6022,
+      "step": 5691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0164644631521598,
+      "learning_rate": 3.7711983780196006e-06,
+      "loss": 0.5844,
+      "step": 5692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.121542379135994,
+      "learning_rate": 3.7695666601202944e-06,
+      "loss": 0.7505,
+      "step": 5693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7159088934028508,
+      "learning_rate": 3.767935081697622e-06,
+      "loss": 0.5846,
+      "step": 5694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7520439948117041,
+      "learning_rate": 3.766303642936529e-06,
+      "loss": 0.5672,
+      "step": 5695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7646205963873725,
+      "learning_rate": 3.76467234402195e-06,
+      "loss": 0.699,
+      "step": 5696
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7997889644069125,
+      "learning_rate": 3.763041185138802e-06,
+      "loss": 0.6302,
+      "step": 5697
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7988871525135666,
+      "learning_rate": 3.7614101664719866e-06,
+      "loss": 0.651,
+      "step": 5698
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8411621906976277,
+      "learning_rate": 3.759779288206388e-06,
+      "loss": 0.6294,
+      "step": 5699
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8931612809567004,
+      "learning_rate": 3.758148550526877e-06,
+      "loss": 0.5979,
+      "step": 5700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8264053952182302,
+      "learning_rate": 3.7565179536183067e-06,
+      "loss": 0.6323,
+      "step": 5701
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8994850468593727,
+      "learning_rate": 3.7548874976655126e-06,
+      "loss": 0.6624,
+      "step": 5702
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8525791579411612,
+      "learning_rate": 3.7532571828533175e-06,
+      "loss": 0.5778,
+      "step": 5703
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.914265211398974,
+      "learning_rate": 3.751627009366527e-06,
+      "loss": 0.6037,
+      "step": 5704
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9860797713158134,
+      "learning_rate": 3.7499969773899304e-06,
+      "loss": 0.6279,
+      "step": 5705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4453142555832756,
+      "learning_rate": 3.748367087108301e-06,
+      "loss": 0.6271,
+      "step": 5706
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3538812133651965,
+      "learning_rate": 3.7467373387063973e-06,
+      "loss": 0.715,
+      "step": 5707
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0233372795899656,
+      "learning_rate": 3.7451077323689587e-06,
+      "loss": 0.5361,
+      "step": 5708
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8810865910977235,
+      "learning_rate": 3.7434782682807137e-06,
+      "loss": 0.6368,
+      "step": 5709
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9864037253071856,
+      "learning_rate": 3.7418489466263663e-06,
+      "loss": 0.6532,
+      "step": 5710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7612719043231966,
+      "learning_rate": 3.740219767590613e-06,
+      "loss": 0.6277,
+      "step": 5711
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9914586483658314,
+      "learning_rate": 3.738590731358129e-06,
+      "loss": 0.6322,
+      "step": 5712
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0569132735077074,
+      "learning_rate": 3.736961838113575e-06,
+      "loss": 0.5606,
+      "step": 5713
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7802065674190661,
+      "learning_rate": 3.7353330880415963e-06,
+      "loss": 0.5631,
+      "step": 5714
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.919709196487885,
+      "learning_rate": 3.7337044813268204e-06,
+      "loss": 0.6351,
+      "step": 5715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.805335375870117,
+      "learning_rate": 3.732076018153861e-06,
+      "loss": 0.5069,
+      "step": 5716
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9844777052067826,
+      "learning_rate": 3.73044769870731e-06,
+      "loss": 0.6736,
+      "step": 5717
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9877835438191704,
+      "learning_rate": 3.728819523171748e-06,
+      "loss": 0.6128,
+      "step": 5718
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8767135977824283,
+      "learning_rate": 3.7271914917317397e-06,
+      "loss": 0.7063,
+      "step": 5719
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.979452378878533,
+      "learning_rate": 3.7255636045718295e-06,
+      "loss": 0.7067,
+      "step": 5720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.88533991292467,
+      "learning_rate": 3.723935861876549e-06,
+      "loss": 0.6882,
+      "step": 5721
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.027102272390913,
+      "learning_rate": 3.722308263830412e-06,
+      "loss": 0.659,
+      "step": 5722
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.023782394536249,
+      "learning_rate": 3.720680810617917e-06,
+      "loss": 0.6214,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.229036069263612,
+      "learning_rate": 3.719053502423543e-06,
+      "loss": 0.6,
+      "step": 5724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8634729871510112,
+      "learning_rate": 3.7174263394317565e-06,
+      "loss": 0.5997,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.162803651867326,
+      "learning_rate": 3.715799321827004e-06,
+      "loss": 0.7181,
+      "step": 5726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7501057047105535,
+      "learning_rate": 3.714172449793718e-06,
+      "loss": 0.6726,
+      "step": 5727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.821385233254919,
+      "learning_rate": 3.7125457235163144e-06,
+      "loss": 0.6426,
+      "step": 5728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.87698360403373,
+      "learning_rate": 3.7109191431791902e-06,
+      "loss": 0.5506,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9541379681456583,
+      "learning_rate": 3.7092927089667293e-06,
+      "loss": 0.6531,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9558109805930095,
+      "learning_rate": 3.7076664210632972e-06,
+      "loss": 0.665,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8327803792621096,
+      "learning_rate": 3.7060402796532414e-06,
+      "loss": 0.5689,
+      "step": 5732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7268754577479737,
+      "learning_rate": 3.7044142849208953e-06,
+      "loss": 0.6506,
+      "step": 5733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8842322171157952,
+      "learning_rate": 3.7027884370505753e-06,
+      "loss": 0.6301,
+      "step": 5734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7694469325931714,
+      "learning_rate": 3.701162736226579e-06,
+      "loss": 0.5931,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9551315711768495,
+      "learning_rate": 3.699537182633189e-06,
+      "loss": 0.6189,
+      "step": 5736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0096838711866436,
+      "learning_rate": 3.6979117764546735e-06,
+      "loss": 0.7111,
+      "step": 5737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.688583682617302,
+      "learning_rate": 3.6962865178752804e-06,
+      "loss": 0.5609,
+      "step": 5738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8099317008081695,
+      "learning_rate": 3.69466140707924e-06,
+      "loss": 0.7394,
+      "step": 5739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0311676895133384,
+      "learning_rate": 3.6930364442507693e-06,
+      "loss": 0.6102,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953417217917916,
+      "learning_rate": 3.6914116295740678e-06,
+      "loss": 0.6152,
+      "step": 5741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.871734797375731,
+      "learning_rate": 3.6897869632333157e-06,
+      "loss": 0.554,
+      "step": 5742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9371609557756013,
+      "learning_rate": 3.6881624454126797e-06,
+      "loss": 0.6307,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9869409405219924,
+      "learning_rate": 3.686538076296307e-06,
+      "loss": 0.6393,
+      "step": 5744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1734737870284913,
+      "learning_rate": 3.6849138560683305e-06,
+      "loss": 0.6261,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9689826803452837,
+      "learning_rate": 3.683289784912866e-06,
+      "loss": 0.5688,
+      "step": 5746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8841430860004258,
+      "learning_rate": 3.681665863014008e-06,
+      "loss": 0.5939,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.068574065898926,
+      "learning_rate": 3.6800420905558378e-06,
+      "loss": 0.6049,
+      "step": 5748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8856912763649123,
+      "learning_rate": 3.6784184677224204e-06,
+      "loss": 0.6656,
+      "step": 5749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.067502585225236,
+      "learning_rate": 3.6767949946978026e-06,
+      "loss": 0.6474,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.041166249135522,
+      "learning_rate": 3.6751716716660146e-06,
+      "loss": 0.5992,
+      "step": 5751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.934731123565553,
+      "learning_rate": 3.673548498811068e-06,
+      "loss": 0.6147,
+      "step": 5752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7182470344213379,
+      "learning_rate": 3.671925476316962e-06,
+      "loss": 0.6081,
+      "step": 5753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2804452109132676,
+      "learning_rate": 3.6703026043676715e-06,
+      "loss": 0.7114,
+      "step": 5754
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.216666784325599,
+      "learning_rate": 3.6686798831471594e-06,
+      "loss": 0.6938,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7060098887404596,
+      "learning_rate": 3.6670573128393704e-06,
+      "loss": 0.562,
+      "step": 5756
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.781597914983373,
+      "learning_rate": 3.6654348936282324e-06,
+      "loss": 0.6731,
+      "step": 5757
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.880276423977131,
+      "learning_rate": 3.663812625697657e-06,
+      "loss": 0.6672,
+      "step": 5758
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0046983496227435,
+      "learning_rate": 3.6621905092315357e-06,
+      "loss": 0.6595,
+      "step": 5759
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8122835785911924,
+      "learning_rate": 3.6605685444137463e-06,
+      "loss": 0.6348,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8979209319451233,
+      "learning_rate": 3.658946731428147e-06,
+      "loss": 0.6303,
+      "step": 5761
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9871359315563444,
+      "learning_rate": 3.6573250704585783e-06,
+      "loss": 0.673,
+      "step": 5762
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.174620750981435,
+      "learning_rate": 3.655703561688867e-06,
+      "loss": 0.7261,
+      "step": 5763
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9509384450832072,
+      "learning_rate": 3.6540822053028185e-06,
+      "loss": 0.603,
+      "step": 5764
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.123960386898783,
+      "learning_rate": 3.6524610014842234e-06,
+      "loss": 0.6584,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8173412461480742,
+      "learning_rate": 3.6508399504168547e-06,
+      "loss": 0.6287,
+      "step": 5766
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9404913184513584,
+      "learning_rate": 3.6492190522844673e-06,
+      "loss": 0.656,
+      "step": 5767
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7743176436764714,
+      "learning_rate": 3.647598307270801e-06,
+      "loss": 0.5844,
+      "step": 5768
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8949196446662082,
+      "learning_rate": 3.6459777155595733e-06,
+      "loss": 0.6704,
+      "step": 5769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2475526530936225,
+      "learning_rate": 3.64435727733449e-06,
+      "loss": 0.6862,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9168800541401423,
+      "learning_rate": 3.6427369927792354e-06,
+      "loss": 0.5947,
+      "step": 5771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9223629195860992,
+      "learning_rate": 3.6411168620774795e-06,
+      "loss": 0.5721,
+      "step": 5772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9231743963862156,
+      "learning_rate": 3.639496885412872e-06,
+      "loss": 0.5919,
+      "step": 5773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6447425060429046,
+      "learning_rate": 3.6378770629690476e-06,
+      "loss": 0.581,
+      "step": 5774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0694344031077168,
+      "learning_rate": 3.6362573949296242e-06,
+      "loss": 0.586,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7199178479200063,
+      "learning_rate": 3.634637881478196e-06,
+      "loss": 0.6447,
+      "step": 5776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1261329402743763,
+      "learning_rate": 3.633018522798346e-06,
+      "loss": 0.5911,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0734597415985627,
+      "learning_rate": 3.6313993190736375e-06,
+      "loss": 0.6408,
+      "step": 5778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0006256078542535,
+      "learning_rate": 3.629780270487617e-06,
+      "loss": 0.6043,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9745647927432075,
+      "learning_rate": 3.628161377223814e-06,
+      "loss": 0.7071,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6602642905256981,
+      "learning_rate": 3.626542639465738e-06,
+      "loss": 0.5602,
+      "step": 5781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9001677004630113,
+      "learning_rate": 3.624924057396883e-06,
+      "loss": 0.6861,
+      "step": 5782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0813915739240447,
+      "learning_rate": 3.6233056312007226e-06,
+      "loss": 0.7129,
+      "step": 5783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0518818520506477,
+      "learning_rate": 3.6216873610607155e-06,
+      "loss": 0.5889,
+      "step": 5784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.726267024963287,
+      "learning_rate": 3.620069247160303e-06,
+      "loss": 0.6046,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6789114214946894,
+      "learning_rate": 3.618451289682905e-06,
+      "loss": 0.6614,
+      "step": 5786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8891420844312576,
+      "learning_rate": 3.6168334888119295e-06,
+      "loss": 0.604,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6271989700635565,
+      "learning_rate": 3.6152158447307607e-06,
+      "loss": 0.5604,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.960085702524464,
+      "learning_rate": 3.6135983576227694e-06,
+      "loss": 0.7115,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.055291396604353,
+      "learning_rate": 3.6119810276713085e-06,
+      "loss": 0.6249,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8777115393478256,
+      "learning_rate": 3.6103638550597074e-06,
+      "loss": 0.6346,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8694877602773852,
+      "learning_rate": 3.6087468399712842e-06,
+      "loss": 0.6529,
+      "step": 5792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.007269671680019,
+      "learning_rate": 3.607129982589337e-06,
+      "loss": 0.5934,
+      "step": 5793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.771068848538593,
+      "learning_rate": 3.6055132830971446e-06,
+      "loss": 0.5829,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9968602046327504,
+      "learning_rate": 3.60389674167797e-06,
+      "loss": 0.6455,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9085289823709048,
+      "learning_rate": 3.6022803585150574e-06,
+      "loss": 0.6621,
+      "step": 5796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8562577460156162,
+      "learning_rate": 3.6006641337916335e-06,
+      "loss": 0.6193,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0116263130134566,
+      "learning_rate": 3.5990480676909055e-06,
+      "loss": 0.7013,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8696463786190631,
+      "learning_rate": 3.597432160396064e-06,
+      "loss": 0.592,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9258705805458398,
+      "learning_rate": 3.5958164120902816e-06,
+      "loss": 0.6927,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.788877244074425,
+      "learning_rate": 3.5942008229567128e-06,
+      "loss": 0.6083,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.819861269764366,
+      "learning_rate": 3.592585393178494e-06,
+      "loss": 0.5593,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.176572903100845,
+      "learning_rate": 3.590970122938742e-06,
+      "loss": 0.6081,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8421415635256122,
+      "learning_rate": 3.589355012420558e-06,
+      "loss": 0.5901,
+      "step": 5804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7830410387814137,
+      "learning_rate": 3.587740061807024e-06,
+      "loss": 0.7421,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.901456241678915,
+      "learning_rate": 3.5861252712812032e-06,
+      "loss": 0.5899,
+      "step": 5806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.740330243536855,
+      "learning_rate": 3.5845106410261417e-06,
+      "loss": 0.6182,
+      "step": 5807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9849175676505417,
+      "learning_rate": 3.5828961712248667e-06,
+      "loss": 0.6465,
+      "step": 5808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7925579462512344,
+      "learning_rate": 3.5812818620603883e-06,
+      "loss": 0.6665,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.963203797068177,
+      "learning_rate": 3.579667713715697e-06,
+      "loss": 0.6477,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.90177521042788,
+      "learning_rate": 3.5780537263737657e-06,
+      "loss": 0.717,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9095440405290003,
+      "learning_rate": 3.576439900217552e-06,
+      "loss": 0.6262,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0776672586275713,
+      "learning_rate": 3.574826235429988e-06,
+      "loss": 0.6042,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.115976450204016,
+      "learning_rate": 3.5732127321939925e-06,
+      "loss": 0.7406,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8460290362187606,
+      "learning_rate": 3.5715993906924663e-06,
+      "loss": 0.5892,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8060615761663243,
+      "learning_rate": 3.5699862111082917e-06,
+      "loss": 0.6119,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8786816582722556,
+      "learning_rate": 3.56837319362433e-06,
+      "loss": 0.6141,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8902089472404162,
+      "learning_rate": 3.566760338423427e-06,
+      "loss": 0.6255,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.859106348664016,
+      "learning_rate": 3.5651476456884103e-06,
+      "loss": 0.6933,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.06543221051437,
+      "learning_rate": 3.5635351156020853e-06,
+      "loss": 0.6413,
+      "step": 5820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.929133796522785,
+      "learning_rate": 3.5619227483472417e-06,
+      "loss": 0.6346,
+      "step": 5821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7355976170586247,
+      "learning_rate": 3.560310544106652e-06,
+      "loss": 0.5988,
+      "step": 5822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.876723408507209,
+      "learning_rate": 3.5586985030630685e-06,
+      "loss": 0.5479,
+      "step": 5823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7976909367236422,
+      "learning_rate": 3.557086625399224e-06,
+      "loss": 0.6151,
+      "step": 5824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7457066314116323,
+      "learning_rate": 3.555474911297835e-06,
+      "loss": 0.6609,
+      "step": 5825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8359401944106575,
+      "learning_rate": 3.553863360941598e-06,
+      "loss": 0.6886,
+      "step": 5826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9190157263515015,
+      "learning_rate": 3.552251974513194e-06,
+      "loss": 0.6556,
+      "step": 5827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7954112329008878,
+      "learning_rate": 3.5506407521952783e-06,
+      "loss": 0.598,
+      "step": 5828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7418220797119244,
+      "learning_rate": 3.5490296941704948e-06,
+      "loss": 0.6735,
+      "step": 5829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8969973614397408,
+      "learning_rate": 3.547418800621466e-06,
+      "loss": 0.6207,
+      "step": 5830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.950442180880578,
+      "learning_rate": 3.545808071730795e-06,
+      "loss": 0.604,
+      "step": 5831
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6785701177368428,
+      "learning_rate": 3.544197507681068e-06,
+      "loss": 0.6512,
+      "step": 5832
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7768027074664066,
+      "learning_rate": 3.5425871086548513e-06,
+      "loss": 0.5867,
+      "step": 5833
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8996288570794853,
+      "learning_rate": 3.540976874834694e-06,
+      "loss": 0.6475,
+      "step": 5834
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1207513817114783,
+      "learning_rate": 3.539366806403123e-06,
+      "loss": 0.6733,
+      "step": 5835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8518276147104666,
+      "learning_rate": 3.5377569035426494e-06,
+      "loss": 0.6066,
+      "step": 5836
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9845097936191345,
+      "learning_rate": 3.536147166435765e-06,
+      "loss": 0.7186,
+      "step": 5837
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8353019441328098,
+      "learning_rate": 3.534537595264944e-06,
+      "loss": 0.6514,
+      "step": 5838
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7635849381509756,
+      "learning_rate": 3.532928190212639e-06,
+      "loss": 0.6148,
+      "step": 5839
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.650677732272397,
+      "learning_rate": 3.5313189514612867e-06,
+      "loss": 0.5851,
+      "step": 5840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9688433425039455,
+      "learning_rate": 3.529709879193301e-06,
+      "loss": 0.607,
+      "step": 5841
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7671648226840087,
+      "learning_rate": 3.5281009735910822e-06,
+      "loss": 0.5911,
+      "step": 5842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4477526780203385,
+      "learning_rate": 3.5264922348370066e-06,
+      "loss": 0.6202,
+      "step": 5843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7697614181136327,
+      "learning_rate": 3.524883663113435e-06,
+      "loss": 0.6076,
+      "step": 5844
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8436384469037794,
+      "learning_rate": 3.523275258602708e-06,
+      "loss": 0.667,
+      "step": 5845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7263835471812945,
+      "learning_rate": 3.5216670214871475e-06,
+      "loss": 0.6649,
+      "step": 5846
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7266754034601193,
+      "learning_rate": 3.520058951949056e-06,
+      "loss": 0.5832,
+      "step": 5847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9772191883626575,
+      "learning_rate": 3.518451050170718e-06,
+      "loss": 0.6297,
+      "step": 5848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.903340236332747,
+      "learning_rate": 3.5168433163344005e-06,
+      "loss": 0.6192,
+      "step": 5849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7564343946683172,
+      "learning_rate": 3.5152357506223444e-06,
+      "loss": 0.5977,
+      "step": 5850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7624806514627709,
+      "learning_rate": 3.5136283532167786e-06,
+      "loss": 0.578,
+      "step": 5851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7577685153115692,
+      "learning_rate": 3.5120211242999115e-06,
+      "loss": 0.623,
+      "step": 5852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0273575366018726,
+      "learning_rate": 3.5104140640539302e-06,
+      "loss": 0.5983,
+      "step": 5853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.945975183296312,
+      "learning_rate": 3.508807172661006e-06,
+      "loss": 0.6439,
+      "step": 5854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9502084132528867,
+      "learning_rate": 3.5072004503032876e-06,
+      "loss": 0.6141,
+      "step": 5855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7862146823553806,
+      "learning_rate": 3.5055938971629096e-06,
+      "loss": 0.587,
+      "step": 5856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8574712577859098,
+      "learning_rate": 3.5039875134219784e-06,
+      "loss": 0.6534,
+      "step": 5857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8248444230202323,
+      "learning_rate": 3.5023812992625905e-06,
+      "loss": 0.625,
+      "step": 5858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.85422148267425,
+      "learning_rate": 3.5007752548668173e-06,
+      "loss": 0.5738,
+      "step": 5859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.638592929436304,
+      "learning_rate": 3.499169380416715e-06,
+      "loss": 0.5042,
+      "step": 5860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0112792439563982,
+      "learning_rate": 3.4975636760943177e-06,
+      "loss": 0.5882,
+      "step": 5861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7295258052488887,
+      "learning_rate": 3.4959581420816413e-06,
+      "loss": 0.6108,
+      "step": 5862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9019332337346095,
+      "learning_rate": 3.4943527785606824e-06,
+      "loss": 0.6732,
+      "step": 5863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9081740044039712,
+      "learning_rate": 3.49274758571342e-06,
+      "loss": 0.5051,
+      "step": 5864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8635557360728945,
+      "learning_rate": 3.491142563721808e-06,
+      "loss": 0.6049,
+      "step": 5865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.583335793635603,
+      "learning_rate": 3.489537712767786e-06,
+      "loss": 0.5906,
+      "step": 5866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7730511646734124,
+      "learning_rate": 3.487933033033274e-06,
+      "loss": 0.6527,
+      "step": 5867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125335096700481,
+      "learning_rate": 3.486328524700171e-06,
+      "loss": 0.5927,
+      "step": 5868
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0856948814252436,
+      "learning_rate": 3.4847241879503574e-06,
+      "loss": 0.7139,
+      "step": 5869
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8465188230889573,
+      "learning_rate": 3.4831200229656935e-06,
+      "loss": 0.6565,
+      "step": 5870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8051867760992228,
+      "learning_rate": 3.4815160299280225e-06,
+      "loss": 0.5733,
+      "step": 5871
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9613480970154895,
+      "learning_rate": 3.4799122090191638e-06,
+      "loss": 0.633,
+      "step": 5872
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7007000511001091,
+      "learning_rate": 3.47830856042092e-06,
+      "loss": 0.5902,
+      "step": 5873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7704526339579787,
+      "learning_rate": 3.476705084315074e-06,
+      "loss": 0.5984,
+      "step": 5874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0012874669262692,
+      "learning_rate": 3.47510178088339e-06,
+      "loss": 0.6879,
+      "step": 5875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8050110100253667,
+      "learning_rate": 3.4734986503076096e-06,
+      "loss": 0.5875,
+      "step": 5876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.132743737087264,
+      "learning_rate": 3.4718956927694593e-06,
+      "loss": 0.6638,
+      "step": 5877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.848961511546181,
+      "learning_rate": 3.4702929084506433e-06,
+      "loss": 0.6154,
+      "step": 5878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7426350866830753,
+      "learning_rate": 3.468690297532843e-06,
+      "loss": 0.57,
+      "step": 5879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9103389699630549,
+      "learning_rate": 3.467087860197726e-06,
+      "loss": 0.6717,
+      "step": 5880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9028495231307951,
+      "learning_rate": 3.4654855966269373e-06,
+      "loss": 0.625,
+      "step": 5881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.888079361566811,
+      "learning_rate": 3.4638835070021027e-06,
+      "loss": 0.6302,
+      "step": 5882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.845351370141749,
+      "learning_rate": 3.462281591504828e-06,
+      "loss": 0.6333,
+      "step": 5883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.732023585322755,
+      "learning_rate": 3.4606798503166994e-06,
+      "loss": 0.6475,
+      "step": 5884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8978626563200731,
+      "learning_rate": 3.4590782836192837e-06,
+      "loss": 0.6542,
+      "step": 5885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.869401160752364,
+      "learning_rate": 3.45747689159413e-06,
+      "loss": 0.5564,
+      "step": 5886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.868999390580027,
+      "learning_rate": 3.455875674422761e-06,
+      "loss": 0.579,
+      "step": 5887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.832886798253401,
+      "learning_rate": 3.4542746322866842e-06,
+      "loss": 0.4929,
+      "step": 5888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.913784743810917,
+      "learning_rate": 3.452673765367389e-06,
+      "loss": 0.6379,
+      "step": 5889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9179201733844766,
+      "learning_rate": 3.4510730738463417e-06,
+      "loss": 0.6739,
+      "step": 5890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8110837161807765,
+      "learning_rate": 3.4494725579049904e-06,
+      "loss": 0.6973,
+      "step": 5891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.987201446085457,
+      "learning_rate": 3.4478722177247624e-06,
+      "loss": 0.6087,
+      "step": 5892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9980930635109033,
+      "learning_rate": 3.4462720534870673e-06,
+      "loss": 0.8181,
+      "step": 5893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8706741387059407,
+      "learning_rate": 3.4446720653732883e-06,
+      "loss": 0.6902,
+      "step": 5894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9409228527860711,
+      "learning_rate": 3.4430722535647966e-06,
+      "loss": 0.6133,
+      "step": 5895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0029309963359085,
+      "learning_rate": 3.4414726182429388e-06,
+      "loss": 0.6602,
+      "step": 5896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8613225733512109,
+      "learning_rate": 3.439873159589043e-06,
+      "loss": 0.6484,
+      "step": 5897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.332715989410047,
+      "learning_rate": 3.438273877784417e-06,
+      "loss": 0.6361,
+      "step": 5898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.988217349346568,
+      "learning_rate": 3.4366747730103486e-06,
+      "loss": 0.5936,
+      "step": 5899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8721773260571348,
+      "learning_rate": 3.435075845448105e-06,
+      "loss": 0.5663,
+      "step": 5900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9144622101502546,
+      "learning_rate": 3.4334770952789354e-06,
+      "loss": 0.5372,
+      "step": 5901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1063634623324843,
+      "learning_rate": 3.4318785226840646e-06,
+      "loss": 0.5908,
+      "step": 5902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.920747599443631,
+      "learning_rate": 3.4302801278447028e-06,
+      "loss": 0.6181,
+      "step": 5903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3644717939435353,
+      "learning_rate": 3.4286819109420346e-06,
+      "loss": 0.7202,
+      "step": 5904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9060094333028947,
+      "learning_rate": 3.4270838721572277e-06,
+      "loss": 0.6332,
+      "step": 5905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8361097469958723,
+      "learning_rate": 3.4254860116714284e-06,
+      "loss": 0.6645,
+      "step": 5906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9388690291273964,
+      "learning_rate": 3.4238883296657656e-06,
+      "loss": 0.5735,
+      "step": 5907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9303120106787002,
+      "learning_rate": 3.4222908263213438e-06,
+      "loss": 0.6181,
+      "step": 5908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8319157145588998,
+      "learning_rate": 3.4206935018192496e-06,
+      "loss": 0.5852,
+      "step": 5909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8890578866580727,
+      "learning_rate": 3.4190963563405482e-06,
+      "loss": 0.6554,
+      "step": 5910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.719008543508964,
+      "learning_rate": 3.4174993900662854e-06,
+      "loss": 0.5783,
+      "step": 5911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9381651528278054,
+      "learning_rate": 3.4159026031774873e-06,
+      "loss": 0.6329,
+      "step": 5912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9071649235590062,
+      "learning_rate": 3.4143059958551576e-06,
+      "loss": 0.5583,
+      "step": 5913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6027346173908796,
+      "learning_rate": 3.4127095682802823e-06,
+      "loss": 0.6465,
+      "step": 5914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7520728181991414,
+      "learning_rate": 3.4111133206338257e-06,
+      "loss": 0.615,
+      "step": 5915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.061567472873294,
+      "learning_rate": 3.409517253096729e-06,
+      "loss": 0.6207,
+      "step": 5916
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.017959712602565,
+      "learning_rate": 3.407921365849917e-06,
+      "loss": 0.6219,
+      "step": 5917
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6541870573418043,
+      "learning_rate": 3.406325659074293e-06,
+      "loss": 0.5519,
+      "step": 5918
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0052219645499934,
+      "learning_rate": 3.404730132950739e-06,
+      "loss": 0.6655,
+      "step": 5919
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.373349253904208,
+      "learning_rate": 3.403134787660117e-06,
+      "loss": 0.638,
+      "step": 5920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9857596614742279,
+      "learning_rate": 3.4015396233832687e-06,
+      "loss": 0.6129,
+      "step": 5921
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9865545357725798,
+      "learning_rate": 3.3999446403010156e-06,
+      "loss": 0.6958,
+      "step": 5922
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2225225785812492,
+      "learning_rate": 3.398349838594159e-06,
+      "loss": 0.6527,
+      "step": 5923
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.965865099296652,
+      "learning_rate": 3.3967552184434753e-06,
+      "loss": 0.6616,
+      "step": 5924
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7599925725199548,
+      "learning_rate": 3.395160780029726e-06,
+      "loss": 0.6546,
+      "step": 5925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7892999068048223,
+      "learning_rate": 3.393566523533649e-06,
+      "loss": 0.5618,
+      "step": 5926
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.907038190708332,
+      "learning_rate": 3.391972449135964e-06,
+      "loss": 0.6594,
+      "step": 5927
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0532475771298313,
+      "learning_rate": 3.3903785570173665e-06,
+      "loss": 0.7017,
+      "step": 5928
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8848373537785805,
+      "learning_rate": 3.388784847358534e-06,
+      "loss": 0.636,
+      "step": 5929
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.710821951197328,
+      "learning_rate": 3.387191320340125e-06,
+      "loss": 0.6399,
+      "step": 5930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8487227564053192,
+      "learning_rate": 3.3855979761427705e-06,
+      "loss": 0.6817,
+      "step": 5931
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.907129570306386,
+      "learning_rate": 3.384004814947087e-06,
+      "loss": 0.6738,
+      "step": 5932
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8995663612234321,
+      "learning_rate": 3.382411836933669e-06,
+      "loss": 0.5573,
+      "step": 5933
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.756986647365497,
+      "learning_rate": 3.3808190422830887e-06,
+      "loss": 0.6431,
+      "step": 5934
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7715018375793272,
+      "learning_rate": 3.379226431175899e-06,
+      "loss": 0.7003,
+      "step": 5935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.035648914953771,
+      "learning_rate": 3.377634003792632e-06,
+      "loss": 0.6059,
+      "step": 5936
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.014535610980337,
+      "learning_rate": 3.3760417603137976e-06,
+      "loss": 0.5789,
+      "step": 5937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7341296370744583,
+      "learning_rate": 3.374449700919887e-06,
+      "loss": 0.6037,
+      "step": 5938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9227588740393737,
+      "learning_rate": 3.372857825791367e-06,
+      "loss": 0.7159,
+      "step": 5939
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8532934737084343,
+      "learning_rate": 3.371266135108687e-06,
+      "loss": 0.6038,
+      "step": 5940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8469207224534798,
+      "learning_rate": 3.3696746290522737e-06,
+      "loss": 0.6791,
+      "step": 5941
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9161878551856997,
+      "learning_rate": 3.368083307802535e-06,
+      "loss": 0.5931,
+      "step": 5942
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7666500679293722,
+      "learning_rate": 3.3664921715398534e-06,
+      "loss": 0.5605,
+      "step": 5943
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9254176807282661,
+      "learning_rate": 3.3649012204445953e-06,
+      "loss": 0.564,
+      "step": 5944
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9573810160395808,
+      "learning_rate": 3.3633104546971052e-06,
+      "loss": 0.6968,
+      "step": 5945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0044533425543785,
+      "learning_rate": 3.3617198744777023e-06,
+      "loss": 0.6491,
+      "step": 5946
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.763232630775274,
+      "learning_rate": 3.3601294799666896e-06,
+      "loss": 0.6697,
+      "step": 5947
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.098459423812884,
+      "learning_rate": 3.3585392713443464e-06,
+      "loss": 0.572,
+      "step": 5948
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6914031236102578,
+      "learning_rate": 3.356949248790934e-06,
+      "loss": 0.5091,
+      "step": 5949
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1716317317901366,
+      "learning_rate": 3.3553594124866897e-06,
+      "loss": 0.5757,
+      "step": 5950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0048322695335354,
+      "learning_rate": 3.3537697626118286e-06,
+      "loss": 0.5663,
+      "step": 5951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9550871662432008,
+      "learning_rate": 3.3521802993465513e-06,
+      "loss": 0.6769,
+      "step": 5952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.94484331544079,
+      "learning_rate": 3.350591022871027e-06,
+      "loss": 0.6025,
+      "step": 5953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8226399499008687,
+      "learning_rate": 3.349001933365411e-06,
+      "loss": 0.6214,
+      "step": 5954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8242034051010882,
+      "learning_rate": 3.3474130310098373e-06,
+      "loss": 0.646,
+      "step": 5955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9622212853904062,
+      "learning_rate": 3.345824315984415e-06,
+      "loss": 0.6289,
+      "step": 5956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.120458213494939,
+      "learning_rate": 3.3442357884692354e-06,
+      "loss": 0.6833,
+      "step": 5957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1252064549314236,
+      "learning_rate": 3.3426474486443673e-06,
+      "loss": 0.5385,
+      "step": 5958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9993862933069262,
+      "learning_rate": 3.3410592966898565e-06,
+      "loss": 0.5866,
+      "step": 5959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.794889673681397,
+      "learning_rate": 3.3394713327857325e-06,
+      "loss": 0.5263,
+      "step": 5960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.856636747820928,
+      "learning_rate": 3.3378835571119953e-06,
+      "loss": 0.5611,
+      "step": 5961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7722291953552272,
+      "learning_rate": 3.3362959698486307e-06,
+      "loss": 0.6446,
+      "step": 5962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0901179185866945,
+      "learning_rate": 3.3347085711756012e-06,
+      "loss": 0.5646,
+      "step": 5963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.836967165172361,
+      "learning_rate": 3.333121361272847e-06,
+      "loss": 0.5881,
+      "step": 5964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.102039726838115,
+      "learning_rate": 3.331534340320287e-06,
+      "loss": 0.644,
+      "step": 5965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9283561150549664,
+      "learning_rate": 3.3299475084978195e-06,
+      "loss": 0.67,
+      "step": 5966
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9259529340587442,
+      "learning_rate": 3.328360865985323e-06,
+      "loss": 0.5941,
+      "step": 5967
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.943892691297661,
+      "learning_rate": 3.3267744129626483e-06,
+      "loss": 0.6427,
+      "step": 5968
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9953597123623206,
+      "learning_rate": 3.3251881496096313e-06,
+      "loss": 0.6807,
+      "step": 5969
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6739874388513096,
+      "learning_rate": 3.3236020761060834e-06,
+      "loss": 0.4788,
+      "step": 5970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0506706084869584,
+      "learning_rate": 3.322016192631795e-06,
+      "loss": 0.6735,
+      "step": 5971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9665235766949143,
+      "learning_rate": 3.320430499366536e-06,
+      "loss": 0.6661,
+      "step": 5972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7042912287859477,
+      "learning_rate": 3.3188449964900527e-06,
+      "loss": 0.5868,
+      "step": 5973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7429369415654616,
+      "learning_rate": 3.3172596841820713e-06,
+      "loss": 0.5986,
+      "step": 5974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.846606375173926,
+      "learning_rate": 3.315674562622297e-06,
+      "loss": 0.604,
+      "step": 5975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8538542112748246,
+      "learning_rate": 3.31408963199041e-06,
+      "loss": 0.6191,
+      "step": 5976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9148060253706294,
+      "learning_rate": 3.312504892466073e-06,
+      "loss": 0.6505,
+      "step": 5977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.78279362426924,
+      "learning_rate": 3.310920344228925e-06,
+      "loss": 0.6188,
+      "step": 5978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8717477405727647,
+      "learning_rate": 3.3093359874585832e-06,
+      "loss": 0.6894,
+      "step": 5979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.749990881968654,
+      "learning_rate": 3.3077518223346448e-06,
+      "loss": 0.6087,
+      "step": 5980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7233031652705608,
+      "learning_rate": 3.3061678490366824e-06,
+      "loss": 0.649,
+      "step": 5981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.962583380637211,
+      "learning_rate": 3.3045840677442485e-06,
+      "loss": 0.5858,
+      "step": 5982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1968870344761626,
+      "learning_rate": 3.303000478636874e-06,
+      "loss": 0.6908,
+      "step": 5983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9286378413854515,
+      "learning_rate": 3.3014170818940677e-06,
+      "loss": 0.5823,
+      "step": 5984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9232889386442449,
+      "learning_rate": 3.2998338776953163e-06,
+      "loss": 0.6119,
+      "step": 5985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8910875657751063,
+      "learning_rate": 3.2982508662200864e-06,
+      "loss": 0.6875,
+      "step": 5986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9224963141768185,
+      "learning_rate": 3.2966680476478196e-06,
+      "loss": 0.6215,
+      "step": 5987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2111799563581465,
+      "learning_rate": 3.295085422157939e-06,
+      "loss": 0.6311,
+      "step": 5988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7623188008375033,
+      "learning_rate": 3.2935029899298444e-06,
+      "loss": 0.5608,
+      "step": 5989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8634780884349817,
+      "learning_rate": 3.291920751142912e-06,
+      "loss": 0.669,
+      "step": 5990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0076300266293696,
+      "learning_rate": 3.290338705976497e-06,
+      "loss": 0.6168,
+      "step": 5991
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9389766540870652,
+      "learning_rate": 3.2887568546099346e-06,
+      "loss": 0.6409,
+      "step": 5992
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8669408346664602,
+      "learning_rate": 3.287175197222537e-06,
+      "loss": 0.6464,
+      "step": 5993
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1106017656825755,
+      "learning_rate": 3.2855937339935933e-06,
+      "loss": 0.6207,
+      "step": 5994
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8279681125103633,
+      "learning_rate": 3.284012465102372e-06,
+      "loss": 0.5878,
+      "step": 5995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.096349156684369,
+      "learning_rate": 3.282431390728118e-06,
+      "loss": 0.6014,
+      "step": 5996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.032892431253479,
+      "learning_rate": 3.280850511050058e-06,
+      "loss": 0.689,
+      "step": 5997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9785676746663783,
+      "learning_rate": 3.279269826247389e-06,
+      "loss": 0.4741,
+      "step": 5998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.676119831126687,
+      "learning_rate": 3.2776893364992936e-06,
+      "loss": 0.5467,
+      "step": 5999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.90052864684741,
+      "learning_rate": 3.2761090419849286e-06,
+      "loss": 0.5841,
+      "step": 6000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.229316068199629,
+      "learning_rate": 3.2745289428834294e-06,
+      "loss": 0.5785,
+      "step": 6001
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0934347201144146,
+      "learning_rate": 3.2729490393739093e-06,
+      "loss": 0.6735,
+      "step": 6002
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.048084478335236,
+      "learning_rate": 3.2713693316354593e-06,
+      "loss": 0.6202,
+      "step": 6003
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9404776557417234,
+      "learning_rate": 3.269789819847151e-06,
+      "loss": 0.6798,
+      "step": 6004
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9162170778231924,
+      "learning_rate": 3.2682105041880264e-06,
+      "loss": 0.5845,
+      "step": 6005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9545142603110053,
+      "learning_rate": 3.2666313848371113e-06,
+      "loss": 0.6539,
+      "step": 6006
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.794309367132827,
+      "learning_rate": 3.265052461973409e-06,
+      "loss": 0.5696,
+      "step": 6007
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.842487367019058,
+      "learning_rate": 3.2634737357758994e-06,
+      "loss": 0.6149,
+      "step": 6008
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.723064752719736,
+      "learning_rate": 3.26189520642354e-06,
+      "loss": 0.5282,
+      "step": 6009
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.096348196467705,
+      "learning_rate": 3.2603168740952645e-06,
+      "loss": 0.5349,
+      "step": 6010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0586495432214025,
+      "learning_rate": 3.2587387389699895e-06,
+      "loss": 0.6791,
+      "step": 6011
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8244663325194614,
+      "learning_rate": 3.257160801226601e-06,
+      "loss": 0.6393,
+      "step": 6012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8902371345502247,
+      "learning_rate": 3.255583061043971e-06,
+      "loss": 0.4821,
+      "step": 6013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.719480343071909,
+      "learning_rate": 3.2540055186009428e-06,
+      "loss": 0.5497,
+      "step": 6014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.876300759582372,
+      "learning_rate": 3.252428174076341e-06,
+      "loss": 0.6434,
+      "step": 6015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3262050183986918,
+      "learning_rate": 3.250851027648967e-06,
+      "loss": 0.6501,
+      "step": 6016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8741006195400267,
+      "learning_rate": 3.2492740794975985e-06,
+      "loss": 0.5569,
+      "step": 6017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.992043370643798,
+      "learning_rate": 3.247697329800992e-06,
+      "loss": 0.7077,
+      "step": 6018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0062231422855503,
+      "learning_rate": 3.246120778737883e-06,
+      "loss": 0.6345,
+      "step": 6019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8615543687360108,
+      "learning_rate": 3.2445444264869783e-06,
+      "loss": 0.6295,
+      "step": 6020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9316303752921364,
+      "learning_rate": 3.2429682732269685e-06,
+      "loss": 0.5939,
+      "step": 6021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9484002498233846,
+      "learning_rate": 3.2413923191365203e-06,
+      "loss": 0.6531,
+      "step": 6022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0450463816269737,
+      "learning_rate": 3.239816564394276e-06,
+      "loss": 0.5847,
+      "step": 6023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8031836096624272,
+      "learning_rate": 3.2382410091788567e-06,
+      "loss": 0.6577,
+      "step": 6024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2460149673113237,
+      "learning_rate": 3.2366656536688614e-06,
+      "loss": 0.5988,
+      "step": 6025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.896275207069309,
+      "learning_rate": 3.235090498042866e-06,
+      "loss": 0.5827,
+      "step": 6026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9320872061025782,
+      "learning_rate": 3.2335155424794205e-06,
+      "loss": 0.6269,
+      "step": 6027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6624219058989238,
+      "learning_rate": 3.2319407871570574e-06,
+      "loss": 0.663,
+      "step": 6028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8285245189929658,
+      "learning_rate": 3.2303662322542835e-06,
+      "loss": 0.6115,
+      "step": 6029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9162669741798801,
+      "learning_rate": 3.228791877949583e-06,
+      "loss": 0.6262,
+      "step": 6030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9680777538685632,
+      "learning_rate": 3.2272177244214198e-06,
+      "loss": 0.5613,
+      "step": 6031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0801896158128046,
+      "learning_rate": 3.2256437718482312e-06,
+      "loss": 0.5686,
+      "step": 6032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.217100063165203,
+      "learning_rate": 3.2240700204084353e-06,
+      "loss": 0.6403,
+      "step": 6033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8945485168220204,
+      "learning_rate": 3.222496470280427e-06,
+      "loss": 0.6851,
+      "step": 6034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8383948040993052,
+      "learning_rate": 3.220923121642573e-06,
+      "loss": 0.6221,
+      "step": 6035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9516341763000284,
+      "learning_rate": 3.219349974673223e-06,
+      "loss": 0.62,
+      "step": 6036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8320310813825704,
+      "learning_rate": 3.217777029550703e-06,
+      "loss": 0.6195,
+      "step": 6037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.847472645041628,
+      "learning_rate": 3.2162042864533154e-06,
+      "loss": 0.56,
+      "step": 6038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.978049182310532,
+      "learning_rate": 3.214631745559339e-06,
+      "loss": 0.6112,
+      "step": 6039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8176012495141787,
+      "learning_rate": 3.2130594070470307e-06,
+      "loss": 0.5219,
+      "step": 6040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8666199364201022,
+      "learning_rate": 3.2114872710946243e-06,
+      "loss": 0.6095,
+      "step": 6041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.969500493585429,
+      "learning_rate": 3.2099153378803294e-06,
+      "loss": 0.6168,
+      "step": 6042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7974069563510562,
+      "learning_rate": 3.2083436075823353e-06,
+      "loss": 0.6153,
+      "step": 6043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8639577027509828,
+      "learning_rate": 3.206772080378804e-06,
+      "loss": 0.612,
+      "step": 6044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9270717940972086,
+      "learning_rate": 3.205200756447878e-06,
+      "loss": 0.6043,
+      "step": 6045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7652996412307986,
+      "learning_rate": 3.2036296359676777e-06,
+      "loss": 0.6276,
+      "step": 6046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.995920741483894,
+      "learning_rate": 3.2020587191162956e-06,
+      "loss": 0.482,
+      "step": 6047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9227514888406252,
+      "learning_rate": 3.2004880060718072e-06,
+      "loss": 0.6573,
+      "step": 6048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7486882907902446,
+      "learning_rate": 3.1989174970122594e-06,
+      "loss": 0.6222,
+      "step": 6049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.138887268029901,
+      "learning_rate": 3.197347192115679e-06,
+      "loss": 0.5945,
+      "step": 6050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109380295021703,
+      "learning_rate": 3.1957770915600696e-06,
+      "loss": 0.6232,
+      "step": 6051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7196849956781948,
+      "learning_rate": 3.19420719552341e-06,
+      "loss": 0.4851,
+      "step": 6052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.917722969325608,
+      "learning_rate": 3.1926375041836573e-06,
+      "loss": 0.5262,
+      "step": 6053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8732009281586104,
+      "learning_rate": 3.1910680177187453e-06,
+      "loss": 0.6462,
+      "step": 6054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9473297187009115,
+      "learning_rate": 3.189498736306584e-06,
+      "loss": 0.6695,
+      "step": 6055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7892239645860206,
+      "learning_rate": 3.187929660125063e-06,
+      "loss": 0.5702,
+      "step": 6056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.888941220834544,
+      "learning_rate": 3.186360789352041e-06,
+      "loss": 0.6687,
+      "step": 6057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0322577338706784,
+      "learning_rate": 3.1847921241653614e-06,
+      "loss": 0.6612,
+      "step": 6058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9369396393065463,
+      "learning_rate": 3.18322366474284e-06,
+      "loss": 0.6982,
+      "step": 6059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.277936713753932,
+      "learning_rate": 3.181655411262272e-06,
+      "loss": 0.5452,
+      "step": 6060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9074840722455486,
+      "learning_rate": 3.1800873639014276e-06,
+      "loss": 0.4769,
+      "step": 6061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9101919714628626,
+      "learning_rate": 3.1785195228380527e-06,
+      "loss": 0.6405,
+      "step": 6062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9123614225065813,
+      "learning_rate": 3.176951888249875e-06,
+      "loss": 0.6504,
+      "step": 6063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2213263420816998,
+      "learning_rate": 3.1753844603145894e-06,
+      "loss": 0.6375,
+      "step": 6064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9608161818046423,
+      "learning_rate": 3.1738172392098752e-06,
+      "loss": 0.5604,
+      "step": 6065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.995234308228533,
+      "learning_rate": 3.172250225113386e-06,
+      "loss": 0.6577,
+      "step": 6066
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0377980498966806,
+      "learning_rate": 3.170683418202751e-06,
+      "loss": 0.6402,
+      "step": 6067
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8853013788369504,
+      "learning_rate": 3.1691168186555778e-06,
+      "loss": 0.5516,
+      "step": 6068
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.71003810845978,
+      "learning_rate": 3.1675504266494493e-06,
+      "loss": 0.6465,
+      "step": 6069
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6686351788488085,
+      "learning_rate": 3.1659842423619237e-06,
+      "loss": 0.5669,
+      "step": 6070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.061531475156852,
+      "learning_rate": 3.1644182659705403e-06,
+      "loss": 0.5795,
+      "step": 6071
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.762438560341012,
+      "learning_rate": 3.162852497652807e-06,
+      "loss": 0.6474,
+      "step": 6072
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9593057376669718,
+      "learning_rate": 3.161286937586214e-06,
+      "loss": 0.6712,
+      "step": 6073
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0318391689656377,
+      "learning_rate": 3.159721585948228e-06,
+      "loss": 0.6308,
+      "step": 6074
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.931145018821637,
+      "learning_rate": 3.158156442916288e-06,
+      "loss": 0.7147,
+      "step": 6075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.337635038951802,
+      "learning_rate": 3.156591508667814e-06,
+      "loss": 0.734,
+      "step": 6076
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9031520641680848,
+      "learning_rate": 3.1550267833801993e-06,
+      "loss": 0.6095,
+      "step": 6077
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7487631178062244,
+      "learning_rate": 3.1534622672308165e-06,
+      "loss": 0.6306,
+      "step": 6078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9160299182081961,
+      "learning_rate": 3.151897960397009e-06,
+      "loss": 0.6177,
+      "step": 6079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9860175732897307,
+      "learning_rate": 3.150333863056102e-06,
+      "loss": 0.7167,
+      "step": 6080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7346053348798447,
+      "learning_rate": 3.148769975385394e-06,
+      "loss": 0.5572,
+      "step": 6081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7993775431735983,
+      "learning_rate": 3.147206297562162e-06,
+      "loss": 0.576,
+      "step": 6082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8139002403350746,
+      "learning_rate": 3.1456428297636555e-06,
+      "loss": 0.5244,
+      "step": 6083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6929454845149765,
+      "learning_rate": 3.1440795721671036e-06,
+      "loss": 0.7666,
+      "step": 6084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7528536995455826,
+      "learning_rate": 3.1425165249497118e-06,
+      "loss": 0.599,
+      "step": 6085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0191237440247836,
+      "learning_rate": 3.140953688288658e-06,
+      "loss": 0.6796,
+      "step": 6086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9680374968614391,
+      "learning_rate": 3.1393910623611007e-06,
+      "loss": 0.6019,
+      "step": 6087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9442953822047284,
+      "learning_rate": 3.137828647344171e-06,
+      "loss": 0.6288,
+      "step": 6088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.981265816746576,
+      "learning_rate": 3.1362664434149782e-06,
+      "loss": 0.5979,
+      "step": 6089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.726191123392917,
+      "learning_rate": 3.134704450750607e-06,
+      "loss": 0.6121,
+      "step": 6090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6721715460247808,
+      "learning_rate": 3.133142669528118e-06,
+      "loss": 0.5627,
+      "step": 6091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9570130159305819,
+      "learning_rate": 3.1315810999245483e-06,
+      "loss": 0.6396,
+      "step": 6092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.995162107251734,
+      "learning_rate": 3.1300197421169125e-06,
+      "loss": 0.7216,
+      "step": 6093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1451911595083795,
+      "learning_rate": 3.1284585962821957e-06,
+      "loss": 0.6261,
+      "step": 6094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.901859286098706,
+      "learning_rate": 3.126897662597364e-06,
+      "loss": 0.677,
+      "step": 6095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.889639426022135,
+      "learning_rate": 3.1253369412393584e-06,
+      "loss": 0.6535,
+      "step": 6096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8686229215951038,
+      "learning_rate": 3.1237764323850964e-06,
+      "loss": 0.6528,
+      "step": 6097
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.311920832241536,
+      "learning_rate": 3.12221613621147e-06,
+      "loss": 0.6193,
+      "step": 6098
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9239392502057686,
+      "learning_rate": 3.1206560528953467e-06,
+      "loss": 0.6082,
+      "step": 6099
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8118553614299049,
+      "learning_rate": 3.1190961826135744e-06,
+      "loss": 0.635,
+      "step": 6100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7947350380331701,
+      "learning_rate": 3.1175365255429685e-06,
+      "loss": 0.572,
+      "step": 6101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.840175405623914,
+      "learning_rate": 3.115977081860327e-06,
+      "loss": 0.56,
+      "step": 6102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.014363206044062,
+      "learning_rate": 3.1144178517424217e-06,
+      "loss": 0.5143,
+      "step": 6103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.692597253312618,
+      "learning_rate": 3.1128588353660006e-06,
+      "loss": 0.5608,
+      "step": 6104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2750803784856393,
+      "learning_rate": 3.111300032907787e-06,
+      "loss": 0.5791,
+      "step": 6105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7369305752415842,
+      "learning_rate": 3.1097414445444796e-06,
+      "loss": 0.5949,
+      "step": 6106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.220660084433447,
+      "learning_rate": 3.1081830704527535e-06,
+      "loss": 0.6184,
+      "step": 6107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8162843137045013,
+      "learning_rate": 3.1066249108092616e-06,
+      "loss": 0.6274,
+      "step": 6108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7452644073711159,
+      "learning_rate": 3.1050669657906257e-06,
+      "loss": 0.5825,
+      "step": 6109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109057932821261,
+      "learning_rate": 3.10350923557345e-06,
+      "loss": 0.6562,
+      "step": 6110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.07479718970697,
+      "learning_rate": 3.101951720334312e-06,
+      "loss": 0.5884,
+      "step": 6111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9154797157879695,
+      "learning_rate": 3.1003944202497655e-06,
+      "loss": 0.5808,
+      "step": 6112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0766652476271794,
+      "learning_rate": 3.0988373354963387e-06,
+      "loss": 0.6579,
+      "step": 6113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0534225279153886,
+      "learning_rate": 3.097280466250536e-06,
+      "loss": 0.6693,
+      "step": 6114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9602854542078891,
+      "learning_rate": 3.0957238126888384e-06,
+      "loss": 0.5914,
+      "step": 6115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8801721093726407,
+      "learning_rate": 3.0941673749877e-06,
+      "loss": 0.6304,
+      "step": 6116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0461787180132682,
+      "learning_rate": 3.0926111533235526e-06,
+      "loss": 0.6503,
+      "step": 6117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9878343487415226,
+      "learning_rate": 3.091055147872802e-06,
+      "loss": 0.7512,
+      "step": 6118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1283625476528787,
+      "learning_rate": 3.0894993588118318e-06,
+      "loss": 0.5694,
+      "step": 6119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8045352095592917,
+      "learning_rate": 3.087943786316999e-06,
+      "loss": 0.6016,
+      "step": 6120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8931603547365077,
+      "learning_rate": 3.0863884305646364e-06,
+      "loss": 0.5812,
+      "step": 6121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.804868568230148,
+      "learning_rate": 3.0848332917310532e-06,
+      "loss": 0.5961,
+      "step": 6122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9803781360337822,
+      "learning_rate": 3.0832783699925307e-06,
+      "loss": 0.7075,
+      "step": 6123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8810190808368576,
+      "learning_rate": 3.081723665525331e-06,
+      "loss": 0.6251,
+      "step": 6124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8230914084353704,
+      "learning_rate": 3.0801691785056863e-06,
+      "loss": 0.6425,
+      "step": 6125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.801356178216089,
+      "learning_rate": 3.0786149091098087e-06,
+      "loss": 0.5365,
+      "step": 6126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8914863670162068,
+      "learning_rate": 3.0770608575138825e-06,
+      "loss": 0.6569,
+      "step": 6127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8896895490041579,
+      "learning_rate": 3.075507023894069e-06,
+      "loss": 0.6015,
+      "step": 6128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1105225623734287,
+      "learning_rate": 3.0739534084265032e-06,
+      "loss": 0.6685,
+      "step": 6129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9830540304838769,
+      "learning_rate": 3.072400011287299e-06,
+      "loss": 0.6105,
+      "step": 6130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1810675903640586,
+      "learning_rate": 3.0708468326525376e-06,
+      "loss": 0.6237,
+      "step": 6131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8404334794364534,
+      "learning_rate": 3.069293872698284e-06,
+      "loss": 0.6088,
+      "step": 6132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.772357721649005,
+      "learning_rate": 3.0677411316005744e-06,
+      "loss": 0.6274,
+      "step": 6133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8695683122224707,
+      "learning_rate": 3.066188609535421e-06,
+      "loss": 0.7137,
+      "step": 6134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7470798861679386,
+      "learning_rate": 3.0646363066788114e-06,
+      "loss": 0.5847,
+      "step": 6135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8760125269130379,
+      "learning_rate": 3.063084223206708e-06,
+      "loss": 0.6395,
+      "step": 6136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.879803637103201,
+      "learning_rate": 3.0615323592950495e-06,
+      "loss": 0.5537,
+      "step": 6137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.02465353465098,
+      "learning_rate": 3.0599807151197446e-06,
+      "loss": 0.6646,
+      "step": 6138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.013051093932397,
+      "learning_rate": 3.0584292908566836e-06,
+      "loss": 0.601,
+      "step": 6139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8921168905387726,
+      "learning_rate": 3.056878086681729e-06,
+      "loss": 0.5706,
+      "step": 6140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6381413943978955,
+      "learning_rate": 3.055327102770719e-06,
+      "loss": 0.5007,
+      "step": 6141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9367879723399168,
+      "learning_rate": 3.053776339299467e-06,
+      "loss": 0.5682,
+      "step": 6142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8429248685663757,
+      "learning_rate": 3.0522257964437586e-06,
+      "loss": 0.5974,
+      "step": 6143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9168065728136365,
+      "learning_rate": 3.050675474379361e-06,
+      "loss": 0.5974,
+      "step": 6144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.749866080818118,
+      "learning_rate": 3.0491253732820063e-06,
+      "loss": 0.6956,
+      "step": 6145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9739705708202513,
+      "learning_rate": 3.0475754933274106e-06,
+      "loss": 0.6371,
+      "step": 6146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7874894202016325,
+      "learning_rate": 3.0460258346912615e-06,
+      "loss": 0.6762,
+      "step": 6147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9124508607094237,
+      "learning_rate": 3.044476397549221e-06,
+      "loss": 0.6693,
+      "step": 6148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7300970486859781,
+      "learning_rate": 3.042927182076927e-06,
+      "loss": 0.5399,
+      "step": 6149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9977396120353856,
+      "learning_rate": 3.0413781884499916e-06,
+      "loss": 0.6662,
+      "step": 6150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7582271936489247,
+      "learning_rate": 3.0398294168440023e-06,
+      "loss": 0.6091,
+      "step": 6151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0479440125108272,
+      "learning_rate": 3.0382808674345228e-06,
+      "loss": 0.6703,
+      "step": 6152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9526668255453477,
+      "learning_rate": 3.036732540397087e-06,
+      "loss": 0.527,
+      "step": 6153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7155652955499163,
+      "learning_rate": 3.035184435907208e-06,
+      "loss": 0.5533,
+      "step": 6154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9812973705820782,
+      "learning_rate": 3.0336365541403723e-06,
+      "loss": 0.5941,
+      "step": 6155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8769962170542873,
+      "learning_rate": 3.0320888952720414e-06,
+      "loss": 0.596,
+      "step": 6156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.923258773106532,
+      "learning_rate": 3.0305414594776505e-06,
+      "loss": 0.6878,
+      "step": 6157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3812002253021145,
+      "learning_rate": 3.0289942469326106e-06,
+      "loss": 0.5918,
+      "step": 6158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7311215954566284,
+      "learning_rate": 3.0274472578123095e-06,
+      "loss": 0.5567,
+      "step": 6159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8309946265303754,
+      "learning_rate": 3.0259004922921033e-06,
+      "loss": 0.5843,
+      "step": 6160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8394554565201098,
+      "learning_rate": 3.0243539505473275e-06,
+      "loss": 0.529,
+      "step": 6161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8519372361597473,
+      "learning_rate": 3.0228076327532925e-06,
+      "loss": 0.5625,
+      "step": 6162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8204661064383223,
+      "learning_rate": 3.021261539085282e-06,
+      "loss": 0.5559,
+      "step": 6163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.782344163712384,
+      "learning_rate": 3.019715669718554e-06,
+      "loss": 0.604,
+      "step": 6164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6640325517922043,
+      "learning_rate": 3.018170024828343e-06,
+      "loss": 0.5919,
+      "step": 6165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.037145171502773,
+      "learning_rate": 3.016624604589855e-06,
+      "loss": 0.6316,
+      "step": 6166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6769480920552722,
+      "learning_rate": 3.0150794091782753e-06,
+      "loss": 0.6897,
+      "step": 6167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0464765956664976,
+      "learning_rate": 3.013534438768756e-06,
+      "loss": 0.63,
+      "step": 6168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.916989468935177,
+      "learning_rate": 3.0119896935364305e-06,
+      "loss": 0.5742,
+      "step": 6169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9472328083428492,
+      "learning_rate": 3.010445173656405e-06,
+      "loss": 0.6717,
+      "step": 6170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8992859253335288,
+      "learning_rate": 3.0089008793037587e-06,
+      "loss": 0.6509,
+      "step": 6171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.916102090617852,
+      "learning_rate": 3.0073568106535465e-06,
+      "loss": 0.6445,
+      "step": 6172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9064952084707214,
+      "learning_rate": 3.005812967880798e-06,
+      "loss": 0.6296,
+      "step": 6173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.807783199230649,
+      "learning_rate": 3.004269351160518e-06,
+      "loss": 0.6381,
+      "step": 6174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9174580476574532,
+      "learning_rate": 3.00272596066768e-06,
+      "loss": 0.6323,
+      "step": 6175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9492380466914858,
+      "learning_rate": 3.001182796577239e-06,
+      "loss": 0.6298,
+      "step": 6176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6924745865542088,
+      "learning_rate": 2.9996398590641203e-06,
+      "loss": 0.5399,
+      "step": 6177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.033275499357687,
+      "learning_rate": 2.998097148303225e-06,
+      "loss": 0.7353,
+      "step": 6178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9026486542697536,
+      "learning_rate": 2.9965546644694287e-06,
+      "loss": 0.5794,
+      "step": 6179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0228080110290105,
+      "learning_rate": 2.995012407737581e-06,
+      "loss": 0.6467,
+      "step": 6180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7861014751773872,
+      "learning_rate": 2.993470378282505e-06,
+      "loss": 0.6066,
+      "step": 6181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.839263166878586,
+      "learning_rate": 2.9919285762789983e-06,
+      "loss": 0.5448,
+      "step": 6182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7072274410368804,
+      "learning_rate": 2.990387001901834e-06,
+      "loss": 0.5578,
+      "step": 6183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6431673164569158,
+      "learning_rate": 2.988845655325756e-06,
+      "loss": 0.6351,
+      "step": 6184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9010572335147635,
+      "learning_rate": 2.987304536725486e-06,
+      "loss": 0.6327,
+      "step": 6185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6632693290336715,
+      "learning_rate": 2.9857636462757193e-06,
+      "loss": 0.5914,
+      "step": 6186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1089846908801015,
+      "learning_rate": 2.984222984151124e-06,
+      "loss": 0.6263,
+      "step": 6187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.817502491313046,
+      "learning_rate": 2.9826825505263427e-06,
+      "loss": 0.6505,
+      "step": 6188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8606164928292765,
+      "learning_rate": 2.981142345575994e-06,
+      "loss": 0.6494,
+      "step": 6189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8351620719634854,
+      "learning_rate": 2.979602369474667e-06,
+      "loss": 0.522,
+      "step": 6190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6703557627008776,
+      "learning_rate": 2.9780626223969256e-06,
+      "loss": 0.5111,
+      "step": 6191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9164204387222075,
+      "learning_rate": 2.976523104517312e-06,
+      "loss": 0.6429,
+      "step": 6192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9499476254969241,
+      "learning_rate": 2.9749838160103372e-06,
+      "loss": 0.6298,
+      "step": 6193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0213576043893964,
+      "learning_rate": 2.9734447570504898e-06,
+      "loss": 0.6417,
+      "step": 6194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8219746597887712,
+      "learning_rate": 2.97190592781223e-06,
+      "loss": 0.6355,
+      "step": 6195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9916439598009843,
+      "learning_rate": 2.9703673284699945e-06,
+      "loss": 0.7587,
+      "step": 6196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8151195802546238,
+      "learning_rate": 2.9688289591981887e-06,
+      "loss": 0.6119,
+      "step": 6197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9629854400858462,
+      "learning_rate": 2.9672908201711986e-06,
+      "loss": 0.6639,
+      "step": 6198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0632728967170935,
+      "learning_rate": 2.96575291156338e-06,
+      "loss": 0.6161,
+      "step": 6199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0544901998541976,
+      "learning_rate": 2.9642152335490633e-06,
+      "loss": 0.5926,
+      "step": 6200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8887067530096628,
+      "learning_rate": 2.9626777863025535e-06,
+      "loss": 0.6645,
+      "step": 6201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9767964985943947,
+      "learning_rate": 2.961140569998129e-06,
+      "loss": 0.5331,
+      "step": 6202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9912020438130313,
+      "learning_rate": 2.959603584810041e-06,
+      "loss": 0.595,
+      "step": 6203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8860676490378074,
+      "learning_rate": 2.9580668309125203e-06,
+      "loss": 0.6943,
+      "step": 6204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2330331418419878,
+      "learning_rate": 2.95653030847976e-06,
+      "loss": 0.6634,
+      "step": 6205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.081180271310904,
+      "learning_rate": 2.954994017685937e-06,
+      "loss": 0.5837,
+      "step": 6206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8421768186562804,
+      "learning_rate": 2.9534579587051976e-06,
+      "loss": 0.6476,
+      "step": 6207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8404952869659887,
+      "learning_rate": 2.9519221317116644e-06,
+      "loss": 0.6179,
+      "step": 6208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8764757322689638,
+      "learning_rate": 2.9503865368794303e-06,
+      "loss": 0.5937,
+      "step": 6209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0970231465593434,
+      "learning_rate": 2.948851174382565e-06,
+      "loss": 0.6625,
+      "step": 6210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.198985907859825,
+      "learning_rate": 2.947316044395112e-06,
+      "loss": 0.7062,
+      "step": 6211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9215926364945755,
+      "learning_rate": 2.9457811470910837e-06,
+      "loss": 0.6313,
+      "step": 6212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8199210140099955,
+      "learning_rate": 2.944246482644471e-06,
+      "loss": 0.583,
+      "step": 6213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1819976857706016,
+      "learning_rate": 2.9427120512292368e-06,
+      "loss": 0.7136,
+      "step": 6214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0989651592874425,
+      "learning_rate": 2.941177853019318e-06,
+      "loss": 0.6884,
+      "step": 6215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8185220766754597,
+      "learning_rate": 2.9396438881886234e-06,
+      "loss": 0.6199,
+      "step": 6216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8386182832764755,
+      "learning_rate": 2.9381101569110393e-06,
+      "loss": 0.5703,
+      "step": 6217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7960263200181878,
+      "learning_rate": 2.936576659360421e-06,
+      "loss": 0.5296,
+      "step": 6218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0704183498446374,
+      "learning_rate": 2.9350433957105995e-06,
+      "loss": 0.6312,
+      "step": 6219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7147161698387252,
+      "learning_rate": 2.933510366135378e-06,
+      "loss": 0.5166,
+      "step": 6220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0667639388096304,
+      "learning_rate": 2.9319775708085364e-06,
+      "loss": 0.6563,
+      "step": 6221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8954356153973286,
+      "learning_rate": 2.930445009903824e-06,
+      "loss": 0.6817,
+      "step": 6222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1322815169531784,
+      "learning_rate": 2.9289126835949657e-06,
+      "loss": 0.6872,
+      "step": 6223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9819330643264859,
+      "learning_rate": 2.9273805920556586e-06,
+      "loss": 0.6843,
+      "step": 6224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.065487847700474,
+      "learning_rate": 2.9258487354595754e-06,
+      "loss": 0.723,
+      "step": 6225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.928418370629797,
+      "learning_rate": 2.9243171139803617e-06,
+      "loss": 0.6875,
+      "step": 6226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9522013519745338,
+      "learning_rate": 2.9227857277916325e-06,
+      "loss": 0.655,
+      "step": 6227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.082709028367856,
+      "learning_rate": 2.9212545770669814e-06,
+      "loss": 0.6324,
+      "step": 6228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.057952654711287,
+      "learning_rate": 2.919723661979972e-06,
+      "loss": 0.5704,
+      "step": 6229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9576733185893804,
+      "learning_rate": 2.918192982704143e-06,
+      "loss": 0.6425,
+      "step": 6230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6660456041502691,
+      "learning_rate": 2.9166625394130066e-06,
+      "loss": 0.5195,
+      "step": 6231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9031996084090193,
+      "learning_rate": 2.9151323322800433e-06,
+      "loss": 0.7252,
+      "step": 6232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.933943035659668,
+      "learning_rate": 2.913602361478716e-06,
+      "loss": 0.5009,
+      "step": 6233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.235638647690862,
+      "learning_rate": 2.912072627182453e-06,
+      "loss": 0.6169,
+      "step": 6234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7939098430985856,
+      "learning_rate": 2.910543129564658e-06,
+      "loss": 0.6096,
+      "step": 6235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9999371977297142,
+      "learning_rate": 2.9090138687987075e-06,
+      "loss": 0.626,
+      "step": 6236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7553733344385136,
+      "learning_rate": 2.9074848450579545e-06,
+      "loss": 0.6289,
+      "step": 6237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.013563698558594,
+      "learning_rate": 2.9059560585157197e-06,
+      "loss": 0.6329,
+      "step": 6238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9821038414510794,
+      "learning_rate": 2.9044275093453034e-06,
+      "loss": 0.5404,
+      "step": 6239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.880071424325999,
+      "learning_rate": 2.9028991977199705e-06,
+      "loss": 0.6771,
+      "step": 6240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8332617137796834,
+      "learning_rate": 2.9013711238129693e-06,
+      "loss": 0.6157,
+      "step": 6241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9401173115427022,
+      "learning_rate": 2.899843287797513e-06,
+      "loss": 0.645,
+      "step": 6242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0386914173802833,
+      "learning_rate": 2.8983156898467885e-06,
+      "loss": 0.6974,
+      "step": 6243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6509330084534442,
+      "learning_rate": 2.896788330133962e-06,
+      "loss": 0.5447,
+      "step": 6244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.94624189730017,
+      "learning_rate": 2.8952612088321636e-06,
+      "loss": 0.6139,
+      "step": 6245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9680166782133846,
+      "learning_rate": 2.893734326114506e-06,
+      "loss": 0.653,
+      "step": 6246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9087806201292232,
+      "learning_rate": 2.8922076821540657e-06,
+      "loss": 0.5587,
+      "step": 6247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9714706233304224,
+      "learning_rate": 2.8906812771239034e-06,
+      "loss": 0.666,
+      "step": 6248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8099975489747797,
+      "learning_rate": 2.889155111197036e-06,
+      "loss": 0.5629,
+      "step": 6249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8264082657553349,
+      "learning_rate": 2.887629184546471e-06,
+      "loss": 0.5704,
+      "step": 6250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2058717301943824,
+      "learning_rate": 2.8861034973451753e-06,
+      "loss": 0.7257,
+      "step": 6251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8518154957166706,
+      "learning_rate": 2.8845780497660996e-06,
+      "loss": 0.6928,
+      "step": 6252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9484583542830034,
+      "learning_rate": 2.883052841982157e-06,
+      "loss": 0.6393,
+      "step": 6253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7555508796894785,
+      "learning_rate": 2.8815278741662433e-06,
+      "loss": 0.5886,
+      "step": 6254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.994950526398456,
+      "learning_rate": 2.88000314649122e-06,
+      "loss": 0.5842,
+      "step": 6255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8662457743534389,
+      "learning_rate": 2.878478659129923e-06,
+      "loss": 0.584,
+      "step": 6256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9420532601252392,
+      "learning_rate": 2.8769544122551606e-06,
+      "loss": 0.7123,
+      "step": 6257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.835196222874027,
+      "learning_rate": 2.875430406039719e-06,
+      "loss": 0.5903,
+      "step": 6258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8756813320395385,
+      "learning_rate": 2.873906640656348e-06,
+      "loss": 0.5721,
+      "step": 6259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9765833048671797,
+      "learning_rate": 2.8723831162777806e-06,
+      "loss": 0.6045,
+      "step": 6260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.032795441691923,
+      "learning_rate": 2.8708598330767105e-06,
+      "loss": 0.6522,
+      "step": 6261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7505441325588476,
+      "learning_rate": 2.869336791225817e-06,
+      "loss": 0.5361,
+      "step": 6262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9496901968364277,
+      "learning_rate": 2.867813990897742e-06,
+      "loss": 0.6319,
+      "step": 6263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.848210036629578,
+      "learning_rate": 2.8662914322651046e-06,
+      "loss": 0.5804,
+      "step": 6264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0674380320579235,
+      "learning_rate": 2.864769115500493e-06,
+      "loss": 0.5805,
+      "step": 6265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.864513334235877,
+      "learning_rate": 2.8632470407764746e-06,
+      "loss": 0.6263,
+      "step": 6266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9868287714690414,
+      "learning_rate": 2.8617252082655813e-06,
+      "loss": 0.568,
+      "step": 6267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.003702026459403,
+      "learning_rate": 2.860203618140325e-06,
+      "loss": 0.6686,
+      "step": 6268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9781283187382133,
+      "learning_rate": 2.858682270573183e-06,
+      "loss": 0.5742,
+      "step": 6269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8962333723378362,
+      "learning_rate": 2.857161165736613e-06,
+      "loss": 0.562,
+      "step": 6270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9548515931569148,
+      "learning_rate": 2.8556403038030385e-06,
+      "loss": 0.7561,
+      "step": 6271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0789855989958763,
+      "learning_rate": 2.8541196849448582e-06,
+      "loss": 0.5907,
+      "step": 6272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9881196918457014,
+      "learning_rate": 2.8525993093344407e-06,
+      "loss": 0.6955,
+      "step": 6273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.968486206575867,
+      "learning_rate": 2.8510791771441327e-06,
+      "loss": 0.5586,
+      "step": 6274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7880365080779919,
+      "learning_rate": 2.8495592885462476e-06,
+      "loss": 0.5935,
+      "step": 6275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2478113171852816,
+      "learning_rate": 2.848039643713075e-06,
+      "loss": 0.5541,
+      "step": 6276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.888739198842372,
+      "learning_rate": 2.8465202428168753e-06,
+      "loss": 0.6149,
+      "step": 6277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1536997825091837,
+      "learning_rate": 2.8450010860298784e-06,
+      "loss": 0.7143,
+      "step": 6278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.814516674311857,
+      "learning_rate": 2.8434821735242935e-06,
+      "loss": 0.6282,
+      "step": 6279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.01424137795222,
+      "learning_rate": 2.841963505472294e-06,
+      "loss": 0.6298,
+      "step": 6280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9768897380178496,
+      "learning_rate": 2.8404450820460326e-06,
+      "loss": 0.6343,
+      "step": 6281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.327163802642382,
+      "learning_rate": 2.838926903417629e-06,
+      "loss": 0.7132,
+      "step": 6282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.968349224205387,
+      "learning_rate": 2.837408969759181e-06,
+      "loss": 0.5983,
+      "step": 6283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7460529943975944,
+      "learning_rate": 2.8358912812427497e-06,
+      "loss": 0.5936,
+      "step": 6284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0593792416805363,
+      "learning_rate": 2.834373838040382e-06,
+      "loss": 0.6621,
+      "step": 6285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0565360915182045,
+      "learning_rate": 2.832856640324078e-06,
+      "loss": 0.6497,
+      "step": 6286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9303234583627442,
+      "learning_rate": 2.831339688265829e-06,
+      "loss": 0.6376,
+      "step": 6287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1135210848654555,
+      "learning_rate": 2.829822982037585e-06,
+      "loss": 0.71,
+      "step": 6288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0335602868059968,
+      "learning_rate": 2.8283065218112775e-06,
+      "loss": 0.6106,
+      "step": 6289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0336428429947935,
+      "learning_rate": 2.826790307758802e-06,
+      "loss": 0.7293,
+      "step": 6290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9485197680578654,
+      "learning_rate": 2.8252743400520345e-06,
+      "loss": 0.6481,
+      "step": 6291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9136820372082457,
+      "learning_rate": 2.8237586188628153e-06,
+      "loss": 0.5989,
+      "step": 6292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8096490093595976,
+      "learning_rate": 2.8222431443629617e-06,
+      "loss": 0.5179,
+      "step": 6293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9304167213931087,
+      "learning_rate": 2.820727916724257e-06,
+      "loss": 0.6775,
+      "step": 6294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.853944468773223,
+      "learning_rate": 2.8192129361184685e-06,
+      "loss": 0.6527,
+      "step": 6295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9454826200033084,
+      "learning_rate": 2.8176982027173206e-06,
+      "loss": 0.5919,
+      "step": 6296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0221472638370903,
+      "learning_rate": 2.816183716692522e-06,
+      "loss": 0.5724,
+      "step": 6297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.116282365722759,
+      "learning_rate": 2.8146694782157447e-06,
+      "loss": 0.6334,
+      "step": 6298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9071368964119593,
+      "learning_rate": 2.8131554874586396e-06,
+      "loss": 0.612,
+      "step": 6299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9872247909004435,
+      "learning_rate": 2.8116417445928245e-06,
+      "loss": 0.6012,
+      "step": 6300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.976827376780296,
+      "learning_rate": 2.810128249789892e-06,
+      "loss": 0.6022,
+      "step": 6301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.805685551078074,
+      "learning_rate": 2.808615003221401e-06,
+      "loss": 0.5421,
+      "step": 6302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8361096040269862,
+      "learning_rate": 2.8071020050588927e-06,
+      "loss": 0.5828,
+      "step": 6303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.117462557354621,
+      "learning_rate": 2.8055892554738683e-06,
+      "loss": 0.67,
+      "step": 6304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9521166314972613,
+      "learning_rate": 2.804076754637812e-06,
+      "loss": 0.6384,
+      "step": 6305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.887459953486852,
+      "learning_rate": 2.8025645027221704e-06,
+      "loss": 0.5704,
+      "step": 6306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.867313977993055,
+      "learning_rate": 2.801052499898369e-06,
+      "loss": 0.5305,
+      "step": 6307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9997508923479637,
+      "learning_rate": 2.7995407463378004e-06,
+      "loss": 0.6093,
+      "step": 6308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8703093721931023,
+      "learning_rate": 2.7980292422118282e-06,
+      "loss": 0.613,
+      "step": 6309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.061021653564945,
+      "learning_rate": 2.7965179876917946e-06,
+      "loss": 0.5967,
+      "step": 6310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.843007255448845,
+      "learning_rate": 2.795006982949006e-06,
+      "loss": 0.6791,
+      "step": 6311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9714166575628622,
+      "learning_rate": 2.7934962281547422e-06,
+      "loss": 0.6519,
+      "step": 6312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7854552666884103,
+      "learning_rate": 2.7919857234802593e-06,
+      "loss": 0.5518,
+      "step": 6313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9856717506321173,
+      "learning_rate": 2.7904754690967808e-06,
+      "loss": 0.6034,
+      "step": 6314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2295217500421063,
+      "learning_rate": 2.7889654651754987e-06,
+      "loss": 0.53,
+      "step": 6315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6963943668827721,
+      "learning_rate": 2.7874557118875863e-06,
+      "loss": 0.5937,
+      "step": 6316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.053957569068724,
+      "learning_rate": 2.785946209404178e-06,
+      "loss": 0.7016,
+      "step": 6317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.911492913171324,
+      "learning_rate": 2.784436957896388e-06,
+      "loss": 0.7116,
+      "step": 6318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8810697368746152,
+      "learning_rate": 2.7829279575352953e-06,
+      "loss": 0.5795,
+      "step": 6319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9914019124478899,
+      "learning_rate": 2.781419208491958e-06,
+      "loss": 0.5958,
+      "step": 6320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.712755966691504,
+      "learning_rate": 2.7799107109373956e-06,
+      "loss": 0.5469,
+      "step": 6321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8143583254344464,
+      "learning_rate": 2.7784024650426133e-06,
+      "loss": 0.6282,
+      "step": 6322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1419849838155,
+      "learning_rate": 2.7768944709785705e-06,
+      "loss": 0.7249,
+      "step": 6323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7170329232042192,
+      "learning_rate": 2.775386728916212e-06,
+      "loss": 0.5251,
+      "step": 6324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.742228854334704,
+      "learning_rate": 2.7738792390264456e-06,
+      "loss": 0.6825,
+      "step": 6325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8459902103849364,
+      "learning_rate": 2.772372001480159e-06,
+      "loss": 0.6152,
+      "step": 6326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8776493794011735,
+      "learning_rate": 2.7708650164482e-06,
+      "loss": 0.639,
+      "step": 6327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.063971101948811,
+      "learning_rate": 2.7693582841013996e-06,
+      "loss": 0.6364,
+      "step": 6328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.036669543108079,
+      "learning_rate": 2.767851804610552e-06,
+      "loss": 0.64,
+      "step": 6329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6546039954617904,
+      "learning_rate": 2.7663455781464245e-06,
+      "loss": 0.5424,
+      "step": 6330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8036933164626912,
+      "learning_rate": 2.7648396048797554e-06,
+      "loss": 0.5899,
+      "step": 6331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8285545473432434,
+      "learning_rate": 2.7633338849812593e-06,
+      "loss": 0.5668,
+      "step": 6332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8127905625169214,
+      "learning_rate": 2.7618284186216137e-06,
+      "loss": 0.6676,
+      "step": 6333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.153394766265645,
+      "learning_rate": 2.760323205971476e-06,
+      "loss": 0.6618,
+      "step": 6334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9747255781767126,
+      "learning_rate": 2.7588182472014668e-06,
+      "loss": 0.6172,
+      "step": 6335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0717810839011723,
+      "learning_rate": 2.757313542482185e-06,
+      "loss": 0.6448,
+      "step": 6336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9391684607298043,
+      "learning_rate": 2.7558090919841972e-06,
+      "loss": 0.6767,
+      "step": 6337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.900234867377518,
+      "learning_rate": 2.7543048958780395e-06,
+      "loss": 0.567,
+      "step": 6338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.026675318690998,
+      "learning_rate": 2.7528009543342197e-06,
+      "loss": 0.6441,
+      "step": 6339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9912333544995164,
+      "learning_rate": 2.751297267523223e-06,
+      "loss": 0.6838,
+      "step": 6340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8455002654000203,
+      "learning_rate": 2.7497938356154957e-06,
+      "loss": 0.4949,
+      "step": 6341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.944518401536242,
+      "learning_rate": 2.748290658781465e-06,
+      "loss": 0.5808,
+      "step": 6342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8734162282177562,
+      "learning_rate": 2.746787737191521e-06,
+      "loss": 0.6036,
+      "step": 6343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0569689392989594,
+      "learning_rate": 2.7452850710160305e-06,
+      "loss": 0.6238,
+      "step": 6344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.859492968004961,
+      "learning_rate": 2.7437826604253296e-06,
+      "loss": 0.6002,
+      "step": 6345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9239528734480238,
+      "learning_rate": 2.7422805055897224e-06,
+      "loss": 0.6592,
+      "step": 6346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7751066614137307,
+      "learning_rate": 2.74077860667949e-06,
+      "loss": 0.6482,
+      "step": 6347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8529491712765056,
+      "learning_rate": 2.7392769638648775e-06,
+      "loss": 0.6862,
+      "step": 6348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8998307106205168,
+      "learning_rate": 2.73777557731611e-06,
+      "loss": 0.6126,
+      "step": 6349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.118611629913783,
+      "learning_rate": 2.736274447203373e-06,
+      "loss": 0.627,
+      "step": 6350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9556808686347844,
+      "learning_rate": 2.7347735736968318e-06,
+      "loss": 0.6527,
+      "step": 6351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.013462843348238,
+      "learning_rate": 2.733272956966615e-06,
+      "loss": 0.7476,
+      "step": 6352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8135008649655475,
+      "learning_rate": 2.7317725971828302e-06,
+      "loss": 0.6531,
+      "step": 6353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9304868034857994,
+      "learning_rate": 2.7302724945155486e-06,
+      "loss": 0.6678,
+      "step": 6354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9647717480570688,
+      "learning_rate": 2.728772649134818e-06,
+      "loss": 0.6529,
+      "step": 6355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.920259935679486,
+      "learning_rate": 2.7272730612106513e-06,
+      "loss": 0.5923,
+      "step": 6356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1502531709807124,
+      "learning_rate": 2.7257737309130393e-06,
+      "loss": 0.6569,
+      "step": 6357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9442863146247567,
+      "learning_rate": 2.7242746584119364e-06,
+      "loss": 0.6219,
+      "step": 6358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8695957383517858,
+      "learning_rate": 2.722775843877275e-06,
+      "loss": 0.5603,
+      "step": 6359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1452012554076223,
+      "learning_rate": 2.7212772874789484e-06,
+      "loss": 0.6129,
+      "step": 6360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7405043357807028,
+      "learning_rate": 2.7197789893868307e-06,
+      "loss": 0.6054,
+      "step": 6361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8611986630402495,
+      "learning_rate": 2.7182809497707594e-06,
+      "loss": 0.647,
+      "step": 6362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8830375405003674,
+      "learning_rate": 2.7167831688005502e-06,
+      "loss": 0.5694,
+      "step": 6363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7562835116328723,
+      "learning_rate": 2.715285646645981e-06,
+      "loss": 0.6591,
+      "step": 6364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7368453184663584,
+      "learning_rate": 2.7137883834768076e-06,
+      "loss": 0.5398,
+      "step": 6365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9427737888494707,
+      "learning_rate": 2.7122913794627528e-06,
+      "loss": 0.6606,
+      "step": 6366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8309498526643528,
+      "learning_rate": 2.710794634773508e-06,
+      "loss": 0.5763,
+      "step": 6367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8981317509223192,
+      "learning_rate": 2.7092981495787387e-06,
+      "loss": 0.5609,
+      "step": 6368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0381064382923353,
+      "learning_rate": 2.7078019240480826e-06,
+      "loss": 0.5834,
+      "step": 6369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0360199613465335,
+      "learning_rate": 2.706305958351141e-06,
+      "loss": 0.5999,
+      "step": 6370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8851392663074493,
+      "learning_rate": 2.7048102526574948e-06,
+      "loss": 0.5561,
+      "step": 6371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0319791070048185,
+      "learning_rate": 2.7033148071366866e-06,
+      "loss": 0.6536,
+      "step": 6372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9323663914875056,
+      "learning_rate": 2.7018196219582404e-06,
+      "loss": 0.59,
+      "step": 6373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.148618386983816,
+      "learning_rate": 2.700324697291634e-06,
+      "loss": 0.6185,
+      "step": 6374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.875273962996191,
+      "learning_rate": 2.698830033306334e-06,
+      "loss": 0.6204,
+      "step": 6375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.050493079361734,
+      "learning_rate": 2.6973356301717633e-06,
+      "loss": 0.6846,
+      "step": 6376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9716423747607381,
+      "learning_rate": 2.6958414880573257e-06,
+      "loss": 0.552,
+      "step": 6377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0713906418446144,
+      "learning_rate": 2.694347607132387e-06,
+      "loss": 0.6723,
+      "step": 6378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9824685803451119,
+      "learning_rate": 2.692853987566291e-06,
+      "loss": 0.5323,
+      "step": 6379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0917417738256425,
+      "learning_rate": 2.6913606295283436e-06,
+      "loss": 0.7374,
+      "step": 6380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8222154905337453,
+      "learning_rate": 2.689867533187829e-06,
+      "loss": 0.6312,
+      "step": 6381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0814088840248615,
+      "learning_rate": 2.688374698713997e-06,
+      "loss": 0.6002,
+      "step": 6382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.193140630954233,
+      "learning_rate": 2.6868821262760673e-06,
+      "loss": 0.732,
+      "step": 6383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.753162209988003,
+      "learning_rate": 2.6853898160432347e-06,
+      "loss": 0.536,
+      "step": 6384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7609052391335116,
+      "learning_rate": 2.683897768184657e-06,
+      "loss": 0.523,
+      "step": 6385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.892836073277086,
+      "learning_rate": 2.6824059828694715e-06,
+      "loss": 0.6081,
+      "step": 6386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.644640787589624,
+      "learning_rate": 2.6809144602667747e-06,
+      "loss": 0.4926,
+      "step": 6387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7857425923134511,
+      "learning_rate": 2.6794232005456468e-06,
+      "loss": 0.6601,
+      "step": 6388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9604816465995976,
+      "learning_rate": 2.6779322038751217e-06,
+      "loss": 0.6702,
+      "step": 6389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.938706881138097,
+      "learning_rate": 2.676441470424218e-06,
+      "loss": 0.5822,
+      "step": 6390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9790553819807086,
+      "learning_rate": 2.674951000361916e-06,
+      "loss": 0.5655,
+      "step": 6391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078457287587751,
+      "learning_rate": 2.673460793857173e-06,
+      "loss": 0.6475,
+      "step": 6392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1570179625592285,
+      "learning_rate": 2.6719708510789077e-06,
+      "loss": 0.651,
+      "step": 6393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.820206500935895,
+      "learning_rate": 2.6704811721960174e-06,
+      "loss": 0.6093,
+      "step": 6394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8501666310944005,
+      "learning_rate": 2.6689917573773615e-06,
+      "loss": 0.6263,
+      "step": 6395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0029446573228453,
+      "learning_rate": 2.6675026067917808e-06,
+      "loss": 0.6917,
+      "step": 6396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8953373502231254,
+      "learning_rate": 2.6660137206080703e-06,
+      "loss": 0.6434,
+      "step": 6397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2750848498069702,
+      "learning_rate": 2.66452509899501e-06,
+      "loss": 0.6846,
+      "step": 6398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7810113358824438,
+      "learning_rate": 2.66303674212134e-06,
+      "loss": 0.6938,
+      "step": 6399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8671929417800113,
+      "learning_rate": 2.6615486501557765e-06,
+      "loss": 0.6661,
+      "step": 6400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8051102463110213,
+      "learning_rate": 2.660060823267001e-06,
+      "loss": 0.5891,
+      "step": 6401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9343515924482275,
+      "learning_rate": 2.6585732616236705e-06,
+      "loss": 0.6584,
+      "step": 6402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8462900458666527,
+      "learning_rate": 2.657085965394406e-06,
+      "loss": 0.6426,
+      "step": 6403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9951702962287277,
+      "learning_rate": 2.655598934747801e-06,
+      "loss": 0.5901,
+      "step": 6404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9579632915860867,
+      "learning_rate": 2.654112169852418e-06,
+      "loss": 0.6673,
+      "step": 6405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8767384202403952,
+      "learning_rate": 2.652625670876794e-06,
+      "loss": 0.6249,
+      "step": 6406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.965704158295992,
+      "learning_rate": 2.6511394379894274e-06,
+      "loss": 0.632,
+      "step": 6407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1880462552255833,
+      "learning_rate": 2.6496534713587952e-06,
+      "loss": 0.8164,
+      "step": 6408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9072224748583726,
+      "learning_rate": 2.648167771153337e-06,
+      "loss": 0.5904,
+      "step": 6409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8003432003879964,
+      "learning_rate": 2.6466823375414686e-06,
+      "loss": 0.5516,
+      "step": 6410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.940253426639748,
+      "learning_rate": 2.6451971706915713e-06,
+      "loss": 0.6669,
+      "step": 6411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.784289714434628,
+      "learning_rate": 2.6437122707719964e-06,
+      "loss": 0.6114,
+      "step": 6412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8651482926009573,
+      "learning_rate": 2.6422276379510635e-06,
+      "loss": 0.564,
+      "step": 6413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7832427381408453,
+      "learning_rate": 2.6407432723970694e-06,
+      "loss": 0.6461,
+      "step": 6414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8443549195554543,
+      "learning_rate": 2.6392591742782704e-06,
+      "loss": 0.6369,
+      "step": 6415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8770648648089452,
+      "learning_rate": 2.637775343762902e-06,
+      "loss": 0.6854,
+      "step": 6416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7688361860428412,
+      "learning_rate": 2.6362917810191597e-06,
+      "loss": 0.5941,
+      "step": 6417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.052556551651509,
+      "learning_rate": 2.634808486215219e-06,
+      "loss": 0.6625,
+      "step": 6418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8203041172722234,
+      "learning_rate": 2.633325459519218e-06,
+      "loss": 0.5459,
+      "step": 6419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9295198108217466,
+      "learning_rate": 2.6318427010992644e-06,
+      "loss": 0.6982,
+      "step": 6420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8076841531615386,
+      "learning_rate": 2.6303602111234394e-06,
+      "loss": 0.6608,
+      "step": 6421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9542372550586,
+      "learning_rate": 2.6288779897597894e-06,
+      "loss": 0.6433,
+      "step": 6422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7732041112945003,
+      "learning_rate": 2.627396037176336e-06,
+      "loss": 0.6528,
+      "step": 6423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6589412606005547,
+      "learning_rate": 2.6259143535410635e-06,
+      "loss": 0.6287,
+      "step": 6424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9124574288954266,
+      "learning_rate": 2.6244329390219347e-06,
+      "loss": 0.6141,
+      "step": 6425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0180214984374745,
+      "learning_rate": 2.6229517937868687e-06,
+      "loss": 0.6131,
+      "step": 6426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.402149675102522,
+      "learning_rate": 2.621470918003768e-06,
+      "loss": 0.6226,
+      "step": 6427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0034872878783734,
+      "learning_rate": 2.6199903118404934e-06,
+      "loss": 0.6166,
+      "step": 6428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9773378931296095,
+      "learning_rate": 2.6185099754648846e-06,
+      "loss": 0.6453,
+      "step": 6429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7467829991028916,
+      "learning_rate": 2.617029909044742e-06,
+      "loss": 0.5366,
+      "step": 6430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.802869220129952,
+      "learning_rate": 2.615550112747844e-06,
+      "loss": 0.6091,
+      "step": 6431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0939771876409927,
+      "learning_rate": 2.614070586741929e-06,
+      "loss": 0.666,
+      "step": 6432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8675237371257338,
+      "learning_rate": 2.612591331194717e-06,
+      "loss": 0.6055,
+      "step": 6433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8547868852263747,
+      "learning_rate": 2.611112346273881e-06,
+      "loss": 0.7287,
+      "step": 6434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125456167041699,
+      "learning_rate": 2.6096336321470796e-06,
+      "loss": 0.6753,
+      "step": 6435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7754033248669119,
+      "learning_rate": 2.608155188981927e-06,
+      "loss": 0.6194,
+      "step": 6436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9086265288313402,
+      "learning_rate": 2.6066770169460198e-06,
+      "loss": 0.6433,
+      "step": 6437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0396739748065693,
+      "learning_rate": 2.605199116206912e-06,
+      "loss": 0.632,
+      "step": 6438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0428624717964166,
+      "learning_rate": 2.603721486932137e-06,
+      "loss": 0.5821,
+      "step": 6439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9531009300773505,
+      "learning_rate": 2.602244129289189e-06,
+      "loss": 0.6206,
+      "step": 6440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6950662619278845,
+      "learning_rate": 2.6007670434455357e-06,
+      "loss": 0.6074,
+      "step": 6441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2333214792807508,
+      "learning_rate": 2.599290229568612e-06,
+      "loss": 0.6618,
+      "step": 6442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8847057365906073,
+      "learning_rate": 2.5978136878258255e-06,
+      "loss": 0.5443,
+      "step": 6443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9719360697628705,
+      "learning_rate": 2.596337418384548e-06,
+      "loss": 0.568,
+      "step": 6444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078216650575824,
+      "learning_rate": 2.594861421412126e-06,
+      "loss": 0.714,
+      "step": 6445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7669285768038225,
+      "learning_rate": 2.5933856970758693e-06,
+      "loss": 0.6496,
+      "step": 6446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7929630099134348,
+      "learning_rate": 2.591910245543063e-06,
+      "loss": 0.5728,
+      "step": 6447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9304811982301202,
+      "learning_rate": 2.5904350669809554e-06,
+      "loss": 0.6409,
+      "step": 6448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9345487202393299,
+      "learning_rate": 2.5889601615567657e-06,
+      "loss": 0.6752,
+      "step": 6449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2505125627362528,
+      "learning_rate": 2.5874855294376853e-06,
+      "loss": 0.5156,
+      "step": 6450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0098075540433054,
+      "learning_rate": 2.586011170790872e-06,
+      "loss": 0.6312,
+      "step": 6451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7605667785806696,
+      "learning_rate": 2.5845370857834497e-06,
+      "loss": 0.5879,
+      "step": 6452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9979475108404847,
+      "learning_rate": 2.583063274582518e-06,
+      "loss": 0.669,
+      "step": 6453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9387775598018375,
+      "learning_rate": 2.581589737355138e-06,
+      "loss": 0.6661,
+      "step": 6454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.017076403377001,
+      "learning_rate": 2.5801164742683484e-06,
+      "loss": 0.625,
+      "step": 6455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9337213692942614,
+      "learning_rate": 2.5786434854891482e-06,
+      "loss": 0.719,
+      "step": 6456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0865829156729423,
+      "learning_rate": 2.5771707711845096e-06,
+      "loss": 0.6503,
+      "step": 6457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2744014743252934,
+      "learning_rate": 2.5756983315213748e-06,
+      "loss": 0.7139,
+      "step": 6458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6805694624059913,
+      "learning_rate": 2.5742261666666506e-06,
+      "loss": 0.6167,
+      "step": 6459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8025014515127096,
+      "learning_rate": 2.5727542767872188e-06,
+      "loss": 0.6049,
+      "step": 6460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8839136762795483,
+      "learning_rate": 2.5712826620499227e-06,
+      "loss": 0.5673,
+      "step": 6461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.997611542367455,
+      "learning_rate": 2.569811322621584e-06,
+      "loss": 0.6339,
+      "step": 6462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.107752055939515,
+      "learning_rate": 2.5683402586689788e-06,
+      "loss": 0.6669,
+      "step": 6463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1043518474334606,
+      "learning_rate": 2.5668694703588683e-06,
+      "loss": 0.626,
+      "step": 6464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.904909097519436,
+      "learning_rate": 2.565398957857969e-06,
+      "loss": 0.5783,
+      "step": 6465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9161233226944596,
+      "learning_rate": 2.5639287213329767e-06,
+      "loss": 0.6308,
+      "step": 6466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0568730245550317,
+      "learning_rate": 2.5624587609505475e-06,
+      "loss": 0.6569,
+      "step": 6467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8986904541545218,
+      "learning_rate": 2.5609890768773126e-06,
+      "loss": 0.5687,
+      "step": 6468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.925080769355097,
+      "learning_rate": 2.5595196692798664e-06,
+      "loss": 0.6665,
+      "step": 6469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1620329086880075,
+      "learning_rate": 2.5580505383247796e-06,
+      "loss": 0.6607,
+      "step": 6470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8802935287691611,
+      "learning_rate": 2.5565816841785785e-06,
+      "loss": 0.6097,
+      "step": 6471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046944025991196,
+      "learning_rate": 2.555113107007773e-06,
+      "loss": 0.6806,
+      "step": 6472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8878402057696602,
+      "learning_rate": 2.55364480697883e-06,
+      "loss": 0.6483,
+      "step": 6473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8425257728641522,
+      "learning_rate": 2.5521767842581947e-06,
+      "loss": 0.5859,
+      "step": 6474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6780569604641604,
+      "learning_rate": 2.5507090390122704e-06,
+      "loss": 0.657,
+      "step": 6475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9087001854008494,
+      "learning_rate": 2.5492415714074387e-06,
+      "loss": 0.7312,
+      "step": 6476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1493828843849827,
+      "learning_rate": 2.5477743816100443e-06,
+      "loss": 0.7286,
+      "step": 6477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8397832863127606,
+      "learning_rate": 2.5463074697864006e-06,
+      "loss": 0.5727,
+      "step": 6478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9416792168693122,
+      "learning_rate": 2.544840836102789e-06,
+      "loss": 0.6589,
+      "step": 6479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7431185986584572,
+      "learning_rate": 2.543374480725464e-06,
+      "loss": 0.5923,
+      "step": 6480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8136659982858228,
+      "learning_rate": 2.5419084038206422e-06,
+      "loss": 0.616,
+      "step": 6481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.793225841672104,
+      "learning_rate": 2.540442605554516e-06,
+      "loss": 0.6643,
+      "step": 6482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6887884282795877,
+      "learning_rate": 2.5389770860932374e-06,
+      "loss": 0.6283,
+      "step": 6483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0298876307442724,
+      "learning_rate": 2.5375118456029345e-06,
+      "loss": 0.5921,
+      "step": 6484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.861398698981882,
+      "learning_rate": 2.5360468842497004e-06,
+      "loss": 0.5887,
+      "step": 6485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.031756505301046,
+      "learning_rate": 2.5345822021995934e-06,
+      "loss": 0.7035,
+      "step": 6486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9440950572916977,
+      "learning_rate": 2.5331177996186494e-06,
+      "loss": 0.6454,
+      "step": 6487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.018185652460288,
+      "learning_rate": 2.5316536766728605e-06,
+      "loss": 0.6834,
+      "step": 6488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.786345392863452,
+      "learning_rate": 2.5301898335281994e-06,
+      "loss": 0.6012,
+      "step": 6489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.094253077479235,
+      "learning_rate": 2.5287262703505973e-06,
+      "loss": 0.6962,
+      "step": 6490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.016490719493658,
+      "learning_rate": 2.5272629873059564e-06,
+      "loss": 0.7104,
+      "step": 6491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.942400372969221,
+      "learning_rate": 2.525799984560152e-06,
+      "loss": 0.6432,
+      "step": 6492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.841903429983575,
+      "learning_rate": 2.524337262279022e-06,
+      "loss": 0.621,
+      "step": 6493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8022659474271276,
+      "learning_rate": 2.5228748206283716e-06,
+      "loss": 0.6844,
+      "step": 6494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9084102317140246,
+      "learning_rate": 2.521412659773982e-06,
+      "loss": 0.653,
+      "step": 6495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.003649909793403,
+      "learning_rate": 2.5199507798815926e-06,
+      "loss": 0.6809,
+      "step": 6496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.806198288915302,
+      "learning_rate": 2.5184891811169203e-06,
+      "loss": 0.513,
+      "step": 6497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0289321659239388,
+      "learning_rate": 2.5170278636456413e-06,
+      "loss": 0.6757,
+      "step": 6498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1476325520610784,
+      "learning_rate": 2.515566827633411e-06,
+      "loss": 0.6057,
+      "step": 6499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1238370270123315,
+      "learning_rate": 2.5141060732458366e-06,
+      "loss": 0.6114,
+      "step": 6500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8882367928318948,
+      "learning_rate": 2.512645600648511e-06,
+      "loss": 0.6919,
+      "step": 6501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7336732581718404,
+      "learning_rate": 2.511185410006981e-06,
+      "loss": 0.6119,
+      "step": 6502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8563482336968113,
+      "learning_rate": 2.5097255014867733e-06,
+      "loss": 0.6606,
+      "step": 6503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1097536526043412,
+      "learning_rate": 2.508265875253372e-06,
+      "loss": 0.6581,
+      "step": 6504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0103292190475575,
+      "learning_rate": 2.5068065314722378e-06,
+      "loss": 0.5797,
+      "step": 6505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6519741874999505,
+      "learning_rate": 2.5053474703087943e-06,
+      "loss": 0.5882,
+      "step": 6506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0533959837032523,
+      "learning_rate": 2.5038886919284333e-06,
+      "loss": 0.7001,
+      "step": 6507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8805266087401824,
+      "learning_rate": 2.5024301964965157e-06,
+      "loss": 0.5486,
+      "step": 6508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1279785478582482,
+      "learning_rate": 2.500971984178372e-06,
+      "loss": 0.6061,
+      "step": 6509
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.938630985747579,
+      "learning_rate": 2.4995140551392965e-06,
+      "loss": 0.7083,
+      "step": 6510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8335443599118721,
+      "learning_rate": 2.4980564095445562e-06,
+      "loss": 0.5235,
+      "step": 6511
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.148717212825636,
+      "learning_rate": 2.4965990475593814e-06,
+      "loss": 0.5978,
+      "step": 6512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.824758821910578,
+      "learning_rate": 2.495141969348975e-06,
+      "loss": 0.5902,
+      "step": 6513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7780524543423788,
+      "learning_rate": 2.493685175078504e-06,
+      "loss": 0.6284,
+      "step": 6514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8609798057641382,
+      "learning_rate": 2.492228664913104e-06,
+      "loss": 0.6661,
+      "step": 6515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8370681072012822,
+      "learning_rate": 2.4907724390178762e-06,
+      "loss": 0.5885,
+      "step": 6516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.895445022270483,
+      "learning_rate": 2.489316497557897e-06,
+      "loss": 0.5775,
+      "step": 6517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7219155092631773,
+      "learning_rate": 2.487860840698201e-06,
+      "loss": 0.6289,
+      "step": 6518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0460365604603816,
+      "learning_rate": 2.4864054686037993e-06,
+      "loss": 0.6826,
+      "step": 6519
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7147425195038142,
+      "learning_rate": 2.4849503814396624e-06,
+      "loss": 0.5136,
+      "step": 6520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6858038167010478,
+      "learning_rate": 2.4834955793707376e-06,
+      "loss": 0.6515,
+      "step": 6521
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.021509740545536,
+      "learning_rate": 2.4820410625619325e-06,
+      "loss": 0.5946,
+      "step": 6522
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.044099212949625,
+      "learning_rate": 2.4805868311781228e-06,
+      "loss": 0.6226,
+      "step": 6523
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0554872349030506,
+      "learning_rate": 2.4791328853841577e-06,
+      "loss": 0.6877,
+      "step": 6524
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7756247535084098,
+      "learning_rate": 2.4776792253448465e-06,
+      "loss": 0.6656,
+      "step": 6525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9002880458700975,
+      "learning_rate": 2.4762258512249745e-06,
+      "loss": 0.721,
+      "step": 6526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7127361731051247,
+      "learning_rate": 2.4747727631892847e-06,
+      "loss": 0.5513,
+      "step": 6527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.804770412332777,
+      "learning_rate": 2.4733199614024978e-06,
+      "loss": 0.5056,
+      "step": 6528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9003328353324378,
+      "learning_rate": 2.4718674460292945e-06,
+      "loss": 0.5586,
+      "step": 6529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8474351176816664,
+      "learning_rate": 2.470415217234326e-06,
+      "loss": 0.5364,
+      "step": 6530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8999533795354315,
+      "learning_rate": 2.468963275182209e-06,
+      "loss": 0.6509,
+      "step": 6531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7821564463248776,
+      "learning_rate": 2.467511620037533e-06,
+      "loss": 0.6039,
+      "step": 6532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.815283031049219,
+      "learning_rate": 2.466060251964848e-06,
+      "loss": 0.5468,
+      "step": 6533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7976932655952507,
+      "learning_rate": 2.4646091711286783e-06,
+      "loss": 0.6342,
+      "step": 6534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1496032676946073,
+      "learning_rate": 2.4631583776935087e-06,
+      "loss": 0.6687,
+      "step": 6535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9563009201475414,
+      "learning_rate": 2.4617078718237996e-06,
+      "loss": 0.5885,
+      "step": 6536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9025425900601731,
+      "learning_rate": 2.4602576536839672e-06,
+      "loss": 0.7126,
+      "step": 6537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2229623466585378,
+      "learning_rate": 2.4588077234384084e-06,
+      "loss": 0.6516,
+      "step": 6538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078758706067484,
+      "learning_rate": 2.457358081251476e-06,
+      "loss": 0.6454,
+      "step": 6539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.206863288113695,
+      "learning_rate": 2.4559087272875e-06,
+      "loss": 0.5985,
+      "step": 6540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7974189151535742,
+      "learning_rate": 2.454459661710768e-06,
+      "loss": 0.6699,
+      "step": 6541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.183890569895723,
+      "learning_rate": 2.453010884685545e-06,
+      "loss": 0.6428,
+      "step": 6542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1368283835643664,
+      "learning_rate": 2.451562396376055e-06,
+      "loss": 0.6995,
+      "step": 6543
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8922048559177875,
+      "learning_rate": 2.4501141969464936e-06,
+      "loss": 0.7507,
+      "step": 6544
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.044874405618413,
+      "learning_rate": 2.4486662865610194e-06,
+      "loss": 0.682,
+      "step": 6545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.903529162451403,
+      "learning_rate": 2.447218665383766e-06,
+      "loss": 0.5761,
+      "step": 6546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7651894250163067,
+      "learning_rate": 2.445771333578825e-06,
+      "loss": 0.5569,
+      "step": 6547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.930097576630646,
+      "learning_rate": 2.4443242913102645e-06,
+      "loss": 0.6176,
+      "step": 6548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9430342493397708,
+      "learning_rate": 2.44287753874211e-06,
+      "loss": 0.5452,
+      "step": 6549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7755635352299155,
+      "learning_rate": 2.4414310760383635e-06,
+      "loss": 0.5851,
+      "step": 6550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8019542492761174,
+      "learning_rate": 2.439984903362988e-06,
+      "loss": 0.5942,
+      "step": 6551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8952673414148868,
+      "learning_rate": 2.4385390208799153e-06,
+      "loss": 0.5434,
+      "step": 6552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6146056407936027,
+      "learning_rate": 2.437093428753042e-06,
+      "loss": 0.53,
+      "step": 6553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0490607931816394,
+      "learning_rate": 2.4356481271462396e-06,
+      "loss": 0.6576,
+      "step": 6554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0066381095536445,
+      "learning_rate": 2.434203116223336e-06,
+      "loss": 0.6539,
+      "step": 6555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9100614870915273,
+      "learning_rate": 2.4327583961481356e-06,
+      "loss": 0.6054,
+      "step": 6556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9574341823528867,
+      "learning_rate": 2.4313139670844016e-06,
+      "loss": 0.618,
+      "step": 6557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8779301828805786,
+      "learning_rate": 2.429869829195872e-06,
+      "loss": 0.5621,
+      "step": 6558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.06963245031177,
+      "learning_rate": 2.4284259826462475e-06,
+      "loss": 0.6054,
+      "step": 6559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.074995237798964,
+      "learning_rate": 2.4269824275991925e-06,
+      "loss": 0.539,
+      "step": 6560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8641536750051724,
+      "learning_rate": 2.425539164218348e-06,
+      "loss": 0.5851,
+      "step": 6561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7855445771394514,
+      "learning_rate": 2.4240961926673107e-06,
+      "loss": 0.6371,
+      "step": 6562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.033972613929812,
+      "learning_rate": 2.422653513109654e-06,
+      "loss": 0.6142,
+      "step": 6563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1191051352226,
+      "learning_rate": 2.42121112570891e-06,
+      "loss": 0.5777,
+      "step": 6564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8636220501849985,
+      "learning_rate": 2.4197690306285855e-06,
+      "loss": 0.6975,
+      "step": 6565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.255684862296978,
+      "learning_rate": 2.4183272280321477e-06,
+      "loss": 0.6607,
+      "step": 6566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9075617579966504,
+      "learning_rate": 2.416885718083035e-06,
+      "loss": 0.5841,
+      "step": 6567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9785565108430303,
+      "learning_rate": 2.4154445009446457e-06,
+      "loss": 0.6007,
+      "step": 6568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046547103393187,
+      "learning_rate": 2.414003576780357e-06,
+      "loss": 0.6176,
+      "step": 6569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.021562045047473,
+      "learning_rate": 2.4125629457535003e-06,
+      "loss": 0.6506,
+      "step": 6570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9063754184105355,
+      "learning_rate": 2.4111226080273832e-06,
+      "loss": 0.6206,
+      "step": 6571
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7439814488302037,
+      "learning_rate": 2.409682563765273e-06,
+      "loss": 0.6125,
+      "step": 6572
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7171532890783108,
+      "learning_rate": 2.408242813130412e-06,
+      "loss": 0.5728,
+      "step": 6573
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9647678953123011,
+      "learning_rate": 2.406803356285997e-06,
+      "loss": 0.661,
+      "step": 6574
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.76511392894678,
+      "learning_rate": 2.4053641933952043e-06,
+      "loss": 0.4648,
+      "step": 6575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9107782987664588,
+      "learning_rate": 2.4039253246211673e-06,
+      "loss": 0.5542,
+      "step": 6576
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.912398300444987,
+      "learning_rate": 2.402486750126994e-06,
+      "loss": 0.6348,
+      "step": 6577
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7869081393535953,
+      "learning_rate": 2.401048470075751e-06,
+      "loss": 0.5759,
+      "step": 6578
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8508351567029224,
+      "learning_rate": 2.399610484630479e-06,
+      "loss": 0.6337,
+      "step": 6579
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9020644166463498,
+      "learning_rate": 2.3981727939541806e-06,
+      "loss": 0.5074,
+      "step": 6580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8625920857309566,
+      "learning_rate": 2.396735398209825e-06,
+      "loss": 0.5701,
+      "step": 6581
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8707396576414643,
+      "learning_rate": 2.3952982975603494e-06,
+      "loss": 0.6144,
+      "step": 6582
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7674720029464268,
+      "learning_rate": 2.3938614921686592e-06,
+      "loss": 0.4661,
+      "step": 6583
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1866573593617296,
+      "learning_rate": 2.392424982197622e-06,
+      "loss": 0.7057,
+      "step": 6584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5827145809561256,
+      "learning_rate": 2.3909887678100774e-06,
+      "loss": 0.494,
+      "step": 6585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9495572070111211,
+      "learning_rate": 2.3895528491688246e-06,
+      "loss": 0.6217,
+      "step": 6586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.840105029036223,
+      "learning_rate": 2.388117226436638e-06,
+      "loss": 0.6843,
+      "step": 6587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1018687611008997,
+      "learning_rate": 2.3866818997762507e-06,
+      "loss": 0.6491,
+      "step": 6588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9982661992960102,
+      "learning_rate": 2.3852468693503635e-06,
+      "loss": 0.7238,
+      "step": 6589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8010975053442853,
+      "learning_rate": 2.3838121353216494e-06,
+      "loss": 0.6081,
+      "step": 6590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0237199118531266,
+      "learning_rate": 2.3823776978527412e-06,
+      "loss": 0.6033,
+      "step": 6591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8954053993498832,
+      "learning_rate": 2.380943557106239e-06,
+      "loss": 0.6438,
+      "step": 6592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8578242402001044,
+      "learning_rate": 2.379509713244715e-06,
+      "loss": 0.6441,
+      "step": 6593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.913753592633508,
+      "learning_rate": 2.3780761664306988e-06,
+      "loss": 0.5426,
+      "step": 6594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8191071740448894,
+      "learning_rate": 2.3766429168266958e-06,
+      "loss": 0.5654,
+      "step": 6595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.775553219293777,
+      "learning_rate": 2.375209964595171e-06,
+      "loss": 0.601,
+      "step": 6596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0852233066631856,
+      "learning_rate": 2.3737773098985556e-06,
+      "loss": 0.5894,
+      "step": 6597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.811993129732441,
+      "learning_rate": 2.3723449528992527e-06,
+      "loss": 0.5774,
+      "step": 6598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3368875265028968,
+      "learning_rate": 2.3709128937596248e-06,
+      "loss": 0.6608,
+      "step": 6599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.145791013358563,
+      "learning_rate": 2.3694811326420074e-06,
+      "loss": 0.6564,
+      "step": 6600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0766083446244052,
+      "learning_rate": 2.3680496697086956e-06,
+      "loss": 0.6666,
+      "step": 6601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9026813857247935,
+      "learning_rate": 2.366618505121957e-06,
+      "loss": 0.6696,
+      "step": 6602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7019084860970937,
+      "learning_rate": 2.365187639044021e-06,
+      "loss": 0.574,
+      "step": 6603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0689299845262132,
+      "learning_rate": 2.3637570716370835e-06,
+      "loss": 0.6265,
+      "step": 6604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0029581171015454,
+      "learning_rate": 2.362326803063306e-06,
+      "loss": 0.5978,
+      "step": 6605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7689417076989542,
+      "learning_rate": 2.360896833484822e-06,
+      "loss": 0.5353,
+      "step": 6606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0905140765067185,
+      "learning_rate": 2.3594671630637223e-06,
+      "loss": 0.6662,
+      "step": 6607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.768719947566444,
+      "learning_rate": 2.3580377919620716e-06,
+      "loss": 0.6265,
+      "step": 6608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1019145030463755,
+      "learning_rate": 2.3566087203418946e-06,
+      "loss": 0.6763,
+      "step": 6609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8952776750300382,
+      "learning_rate": 2.3551799483651894e-06,
+      "loss": 0.5889,
+      "step": 6610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1165276188915803,
+      "learning_rate": 2.3537514761939083e-06,
+      "loss": 0.7131,
+      "step": 6611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8468004323684801,
+      "learning_rate": 2.3523233039899827e-06,
+      "loss": 0.6277,
+      "step": 6612
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9635660498133092,
+      "learning_rate": 2.3508954319153e-06,
+      "loss": 0.577,
+      "step": 6613
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.173612272307502,
+      "learning_rate": 2.3494678601317204e-06,
+      "loss": 0.5449,
+      "step": 6614
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09870439358293,
+      "learning_rate": 2.3480405888010654e-06,
+      "loss": 0.5512,
+      "step": 6615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9973381551511349,
+      "learning_rate": 2.3466136180851274e-06,
+      "loss": 0.5502,
+      "step": 6616
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0266125822500776,
+      "learning_rate": 2.345186948145659e-06,
+      "loss": 0.5873,
+      "step": 6617
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7407141658103176,
+      "learning_rate": 2.343760579144382e-06,
+      "loss": 0.6076,
+      "step": 6618
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1528326826833104,
+      "learning_rate": 2.342334511242982e-06,
+      "loss": 0.6227,
+      "step": 6619
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.877769142307935,
+      "learning_rate": 2.3409087446031144e-06,
+      "loss": 0.5574,
+      "step": 6620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.970392303151823,
+      "learning_rate": 2.3394832793863955e-06,
+      "loss": 0.5651,
+      "step": 6621
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.896484797097487,
+      "learning_rate": 2.338058115754413e-06,
+      "loss": 0.6181,
+      "step": 6622
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8879614504411355,
+      "learning_rate": 2.336633253868714e-06,
+      "loss": 0.5099,
+      "step": 6623
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9349750897671085,
+      "learning_rate": 2.335208693890819e-06,
+      "loss": 0.6151,
+      "step": 6624
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9611926446015844,
+      "learning_rate": 2.333784435982206e-06,
+      "loss": 0.5171,
+      "step": 6625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0197304863197316,
+      "learning_rate": 2.3323604803043225e-06,
+      "loss": 0.6571,
+      "step": 6626
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9238997930011825,
+      "learning_rate": 2.3309368270185863e-06,
+      "loss": 0.6127,
+      "step": 6627
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9924563648365212,
+      "learning_rate": 2.3295134762863713e-06,
+      "loss": 0.6,
+      "step": 6628
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1233139368564826,
+      "learning_rate": 2.3280904282690268e-06,
+      "loss": 0.6467,
+      "step": 6629
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6964172462449443,
+      "learning_rate": 2.3266676831278625e-06,
+      "loss": 0.5363,
+      "step": 6630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.111431211506792,
+      "learning_rate": 2.325245241024151e-06,
+      "loss": 0.5895,
+      "step": 6631
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8468775750315043,
+      "learning_rate": 2.3238231021191392e-06,
+      "loss": 0.5522,
+      "step": 6632
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.865844535939304,
+      "learning_rate": 2.3224012665740327e-06,
+      "loss": 0.5739,
+      "step": 6633
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0846033087576568,
+      "learning_rate": 2.3209797345500025e-06,
+      "loss": 0.7064,
+      "step": 6634
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9161200839699233,
+      "learning_rate": 2.3195585062081904e-06,
+      "loss": 0.6521,
+      "step": 6635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0009721784284826,
+      "learning_rate": 2.3181375817096986e-06,
+      "loss": 0.6137,
+      "step": 6636
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9519332289640217,
+      "learning_rate": 2.3167169612155997e-06,
+      "loss": 0.5762,
+      "step": 6637
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.837292165340319,
+      "learning_rate": 2.315296644886926e-06,
+      "loss": 0.6546,
+      "step": 6638
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9042025982700623,
+      "learning_rate": 2.313876632884683e-06,
+      "loss": 0.6275,
+      "step": 6639
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8454885332773772,
+      "learning_rate": 2.3124569253698305e-06,
+      "loss": 0.6006,
+      "step": 6640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7636093008937368,
+      "learning_rate": 2.3110375225033056e-06,
+      "loss": 0.5944,
+      "step": 6641
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7390161950750196,
+      "learning_rate": 2.3096184244460025e-06,
+      "loss": 0.6379,
+      "step": 6642
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.069866503330104,
+      "learning_rate": 2.3081996313587873e-06,
+      "loss": 0.6273,
+      "step": 6643
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.752133432566502,
+      "learning_rate": 2.306781143402485e-06,
+      "loss": 0.5812,
+      "step": 6644
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8767393290787764,
+      "learning_rate": 2.305362960737893e-06,
+      "loss": 0.563,
+      "step": 6645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.088228977799822,
+      "learning_rate": 2.3039450835257663e-06,
+      "loss": 0.6123,
+      "step": 6646
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8447026144205156,
+      "learning_rate": 2.3025275119268352e-06,
+      "loss": 0.5872,
+      "step": 6647
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9480786651098552,
+      "learning_rate": 2.3011102461017816e-06,
+      "loss": 0.6293,
+      "step": 6648
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9232095483162872,
+      "learning_rate": 2.299693286211267e-06,
+      "loss": 0.7181,
+      "step": 6649
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9315203503161686,
+      "learning_rate": 2.298276632415908e-06,
+      "loss": 0.6642,
+      "step": 6650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.075640982086266,
+      "learning_rate": 2.296860284876293e-06,
+      "loss": 0.6924,
+      "step": 6651
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9253684895073375,
+      "learning_rate": 2.2954442437529705e-06,
+      "loss": 0.5573,
+      "step": 6652
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9033756169578906,
+      "learning_rate": 2.294028509206461e-06,
+      "loss": 0.5732,
+      "step": 6653
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.097629789333268,
+      "learning_rate": 2.292613081397243e-06,
+      "loss": 0.7167,
+      "step": 6654
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.047506195550759,
+      "learning_rate": 2.2911979604857636e-06,
+      "loss": 0.5737,
+      "step": 6655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1135148834024093,
+      "learning_rate": 2.289783146632434e-06,
+      "loss": 0.6373,
+      "step": 6656
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.153401373307362,
+      "learning_rate": 2.2883686399976335e-06,
+      "loss": 0.6147,
+      "step": 6657
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8547858747646147,
+      "learning_rate": 2.2869544407417016e-06,
+      "loss": 0.6559,
+      "step": 6658
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.871073109375712,
+      "learning_rate": 2.2855405490249498e-06,
+      "loss": 0.6087,
+      "step": 6659
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9986690467749073,
+      "learning_rate": 2.2841269650076468e-06,
+      "loss": 0.6595,
+      "step": 6660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9087678195779618,
+      "learning_rate": 2.282713688850034e-06,
+      "loss": 0.5527,
+      "step": 6661
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.014305188228896,
+      "learning_rate": 2.281300720712313e-06,
+      "loss": 0.7236,
+      "step": 6662
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9398579036796748,
+      "learning_rate": 2.2798880607546486e-06,
+      "loss": 0.5963,
+      "step": 6663
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3211520585069536,
+      "learning_rate": 2.2784757091371797e-06,
+      "loss": 0.5343,
+      "step": 6664
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0825157206727343,
+      "learning_rate": 2.2770636660199983e-06,
+      "loss": 0.6533,
+      "step": 6665
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8372912903103717,
+      "learning_rate": 2.275651931563173e-06,
+      "loss": 0.7069,
+      "step": 6666
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8541732518907224,
+      "learning_rate": 2.274240505926728e-06,
+      "loss": 0.6389,
+      "step": 6667
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.049805210637795,
+      "learning_rate": 2.2728293892706595e-06,
+      "loss": 0.6054,
+      "step": 6668
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.13594935059443,
+      "learning_rate": 2.271418581754924e-06,
+      "loss": 0.6606,
+      "step": 6669
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0250909170956506,
+      "learning_rate": 2.2700080835394444e-06,
+      "loss": 0.6581,
+      "step": 6670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.056761095650911,
+      "learning_rate": 2.2685978947841077e-06,
+      "loss": 0.6361,
+      "step": 6671
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265738878544974,
+      "learning_rate": 2.2671880156487695e-06,
+      "loss": 0.7253,
+      "step": 6672
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8661634959294517,
+      "learning_rate": 2.265778446293245e-06,
+      "loss": 0.6385,
+      "step": 6673
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8345999506199009,
+      "learning_rate": 2.264369186877319e-06,
+      "loss": 0.6667,
+      "step": 6674
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1188435197091833,
+      "learning_rate": 2.2629602375607373e-06,
+      "loss": 0.6107,
+      "step": 6675
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.766286234357666,
+      "learning_rate": 2.2615515985032164e-06,
+      "loss": 0.5546,
+      "step": 6676
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9151920493197223,
+      "learning_rate": 2.260143269864427e-06,
+      "loss": 0.6048,
+      "step": 6677
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7001073092306225,
+      "learning_rate": 2.258735251804017e-06,
+      "loss": 0.4986,
+      "step": 6678
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.995324748961039,
+      "learning_rate": 2.2573275444815886e-06,
+      "loss": 0.6454,
+      "step": 6679
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.121005092218295,
+      "learning_rate": 2.255920148056717e-06,
+      "loss": 0.7024,
+      "step": 6680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1720784642317454,
+      "learning_rate": 2.2545130626889363e-06,
+      "loss": 0.5723,
+      "step": 6681
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0099728903305523,
+      "learning_rate": 2.25310628853775e-06,
+      "loss": 0.6592,
+      "step": 6682
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1435430270674973,
+      "learning_rate": 2.251699825762621e-06,
+      "loss": 0.5571,
+      "step": 6683
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9626458440348495,
+      "learning_rate": 2.2502936745229852e-06,
+      "loss": 0.6454,
+      "step": 6684
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7483370034305972,
+      "learning_rate": 2.2488878349782306e-06,
+      "loss": 0.5895,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9414199777315755,
+      "learning_rate": 2.2474823072877226e-06,
+      "loss": 0.6565,
+      "step": 6686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9913410893273478,
+      "learning_rate": 2.2460770916107823e-06,
+      "loss": 0.6351,
+      "step": 6687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.04831872577914,
+      "learning_rate": 2.244672188106702e-06,
+      "loss": 0.5214,
+      "step": 6688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3021743565866615,
+      "learning_rate": 2.243267596934732e-06,
+      "loss": 0.6176,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9373025833593858,
+      "learning_rate": 2.241863318254095e-06,
+      "loss": 0.6219,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8598534415891308,
+      "learning_rate": 2.2404593522239715e-06,
+      "loss": 0.6604,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6833312249023875,
+      "learning_rate": 2.239055699003509e-06,
+      "loss": 0.5226,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.963731738827018,
+      "learning_rate": 2.2376523587518184e-06,
+      "loss": 0.6036,
+      "step": 6693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.017374029375341,
+      "learning_rate": 2.23624933162798e-06,
+      "loss": 0.7601,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8393505053809505,
+      "learning_rate": 2.23484661779103e-06,
+      "loss": 0.602,
+      "step": 6695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.012100195733019,
+      "learning_rate": 2.2334442173999794e-06,
+      "loss": 0.7229,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.829512485563068,
+      "learning_rate": 2.232042130613793e-06,
+      "loss": 0.6148,
+      "step": 6697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.662329440268267,
+      "learning_rate": 2.2306403575914103e-06,
+      "loss": 0.6035,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.926230971491137,
+      "learning_rate": 2.229238898491728e-06,
+      "loss": 0.5508,
+      "step": 6699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9284111002937832,
+      "learning_rate": 2.2278377534736067e-06,
+      "loss": 0.652,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.023252463481603,
+      "learning_rate": 2.2264369226958794e-06,
+      "loss": 0.5936,
+      "step": 6701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0858121565124996,
+      "learning_rate": 2.225036406317334e-06,
+      "loss": 0.687,
+      "step": 6702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0405317157720906,
+      "learning_rate": 2.2236362044967304e-06,
+      "loss": 0.6622,
+      "step": 6703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8435761932262933,
+      "learning_rate": 2.2222363173927853e-06,
+      "loss": 0.6013,
+      "step": 6704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0622278757264785,
+      "learning_rate": 2.2208367451641886e-06,
+      "loss": 0.5633,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8044115147358368,
+      "learning_rate": 2.219437487969588e-06,
+      "loss": 0.5886,
+      "step": 6706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.869590527714706,
+      "learning_rate": 2.2180385459675964e-06,
+      "loss": 0.598,
+      "step": 6707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.991125138912976,
+      "learning_rate": 2.2166399193167905e-06,
+      "loss": 0.6672,
+      "step": 6708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078202290454276,
+      "learning_rate": 2.2152416081757154e-06,
+      "loss": 0.5946,
+      "step": 6709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9841789691260552,
+      "learning_rate": 2.213843612702876e-06,
+      "loss": 0.624,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9292348217406463,
+      "learning_rate": 2.212445933056745e-06,
+      "loss": 0.5677,
+      "step": 6711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7298685570499033,
+      "learning_rate": 2.211048569395754e-06,
+      "loss": 0.6022,
+      "step": 6712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9413989348010925,
+      "learning_rate": 2.2096515218783084e-06,
+      "loss": 0.5198,
+      "step": 6713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.106895321722421,
+      "learning_rate": 2.208254790662763e-06,
+      "loss": 0.6963,
+      "step": 6714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.899806937788191,
+      "learning_rate": 2.2068583759074513e-06,
+      "loss": 0.6799,
+      "step": 6715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.919074168896033,
+      "learning_rate": 2.2054622777706612e-06,
+      "loss": 0.6711,
+      "step": 6716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8067940485711798,
+      "learning_rate": 2.204066496410653e-06,
+      "loss": 0.5445,
+      "step": 6717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9616475474941888,
+      "learning_rate": 2.2026710319856407e-06,
+      "loss": 0.6405,
+      "step": 6718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1720562157049415,
+      "learning_rate": 2.2012758846538135e-06,
+      "loss": 0.5855,
+      "step": 6719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9905847516723316,
+      "learning_rate": 2.199881054573315e-06,
+      "loss": 0.6609,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8413328911147004,
+      "learning_rate": 2.1984865419022633e-06,
+      "loss": 0.515,
+      "step": 6721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7123430797565387,
+      "learning_rate": 2.197092346798726e-06,
+      "loss": 0.5539,
+      "step": 6722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.809983657480813,
+      "learning_rate": 2.1956984694207495e-06,
+      "loss": 0.5613,
+      "step": 6723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9263268915106502,
+      "learning_rate": 2.1943049099263333e-06,
+      "loss": 0.6399,
+      "step": 6724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.005700997387467,
+      "learning_rate": 2.1929116684734493e-06,
+      "loss": 0.746,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.057704807111089,
+      "learning_rate": 2.1915187452200255e-06,
+      "loss": 0.5946,
+      "step": 6726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9196933178433027,
+      "learning_rate": 2.190126140323962e-06,
+      "loss": 0.601,
+      "step": 6727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.10808635939514,
+      "learning_rate": 2.188733853943116e-06,
+      "loss": 0.683,
+      "step": 6728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9521402765230733,
+      "learning_rate": 2.1873418862353095e-06,
+      "loss": 0.5989,
+      "step": 6729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9980956231048448,
+      "learning_rate": 2.1859502373583336e-06,
+      "loss": 0.6752,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9240445744428603,
+      "learning_rate": 2.184558907469938e-06,
+      "loss": 0.739,
+      "step": 6731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7796759312492878,
+      "learning_rate": 2.1831678967278356e-06,
+      "loss": 0.6363,
+      "step": 6732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8523858957919184,
+      "learning_rate": 2.18177720528971e-06,
+      "loss": 0.6039,
+      "step": 6733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.895909602685592,
+      "learning_rate": 2.1803868333131996e-06,
+      "loss": 0.5693,
+      "step": 6734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0462928765800226,
+      "learning_rate": 2.1789967809559144e-06,
+      "loss": 0.6011,
+      "step": 6735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8139579333458893,
+      "learning_rate": 2.177607048375423e-06,
+      "loss": 0.5731,
+      "step": 6736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9562113590511148,
+      "learning_rate": 2.1762176357292582e-06,
+      "loss": 0.6779,
+      "step": 6737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015963062808654,
+      "learning_rate": 2.174828543174921e-06,
+      "loss": 0.7395,
+      "step": 6738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7438058430933063,
+      "learning_rate": 2.17343977086987e-06,
+      "loss": 0.5682,
+      "step": 6739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3548843240144435,
+      "learning_rate": 2.172051318971533e-06,
+      "loss": 0.7333,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8381592551491213,
+      "learning_rate": 2.170663187637297e-06,
+      "loss": 0.6598,
+      "step": 6741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.153703994101113,
+      "learning_rate": 2.169275377024516e-06,
+      "loss": 0.713,
+      "step": 6742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9617566205072865,
+      "learning_rate": 2.1678878872905063e-06,
+      "loss": 0.6191,
+      "step": 6743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9815659603480604,
+      "learning_rate": 2.1665007185925468e-06,
+      "loss": 0.58,
+      "step": 6744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8177032013635215,
+      "learning_rate": 2.16511387108788e-06,
+      "loss": 0.4916,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8158442719708523,
+      "learning_rate": 2.1637273449337156e-06,
+      "loss": 0.6302,
+      "step": 6746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.027352451014181,
+      "learning_rate": 2.1623411402872206e-06,
+      "loss": 0.5283,
+      "step": 6747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4065306653749294,
+      "learning_rate": 2.160955257305534e-06,
+      "loss": 0.7473,
+      "step": 6748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.218280632800332,
+      "learning_rate": 2.159569696145749e-06,
+      "loss": 0.6691,
+      "step": 6749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8586197979012782,
+      "learning_rate": 2.158184456964932e-06,
+      "loss": 0.5736,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7866938651306266,
+      "learning_rate": 2.1567995399201018e-06,
+      "loss": 0.6215,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.135629174836658,
+      "learning_rate": 2.155414945168251e-06,
+      "loss": 0.6011,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9452854659735015,
+      "learning_rate": 2.1540306728663274e-06,
+      "loss": 0.5866,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.086410546915578,
+      "learning_rate": 2.152646723171251e-06,
+      "loss": 0.5784,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8368795960451758,
+      "learning_rate": 2.1512630962398954e-06,
+      "loss": 0.6354,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.070655836107693,
+      "learning_rate": 2.1498797922291075e-06,
+      "loss": 0.5932,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7569569690003866,
+      "learning_rate": 2.1484968112956884e-06,
+      "loss": 0.5498,
+      "step": 6757
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.821829426027513,
+      "learning_rate": 2.1471141535964126e-06,
+      "loss": 0.6307,
+      "step": 6758
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.898338128453966,
+      "learning_rate": 2.1457318192880043e-06,
+      "loss": 0.6396,
+      "step": 6759
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8115570550325248,
+      "learning_rate": 2.144349808527165e-06,
+      "loss": 0.6435,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8336892208081155,
+      "learning_rate": 2.14296812147055e-06,
+      "loss": 0.5665,
+      "step": 6761
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8332267785810863,
+      "learning_rate": 2.1415867582747847e-06,
+      "loss": 0.5371,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140615577561217,
+      "learning_rate": 2.1402057190964503e-06,
+      "loss": 0.5874,
+      "step": 6763
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0702064176398993,
+      "learning_rate": 2.1388250040921007e-06,
+      "loss": 0.6677,
+      "step": 6764
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6913066775000711,
+      "learning_rate": 2.137444613418244e-06,
+      "loss": 0.5909,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8374709309177701,
+      "learning_rate": 2.1360645472313556e-06,
+      "loss": 0.5114,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9940985284039772,
+      "learning_rate": 2.134684805687876e-06,
+      "loss": 0.7109,
+      "step": 6767
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8285598217429264,
+      "learning_rate": 2.1333053889442033e-06,
+      "loss": 0.6866,
+      "step": 6768
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8855850725628227,
+      "learning_rate": 2.131926297156707e-06,
+      "loss": 0.5933,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.004497545118367,
+      "learning_rate": 2.130547530481712e-06,
+      "loss": 0.6405,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7597570513215508,
+      "learning_rate": 2.1291690890755078e-06,
+      "loss": 0.6394,
+      "step": 6771
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0693726659785225,
+      "learning_rate": 2.1277909730943526e-06,
+      "loss": 0.6361,
+      "step": 6772
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8601120817243084,
+      "learning_rate": 2.126413182694461e-06,
+      "loss": 0.5867,
+      "step": 6773
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0572739362735324,
+      "learning_rate": 2.125035718032013e-06,
+      "loss": 0.7085,
+      "step": 6774
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9037558097151903,
+      "learning_rate": 2.123658579263155e-06,
+      "loss": 0.6568,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0402613319469007,
+      "learning_rate": 2.1222817665439893e-06,
+      "loss": 0.7298,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9559818195133252,
+      "learning_rate": 2.1209052800305897e-06,
+      "loss": 0.6301,
+      "step": 6777
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8799592134451022,
+      "learning_rate": 2.119529119878985e-06,
+      "loss": 0.5894,
+      "step": 6778
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8268326940086235,
+      "learning_rate": 2.1181532862451746e-06,
+      "loss": 0.6109,
+      "step": 6779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.115331740345243,
+      "learning_rate": 2.1167777792851153e-06,
+      "loss": 0.5923,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0187530099260917,
+      "learning_rate": 2.1154025991547283e-06,
+      "loss": 0.5897,
+      "step": 6781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9106316038087288,
+      "learning_rate": 2.114027746009897e-06,
+      "loss": 0.6144,
+      "step": 6782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8569111712932218,
+      "learning_rate": 2.112653220006472e-06,
+      "loss": 0.6748,
+      "step": 6783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8045272023207288,
+      "learning_rate": 2.1112790213002592e-06,
+      "loss": 0.6509,
+      "step": 6784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1234892959481653,
+      "learning_rate": 2.1099051500470368e-06,
+      "loss": 0.78,
+      "step": 6785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.77970791042375,
+      "learning_rate": 2.1085316064025375e-06,
+      "loss": 0.5062,
+      "step": 6786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9188195927515572,
+      "learning_rate": 2.1071583905224643e-06,
+      "loss": 0.5566,
+      "step": 6787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9184818718964722,
+      "learning_rate": 2.105785502562472e-06,
+      "loss": 0.6935,
+      "step": 6788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1765053123488634,
+      "learning_rate": 2.1044129426781925e-06,
+      "loss": 0.6602,
+      "step": 6789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0392376211455963,
+      "learning_rate": 2.1030407110252077e-06,
+      "loss": 0.6014,
+      "step": 6790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.779595008058516,
+      "learning_rate": 2.1016688077590726e-06,
+      "loss": 0.6121,
+      "step": 6791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8969276425103183,
+      "learning_rate": 2.100297233035296e-06,
+      "loss": 0.6356,
+      "step": 6792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.090929508806987,
+      "learning_rate": 2.0989259870093575e-06,
+      "loss": 0.7559,
+      "step": 6793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0298868348206267,
+      "learning_rate": 2.0975550698366924e-06,
+      "loss": 0.669,
+      "step": 6794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9494565795464749,
+      "learning_rate": 2.096184481672707e-06,
+      "loss": 0.5928,
+      "step": 6795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9858014912834072,
+      "learning_rate": 2.0948142226727584e-06,
+      "loss": 0.67,
+      "step": 6796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9407007847129225,
+      "learning_rate": 2.0934442929921783e-06,
+      "loss": 0.539,
+      "step": 6797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0414032198405,
+      "learning_rate": 2.0920746927862523e-06,
+      "loss": 0.6952,
+      "step": 6798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.959079680325511,
+      "learning_rate": 2.0907054222102367e-06,
+      "loss": 0.5815,
+      "step": 6799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8445652931836316,
+      "learning_rate": 2.0893364814193424e-06,
+      "loss": 0.5937,
+      "step": 6800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8116444330336892,
+      "learning_rate": 2.0879678705687495e-06,
+      "loss": 0.667,
+      "step": 6801
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8516217311768648,
+      "learning_rate": 2.0865995898135965e-06,
+      "loss": 0.5826,
+      "step": 6802
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9105961253495354,
+      "learning_rate": 2.0852316393089837e-06,
+      "loss": 0.5449,
+      "step": 6803
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1050591561753578,
+      "learning_rate": 2.083864019209981e-06,
+      "loss": 0.6291,
+      "step": 6804
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8619748597791044,
+      "learning_rate": 2.08249672967161e-06,
+      "loss": 0.5646,
+      "step": 6805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8444285783299936,
+      "learning_rate": 2.081129770848867e-06,
+      "loss": 0.6131,
+      "step": 6806
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.271770773829363,
+      "learning_rate": 2.079763142896699e-06,
+      "loss": 0.5903,
+      "step": 6807
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.052492209017752,
+      "learning_rate": 2.0783968459700253e-06,
+      "loss": 0.607,
+      "step": 6808
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8228278521126375,
+      "learning_rate": 2.077030880223722e-06,
+      "loss": 0.6362,
+      "step": 6809
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1175619753666206,
+      "learning_rate": 2.0756652458126285e-06,
+      "loss": 0.635,
+      "step": 6810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.85016309421988,
+      "learning_rate": 2.074299942891546e-06,
+      "loss": 0.6413,
+      "step": 6811
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8552690267386498,
+      "learning_rate": 2.0729349716152424e-06,
+      "loss": 0.5696,
+      "step": 6812
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1779947285160848,
+      "learning_rate": 2.071570332138442e-06,
+      "loss": 0.6923,
+      "step": 6813
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8964953958247945,
+      "learning_rate": 2.0702060246158378e-06,
+      "loss": 0.6115,
+      "step": 6814
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.799391893677241,
+      "learning_rate": 2.068842049202078e-06,
+      "loss": 0.6503,
+      "step": 6815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8439040043393595,
+      "learning_rate": 2.0674784060517803e-06,
+      "loss": 0.6019,
+      "step": 6816
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0442863866035004,
+      "learning_rate": 2.066115095319521e-06,
+      "loss": 0.5634,
+      "step": 6817
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9730979415829402,
+      "learning_rate": 2.0647521171598376e-06,
+      "loss": 0.6527,
+      "step": 6818
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.965494335587785,
+      "learning_rate": 2.0633894717272308e-06,
+      "loss": 0.5889,
+      "step": 6819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0617167262437013,
+      "learning_rate": 2.0620271591761666e-06,
+      "loss": 0.6645,
+      "step": 6820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0622153316834453,
+      "learning_rate": 2.060665179661068e-06,
+      "loss": 0.6241,
+      "step": 6821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9657200879940147,
+      "learning_rate": 2.0593035333363275e-06,
+      "loss": 0.6282,
+      "step": 6822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1417828950209485,
+      "learning_rate": 2.0579422203562905e-06,
+      "loss": 0.6605,
+      "step": 6823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8466583378272763,
+      "learning_rate": 2.056581240875276e-06,
+      "loss": 0.5548,
+      "step": 6824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9783061847720764,
+      "learning_rate": 2.055220595047551e-06,
+      "loss": 0.5555,
+      "step": 6825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7190514762365583,
+      "learning_rate": 2.053860283027358e-06,
+      "loss": 0.5731,
+      "step": 6826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2693999624623724,
+      "learning_rate": 2.0525003049688923e-06,
+      "loss": 0.521,
+      "step": 6827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.028005261832898,
+      "learning_rate": 2.0511406610263196e-06,
+      "loss": 0.6534,
+      "step": 6828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8211798337378464,
+      "learning_rate": 2.0497813513537583e-06,
+      "loss": 0.5156,
+      "step": 6829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0316985269130874,
+      "learning_rate": 2.048422376105299e-06,
+      "loss": 0.5355,
+      "step": 6830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.035569698591049,
+      "learning_rate": 2.047063735434985e-06,
+      "loss": 0.5921,
+      "step": 6831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8627838823785647,
+      "learning_rate": 2.045705429496831e-06,
+      "loss": 0.5364,
+      "step": 6832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9083218950041783,
+      "learning_rate": 2.044347458444802e-06,
+      "loss": 0.6016,
+      "step": 6833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.925848000350107,
+      "learning_rate": 2.042989822432837e-06,
+      "loss": 0.624,
+      "step": 6834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.098946597015218,
+      "learning_rate": 2.041632521614828e-06,
+      "loss": 0.5922,
+      "step": 6835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.100563969003895,
+      "learning_rate": 2.040275556144637e-06,
+      "loss": 0.6121,
+      "step": 6836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9737206038407717,
+      "learning_rate": 2.03891892617608e-06,
+      "loss": 0.5919,
+      "step": 6837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9297318075312375,
+      "learning_rate": 2.0375626318629418e-06,
+      "loss": 0.6429,
+      "step": 6838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4570564234245285,
+      "learning_rate": 2.036206673358964e-06,
+      "loss": 0.7009,
+      "step": 6839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7813614587881377,
+      "learning_rate": 2.034851050817852e-06,
+      "loss": 0.5578,
+      "step": 6840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.100717069970721,
+      "learning_rate": 2.0334957643932757e-06,
+      "loss": 0.5658,
+      "step": 6841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.02686330607163,
+      "learning_rate": 2.032140814238861e-06,
+      "loss": 0.5931,
+      "step": 6842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6161797864253376,
+      "learning_rate": 2.030786200508203e-06,
+      "loss": 0.4844,
+      "step": 6843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9412498996165422,
+      "learning_rate": 2.0294319233548516e-06,
+      "loss": 0.5354,
+      "step": 6844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.986315746322376,
+      "learning_rate": 2.028077982932325e-06,
+      "loss": 0.6693,
+      "step": 6845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.86339065340936,
+      "learning_rate": 2.026724379394098e-06,
+      "loss": 0.5476,
+      "step": 6846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7315228539679266,
+      "learning_rate": 2.0253711128936104e-06,
+      "loss": 0.5505,
+      "step": 6847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0758485452326267,
+      "learning_rate": 2.0240181835842605e-06,
+      "loss": 0.629,
+      "step": 6848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.98434925798822,
+      "learning_rate": 2.0226655916194127e-06,
+      "loss": 0.555,
+      "step": 6849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.963022411365478,
+      "learning_rate": 2.0213133371523893e-06,
+      "loss": 0.6524,
+      "step": 6850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0305066724656267,
+      "learning_rate": 2.0199614203364787e-06,
+      "loss": 0.6561,
+      "step": 6851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9822291874774338,
+      "learning_rate": 2.018609841324925e-06,
+      "loss": 0.584,
+      "step": 6852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2247395503229326,
+      "learning_rate": 2.0172586002709403e-06,
+      "loss": 0.6224,
+      "step": 6853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.364632538967232,
+      "learning_rate": 2.0159076973276954e-06,
+      "loss": 0.6781,
+      "step": 6854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2066032797604134,
+      "learning_rate": 2.014557132648321e-06,
+      "loss": 0.7244,
+      "step": 6855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.854399879858053,
+      "learning_rate": 2.0132069063859107e-06,
+      "loss": 0.5404,
+      "step": 6856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.198675457264703,
+      "learning_rate": 2.0118570186935234e-06,
+      "loss": 0.679,
+      "step": 6857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9050712289570522,
+      "learning_rate": 2.010507469724173e-06,
+      "loss": 0.702,
+      "step": 6858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0735079350659302,
+      "learning_rate": 2.0091582596308423e-06,
+      "loss": 0.6322,
+      "step": 6859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.010600304322372,
+      "learning_rate": 2.007809388566468e-06,
+      "loss": 0.7113,
+      "step": 6860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7636218721247439,
+      "learning_rate": 2.0064608566839584e-06,
+      "loss": 0.5361,
+      "step": 6861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9501623701778918,
+      "learning_rate": 2.0051126641361697e-06,
+      "loss": 0.6234,
+      "step": 6862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9960117280272525,
+      "learning_rate": 2.0037648110759324e-06,
+      "loss": 0.6327,
+      "step": 6863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.817735464885767,
+      "learning_rate": 2.0024172976560296e-06,
+      "loss": 0.5358,
+      "step": 6864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1683050481880564,
+      "learning_rate": 2.001070124029214e-06,
+      "loss": 0.5889,
+      "step": 6865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8559272054114377,
+      "learning_rate": 1.9997232903481916e-06,
+      "loss": 0.5864,
+      "step": 6866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.101506830003224,
+      "learning_rate": 1.9983767967656364e-06,
+      "loss": 0.6091,
+      "step": 6867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9674756726064457,
+      "learning_rate": 1.9970306434341806e-06,
+      "loss": 0.6021,
+      "step": 6868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9449109915690936,
+      "learning_rate": 1.9956848305064156e-06,
+      "loss": 0.5826,
+      "step": 6869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.312848931186964,
+      "learning_rate": 1.994339358134901e-06,
+      "loss": 0.5566,
+      "step": 6870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8944276535814153,
+      "learning_rate": 1.992994226472152e-06,
+      "loss": 0.6597,
+      "step": 6871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.045736585954296,
+      "learning_rate": 1.9916494356706447e-06,
+      "loss": 0.6818,
+      "step": 6872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8958380409998368,
+      "learning_rate": 1.9903049858828226e-06,
+      "loss": 0.5696,
+      "step": 6873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0705398801996773,
+      "learning_rate": 1.9889608772610837e-06,
+      "loss": 0.5976,
+      "step": 6874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0054209308593265,
+      "learning_rate": 1.987617109957793e-06,
+      "loss": 0.5429,
+      "step": 6875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.041343124527013,
+      "learning_rate": 1.9862736841252734e-06,
+      "loss": 0.6029,
+      "step": 6876
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.918282480016842,
+      "learning_rate": 1.984930599915807e-06,
+      "loss": 0.6197,
+      "step": 6877
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.75857226275821,
+      "learning_rate": 1.983587857481645e-06,
+      "loss": 0.5959,
+      "step": 6878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.918945998094483,
+      "learning_rate": 1.9822454569749895e-06,
+      "loss": 0.5751,
+      "step": 6879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.102556873591886,
+      "learning_rate": 1.980903398548015e-06,
+      "loss": 0.6102,
+      "step": 6880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.894996073709637,
+      "learning_rate": 1.9795616823528457e-06,
+      "loss": 0.6569,
+      "step": 6881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9034762382422525,
+      "learning_rate": 1.978220308541578e-06,
+      "loss": 0.6017,
+      "step": 6882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8933476962584155,
+      "learning_rate": 1.9768792772662616e-06,
+      "loss": 0.5561,
+      "step": 6883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.804196672010393,
+      "learning_rate": 1.9755385886789107e-06,
+      "loss": 0.6527,
+      "step": 6884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8793796521053991,
+      "learning_rate": 1.9741982429314977e-06,
+      "loss": 0.6567,
+      "step": 6885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8753444491107256,
+      "learning_rate": 1.972858240175962e-06,
+      "loss": 0.6202,
+      "step": 6886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0290386500817785,
+      "learning_rate": 1.9715185805641974e-06,
+      "loss": 0.5861,
+      "step": 6887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.847889462311458,
+      "learning_rate": 1.970179264248065e-06,
+      "loss": 0.6332,
+      "step": 6888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.080563361469307,
+      "learning_rate": 1.9688402913793804e-06,
+      "loss": 0.6881,
+      "step": 6889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9954194703008998,
+      "learning_rate": 1.967501662109928e-06,
+      "loss": 0.6653,
+      "step": 6890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.077546547685542,
+      "learning_rate": 1.9661633765914467e-06,
+      "loss": 0.6304,
+      "step": 6891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2298547879318162,
+      "learning_rate": 1.964825434975639e-06,
+      "loss": 0.6967,
+      "step": 6892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.201827109812438,
+      "learning_rate": 1.9634878374141662e-06,
+      "loss": 0.5941,
+      "step": 6893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2384917852547126,
+      "learning_rate": 1.962150584058657e-06,
+      "loss": 0.666,
+      "step": 6894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.113834279753955,
+      "learning_rate": 1.9608136750606917e-06,
+      "loss": 0.6328,
+      "step": 6895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.117062026332053,
+      "learning_rate": 1.959477110571821e-06,
+      "loss": 0.6106,
+      "step": 6896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8239442513943565,
+      "learning_rate": 1.958140890743549e-06,
+      "loss": 0.5488,
+      "step": 6897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8460499061498403,
+      "learning_rate": 1.956805015727348e-06,
+      "loss": 0.4865,
+      "step": 6898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1680120538340493,
+      "learning_rate": 1.955469485674641e-06,
+      "loss": 0.6363,
+      "step": 6899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0433806599917523,
+      "learning_rate": 1.9541343007368225e-06,
+      "loss": 0.5699,
+      "step": 6900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9010351880597207,
+      "learning_rate": 1.95279946106524e-06,
+      "loss": 0.6144,
+      "step": 6901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015347226433184,
+      "learning_rate": 1.951464966811209e-06,
+      "loss": 0.5687,
+      "step": 6902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8436595068007384,
+      "learning_rate": 1.9501308181259986e-06,
+      "loss": 0.5836,
+      "step": 6903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8664745085447245,
+      "learning_rate": 1.948797015160845e-06,
+      "loss": 0.6421,
+      "step": 6904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125979204251544,
+      "learning_rate": 1.947463558066941e-06,
+      "loss": 0.6685,
+      "step": 6905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9993794382514192,
+      "learning_rate": 1.94613044699544e-06,
+      "loss": 0.628,
+      "step": 6906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7669592256552917,
+      "learning_rate": 1.944797682097461e-06,
+      "loss": 0.5624,
+      "step": 6907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8143771752464506,
+      "learning_rate": 1.9434652635240775e-06,
+      "loss": 0.6319,
+      "step": 6908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6875843744992083,
+      "learning_rate": 1.9421331914263293e-06,
+      "loss": 0.5459,
+      "step": 6909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0810872582365767,
+      "learning_rate": 1.9408014659552133e-06,
+      "loss": 0.6141,
+      "step": 6910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8317397564624678,
+      "learning_rate": 1.9394700872616856e-06,
+      "loss": 0.569,
+      "step": 6911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.807931318726164,
+      "learning_rate": 1.9381390554966705e-06,
+      "loss": 0.6178,
+      "step": 6912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.014838295762959,
+      "learning_rate": 1.9368083708110454e-06,
+      "loss": 0.5852,
+      "step": 6913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7712989338761949,
+      "learning_rate": 1.935478033355649e-06,
+      "loss": 0.6652,
+      "step": 6914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.899489708641119,
+      "learning_rate": 1.9341480432812867e-06,
+      "loss": 0.5458,
+      "step": 6915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.719903119059106,
+      "learning_rate": 1.9328184007387163e-06,
+      "loss": 0.5132,
+      "step": 6916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2829620839907903,
+      "learning_rate": 1.9314891058786644e-06,
+      "loss": 0.4828,
+      "step": 6917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8957640141264724,
+      "learning_rate": 1.930160158851811e-06,
+      "loss": 0.5468,
+      "step": 6918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.975808639599023,
+      "learning_rate": 1.9288315598088024e-06,
+      "loss": 0.6222,
+      "step": 6919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9159184447005044,
+      "learning_rate": 1.9275033089002413e-06,
+      "loss": 0.5843,
+      "step": 6920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.865664172333486,
+      "learning_rate": 1.9261754062766937e-06,
+      "loss": 0.667,
+      "step": 6921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8030021224323394,
+      "learning_rate": 1.9248478520886815e-06,
+      "loss": 0.5762,
+      "step": 6922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9875737286149218,
+      "learning_rate": 1.923520646486695e-06,
+      "loss": 0.5619,
+      "step": 6923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8218277493472246,
+      "learning_rate": 1.9221937896211773e-06,
+      "loss": 0.602,
+      "step": 6924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9715617137675558,
+      "learning_rate": 1.920867281642538e-06,
+      "loss": 0.6224,
+      "step": 6925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.131266043950465,
+      "learning_rate": 1.9195411227011403e-06,
+      "loss": 0.5832,
+      "step": 6926
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0327060331736404,
+      "learning_rate": 1.9182153129473167e-06,
+      "loss": 0.7407,
+      "step": 6927
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8922822746296175,
+      "learning_rate": 1.916889852531353e-06,
+      "loss": 0.5791,
+      "step": 6928
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9156447848338625,
+      "learning_rate": 1.9155647416034972e-06,
+      "loss": 0.5832,
+      "step": 6929
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8836997633797954,
+      "learning_rate": 1.914239980313958e-06,
+      "loss": 0.6571,
+      "step": 6930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.041638407251988,
+      "learning_rate": 1.912915568812906e-06,
+      "loss": 0.5327,
+      "step": 6931
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.958615762904026,
+      "learning_rate": 1.9115915072504683e-06,
+      "loss": 0.6959,
+      "step": 6932
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9335752293452386,
+      "learning_rate": 1.9102677957767384e-06,
+      "loss": 0.6213,
+      "step": 6933
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8751908142036637,
+      "learning_rate": 1.9089444345417636e-06,
+      "loss": 0.6393,
+      "step": 6934
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8739242366864037,
+      "learning_rate": 1.9076214236955585e-06,
+      "loss": 0.6625,
+      "step": 6935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.862653258537508,
+      "learning_rate": 1.9062987633880876e-06,
+      "loss": 0.6272,
+      "step": 6936
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9757417242906068,
+      "learning_rate": 1.9049764537692872e-06,
+      "loss": 0.5792,
+      "step": 6937
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0341787737940793,
+      "learning_rate": 1.903654494989045e-06,
+      "loss": 0.66,
+      "step": 6938
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8427766740494667,
+      "learning_rate": 1.9023328871972163e-06,
+      "loss": 0.5716,
+      "step": 6939
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.024770792040896,
+      "learning_rate": 1.9010116305436094e-06,
+      "loss": 0.6085,
+      "step": 6940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8948892940354372,
+      "learning_rate": 1.8996907251779988e-06,
+      "loss": 0.5694,
+      "step": 6941
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.889140157202737,
+      "learning_rate": 1.8983701712501163e-06,
+      "loss": 0.6102,
+      "step": 6942
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0194435530553716,
+      "learning_rate": 1.8970499689096516e-06,
+      "loss": 0.6385,
+      "step": 6943
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9671159856485523,
+      "learning_rate": 1.895730118306261e-06,
+      "loss": 0.6185,
+      "step": 6944
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.906154410239305,
+      "learning_rate": 1.8944106195895535e-06,
+      "loss": 0.5676,
+      "step": 6945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1301700751404984,
+      "learning_rate": 1.8930914729091055e-06,
+      "loss": 0.6528,
+      "step": 6946
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8831602503537106,
+      "learning_rate": 1.8917726784144458e-06,
+      "loss": 0.5776,
+      "step": 6947
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.062453945989891,
+      "learning_rate": 1.890454236255071e-06,
+      "loss": 0.61,
+      "step": 6948
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9660287314255194,
+      "learning_rate": 1.8891361465804326e-06,
+      "loss": 0.6367,
+      "step": 6949
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9541472768077464,
+      "learning_rate": 1.8878184095399428e-06,
+      "loss": 0.5979,
+      "step": 6950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7630023901018252,
+      "learning_rate": 1.886501025282974e-06,
+      "loss": 0.5508,
+      "step": 6951
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.693350200238252,
+      "learning_rate": 1.8851839939588617e-06,
+      "loss": 0.5917,
+      "step": 6952
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9405268803538889,
+      "learning_rate": 1.8838673157168956e-06,
+      "loss": 0.6026,
+      "step": 6953
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8552942459894206,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.6218,
+      "step": 6954
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8592020844670467,
+      "learning_rate": 1.8812350190763822e-06,
+      "loss": 0.6586,
+      "step": 6955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0291508697413643,
+      "learning_rate": 1.8799194009762201e-06,
+      "loss": 0.6224,
+      "step": 6956
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0727455489975797,
+      "learning_rate": 1.8786041365549784e-06,
+      "loss": 0.7172,
+      "step": 6957
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8228323748167872,
+      "learning_rate": 1.8772892259617487e-06,
+      "loss": 0.5831,
+      "step": 6958
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8902269228729196,
+      "learning_rate": 1.875974669345582e-06,
+      "loss": 0.6655,
+      "step": 6959
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.962322524254648,
+      "learning_rate": 1.8746604668554952e-06,
+      "loss": 0.5862,
+      "step": 6960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2883723322771603,
+      "learning_rate": 1.8733466186404565e-06,
+      "loss": 0.6552,
+      "step": 6961
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8072852247050788,
+      "learning_rate": 1.8720331248494012e-06,
+      "loss": 0.5782,
+      "step": 6962
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.123278381013698,
+      "learning_rate": 1.8707199856312186e-06,
+      "loss": 0.7116,
+      "step": 6963
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0807833558357838,
+      "learning_rate": 1.8694072011347636e-06,
+      "loss": 0.7126,
+      "step": 6964
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1994360140008165,
+      "learning_rate": 1.8680947715088465e-06,
+      "loss": 0.6547,
+      "step": 6965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3923087716494797,
+      "learning_rate": 1.8667826969022379e-06,
+      "loss": 0.6168,
+      "step": 6966
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2628861411700947,
+      "learning_rate": 1.8654709774636676e-06,
+      "loss": 0.6133,
+      "step": 6967
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9026703854680282,
+      "learning_rate": 1.8641596133418305e-06,
+      "loss": 0.5893,
+      "step": 6968
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.024038972343844,
+      "learning_rate": 1.8628486046853728e-06,
+      "loss": 0.6062,
+      "step": 6969
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8290810303514669,
+      "learning_rate": 1.8615379516429084e-06,
+      "loss": 0.6898,
+      "step": 6970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9372867721866058,
+      "learning_rate": 1.8602276543630044e-06,
+      "loss": 0.6593,
+      "step": 6971
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0472330529227487,
+      "learning_rate": 1.858917712994195e-06,
+      "loss": 0.5912,
+      "step": 6972
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0346582715337576,
+      "learning_rate": 1.8576081276849633e-06,
+      "loss": 0.6109,
+      "step": 6973
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.881561791570671,
+      "learning_rate": 1.8562988985837632e-06,
+      "loss": 0.6301,
+      "step": 6974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9364107609270342,
+      "learning_rate": 1.8549900258389992e-06,
+      "loss": 0.5445,
+      "step": 6975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.764202402624153,
+      "learning_rate": 1.853681509599044e-06,
+      "loss": 0.5492,
+      "step": 6976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9477673427273048,
+      "learning_rate": 1.852373350012221e-06,
+      "loss": 0.5653,
+      "step": 6977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0400789263657946,
+      "learning_rate": 1.8510655472268212e-06,
+      "loss": 0.5676,
+      "step": 6978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9127120678077745,
+      "learning_rate": 1.8497581013910904e-06,
+      "loss": 0.6438,
+      "step": 6979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9248975570888085,
+      "learning_rate": 1.848451012653233e-06,
+      "loss": 0.6097,
+      "step": 6980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9422713468844135,
+      "learning_rate": 1.8471442811614177e-06,
+      "loss": 0.6654,
+      "step": 6981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.160167412360216,
+      "learning_rate": 1.8458379070637678e-06,
+      "loss": 0.6017,
+      "step": 6982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0222848743451896,
+      "learning_rate": 1.8445318905083703e-06,
+      "loss": 0.7048,
+      "step": 6983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.057612384950042,
+      "learning_rate": 1.843226231643267e-06,
+      "loss": 0.6671,
+      "step": 6984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7047056832519873,
+      "learning_rate": 1.8419209306164653e-06,
+      "loss": 0.6048,
+      "step": 6985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.060269652286786,
+      "learning_rate": 1.8406159875759266e-06,
+      "loss": 0.6438,
+      "step": 6986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.159682548383432,
+      "learning_rate": 1.8393114026695736e-06,
+      "loss": 0.5876,
+      "step": 6987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8583493961058244,
+      "learning_rate": 1.8380071760452862e-06,
+      "loss": 0.7386,
+      "step": 6988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7793109875706183,
+      "learning_rate": 1.83670330785091e-06,
+      "loss": 0.5908,
+      "step": 6989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1854173295401274,
+      "learning_rate": 1.8353997982342425e-06,
+      "loss": 0.6927,
+      "step": 6990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6687385136463946,
+      "learning_rate": 1.8340966473430477e-06,
+      "loss": 0.5707,
+      "step": 6991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.817059248514043,
+      "learning_rate": 1.8327938553250407e-06,
+      "loss": 0.5635,
+      "step": 6992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9641887104756441,
+      "learning_rate": 1.8314914223279046e-06,
+      "loss": 0.7102,
+      "step": 6993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7640602296208145,
+      "learning_rate": 1.8301893484992755e-06,
+      "loss": 0.6463,
+      "step": 6994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2745339932755324,
+      "learning_rate": 1.8288876339867511e-06,
+      "loss": 0.7781,
+      "step": 6995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8716453246074705,
+      "learning_rate": 1.8275862789378862e-06,
+      "loss": 0.6602,
+      "step": 6996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.353529730295554,
+      "learning_rate": 1.826285283500201e-06,
+      "loss": 0.7568,
+      "step": 6997
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9426701427221278,
+      "learning_rate": 1.8249846478211663e-06,
+      "loss": 0.7043,
+      "step": 6998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9052478319387323,
+      "learning_rate": 1.8236843720482206e-06,
+      "loss": 0.631,
+      "step": 6999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.884005001155357,
+      "learning_rate": 1.822384456328754e-06,
+      "loss": 0.5732,
+      "step": 7000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.906581597726097,
+      "learning_rate": 1.8210849008101244e-06,
+      "loss": 0.6591,
+      "step": 7001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.032360075463133,
+      "learning_rate": 1.8197857056396372e-06,
+      "loss": 0.5087,
+      "step": 7002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.072367018499558,
+      "learning_rate": 1.8184868709645686e-06,
+      "loss": 0.6859,
+      "step": 7003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9137751075131972,
+      "learning_rate": 1.8171883969321458e-06,
+      "loss": 0.6251,
+      "step": 7004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.016277252510613,
+      "learning_rate": 1.815890283689561e-06,
+      "loss": 0.7028,
+      "step": 7005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8514033044622462,
+      "learning_rate": 1.8145925313839597e-06,
+      "loss": 0.6176,
+      "step": 7006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.707417328439556,
+      "learning_rate": 1.8132951401624527e-06,
+      "loss": 0.5991,
+      "step": 7007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.05162512516868,
+      "learning_rate": 1.811998110172104e-06,
+      "loss": 0.6503,
+      "step": 7008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8997173006855768,
+      "learning_rate": 1.8107014415599416e-06,
+      "loss": 0.6086,
+      "step": 7009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9842498026525295,
+      "learning_rate": 1.8094051344729497e-06,
+      "loss": 0.6816,
+      "step": 7010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8316376168242585,
+      "learning_rate": 1.808109189058071e-06,
+      "loss": 0.5557,
+      "step": 7011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.841851848534049,
+      "learning_rate": 1.8068136054622076e-06,
+      "loss": 0.6038,
+      "step": 7012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7337353429307356,
+      "learning_rate": 1.8055183838322243e-06,
+      "loss": 0.6381,
+      "step": 7013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.066314680142265,
+      "learning_rate": 1.804223524314938e-06,
+      "loss": 0.6052,
+      "step": 7014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9221531255760698,
+      "learning_rate": 1.802929027057133e-06,
+      "loss": 0.6784,
+      "step": 7015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8537567633360141,
+      "learning_rate": 1.8016348922055448e-06,
+      "loss": 0.6299,
+      "step": 7016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9730320203082514,
+      "learning_rate": 1.8003411199068704e-06,
+      "loss": 0.5932,
+      "step": 7017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9237631065008112,
+      "learning_rate": 1.7990477103077691e-06,
+      "loss": 0.6454,
+      "step": 7018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9063776911922903,
+      "learning_rate": 1.7977546635548527e-06,
+      "loss": 0.5453,
+      "step": 7019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7272938089567267,
+      "learning_rate": 1.7964619797946991e-06,
+      "loss": 0.5609,
+      "step": 7020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9139613548645424,
+      "learning_rate": 1.7951696591738382e-06,
+      "loss": 0.6083,
+      "step": 7021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0076928391817583,
+      "learning_rate": 1.7938777018387648e-06,
+      "loss": 0.6729,
+      "step": 7022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8414781413829706,
+      "learning_rate": 1.7925861079359268e-06,
+      "loss": 0.6657,
+      "step": 7023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9926494137956807,
+      "learning_rate": 1.7912948776117385e-06,
+      "loss": 0.5832,
+      "step": 7024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.047037722371666,
+      "learning_rate": 1.7900040110125611e-06,
+      "loss": 0.5773,
+      "step": 7025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9110124757511402,
+      "learning_rate": 1.7887135082847274e-06,
+      "loss": 0.6048,
+      "step": 7026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9091170712990175,
+      "learning_rate": 1.7874233695745191e-06,
+      "loss": 0.5744,
+      "step": 7027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1501681795544605,
+      "learning_rate": 1.786133595028185e-06,
+      "loss": 0.6717,
+      "step": 7028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8727818568701433,
+      "learning_rate": 1.7848441847919246e-06,
+      "loss": 0.6262,
+      "step": 7029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0302444195043146,
+      "learning_rate": 1.7835551390119033e-06,
+      "loss": 0.6903,
+      "step": 7030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8811800640863703,
+      "learning_rate": 1.7822664578342403e-06,
+      "loss": 0.5732,
+      "step": 7031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0044708198114662,
+      "learning_rate": 1.7809781414050147e-06,
+      "loss": 0.6878,
+      "step": 7032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0627253959610394,
+      "learning_rate": 1.7796901898702628e-06,
+      "loss": 0.6258,
+      "step": 7033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9855655422507894,
+      "learning_rate": 1.7784026033759844e-06,
+      "loss": 0.7077,
+      "step": 7034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1111295924164546,
+      "learning_rate": 1.777115382068132e-06,
+      "loss": 0.6692,
+      "step": 7035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0552672782335533,
+      "learning_rate": 1.7758285260926228e-06,
+      "loss": 0.6351,
+      "step": 7036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.790621899903027,
+      "learning_rate": 1.7745420355953253e-06,
+      "loss": 0.5409,
+      "step": 7037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0282472564011536,
+      "learning_rate": 1.7732559107220765e-06,
+      "loss": 0.6969,
+      "step": 7038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8675459459068524,
+      "learning_rate": 1.7719701516186578e-06,
+      "loss": 0.5922,
+      "step": 7039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2556574076998035,
+      "learning_rate": 1.770684758430824e-06,
+      "loss": 0.6478,
+      "step": 7040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4928599008992904,
+      "learning_rate": 1.769399731304277e-06,
+      "loss": 0.5609,
+      "step": 7041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8371152290767596,
+      "learning_rate": 1.7681150703846867e-06,
+      "loss": 0.6117,
+      "step": 7042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.036972378473435,
+      "learning_rate": 1.7668307758176717e-06,
+      "loss": 0.6935,
+      "step": 7043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9198084244760627,
+      "learning_rate": 1.7655468477488191e-06,
+      "loss": 0.5391,
+      "step": 7044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7153890540288939,
+      "learning_rate": 1.7642632863236653e-06,
+      "loss": 0.5468,
+      "step": 7045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9903008100611193,
+      "learning_rate": 1.7629800916877126e-06,
+      "loss": 0.697,
+      "step": 7046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9312849352503563,
+      "learning_rate": 1.7616972639864166e-06,
+      "loss": 0.5932,
+      "step": 7047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9990339340200853,
+      "learning_rate": 1.7604148033651925e-06,
+      "loss": 0.6599,
+      "step": 7048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1694031802547125,
+      "learning_rate": 1.7591327099694167e-06,
+      "loss": 0.6993,
+      "step": 7049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8424469644992956,
+      "learning_rate": 1.7578509839444202e-06,
+      "loss": 0.5265,
+      "step": 7050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.890458213661419,
+      "learning_rate": 1.756569625435493e-06,
+      "loss": 0.6508,
+      "step": 7051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0184026859189967,
+      "learning_rate": 1.7552886345878879e-06,
+      "loss": 0.6978,
+      "step": 7052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0127389742757353,
+      "learning_rate": 1.7540080115468095e-06,
+      "loss": 0.6505,
+      "step": 7053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3335684879728342,
+      "learning_rate": 1.752727756457423e-06,
+      "loss": 0.6635,
+      "step": 7054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9125226367301975,
+      "learning_rate": 1.7514478694648563e-06,
+      "loss": 0.6221,
+      "step": 7055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2096866866360534,
+      "learning_rate": 1.7501683507141876e-06,
+      "loss": 0.6165,
+      "step": 7056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8901019668590993,
+      "learning_rate": 1.7488892003504615e-06,
+      "loss": 0.6364,
+      "step": 7057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.835005177582281,
+      "learning_rate": 1.7476104185186737e-06,
+      "loss": 0.5718,
+      "step": 7058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7368860915767115,
+      "learning_rate": 1.7463320053637844e-06,
+      "loss": 0.6054,
+      "step": 7059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0065600854979793,
+      "learning_rate": 1.745053961030706e-06,
+      "loss": 0.7173,
+      "step": 7060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9553760769747448,
+      "learning_rate": 1.743776285664317e-06,
+      "loss": 0.6248,
+      "step": 7061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9170304004440861,
+      "learning_rate": 1.7424989794094426e-06,
+      "loss": 0.5648,
+      "step": 7062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8843076141856252,
+      "learning_rate": 1.7412220424108778e-06,
+      "loss": 0.5752,
+      "step": 7063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.892460870380063,
+      "learning_rate": 1.739945474813367e-06,
+      "loss": 0.6732,
+      "step": 7064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.064212792184756,
+      "learning_rate": 1.7386692767616204e-06,
+      "loss": 0.6855,
+      "step": 7065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0042471676355587,
+      "learning_rate": 1.737393448400298e-06,
+      "loss": 0.6518,
+      "step": 7066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1605114964526453,
+      "learning_rate": 1.7361179898740265e-06,
+      "loss": 0.5507,
+      "step": 7067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9867984391295228,
+      "learning_rate": 1.7348429013273844e-06,
+      "loss": 0.583,
+      "step": 7068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8676294756248204,
+      "learning_rate": 1.73356818290491e-06,
+      "loss": 0.5919,
+      "step": 7069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0320162519150218,
+      "learning_rate": 1.7322938347510986e-06,
+      "loss": 0.6553,
+      "step": 7070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.996766764699223,
+      "learning_rate": 1.731019857010408e-06,
+      "loss": 0.6392,
+      "step": 7071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0880116579155645,
+      "learning_rate": 1.7297462498272476e-06,
+      "loss": 0.6324,
+      "step": 7072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9530026108454182,
+      "learning_rate": 1.728473013345991e-06,
+      "loss": 0.6077,
+      "step": 7073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.853123715597261,
+      "learning_rate": 1.727200147710964e-06,
+      "loss": 0.6759,
+      "step": 7074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6900306342788634,
+      "learning_rate": 1.7259276530664577e-06,
+      "loss": 0.6573,
+      "step": 7075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.808885919868968,
+      "learning_rate": 1.7246555295567102e-06,
+      "loss": 0.6117,
+      "step": 7076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.962491279043586,
+      "learning_rate": 1.7233837773259288e-06,
+      "loss": 0.563,
+      "step": 7077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0288719441359766,
+      "learning_rate": 1.7221123965182712e-06,
+      "loss": 0.5779,
+      "step": 7078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.122216006034978,
+      "learning_rate": 1.720841387277858e-06,
+      "loss": 0.6406,
+      "step": 7079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8225722448977553,
+      "learning_rate": 1.7195707497487624e-06,
+      "loss": 0.6658,
+      "step": 7080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8260903677576232,
+      "learning_rate": 1.7183004840750223e-06,
+      "loss": 0.5738,
+      "step": 7081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9433119305101705,
+      "learning_rate": 1.7170305904006252e-06,
+      "loss": 0.6746,
+      "step": 7082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.277323368339559,
+      "learning_rate": 1.7157610688695248e-06,
+      "loss": 0.6327,
+      "step": 7083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9858333985133685,
+      "learning_rate": 1.714491919625627e-06,
+      "loss": 0.6856,
+      "step": 7084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.875314227630569,
+      "learning_rate": 1.7132231428127949e-06,
+      "loss": 0.5694,
+      "step": 7085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8591613869115624,
+      "learning_rate": 1.7119547385748552e-06,
+      "loss": 0.6008,
+      "step": 7086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1645944722824653,
+      "learning_rate": 1.710686707055586e-06,
+      "loss": 0.5587,
+      "step": 7087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.934409516211822,
+      "learning_rate": 1.7094190483987282e-06,
+      "loss": 0.556,
+      "step": 7088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5143520523481275,
+      "learning_rate": 1.708151762747977e-06,
+      "loss": 0.5124,
+      "step": 7089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8513917629013414,
+      "learning_rate": 1.7068848502469866e-06,
+      "loss": 0.5633,
+      "step": 7090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0804716730668145,
+      "learning_rate": 1.7056183110393666e-06,
+      "loss": 0.7043,
+      "step": 7091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8489263596056118,
+      "learning_rate": 1.7043521452686902e-06,
+      "loss": 0.6114,
+      "step": 7092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9388587591896767,
+      "learning_rate": 1.7030863530784814e-06,
+      "loss": 0.5371,
+      "step": 7093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8122345033573635,
+      "learning_rate": 1.7018209346122272e-06,
+      "loss": 0.5691,
+      "step": 7094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0118884418469616,
+      "learning_rate": 1.7005558900133678e-06,
+      "loss": 0.6261,
+      "step": 7095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.734039688883811,
+      "learning_rate": 1.6992912194253065e-06,
+      "loss": 0.5549,
+      "step": 7096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1291135005730237,
+      "learning_rate": 1.6980269229913965e-06,
+      "loss": 0.7039,
+      "step": 7097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.040583129036302,
+      "learning_rate": 1.696763000854959e-06,
+      "loss": 0.6216,
+      "step": 7098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7362654789557477,
+      "learning_rate": 1.69549945315926e-06,
+      "loss": 0.5089,
+      "step": 7099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8933279122929687,
+      "learning_rate": 1.6942362800475343e-06,
+      "loss": 0.6009,
+      "step": 7100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0124255601190577,
+      "learning_rate": 1.6929734816629674e-06,
+      "loss": 0.672,
+      "step": 7101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0199051597392446,
+      "learning_rate": 1.6917110581487067e-06,
+      "loss": 0.5459,
+      "step": 7102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.024380319703663,
+      "learning_rate": 1.690449009647853e-06,
+      "loss": 0.6074,
+      "step": 7103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7856784382758137,
+      "learning_rate": 1.6891873363034693e-06,
+      "loss": 0.4908,
+      "step": 7104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8101573491267204,
+      "learning_rate": 1.6879260382585727e-06,
+      "loss": 0.5492,
+      "step": 7105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9753378786530036,
+      "learning_rate": 1.686665115656137e-06,
+      "loss": 0.6705,
+      "step": 7106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9769168029519377,
+      "learning_rate": 1.6854045686390947e-06,
+      "loss": 0.6015,
+      "step": 7107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0534968822940582,
+      "learning_rate": 1.6841443973503384e-06,
+      "loss": 0.5757,
+      "step": 7108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1367136070784816,
+      "learning_rate": 1.6828846019327128e-06,
+      "loss": 0.6714,
+      "step": 7109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9896795722232372,
+      "learning_rate": 1.6816251825290265e-06,
+      "loss": 0.6377,
+      "step": 7110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.051473929576098,
+      "learning_rate": 1.680366139282038e-06,
+      "loss": 0.6382,
+      "step": 7111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.845927910849788,
+      "learning_rate": 1.6791074723344719e-06,
+      "loss": 0.6508,
+      "step": 7112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1883562893787234,
+      "learning_rate": 1.6778491818289995e-06,
+      "loss": 0.5516,
+      "step": 7113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.334556242298653,
+      "learning_rate": 1.6765912679082592e-06,
+      "loss": 0.6544,
+      "step": 7114
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7710546453451088,
+      "learning_rate": 1.67533373071484e-06,
+      "loss": 0.6163,
+      "step": 7115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0008488447691177,
+      "learning_rate": 1.6740765703912942e-06,
+      "loss": 0.6591,
+      "step": 7116
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9710345983015822,
+      "learning_rate": 1.6728197870801244e-06,
+      "loss": 0.6042,
+      "step": 7117
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0038441788739054,
+      "learning_rate": 1.6715633809237974e-06,
+      "loss": 0.5822,
+      "step": 7118
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.987280070196863,
+      "learning_rate": 1.6703073520647316e-06,
+      "loss": 0.5685,
+      "step": 7119
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2229776012908724,
+      "learning_rate": 1.6690517006453071e-06,
+      "loss": 0.6201,
+      "step": 7120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0164190353791467,
+      "learning_rate": 1.6677964268078584e-06,
+      "loss": 0.6305,
+      "step": 7121
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9884428866003194,
+      "learning_rate": 1.6665415306946764e-06,
+      "loss": 0.6417,
+      "step": 7122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7354582036347406,
+      "learning_rate": 1.665287012448013e-06,
+      "loss": 0.5151,
+      "step": 7123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2562445463263816,
+      "learning_rate": 1.6640328722100723e-06,
+      "loss": 0.6702,
+      "step": 7124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8771207954912414,
+      "learning_rate": 1.6627791101230222e-06,
+      "loss": 0.5857,
+      "step": 7125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.507413450280522,
+      "learning_rate": 1.6615257263289809e-06,
+      "loss": 0.5257,
+      "step": 7126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.057813853105924,
+      "learning_rate": 1.6602727209700276e-06,
+      "loss": 0.6122,
+      "step": 7127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.309053131838771,
+      "learning_rate": 1.659020094188195e-06,
+      "loss": 0.5922,
+      "step": 7128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.210830259594895,
+      "learning_rate": 1.6577678461254797e-06,
+      "loss": 0.6566,
+      "step": 7129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8767799929288522,
+      "learning_rate": 1.6565159769238276e-06,
+      "loss": 0.6357,
+      "step": 7130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9208667473650485,
+      "learning_rate": 1.6552644867251483e-06,
+      "loss": 0.5577,
+      "step": 7131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9185506346962085,
+      "learning_rate": 1.6540133756713017e-06,
+      "loss": 0.6611,
+      "step": 7132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8123867213865923,
+      "learning_rate": 1.6527626439041128e-06,
+      "loss": 0.5769,
+      "step": 7133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6904443041818638,
+      "learning_rate": 1.6515122915653564e-06,
+      "loss": 0.5196,
+      "step": 7134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8405788256050637,
+      "learning_rate": 1.6502623187967675e-06,
+      "loss": 0.5961,
+      "step": 7135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.276571617181377,
+      "learning_rate": 1.6490127257400363e-06,
+      "loss": 0.5643,
+      "step": 7136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.08819425335901,
+      "learning_rate": 1.6477635125368136e-06,
+      "loss": 0.6155,
+      "step": 7137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9634742797877143,
+      "learning_rate": 1.6465146793287028e-06,
+      "loss": 0.7454,
+      "step": 7138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9754650148020452,
+      "learning_rate": 1.645266226257269e-06,
+      "loss": 0.5814,
+      "step": 7139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9078401551432211,
+      "learning_rate": 1.6440181534640277e-06,
+      "loss": 0.6458,
+      "step": 7140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9830325654887706,
+      "learning_rate": 1.6427704610904594e-06,
+      "loss": 0.6258,
+      "step": 7141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9722224448626224,
+      "learning_rate": 1.6415231492779942e-06,
+      "loss": 0.6054,
+      "step": 7142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1482120325582414,
+      "learning_rate": 1.640276218168023e-06,
+      "loss": 0.5688,
+      "step": 7143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8700239696691592,
+      "learning_rate": 1.6390296679018909e-06,
+      "loss": 0.6565,
+      "step": 7144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0072980679497032,
+      "learning_rate": 1.637783498620904e-06,
+      "loss": 0.7163,
+      "step": 7145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.744011926537701,
+      "learning_rate": 1.6365377104663206e-06,
+      "loss": 0.5255,
+      "step": 7146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1432552564208316,
+      "learning_rate": 1.63529230357936e-06,
+      "loss": 0.7444,
+      "step": 7147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0570680988223917,
+      "learning_rate": 1.6340472781011935e-06,
+      "loss": 0.6855,
+      "step": 7148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1021390615422724,
+      "learning_rate": 1.6328026341729547e-06,
+      "loss": 0.5668,
+      "step": 7149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1758359200731814,
+      "learning_rate": 1.6315583719357298e-06,
+      "loss": 0.6835,
+      "step": 7150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9660860516559455,
+      "learning_rate": 1.630314491530563e-06,
+      "loss": 0.5572,
+      "step": 7151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0277568894165072,
+      "learning_rate": 1.6290709930984533e-06,
+      "loss": 0.703,
+      "step": 7152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7083389944187983,
+      "learning_rate": 1.6278278767803617e-06,
+      "loss": 0.5905,
+      "step": 7153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7879894519009019,
+      "learning_rate": 1.6265851427171996e-06,
+      "loss": 0.5972,
+      "step": 7154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8278090663267998,
+      "learning_rate": 1.625342791049841e-06,
+      "loss": 0.6401,
+      "step": 7155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.857344031323788,
+      "learning_rate": 1.6241008219191107e-06,
+      "loss": 0.5997,
+      "step": 7156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9243269148131965,
+      "learning_rate": 1.622859235465795e-06,
+      "loss": 0.6302,
+      "step": 7157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9731235103533473,
+      "learning_rate": 1.6216180318306352e-06,
+      "loss": 0.6789,
+      "step": 7158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1769733748993194,
+      "learning_rate": 1.6203772111543247e-06,
+      "loss": 0.5889,
+      "step": 7159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8509294311347555,
+      "learning_rate": 1.6191367735775231e-06,
+      "loss": 0.5998,
+      "step": 7160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8165697626069977,
+      "learning_rate": 1.6178967192408367e-06,
+      "loss": 0.5656,
+      "step": 7161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9585567515408426,
+      "learning_rate": 1.616657048284836e-06,
+      "loss": 0.67,
+      "step": 7162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8907536973431627,
+      "learning_rate": 1.6154177608500415e-06,
+      "loss": 0.6474,
+      "step": 7163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8221553661698418,
+      "learning_rate": 1.6141788570769385e-06,
+      "loss": 0.6146,
+      "step": 7164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8889485316058878,
+      "learning_rate": 1.6129403371059576e-06,
+      "loss": 0.6079,
+      "step": 7165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8222971298134154,
+      "learning_rate": 1.611702201077497e-06,
+      "loss": 0.6486,
+      "step": 7166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0447254833968795,
+      "learning_rate": 1.610464449131902e-06,
+      "loss": 0.6075,
+      "step": 7167
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7665235599646913,
+      "learning_rate": 1.609227081409484e-06,
+      "loss": 0.6228,
+      "step": 7168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.964368754990478,
+      "learning_rate": 1.607990098050501e-06,
+      "loss": 0.6175,
+      "step": 7169
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.029133317798616,
+      "learning_rate": 1.6067534991951754e-06,
+      "loss": 0.6734,
+      "step": 7170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8409734232164388,
+      "learning_rate": 1.6055172849836826e-06,
+      "loss": 0.6309,
+      "step": 7171
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7783753720824762,
+      "learning_rate": 1.6042814555561525e-06,
+      "loss": 0.5715,
+      "step": 7172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.698699527789748,
+      "learning_rate": 1.603046011052673e-06,
+      "loss": 0.6219,
+      "step": 7173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.950831771168677,
+      "learning_rate": 1.6018109516132917e-06,
+      "loss": 0.6084,
+      "step": 7174
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7085437834335682,
+      "learning_rate": 1.600576277378007e-06,
+      "loss": 0.5615,
+      "step": 7175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.803822529222228,
+      "learning_rate": 1.5993419884867783e-06,
+      "loss": 0.5288,
+      "step": 7176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0799801699470155,
+      "learning_rate": 1.5981080850795171e-06,
+      "loss": 0.6573,
+      "step": 7177
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.074780152890361,
+      "learning_rate": 1.5968745672960961e-06,
+      "loss": 0.6823,
+      "step": 7178
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9974810974208348,
+      "learning_rate": 1.59564143527634e-06,
+      "loss": 0.6208,
+      "step": 7179
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.024925483595224,
+      "learning_rate": 1.5944086891600314e-06,
+      "loss": 0.6646,
+      "step": 7180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.182335939230614,
+      "learning_rate": 1.5931763290869073e-06,
+      "loss": 0.6291,
+      "step": 7181
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1038938622619363,
+      "learning_rate": 1.5919443551966662e-06,
+      "loss": 0.5521,
+      "step": 7182
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2416642109686515,
+      "learning_rate": 1.5907127676289564e-06,
+      "loss": 0.6313,
+      "step": 7183
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8541008455898098,
+      "learning_rate": 1.589481566523388e-06,
+      "loss": 0.4875,
+      "step": 7184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9562324429343436,
+      "learning_rate": 1.5882507520195218e-06,
+      "loss": 0.5673,
+      "step": 7185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.064914677261776,
+      "learning_rate": 1.587020324256881e-06,
+      "loss": 0.6345,
+      "step": 7186
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9817317237939902,
+      "learning_rate": 1.5857902833749395e-06,
+      "loss": 0.5916,
+      "step": 7187
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.801570357563867,
+      "learning_rate": 1.5845606295131284e-06,
+      "loss": 0.5671,
+      "step": 7188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.817221361338044,
+      "learning_rate": 1.5833313628108388e-06,
+      "loss": 0.5532,
+      "step": 7189
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.854113231046822,
+      "learning_rate": 1.5821024834074134e-06,
+      "loss": 0.6285,
+      "step": 7190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.165618397080074,
+      "learning_rate": 1.5808739914421512e-06,
+      "loss": 0.5933,
+      "step": 7191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9316848594411447,
+      "learning_rate": 1.5796458870543124e-06,
+      "loss": 0.6431,
+      "step": 7192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0443009173966904,
+      "learning_rate": 1.5784181703831059e-06,
+      "loss": 0.6558,
+      "step": 7193
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0513270749303976,
+      "learning_rate": 1.577190841567704e-06,
+      "loss": 0.6386,
+      "step": 7194
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9096009741492599,
+      "learning_rate": 1.575963900747229e-06,
+      "loss": 0.6275,
+      "step": 7195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0390290313937824,
+      "learning_rate": 1.5747373480607607e-06,
+      "loss": 0.6873,
+      "step": 7196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9505996653243933,
+      "learning_rate": 1.5735111836473393e-06,
+      "loss": 0.6114,
+      "step": 7197
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.010251955965999,
+      "learning_rate": 1.5722854076459538e-06,
+      "loss": 0.6547,
+      "step": 7198
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9576954525895818,
+      "learning_rate": 1.5710600201955567e-06,
+      "loss": 0.6175,
+      "step": 7199
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.971259045738693,
+      "learning_rate": 1.5698350214350483e-06,
+      "loss": 0.6752,
+      "step": 7200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9285497231593764,
+      "learning_rate": 1.5686104115032952e-06,
+      "loss": 0.5635,
+      "step": 7201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.070817821852628,
+      "learning_rate": 1.567386190539107e-06,
+      "loss": 0.6503,
+      "step": 7202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8660756222471904,
+      "learning_rate": 1.5661623586812607e-06,
+      "loss": 0.548,
+      "step": 7203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.941836274898934,
+      "learning_rate": 1.5649389160684813e-06,
+      "loss": 0.526,
+      "step": 7204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.032513766314926,
+      "learning_rate": 1.5637158628394572e-06,
+      "loss": 0.5919,
+      "step": 7205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0981992425710727,
+      "learning_rate": 1.5624931991328246e-06,
+      "loss": 0.6707,
+      "step": 7206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0306085746889875,
+      "learning_rate": 1.5612709250871822e-06,
+      "loss": 0.6382,
+      "step": 7207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9767774040239565,
+      "learning_rate": 1.5600490408410807e-06,
+      "loss": 0.541,
+      "step": 7208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.89306377323277,
+      "learning_rate": 1.5588275465330277e-06,
+      "loss": 0.6256,
+      "step": 7209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.961898068094439,
+      "learning_rate": 1.5576064423014846e-06,
+      "loss": 0.6037,
+      "step": 7210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8167320785570311,
+      "learning_rate": 1.5563857282848738e-06,
+      "loss": 0.6319,
+      "step": 7211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0044052618558608,
+      "learning_rate": 1.555165404621567e-06,
+      "loss": 0.6699,
+      "step": 7212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8192002546413146,
+      "learning_rate": 1.5539454714498985e-06,
+      "loss": 0.5856,
+      "step": 7213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.828513720311586,
+      "learning_rate": 1.5527259289081508e-06,
+      "loss": 0.6237,
+      "step": 7214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.962328010297825,
+      "learning_rate": 1.5515067771345694e-06,
+      "loss": 0.7629,
+      "step": 7215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8235786907604312,
+      "learning_rate": 1.5502880162673506e-06,
+      "loss": 0.6629,
+      "step": 7216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1373109789266316,
+      "learning_rate": 1.5490696464446475e-06,
+      "loss": 0.6834,
+      "step": 7217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.798807028243252,
+      "learning_rate": 1.5478516678045686e-06,
+      "loss": 0.6461,
+      "step": 7218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.073920849769466,
+      "learning_rate": 1.546634080485181e-06,
+      "loss": 0.584,
+      "step": 7219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.143982758312892,
+      "learning_rate": 1.545416884624502e-06,
+      "loss": 0.6263,
+      "step": 7220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.868656609798202,
+      "learning_rate": 1.5442000803605117e-06,
+      "loss": 0.6062,
+      "step": 7221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.102607752104285,
+      "learning_rate": 1.5429836678311382e-06,
+      "loss": 0.6158,
+      "step": 7222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015804353429865,
+      "learning_rate": 1.5417676471742716e-06,
+      "loss": 0.6549,
+      "step": 7223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6390174876555401,
+      "learning_rate": 1.5405520185277533e-06,
+      "loss": 0.553,
+      "step": 7224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0043174338965635,
+      "learning_rate": 1.5393367820293809e-06,
+      "loss": 0.5936,
+      "step": 7225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1450822606543487,
+      "learning_rate": 1.5381219378169103e-06,
+      "loss": 0.6166,
+      "step": 7226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9942427306633814,
+      "learning_rate": 1.5369074860280509e-06,
+      "loss": 0.5527,
+      "step": 7227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9786504142555035,
+      "learning_rate": 1.5356934268004648e-06,
+      "loss": 0.5666,
+      "step": 7228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6346339842485804,
+      "learning_rate": 1.534479760271776e-06,
+      "loss": 0.4102,
+      "step": 7229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1155433193814477,
+      "learning_rate": 1.5332664865795594e-06,
+      "loss": 0.6233,
+      "step": 7230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0461764504626054,
+      "learning_rate": 1.532053605861345e-06,
+      "loss": 0.659,
+      "step": 7231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9075305136903682,
+      "learning_rate": 1.5308411182546224e-06,
+      "loss": 0.6123,
+      "step": 7232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.898062658797633,
+      "learning_rate": 1.5296290238968303e-06,
+      "loss": 0.6636,
+      "step": 7233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.954671511226436,
+      "learning_rate": 1.5284173229253712e-06,
+      "loss": 0.5265,
+      "step": 7234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2142740269246506,
+      "learning_rate": 1.527206015477594e-06,
+      "loss": 0.5625,
+      "step": 7235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1394005619281624,
+      "learning_rate": 1.5259951016908108e-06,
+      "loss": 0.7228,
+      "step": 7236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1596350480306192,
+      "learning_rate": 1.5247845817022827e-06,
+      "loss": 0.6213,
+      "step": 7237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9822202370478568,
+      "learning_rate": 1.5235744556492337e-06,
+      "loss": 0.577,
+      "step": 7238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.849473911231631,
+      "learning_rate": 1.5223647236688317e-06,
+      "loss": 0.5713,
+      "step": 7239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1778919669885215,
+      "learning_rate": 1.5211553858982115e-06,
+      "loss": 0.636,
+      "step": 7240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9120175964258541,
+      "learning_rate": 1.5199464424744553e-06,
+      "loss": 0.6798,
+      "step": 7241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2539495939251353,
+      "learning_rate": 1.5187378935346075e-06,
+      "loss": 0.4582,
+      "step": 7242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0770941489672294,
+      "learning_rate": 1.5175297392156602e-06,
+      "loss": 0.6639,
+      "step": 7243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6863575683493603,
+      "learning_rate": 1.516321979654568e-06,
+      "loss": 0.6796,
+      "step": 7244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.05137600474315,
+      "learning_rate": 1.5151146149882356e-06,
+      "loss": 0.6001,
+      "step": 7245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.04734677979564,
+      "learning_rate": 1.513907645353525e-06,
+      "loss": 0.7466,
+      "step": 7246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7640317190712616,
+      "learning_rate": 1.5127010708872513e-06,
+      "loss": 0.5521,
+      "step": 7247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9082985743121863,
+      "learning_rate": 1.5114948917261896e-06,
+      "loss": 0.6531,
+      "step": 7248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.068888380150388,
+      "learning_rate": 1.510289108007064e-06,
+      "loss": 0.573,
+      "step": 7249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1610981787210704,
+      "learning_rate": 1.5090837198665602e-06,
+      "loss": 0.7273,
+      "step": 7250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8145500032819155,
+      "learning_rate": 1.507878727441313e-06,
+      "loss": 0.5476,
+      "step": 7251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1836728973091852,
+      "learning_rate": 1.5066741308679183e-06,
+      "loss": 0.7212,
+      "step": 7252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.026098366272406,
+      "learning_rate": 1.5054699302829217e-06,
+      "loss": 0.6131,
+      "step": 7253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.958268075225633,
+      "learning_rate": 1.5042661258228268e-06,
+      "loss": 0.7548,
+      "step": 7254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2799034428142844,
+      "learning_rate": 1.5030627176240903e-06,
+      "loss": 0.6088,
+      "step": 7255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8507112671260586,
+      "learning_rate": 1.5018597058231276e-06,
+      "loss": 0.6433,
+      "step": 7256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9787360681863888,
+      "learning_rate": 1.500657090556305e-06,
+      "loss": 0.6084,
+      "step": 7257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0335913118093716,
+      "learning_rate": 1.4994548719599478e-06,
+      "loss": 0.6161,
+      "step": 7258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.913473741873713,
+      "learning_rate": 1.4982530501703325e-06,
+      "loss": 0.599,
+      "step": 7259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.993400563984133,
+      "learning_rate": 1.4970516253236938e-06,
+      "loss": 0.6776,
+      "step": 7260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8738338404676722,
+      "learning_rate": 1.4958505975562205e-06,
+      "loss": 0.6608,
+      "step": 7261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9624680385206787,
+      "learning_rate": 1.4946499670040526e-06,
+      "loss": 0.646,
+      "step": 7262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1031705360635313,
+      "learning_rate": 1.4934497338032926e-06,
+      "loss": 0.6785,
+      "step": 7263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7375430132673397,
+      "learning_rate": 1.4922498980899907e-06,
+      "loss": 0.5263,
+      "step": 7264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.942208339687299,
+      "learning_rate": 1.4910504600001574e-06,
+      "loss": 0.6087,
+      "step": 7265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7733528282248332,
+      "learning_rate": 1.489851419669755e-06,
+      "loss": 0.6096,
+      "step": 7266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9702015428371882,
+      "learning_rate": 1.4886527772347015e-06,
+      "loss": 0.6524,
+      "step": 7267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.821655039369508,
+      "learning_rate": 1.4874545328308681e-06,
+      "loss": 0.6979,
+      "step": 7268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6085812027586697,
+      "learning_rate": 1.486256686594086e-06,
+      "loss": 0.4883,
+      "step": 7269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8103497969052265,
+      "learning_rate": 1.4850592386601342e-06,
+      "loss": 0.6498,
+      "step": 7270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8821065569520434,
+      "learning_rate": 1.4838621891647537e-06,
+      "loss": 0.5811,
+      "step": 7271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.026422332113513,
+      "learning_rate": 1.482665538243634e-06,
+      "loss": 0.5675,
+      "step": 7272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.021526459853876,
+      "learning_rate": 1.4814692860324254e-06,
+      "loss": 0.6165,
+      "step": 7273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2131193609875255,
+      "learning_rate": 1.4802734326667261e-06,
+      "loss": 0.6052,
+      "step": 7274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9863781617996716,
+      "learning_rate": 1.4790779782820991e-06,
+      "loss": 0.7431,
+      "step": 7275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.899030267652968,
+      "learning_rate": 1.4778829230140479e-06,
+      "loss": 0.5902,
+      "step": 7276
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.104359170256935,
+      "learning_rate": 1.4766882669980443e-06,
+      "loss": 0.5782,
+      "step": 7277
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3395206730302807,
+      "learning_rate": 1.4754940103695065e-06,
+      "loss": 0.6701,
+      "step": 7278
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9128557347137265,
+      "learning_rate": 1.4743001532638135e-06,
+      "loss": 0.5963,
+      "step": 7279
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.839960847147049,
+      "learning_rate": 1.473106695816292e-06,
+      "loss": 0.5946,
+      "step": 7280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8804694955721557,
+      "learning_rate": 1.4719136381622307e-06,
+      "loss": 0.5972,
+      "step": 7281
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1064096197759916,
+      "learning_rate": 1.4707209804368683e-06,
+      "loss": 0.6805,
+      "step": 7282
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1508059506711388,
+      "learning_rate": 1.4695287227753984e-06,
+      "loss": 0.6387,
+      "step": 7283
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.934117540641856,
+      "learning_rate": 1.4683368653129698e-06,
+      "loss": 0.6755,
+      "step": 7284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.874609864466516,
+      "learning_rate": 1.4671454081846886e-06,
+      "loss": 0.7757,
+      "step": 7285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8727817163935145,
+      "learning_rate": 1.4659543515256103e-06,
+      "loss": 0.7327,
+      "step": 7286
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9625655507945419,
+      "learning_rate": 1.4647636954707517e-06,
+      "loss": 0.5665,
+      "step": 7287
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9176017722412064,
+      "learning_rate": 1.4635734401550761e-06,
+      "loss": 0.5784,
+      "step": 7288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9382591162652874,
+      "learning_rate": 1.4623835857135099e-06,
+      "loss": 0.6005,
+      "step": 7289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0286029348295123,
+      "learning_rate": 1.4611941322809282e-06,
+      "loss": 0.6977,
+      "step": 7290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9278560441634012,
+      "learning_rate": 1.4600050799921622e-06,
+      "loss": 0.6472,
+      "step": 7291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9786842736043846,
+      "learning_rate": 1.4588164289819956e-06,
+      "loss": 0.7257,
+      "step": 7292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.919764223108238,
+      "learning_rate": 1.4576281793851726e-06,
+      "loss": 0.5598,
+      "step": 7293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0586232862552425,
+      "learning_rate": 1.456440331336385e-06,
+      "loss": 0.5034,
+      "step": 7294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8835266264519421,
+      "learning_rate": 1.4552528849702852e-06,
+      "loss": 0.6227,
+      "step": 7295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9365442859333502,
+      "learning_rate": 1.454065840421473e-06,
+      "loss": 0.6033,
+      "step": 7296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8303763066257501,
+      "learning_rate": 1.452879197824511e-06,
+      "loss": 0.6193,
+      "step": 7297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.714017227116502,
+      "learning_rate": 1.45169295731391e-06,
+      "loss": 0.5045,
+      "step": 7298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9553749197730281,
+      "learning_rate": 1.450507119024135e-06,
+      "loss": 0.5116,
+      "step": 7299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0489024004785206,
+      "learning_rate": 1.4493216830896112e-06,
+      "loss": 0.5511,
+      "step": 7300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9177337191952164,
+      "learning_rate": 1.4481366496447113e-06,
+      "loss": 0.6671,
+      "step": 7301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.242369095193321,
+      "learning_rate": 1.4469520188237684e-06,
+      "loss": 0.6147,
+      "step": 7302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9271683681568583,
+      "learning_rate": 1.4457677907610646e-06,
+      "loss": 0.6747,
+      "step": 7303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9556100302069954,
+      "learning_rate": 1.4445839655908432e-06,
+      "loss": 0.6199,
+      "step": 7304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0446941771338847,
+      "learning_rate": 1.4434005434472914e-06,
+      "loss": 0.7253,
+      "step": 7305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.180261152086753,
+      "learning_rate": 1.4422175244645613e-06,
+      "loss": 0.7472,
+      "step": 7306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9436496640175285,
+      "learning_rate": 1.4410349087767521e-06,
+      "loss": 0.681,
+      "step": 7307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.957008494417684,
+      "learning_rate": 1.4398526965179233e-06,
+      "loss": 0.6194,
+      "step": 7308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9913658996313202,
+      "learning_rate": 1.438670887822081e-06,
+      "loss": 0.5928,
+      "step": 7309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9922918850877875,
+      "learning_rate": 1.437489482823195e-06,
+      "loss": 0.7113,
+      "step": 7310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9105299964949578,
+      "learning_rate": 1.4363084816551798e-06,
+      "loss": 0.5878,
+      "step": 7311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8437267826289236,
+      "learning_rate": 1.4351278844519134e-06,
+      "loss": 0.6025,
+      "step": 7312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2123053516034568,
+      "learning_rate": 1.4339476913472177e-06,
+      "loss": 0.6034,
+      "step": 7313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7734591969685631,
+      "learning_rate": 1.4327679024748785e-06,
+      "loss": 0.521,
+      "step": 7314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0931515989249583,
+      "learning_rate": 1.4315885179686285e-06,
+      "loss": 0.6181,
+      "step": 7315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8904524239638931,
+      "learning_rate": 1.4304095379621607e-06,
+      "loss": 0.5739,
+      "step": 7316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8580069179145808,
+      "learning_rate": 1.4292309625891166e-06,
+      "loss": 0.4972,
+      "step": 7317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8960742519739238,
+      "learning_rate": 1.4280527919830966e-06,
+      "loss": 0.5732,
+      "step": 7318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8541559313701252,
+      "learning_rate": 1.4268750262776526e-06,
+      "loss": 0.6536,
+      "step": 7319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.079477738015932,
+      "learning_rate": 1.42569766560629e-06,
+      "loss": 0.6436,
+      "step": 7320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8833701025509348,
+      "learning_rate": 1.4245207101024684e-06,
+      "loss": 0.6474,
+      "step": 7321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0036752845734265,
+      "learning_rate": 1.4233441598996055e-06,
+      "loss": 0.6448,
+      "step": 7322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9424022901674687,
+      "learning_rate": 1.4221680151310667e-06,
+      "loss": 0.5813,
+      "step": 7323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.776323088501909,
+      "learning_rate": 1.420992275930178e-06,
+      "loss": 0.6075,
+      "step": 7324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8971379539053441,
+      "learning_rate": 1.4198169424302133e-06,
+      "loss": 0.5884,
+      "step": 7325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9758928262230633,
+      "learning_rate": 1.4186420147644053e-06,
+      "loss": 0.6538,
+      "step": 7326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0378828824189505,
+      "learning_rate": 1.4174674930659389e-06,
+      "loss": 0.686,
+      "step": 7327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.100855386114118,
+      "learning_rate": 1.4162933774679494e-06,
+      "loss": 0.5805,
+      "step": 7328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8241299180054797,
+      "learning_rate": 1.4151196681035339e-06,
+      "loss": 0.5597,
+      "step": 7329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3646475695480302,
+      "learning_rate": 1.4139463651057377e-06,
+      "loss": 0.696,
+      "step": 7330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7871636366438264,
+      "learning_rate": 1.4127734686075589e-06,
+      "loss": 0.5538,
+      "step": 7331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7624748734883664,
+      "learning_rate": 1.4116009787419555e-06,
+      "loss": 0.5067,
+      "step": 7332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8830917235961375,
+      "learning_rate": 1.4104288956418326e-06,
+      "loss": 0.6711,
+      "step": 7333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1333917259154473,
+      "learning_rate": 1.4092572194400556e-06,
+      "loss": 0.6778,
+      "step": 7334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8922106726635324,
+      "learning_rate": 1.4080859502694399e-06,
+      "loss": 0.5897,
+      "step": 7335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7979650669786016,
+      "learning_rate": 1.406915088262753e-06,
+      "loss": 0.6342,
+      "step": 7336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8262225902179414,
+      "learning_rate": 1.4057446335527224e-06,
+      "loss": 0.5744,
+      "step": 7337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.939920334697821,
+      "learning_rate": 1.4045745862720227e-06,
+      "loss": 0.6456,
+      "step": 7338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046921032309175,
+      "learning_rate": 1.4034049465532884e-06,
+      "loss": 0.6499,
+      "step": 7339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0279532671601435,
+      "learning_rate": 1.4022357145291022e-06,
+      "loss": 0.6686,
+      "step": 7340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.876032959240246,
+      "learning_rate": 1.4010668903320068e-06,
+      "loss": 0.65,
+      "step": 7341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.010683002947994,
+      "learning_rate": 1.3998984740944898e-06,
+      "loss": 0.6212,
+      "step": 7342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9942519888855668,
+      "learning_rate": 1.3987304659490019e-06,
+      "loss": 0.6769,
+      "step": 7343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3241892030749804,
+      "learning_rate": 1.397562866027941e-06,
+      "loss": 0.7233,
+      "step": 7344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8842048203883948,
+      "learning_rate": 1.3963956744636642e-06,
+      "loss": 0.483,
+      "step": 7345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8132127750394096,
+      "learning_rate": 1.3952288913884754e-06,
+      "loss": 0.6027,
+      "step": 7346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9408602426373183,
+      "learning_rate": 1.3940625169346406e-06,
+      "loss": 0.6706,
+      "step": 7347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9509705460858806,
+      "learning_rate": 1.3928965512343705e-06,
+      "loss": 0.512,
+      "step": 7348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.846187361487381,
+      "learning_rate": 1.3917309944198392e-06,
+      "loss": 0.5747,
+      "step": 7349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8360446999359556,
+      "learning_rate": 1.390565846623163e-06,
+      "loss": 0.5452,
+      "step": 7350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.207526838462662,
+      "learning_rate": 1.389401107976423e-06,
+      "loss": 0.6348,
+      "step": 7351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6972499850348746,
+      "learning_rate": 1.3882367786116458e-06,
+      "loss": 0.6734,
+      "step": 7352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9465037911461067,
+      "learning_rate": 1.3870728586608172e-06,
+      "loss": 0.638,
+      "step": 7353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.99105946713035,
+      "learning_rate": 1.3859093482558717e-06,
+      "loss": 0.4869,
+      "step": 7354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9083214942937838,
+      "learning_rate": 1.3847462475287027e-06,
+      "loss": 0.6243,
+      "step": 7355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8052126687935333,
+      "learning_rate": 1.3835835566111527e-06,
+      "loss": 0.5666,
+      "step": 7356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8247580687949765,
+      "learning_rate": 1.3824212756350196e-06,
+      "loss": 0.5969,
+      "step": 7357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.931041342905744,
+      "learning_rate": 1.3812594047320526e-06,
+      "loss": 0.6333,
+      "step": 7358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.003570523038699,
+      "learning_rate": 1.3800979440339602e-06,
+      "loss": 0.6259,
+      "step": 7359
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9910752759176764,
+      "learning_rate": 1.3789368936723967e-06,
+      "loss": 0.596,
+      "step": 7360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.204108423906913,
+      "learning_rate": 1.3777762537789774e-06,
+      "loss": 0.6022,
+      "step": 7361
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7905234257212006,
+      "learning_rate": 1.3766160244852645e-06,
+      "loss": 0.666,
+      "step": 7362
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.785127840395526,
+      "learning_rate": 1.375456205922779e-06,
+      "loss": 0.6004,
+      "step": 7363
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8382116223905691,
+      "learning_rate": 1.3742967982229915e-06,
+      "loss": 0.6217,
+      "step": 7364
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8116921007155165,
+      "learning_rate": 1.373137801517327e-06,
+      "loss": 0.5808,
+      "step": 7365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0143148074817656,
+      "learning_rate": 1.371979215937166e-06,
+      "loss": 0.5981,
+      "step": 7366
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.945327577763545,
+      "learning_rate": 1.3708210416138395e-06,
+      "loss": 0.7188,
+      "step": 7367
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9918025752529571,
+      "learning_rate": 1.3696632786786328e-06,
+      "loss": 0.6553,
+      "step": 7368
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8128557697830252,
+      "learning_rate": 1.368505927262787e-06,
+      "loss": 0.54,
+      "step": 7369
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8567139003564144,
+      "learning_rate": 1.3673489874974916e-06,
+      "loss": 0.6109,
+      "step": 7370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8570547914301103,
+      "learning_rate": 1.3661924595138953e-06,
+      "loss": 0.5828,
+      "step": 7371
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0112052275541963,
+      "learning_rate": 1.3650363434430957e-06,
+      "loss": 0.643,
+      "step": 7372
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9738896384606492,
+      "learning_rate": 1.363880639416144e-06,
+      "loss": 0.7166,
+      "step": 7373
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9657301144532935,
+      "learning_rate": 1.3627253475640484e-06,
+      "loss": 0.5739,
+      "step": 7374
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8189747674386199,
+      "learning_rate": 1.3615704680177649e-06,
+      "loss": 0.5863,
+      "step": 7375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9980086094799228,
+      "learning_rate": 1.3604160009082084e-06,
+      "loss": 0.6279,
+      "step": 7376
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8661162265592808,
+      "learning_rate": 1.359261946366242e-06,
+      "loss": 0.6268,
+      "step": 7377
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0760584257748347,
+      "learning_rate": 1.3581083045226884e-06,
+      "loss": 0.5509,
+      "step": 7378
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.890930457422948,
+      "learning_rate": 1.3569550755083139e-06,
+      "loss": 0.4827,
+      "step": 7379
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8250307809861648,
+      "learning_rate": 1.3558022594538473e-06,
+      "loss": 0.6133,
+      "step": 7380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9776684236653719,
+      "learning_rate": 1.3546498564899647e-06,
+      "loss": 0.65,
+      "step": 7381
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8375601370048764,
+      "learning_rate": 1.3534978667472998e-06,
+      "loss": 0.6446,
+      "step": 7382
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.882348930178244,
+      "learning_rate": 1.3523462903564344e-06,
+      "loss": 0.6684,
+      "step": 7383
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.009995487209896,
+      "learning_rate": 1.3511951274479096e-06,
+      "loss": 0.7314,
+      "step": 7384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9752306252999285,
+      "learning_rate": 1.3500443781522131e-06,
+      "loss": 0.5498,
+      "step": 7385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.989631043234858,
+      "learning_rate": 1.3488940425997937e-06,
+      "loss": 0.6369,
+      "step": 7386
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0437233289117995,
+      "learning_rate": 1.3477441209210418e-06,
+      "loss": 0.6656,
+      "step": 7387
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8972633241525336,
+      "learning_rate": 1.3465946132463125e-06,
+      "loss": 0.6628,
+      "step": 7388
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8438369722578372,
+      "learning_rate": 1.3454455197059064e-06,
+      "loss": 0.5667,
+      "step": 7389
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0959707834954493,
+      "learning_rate": 1.3442968404300822e-06,
+      "loss": 0.5153,
+      "step": 7390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6795735736077286,
+      "learning_rate": 1.3431485755490464e-06,
+      "loss": 0.5253,
+      "step": 7391
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8489557099642173,
+      "learning_rate": 1.342000725192964e-06,
+      "loss": 0.6495,
+      "step": 7392
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0025930868134916,
+      "learning_rate": 1.3408532894919502e-06,
+      "loss": 0.6117,
+      "step": 7393
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9342212627586495,
+      "learning_rate": 1.3397062685760715e-06,
+      "loss": 0.6195,
+      "step": 7394
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3274325479107087,
+      "learning_rate": 1.3385596625753494e-06,
+      "loss": 0.774,
+      "step": 7395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.007477669025387,
+      "learning_rate": 1.3374134716197602e-06,
+      "loss": 0.5757,
+      "step": 7396
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1464769668013823,
+      "learning_rate": 1.336267695839229e-06,
+      "loss": 0.6546,
+      "step": 7397
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.063072317843152,
+      "learning_rate": 1.3351223353636378e-06,
+      "loss": 0.6652,
+      "step": 7398
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.23065638452988,
+      "learning_rate": 1.3339773903228182e-06,
+      "loss": 0.6505,
+      "step": 7399
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0593425716320235,
+      "learning_rate": 1.3328328608465586e-06,
+      "loss": 0.6715,
+      "step": 7400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8612256997436372,
+      "learning_rate": 1.3316887470645956e-06,
+      "loss": 0.5658,
+      "step": 7401
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.092289554514257,
+      "learning_rate": 1.3305450491066207e-06,
+      "loss": 0.5539,
+      "step": 7402
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1098319168701223,
+      "learning_rate": 1.3294017671022812e-06,
+      "loss": 0.6631,
+      "step": 7403
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.055282905776299,
+      "learning_rate": 1.328258901181172e-06,
+      "loss": 0.638,
+      "step": 7404
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8517537301996196,
+      "learning_rate": 1.3271164514728458e-06,
+      "loss": 0.617,
+      "step": 7405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9401490645931452,
+      "learning_rate": 1.3259744181068041e-06,
+      "loss": 0.6754,
+      "step": 7406
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0376313058684565,
+      "learning_rate": 1.3248328012125022e-06,
+      "loss": 0.5623,
+      "step": 7407
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0528085589228446,
+      "learning_rate": 1.3236916009193517e-06,
+      "loss": 0.6334,
+      "step": 7408
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9909521335765008,
+      "learning_rate": 1.3225508173567125e-06,
+      "loss": 0.6425,
+      "step": 7409
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.075921732696546,
+      "learning_rate": 1.3214104506538971e-06,
+      "loss": 0.5893,
+      "step": 7410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.906763513338803,
+      "learning_rate": 1.320270500940176e-06,
+      "loss": 0.6562,
+      "step": 7411
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9042795537620933,
+      "learning_rate": 1.3191309683447662e-06,
+      "loss": 0.5949,
+      "step": 7412
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.876715715793952,
+      "learning_rate": 1.3179918529968422e-06,
+      "loss": 0.6021,
+      "step": 7413
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8171198061068894,
+      "learning_rate": 1.3168531550255275e-06,
+      "loss": 0.5601,
+      "step": 7414
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8263304397349878,
+      "learning_rate": 1.3157148745599035e-06,
+      "loss": 0.6062,
+      "step": 7415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0517363085919533,
+      "learning_rate": 1.3145770117289957e-06,
+      "loss": 0.6128,
+      "step": 7416
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9454726843452275,
+      "learning_rate": 1.313439566661791e-06,
+      "loss": 0.5095,
+      "step": 7417
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9864994234352504,
+      "learning_rate": 1.3123025394872224e-06,
+      "loss": 0.567,
+      "step": 7418
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0924159820519774,
+      "learning_rate": 1.3111659303341824e-06,
+      "loss": 0.627,
+      "step": 7419
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0441005429459826,
+      "learning_rate": 1.3100297393315077e-06,
+      "loss": 0.6127,
+      "step": 7420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.011198133677825,
+      "learning_rate": 1.3088939666079958e-06,
+      "loss": 0.6564,
+      "step": 7421
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.877266216814998,
+      "learning_rate": 1.3077586122923896e-06,
+      "loss": 0.565,
+      "step": 7422
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1366794824633017,
+      "learning_rate": 1.3066236765133933e-06,
+      "loss": 0.5936,
+      "step": 7423
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9038044812800596,
+      "learning_rate": 1.3054891593996515e-06,
+      "loss": 0.6186,
+      "step": 7424
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1095772614580666,
+      "learning_rate": 1.3043550610797728e-06,
+      "loss": 0.6388,
+      "step": 7425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9489744092099393,
+      "learning_rate": 1.3032213816823113e-06,
+      "loss": 0.5858,
+      "step": 7426
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0758525689119067,
+      "learning_rate": 1.3020881213357783e-06,
+      "loss": 0.6035,
+      "step": 7427
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0709398095894356,
+      "learning_rate": 1.3009552801686331e-06,
+      "loss": 0.554,
+      "step": 7428
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.866922642817398,
+      "learning_rate": 1.299822858309292e-06,
+      "loss": 0.5272,
+      "step": 7429
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8806440265200155,
+      "learning_rate": 1.29869085588612e-06,
+      "loss": 0.5618,
+      "step": 7430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.998869396446172,
+      "learning_rate": 1.2975592730274367e-06,
+      "loss": 0.6345,
+      "step": 7431
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8384656204172125,
+      "learning_rate": 1.296428109861511e-06,
+      "loss": 0.6149,
+      "step": 7432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.233235086672102,
+      "learning_rate": 1.2952973665165703e-06,
+      "loss": 0.6092,
+      "step": 7433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6047023375364677,
+      "learning_rate": 1.2941670431207882e-06,
+      "loss": 0.6329,
+      "step": 7434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0698950270152126,
+      "learning_rate": 1.293037139802295e-06,
+      "loss": 0.655,
+      "step": 7435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.403007806832603,
+      "learning_rate": 1.2919076566891703e-06,
+      "loss": 0.6415,
+      "step": 7436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1414097281187785,
+      "learning_rate": 1.290778593909449e-06,
+      "loss": 0.6067,
+      "step": 7437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2802961243526507,
+      "learning_rate": 1.2896499515911165e-06,
+      "loss": 0.7176,
+      "step": 7438
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8863817546743002,
+      "learning_rate": 1.2885217298621084e-06,
+      "loss": 0.6193,
+      "step": 7439
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.044165537334881,
+      "learning_rate": 1.2873939288503185e-06,
+      "loss": 0.682,
+      "step": 7440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.115113504513422,
+      "learning_rate": 1.2862665486835861e-06,
+      "loss": 0.6313,
+      "step": 7441
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8510730137768605,
+      "learning_rate": 1.2851395894897101e-06,
+      "loss": 0.676,
+      "step": 7442
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0138120838329927,
+      "learning_rate": 1.2840130513964338e-06,
+      "loss": 0.5752,
+      "step": 7443
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8887083592061393,
+      "learning_rate": 1.2828869345314599e-06,
+      "loss": 0.5755,
+      "step": 7444
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0255619098514455,
+      "learning_rate": 1.2817612390224388e-06,
+      "loss": 0.6166,
+      "step": 7445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9683394171262643,
+      "learning_rate": 1.2806359649969746e-06,
+      "loss": 0.5901,
+      "step": 7446
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1220195600084066,
+      "learning_rate": 1.2795111125826221e-06,
+      "loss": 0.6434,
+      "step": 7447
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8294003393097111,
+      "learning_rate": 1.2783866819068923e-06,
+      "loss": 0.6808,
+      "step": 7448
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8189125265698487,
+      "learning_rate": 1.2772626730972437e-06,
+      "loss": 0.4841,
+      "step": 7449
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6978368469146112,
+      "learning_rate": 1.2761390862810907e-06,
+      "loss": 0.6485,
+      "step": 7450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2209030633430737,
+      "learning_rate": 1.2750159215857965e-06,
+      "loss": 0.6692,
+      "step": 7451
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.929125071978997,
+      "learning_rate": 1.2738931791386827e-06,
+      "loss": 0.6442,
+      "step": 7452
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.799506153570008,
+      "learning_rate": 1.2727708590670113e-06,
+      "loss": 0.5547,
+      "step": 7453
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.13151128803415,
+      "learning_rate": 1.2716489614980093e-06,
+      "loss": 0.6322,
+      "step": 7454
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.134629474837925,
+      "learning_rate": 1.2705274865588475e-06,
+      "loss": 0.5887,
+      "step": 7455
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.858068512251421,
+      "learning_rate": 1.2694064343766532e-06,
+      "loss": 0.5537,
+      "step": 7456
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8715567367821404,
+      "learning_rate": 1.2682858050785018e-06,
+      "loss": 0.6063,
+      "step": 7457
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.198960334446777,
+      "learning_rate": 1.2671655987914261e-06,
+      "loss": 0.5857,
+      "step": 7458
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1149071876653958,
+      "learning_rate": 1.266045815642405e-06,
+      "loss": 0.5134,
+      "step": 7459
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3036410022633413,
+      "learning_rate": 1.2649264557583758e-06,
+      "loss": 0.7023,
+      "step": 7460
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.725475153183577,
+      "learning_rate": 1.2638075192662196e-06,
+      "loss": 0.5848,
+      "step": 7461
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7445337006329626,
+      "learning_rate": 1.2626890062927781e-06,
+      "loss": 0.5705,
+      "step": 7462
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9445544003943256,
+      "learning_rate": 1.2615709169648382e-06,
+      "loss": 0.5653,
+      "step": 7463
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.86634440943093,
+      "learning_rate": 1.2604532514091444e-06,
+      "loss": 0.6699,
+      "step": 7464
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.884500167863542,
+      "learning_rate": 1.2593360097523883e-06,
+      "loss": 0.5513,
+      "step": 7465
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0358534462218953,
+      "learning_rate": 1.2582191921212172e-06,
+      "loss": 0.6638,
+      "step": 7466
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9210922659682166,
+      "learning_rate": 1.257102798642229e-06,
+      "loss": 0.5599,
+      "step": 7467
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0339311660440456,
+      "learning_rate": 1.2559868294419702e-06,
+      "loss": 0.6709,
+      "step": 7468
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9172549578337978,
+      "learning_rate": 1.2548712846469469e-06,
+      "loss": 0.5646,
+      "step": 7469
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9444512462886203,
+      "learning_rate": 1.2537561643836087e-06,
+      "loss": 0.6041,
+      "step": 7470
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0064010781768653,
+      "learning_rate": 1.2526414687783616e-06,
+      "loss": 0.6542,
+      "step": 7471
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9136862206054561,
+      "learning_rate": 1.2515271979575645e-06,
+      "loss": 0.6535,
+      "step": 7472
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.791221271933324,
+      "learning_rate": 1.2504133520475237e-06,
+      "loss": 0.615,
+      "step": 7473
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0331463728088317,
+      "learning_rate": 1.249299931174503e-06,
+      "loss": 0.7014,
+      "step": 7474
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.092796442100189,
+      "learning_rate": 1.248186935464713e-06,
+      "loss": 0.6129,
+      "step": 7475
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0088765568010576,
+      "learning_rate": 1.2470743650443167e-06,
+      "loss": 0.6894,
+      "step": 7476
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1889027363813027,
+      "learning_rate": 1.2459622200394344e-06,
+      "loss": 0.6256,
+      "step": 7477
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9459517215941233,
+      "learning_rate": 1.2448505005761297e-06,
+      "loss": 0.6502,
+      "step": 7478
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.958059668778648,
+      "learning_rate": 1.243739206780426e-06,
+      "loss": 0.5418,
+      "step": 7479
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0181150473837106,
+      "learning_rate": 1.2426283387782916e-06,
+      "loss": 0.686,
+      "step": 7480
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8056526635316388,
+      "learning_rate": 1.2415178966956531e-06,
+      "loss": 0.6975,
+      "step": 7481
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9223777564267408,
+      "learning_rate": 1.2404078806583835e-06,
+      "loss": 0.578,
+      "step": 7482
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.87336105244118,
+      "learning_rate": 1.2392982907923096e-06,
+      "loss": 0.5269,
+      "step": 7483
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8268304700340083,
+      "learning_rate": 1.2381891272232083e-06,
+      "loss": 0.6357,
+      "step": 7484
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8174515571326288,
+      "learning_rate": 1.237080390076812e-06,
+      "loss": 0.5843,
+      "step": 7485
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9327780722397199,
+      "learning_rate": 1.2359720794788006e-06,
+      "loss": 0.6239,
+      "step": 7486
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0991839672424932,
+      "learning_rate": 1.2348641955548096e-06,
+      "loss": 0.6605,
+      "step": 7487
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8536105464415018,
+      "learning_rate": 1.2337567384304206e-06,
+      "loss": 0.6537,
+      "step": 7488
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.024081612397799,
+      "learning_rate": 1.2326497082311756e-06,
+      "loss": 0.6295,
+      "step": 7489
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9662894712769203,
+      "learning_rate": 1.231543105082556e-06,
+      "loss": 0.7392,
+      "step": 7490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0201598746186793,
+      "learning_rate": 1.230436929110007e-06,
+      "loss": 0.6617,
+      "step": 7491
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.935461672788772,
+      "learning_rate": 1.2293311804389162e-06,
+      "loss": 0.6007,
+      "step": 7492
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7867631971998374,
+      "learning_rate": 1.2282258591946294e-06,
+      "loss": 0.5533,
+      "step": 7493
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9480879837331495,
+      "learning_rate": 1.2271209655024386e-06,
+      "loss": 0.5748,
+      "step": 7494
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8358355416673764,
+      "learning_rate": 1.2260164994875922e-06,
+      "loss": 0.5568,
+      "step": 7495
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.916660508469549,
+      "learning_rate": 1.224912461275287e-06,
+      "loss": 0.5668,
+      "step": 7496
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1041973386188726,
+      "learning_rate": 1.2238088509906715e-06,
+      "loss": 0.6088,
+      "step": 7497
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8336319921741646,
+      "learning_rate": 1.2227056687588445e-06,
+      "loss": 0.6178,
+      "step": 7498
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9408959767507967,
+      "learning_rate": 1.221602914704862e-06,
+      "loss": 0.6562,
+      "step": 7499
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.982054454174401,
+      "learning_rate": 1.2205005889537231e-06,
+      "loss": 0.6758,
+      "step": 7500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.242903224347222,
+      "learning_rate": 1.2193986916303862e-06,
+      "loss": 0.6497,
+      "step": 7501
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.696679047260719,
+      "learning_rate": 1.2182972228597555e-06,
+      "loss": 0.6423,
+      "step": 7502
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4159106630224647,
+      "learning_rate": 1.2171961827666907e-06,
+      "loss": 0.6576,
+      "step": 7503
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.15668495278231,
+      "learning_rate": 1.2160955714759997e-06,
+      "loss": 0.5918,
+      "step": 7504
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.082528155431938,
+      "learning_rate": 1.2149953891124423e-06,
+      "loss": 0.6091,
+      "step": 7505
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9378865879606009,
+      "learning_rate": 1.2138956358007325e-06,
+      "loss": 0.6194,
+      "step": 7506
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9807419537639468,
+      "learning_rate": 1.2127963116655323e-06,
+      "loss": 0.563,
+      "step": 7507
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9001140039929365,
+      "learning_rate": 1.2116974168314549e-06,
+      "loss": 0.5075,
+      "step": 7508
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9044446353892586,
+      "learning_rate": 1.2105989514230699e-06,
+      "loss": 0.5632,
+      "step": 7509
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9036438238927869,
+      "learning_rate": 1.2095009155648908e-06,
+      "loss": 0.6259,
+      "step": 7510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0289114282767042,
+      "learning_rate": 1.2084033093813897e-06,
+      "loss": 0.6061,
+      "step": 7511
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9769797796713287,
+      "learning_rate": 1.2073061329969843e-06,
+      "loss": 0.5597,
+      "step": 7512
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.829432862973179,
+      "learning_rate": 1.2062093865360458e-06,
+      "loss": 0.5307,
+      "step": 7513
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9966239074516121,
+      "learning_rate": 1.205113070122898e-06,
+      "loss": 0.6177,
+      "step": 7514
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0151026946951824,
+      "learning_rate": 1.2040171838818128e-06,
+      "loss": 0.5806,
+      "step": 7515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9245838703900684,
+      "learning_rate": 1.202921727937017e-06,
+      "loss": 0.5436,
+      "step": 7516
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3269530223457475,
+      "learning_rate": 1.201826702412685e-06,
+      "loss": 0.6014,
+      "step": 7517
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0196480903460583,
+      "learning_rate": 1.2007321074329464e-06,
+      "loss": 0.6542,
+      "step": 7518
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0366225641189617,
+      "learning_rate": 1.1996379431218792e-06,
+      "loss": 0.5832,
+      "step": 7519
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8691559040156394,
+      "learning_rate": 1.1985442096035116e-06,
+      "loss": 0.5792,
+      "step": 7520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.97785016962828,
+      "learning_rate": 1.1974509070018242e-06,
+      "loss": 0.6102,
+      "step": 7521
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.909391688954166,
+      "learning_rate": 1.1963580354407523e-06,
+      "loss": 0.5965,
+      "step": 7522
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8397548132442958,
+      "learning_rate": 1.195265595044175e-06,
+      "loss": 0.5612,
+      "step": 7523
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8519815774481667,
+      "learning_rate": 1.1941735859359305e-06,
+      "loss": 0.6289,
+      "step": 7524
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9325455805100002,
+      "learning_rate": 1.193082008239801e-06,
+      "loss": 0.7232,
+      "step": 7525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7899558628801764,
+      "learning_rate": 1.1919908620795274e-06,
+      "loss": 0.6425,
+      "step": 7526
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7674583673545878,
+      "learning_rate": 1.1909001475787917e-06,
+      "loss": 0.6329,
+      "step": 7527
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9014708835524068,
+      "learning_rate": 1.189809864861237e-06,
+      "loss": 0.5598,
+      "step": 7528
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9850467848833522,
+      "learning_rate": 1.1887200140504496e-06,
+      "loss": 0.6087,
+      "step": 7529
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8719486639364562,
+      "learning_rate": 1.187630595269974e-06,
+      "loss": 0.6111,
+      "step": 7530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8776002051132321,
+      "learning_rate": 1.186541608643299e-06,
+      "loss": 0.6164,
+      "step": 7531
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.10496286194707,
+      "learning_rate": 1.1854530542938697e-06,
+      "loss": 0.6653,
+      "step": 7532
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.792696936223161,
+      "learning_rate": 1.184364932345079e-06,
+      "loss": 0.6264,
+      "step": 7533
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.062502687433275,
+      "learning_rate": 1.1832772429202716e-06,
+      "loss": 0.6956,
+      "step": 7534
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.051811267436275,
+      "learning_rate": 1.1821899861427415e-06,
+      "loss": 0.6218,
+      "step": 7535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8590354079413396,
+      "learning_rate": 1.1811031621357388e-06,
+      "loss": 0.5887,
+      "step": 7536
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8694047326150192,
+      "learning_rate": 1.1800167710224585e-06,
+      "loss": 0.5381,
+      "step": 7537
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0369985151819887,
+      "learning_rate": 1.1789308129260518e-06,
+      "loss": 0.6614,
+      "step": 7538
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.024847817291823,
+      "learning_rate": 1.1778452879696156e-06,
+      "loss": 0.5751,
+      "step": 7539
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0143477964838166,
+      "learning_rate": 1.1767601962762025e-06,
+      "loss": 0.668,
+      "step": 7540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1439637308422053,
+      "learning_rate": 1.1756755379688133e-06,
+      "loss": 0.6613,
+      "step": 7541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.128249539956141,
+      "learning_rate": 1.1745913131703983e-06,
+      "loss": 0.594,
+      "step": 7542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.042668732426268,
+      "learning_rate": 1.1735075220038634e-06,
+      "loss": 0.6774,
+      "step": 7543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7655693783203028,
+      "learning_rate": 1.1724241645920597e-06,
+      "loss": 0.5097,
+      "step": 7544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9980257594538535,
+      "learning_rate": 1.1713412410577947e-06,
+      "loss": 0.5953,
+      "step": 7545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9020372671647598,
+      "learning_rate": 1.1702587515238228e-06,
+      "loss": 0.6078,
+      "step": 7546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.84747773445059,
+      "learning_rate": 1.1691766961128486e-06,
+      "loss": 0.6325,
+      "step": 7547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1840723467663334,
+      "learning_rate": 1.1680950749475328e-06,
+      "loss": 0.4747,
+      "step": 7548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.01935125924819,
+      "learning_rate": 1.1670138881504811e-06,
+      "loss": 0.6862,
+      "step": 7549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.790776100057919,
+      "learning_rate": 1.165933135844251e-06,
+      "loss": 0.6148,
+      "step": 7550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9127701595519246,
+      "learning_rate": 1.1648528181513546e-06,
+      "loss": 0.6695,
+      "step": 7551
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8752750795920698,
+      "learning_rate": 1.1637729351942496e-06,
+      "loss": 0.6208,
+      "step": 7552
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0838979472639543,
+      "learning_rate": 1.16269348709535e-06,
+      "loss": 0.6042,
+      "step": 7553
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.870248031496323,
+      "learning_rate": 1.1616144739770134e-06,
+      "loss": 0.5422,
+      "step": 7554
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.985975392199706,
+      "learning_rate": 1.1605358959615559e-06,
+      "loss": 0.6439,
+      "step": 7555
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.142807668782956,
+      "learning_rate": 1.1594577531712392e-06,
+      "loss": 0.638,
+      "step": 7556
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8963683455946494,
+      "learning_rate": 1.1583800457282763e-06,
+      "loss": 0.641,
+      "step": 7557
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8549487285628472,
+      "learning_rate": 1.1573027737548304e-06,
+      "loss": 0.5158,
+      "step": 7558
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7469422801211996,
+      "learning_rate": 1.156225937373019e-06,
+      "loss": 0.5104,
+      "step": 7559
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0926111621878314,
+      "learning_rate": 1.1551495367049047e-06,
+      "loss": 0.7087,
+      "step": 7560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.126301992165716,
+      "learning_rate": 1.154073571872507e-06,
+      "loss": 0.6058,
+      "step": 7561
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.012168076568665,
+      "learning_rate": 1.1529980429977899e-06,
+      "loss": 0.6645,
+      "step": 7562
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7738548550605582,
+      "learning_rate": 1.151922950202674e-06,
+      "loss": 0.6511,
+      "step": 7563
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8809818116839208,
+      "learning_rate": 1.1508482936090226e-06,
+      "loss": 0.6599,
+      "step": 7564
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9829487911729462,
+      "learning_rate": 1.149774073338658e-06,
+      "loss": 0.6738,
+      "step": 7565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1298374139158023,
+      "learning_rate": 1.1487002895133458e-06,
+      "loss": 0.6807,
+      "step": 7566
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7396917186181862,
+      "learning_rate": 1.1476269422548097e-06,
+      "loss": 0.5649,
+      "step": 7567
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.163951307025049,
+      "learning_rate": 1.1465540316847158e-06,
+      "loss": 0.6927,
+      "step": 7568
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7608713500076538,
+      "learning_rate": 1.1454815579246874e-06,
+      "loss": 0.5432,
+      "step": 7569
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.701479430283739,
+      "learning_rate": 1.1444095210962946e-06,
+      "loss": 0.5278,
+      "step": 7570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9439676814040012,
+      "learning_rate": 1.1433379213210589e-06,
+      "loss": 0.6163,
+      "step": 7571
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.041557258404316,
+      "learning_rate": 1.14226675872045e-06,
+      "loss": 0.5878,
+      "step": 7572
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8275477847578068,
+      "learning_rate": 1.1411960334158945e-06,
+      "loss": 0.5765,
+      "step": 7573
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9882523933445255,
+      "learning_rate": 1.1401257455287612e-06,
+      "loss": 0.6118,
+      "step": 7574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1243807472913923,
+      "learning_rate": 1.1390558951803765e-06,
+      "loss": 0.5873,
+      "step": 7575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.036793715513729,
+      "learning_rate": 1.1379864824920116e-06,
+      "loss": 0.6901,
+      "step": 7576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0003545707561745,
+      "learning_rate": 1.1369175075848931e-06,
+      "loss": 0.6905,
+      "step": 7577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8163317896843112,
+      "learning_rate": 1.135848970580194e-06,
+      "loss": 0.5174,
+      "step": 7578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9890937817343028,
+      "learning_rate": 1.1347808715990377e-06,
+      "loss": 0.6099,
+      "step": 7579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0858479558113454,
+      "learning_rate": 1.1337132107625015e-06,
+      "loss": 0.5826,
+      "step": 7580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9155881518121252,
+      "learning_rate": 1.1326459881916091e-06,
+      "loss": 0.6876,
+      "step": 7581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9989956883343196,
+      "learning_rate": 1.1315792040073381e-06,
+      "loss": 0.6681,
+      "step": 7582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9550742344176189,
+      "learning_rate": 1.1305128583306125e-06,
+      "loss": 0.6161,
+      "step": 7583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0103579349657963,
+      "learning_rate": 1.1294469512823109e-06,
+      "loss": 0.6299,
+      "step": 7584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9054575542751568,
+      "learning_rate": 1.128381482983259e-06,
+      "loss": 0.6079,
+      "step": 7585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1391755475909,
+      "learning_rate": 1.1273164535542336e-06,
+      "loss": 0.6313,
+      "step": 7586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.981591108859556,
+      "learning_rate": 1.1262518631159602e-06,
+      "loss": 0.5913,
+      "step": 7587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1379326081176777,
+      "learning_rate": 1.125187711789119e-06,
+      "loss": 0.6499,
+      "step": 7588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.982491035473915,
+      "learning_rate": 1.1241239996943348e-06,
+      "loss": 0.6315,
+      "step": 7589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2170410264953224,
+      "learning_rate": 1.1230607269521886e-06,
+      "loss": 0.6647,
+      "step": 7590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1017656060939376,
+      "learning_rate": 1.1219978936832054e-06,
+      "loss": 0.6512,
+      "step": 7591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9103803015382639,
+      "learning_rate": 1.1209355000078664e-06,
+      "loss": 0.5763,
+      "step": 7592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8881668957666067,
+      "learning_rate": 1.1198735460465987e-06,
+      "loss": 0.5324,
+      "step": 7593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1092691410374727,
+      "learning_rate": 1.1188120319197798e-06,
+      "loss": 0.7296,
+      "step": 7594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7952823369616315,
+      "learning_rate": 1.117750957747738e-06,
+      "loss": 0.5827,
+      "step": 7595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9802795486580884,
+      "learning_rate": 1.1166903236507549e-06,
+      "loss": 0.6069,
+      "step": 7596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9392038562759872,
+      "learning_rate": 1.1156301297490563e-06,
+      "loss": 0.6014,
+      "step": 7597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0103638691186676,
+      "learning_rate": 1.1145703761628234e-06,
+      "loss": 0.6243,
+      "step": 7598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.854549954527706,
+      "learning_rate": 1.1135110630121837e-06,
+      "loss": 0.5502,
+      "step": 7599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.021354607363837,
+      "learning_rate": 1.1124521904172202e-06,
+      "loss": 0.5726,
+      "step": 7600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9661146192099226,
+      "learning_rate": 1.1113937584979561e-06,
+      "loss": 0.6839,
+      "step": 7601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7591677360991955,
+      "learning_rate": 1.1103357673743752e-06,
+      "loss": 0.6131,
+      "step": 7602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9398835866330275,
+      "learning_rate": 1.109278217166404e-06,
+      "loss": 0.6187,
+      "step": 7603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9393774468410334,
+      "learning_rate": 1.1082211079939248e-06,
+      "loss": 0.5964,
+      "step": 7604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.067156283501053,
+      "learning_rate": 1.107164439976764e-06,
+      "loss": 0.7105,
+      "step": 7605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9167409651719103,
+      "learning_rate": 1.106108213234704e-06,
+      "loss": 0.656,
+      "step": 7606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1560390878979914,
+      "learning_rate": 1.105052427887472e-06,
+      "loss": 0.6422,
+      "step": 7607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.118078967090903,
+      "learning_rate": 1.1039970840547464e-06,
+      "loss": 0.6541,
+      "step": 7608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8508371466187719,
+      "learning_rate": 1.1029421818561592e-06,
+      "loss": 0.4905,
+      "step": 7609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9440624024422692,
+      "learning_rate": 1.1018877214112883e-06,
+      "loss": 0.6826,
+      "step": 7610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.849950642377605,
+      "learning_rate": 1.1008337028396616e-06,
+      "loss": 0.5475,
+      "step": 7611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9160204717719853,
+      "learning_rate": 1.0997801262607599e-06,
+      "loss": 0.6415,
+      "step": 7612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9539927844434095,
+      "learning_rate": 1.0987269917940107e-06,
+      "loss": 0.6261,
+      "step": 7613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1191670091655235,
+      "learning_rate": 1.0976742995587941e-06,
+      "loss": 0.6103,
+      "step": 7614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2618502739095514,
+      "learning_rate": 1.096622049674439e-06,
+      "loss": 0.5837,
+      "step": 7615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8741698346992601,
+      "learning_rate": 1.095570242260221e-06,
+      "loss": 0.6731,
+      "step": 7616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9071579986938094,
+      "learning_rate": 1.094518877435372e-06,
+      "loss": 0.5906,
+      "step": 7617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8615446372298599,
+      "learning_rate": 1.093467955319068e-06,
+      "loss": 0.5217,
+      "step": 7618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9575245781215822,
+      "learning_rate": 1.0924174760304385e-06,
+      "loss": 0.5396,
+      "step": 7619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8968632493253208,
+      "learning_rate": 1.0913674396885598e-06,
+      "loss": 0.5562,
+      "step": 7620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9487478206224738,
+      "learning_rate": 1.090317846412461e-06,
+      "loss": 0.6184,
+      "step": 7621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0424190189706133,
+      "learning_rate": 1.0892686963211191e-06,
+      "loss": 0.6212,
+      "step": 7622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.174634093253347,
+      "learning_rate": 1.0882199895334605e-06,
+      "loss": 0.641,
+      "step": 7623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8967766592892383,
+      "learning_rate": 1.0871717261683619e-06,
+      "loss": 0.5959,
+      "step": 7624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0868235703345177,
+      "learning_rate": 1.0861239063446511e-06,
+      "loss": 0.6228,
+      "step": 7625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8227671520819473,
+      "learning_rate": 1.0850765301811028e-06,
+      "loss": 0.5928,
+      "step": 7626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9844970398015163,
+      "learning_rate": 1.0840295977964454e-06,
+      "loss": 0.7122,
+      "step": 7627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8711686783888133,
+      "learning_rate": 1.0829831093093524e-06,
+      "loss": 0.5506,
+      "step": 7628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.045619614830849,
+      "learning_rate": 1.0819370648384525e-06,
+      "loss": 0.7085,
+      "step": 7629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7682681907229623,
+      "learning_rate": 1.080891464502316e-06,
+      "loss": 0.6398,
+      "step": 7630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7360472338932105,
+      "learning_rate": 1.0798463084194715e-06,
+      "loss": 0.5867,
+      "step": 7631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.167967567642517,
+      "learning_rate": 1.0788015967083904e-06,
+      "loss": 0.693,
+      "step": 7632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8159747786768616,
+      "learning_rate": 1.0777573294875005e-06,
+      "loss": 0.5595,
+      "step": 7633
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8337688164202244,
+      "learning_rate": 1.076713506875171e-06,
+      "loss": 0.6175,
+      "step": 7634
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7006940068561807,
+      "learning_rate": 1.0756701289897298e-06,
+      "loss": 0.5631,
+      "step": 7635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8017376813538655,
+      "learning_rate": 1.0746271959494453e-06,
+      "loss": 0.5882,
+      "step": 7636
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.789860603798531,
+      "learning_rate": 1.0735847078725452e-06,
+      "loss": 0.6337,
+      "step": 7637
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.973652062162871,
+      "learning_rate": 1.0725426648771952e-06,
+      "loss": 0.6533,
+      "step": 7638
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.978809763923317,
+      "learning_rate": 1.0715010670815212e-06,
+      "loss": 0.5568,
+      "step": 7639
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7844382460604202,
+      "learning_rate": 1.070459914603592e-06,
+      "loss": 0.5602,
+      "step": 7640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8403111784857593,
+      "learning_rate": 1.0694192075614302e-06,
+      "loss": 0.6194,
+      "step": 7641
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7334240782019665,
+      "learning_rate": 1.0683789460730037e-06,
+      "loss": 0.5426,
+      "step": 7642
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.685316731739376,
+      "learning_rate": 1.0673391302562342e-06,
+      "loss": 0.4805,
+      "step": 7643
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.869559308757824,
+      "learning_rate": 1.0662997602289899e-06,
+      "loss": 0.7268,
+      "step": 7644
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.16910751822174,
+      "learning_rate": 1.0652608361090877e-06,
+      "loss": 0.6145,
+      "step": 7645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.923836692448067,
+      "learning_rate": 1.0642223580142985e-06,
+      "loss": 0.5563,
+      "step": 7646
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9053312003550178,
+      "learning_rate": 1.0631843260623382e-06,
+      "loss": 0.5602,
+      "step": 7647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9624951568475208,
+      "learning_rate": 1.0621467403708718e-06,
+      "loss": 0.5808,
+      "step": 7648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.887605733437194,
+      "learning_rate": 1.0611096010575196e-06,
+      "loss": 0.632,
+      "step": 7649
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9203647513784456,
+      "learning_rate": 1.0600729082398425e-06,
+      "loss": 0.6598,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8767373807747643,
+      "learning_rate": 1.0590366620353604e-06,
+      "loss": 0.598,
+      "step": 7651
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8056123438665657,
+      "learning_rate": 1.058000862561535e-06,
+      "loss": 0.5871,
+      "step": 7652
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.125561482859971,
+      "learning_rate": 1.0569655099357795e-06,
+      "loss": 0.6351,
+      "step": 7653
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0398054017931213,
+      "learning_rate": 1.0559306042754591e-06,
+      "loss": 0.6699,
+      "step": 7654
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0576078182108186,
+      "learning_rate": 1.0548961456978835e-06,
+      "loss": 0.6127,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9026327912259209,
+      "learning_rate": 1.0538621343203176e-06,
+      "loss": 0.5675,
+      "step": 7656
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0639922430174535,
+      "learning_rate": 1.05282857025997e-06,
+      "loss": 0.6958,
+      "step": 7657
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.112511747733857,
+      "learning_rate": 1.051795453634003e-06,
+      "loss": 0.6881,
+      "step": 7658
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0806728442724602,
+      "learning_rate": 1.0507627845595259e-06,
+      "loss": 0.6517,
+      "step": 7659
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8655768403232256,
+      "learning_rate": 1.049730563153597e-06,
+      "loss": 0.5419,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.107389670056977,
+      "learning_rate": 1.0486987895332229e-06,
+      "loss": 0.6122,
+      "step": 7661
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.001585005845773,
+      "learning_rate": 1.0476674638153638e-06,
+      "loss": 0.6874,
+      "step": 7662
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8022025145340081,
+      "learning_rate": 1.0466365861169242e-06,
+      "loss": 0.5743,
+      "step": 7663
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.25061615492626,
+      "learning_rate": 1.045606156554762e-06,
+      "loss": 0.6585,
+      "step": 7664
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8060804364597933,
+      "learning_rate": 1.0445761752456806e-06,
+      "loss": 0.5762,
+      "step": 7665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9562870642355572,
+      "learning_rate": 1.0435466423064373e-06,
+      "loss": 0.5267,
+      "step": 7666
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8288657824591275,
+      "learning_rate": 1.04251755785373e-06,
+      "loss": 0.5505,
+      "step": 7667
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9565933117918024,
+      "learning_rate": 1.0414889220042163e-06,
+      "loss": 0.6889,
+      "step": 7668
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0077080060650956,
+      "learning_rate": 1.0404607348744943e-06,
+      "loss": 0.6004,
+      "step": 7669
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8100825225512873,
+      "learning_rate": 1.0394329965811178e-06,
+      "loss": 0.5825,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9809824856030736,
+      "learning_rate": 1.038405707240585e-06,
+      "loss": 0.5185,
+      "step": 7671
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9217501283416114,
+      "learning_rate": 1.0373788669693464e-06,
+      "loss": 0.5785,
+      "step": 7672
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.056980482387888,
+      "learning_rate": 1.0363524758837984e-06,
+      "loss": 0.7201,
+      "step": 7673
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0266442614164544,
+      "learning_rate": 1.0353265341002916e-06,
+      "loss": 0.5002,
+      "step": 7674
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0009831878190654,
+      "learning_rate": 1.034301041735118e-06,
+      "loss": 0.5331,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.095824252013779,
+      "learning_rate": 1.0332759989045254e-06,
+      "loss": 0.6838,
+      "step": 7676
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7209335276516577,
+      "learning_rate": 1.0322514057247075e-06,
+      "loss": 0.6151,
+      "step": 7677
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9503500468360193,
+      "learning_rate": 1.031227262311809e-06,
+      "loss": 0.6328,
+      "step": 7678
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9867515395161024,
+      "learning_rate": 1.0302035687819202e-06,
+      "loss": 0.6623,
+      "step": 7679
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9160205255434155,
+      "learning_rate": 1.0291803252510857e-06,
+      "loss": 0.6412,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9044848440465423,
+      "learning_rate": 1.0281575318352937e-06,
+      "loss": 0.5959,
+      "step": 7681
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9017102445371714,
+      "learning_rate": 1.0271351886504832e-06,
+      "loss": 0.6255,
+      "step": 7682
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9617797412488593,
+      "learning_rate": 1.0261132958125452e-06,
+      "loss": 0.5663,
+      "step": 7683
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8326899544229418,
+      "learning_rate": 1.025091853437314e-06,
+      "loss": 0.5486,
+      "step": 7684
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9429118963302403,
+      "learning_rate": 1.0240708616405788e-06,
+      "loss": 0.6459,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9982778451386498,
+      "learning_rate": 1.0230503205380732e-06,
+      "loss": 0.6218,
+      "step": 7686
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0367479976721032,
+      "learning_rate": 1.0220302302454804e-06,
+      "loss": 0.6262,
+      "step": 7687
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1004913104077994,
+      "learning_rate": 1.0210105908784362e-06,
+      "loss": 0.6363,
+      "step": 7688
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.196885005810578,
+      "learning_rate": 1.019991402552521e-06,
+      "loss": 0.6211,
+      "step": 7689
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0510349626748896,
+      "learning_rate": 1.0189726653832637e-06,
+      "loss": 0.5913,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.062389403716988,
+      "learning_rate": 1.017954379486148e-06,
+      "loss": 0.6006,
+      "step": 7691
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.000958354903393,
+      "learning_rate": 1.0169365449765982e-06,
+      "loss": 0.6191,
+      "step": 7692
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1648733827702418,
+      "learning_rate": 1.0159191619699955e-06,
+      "loss": 0.6194,
+      "step": 7693
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7312236576166993,
+      "learning_rate": 1.014902230581663e-06,
+      "loss": 0.5829,
+      "step": 7694
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9281553571908774,
+      "learning_rate": 1.0138857509268784e-06,
+      "loss": 0.6326,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6412211399878196,
+      "learning_rate": 1.012869723120864e-06,
+      "loss": 0.5582,
+      "step": 7696
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.005045612911496,
+      "learning_rate": 1.0118541472787918e-06,
+      "loss": 0.5659,
+      "step": 7697
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9229217799978204,
+      "learning_rate": 1.0108390235157828e-06,
+      "loss": 0.5267,
+      "step": 7698
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.000719179387221,
+      "learning_rate": 1.0098243519469091e-06,
+      "loss": 0.5253,
+      "step": 7699
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0842973123188866,
+      "learning_rate": 1.0088101326871873e-06,
+      "loss": 0.6451,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6323359685648406,
+      "learning_rate": 1.0077963658515872e-06,
+      "loss": 0.531,
+      "step": 7701
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0540576166204314,
+      "learning_rate": 1.0067830515550224e-06,
+      "loss": 0.6194,
+      "step": 7702
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8564346985220948,
+      "learning_rate": 1.0057701899123622e-06,
+      "loss": 0.7364,
+      "step": 7703
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7651410118055617,
+      "learning_rate": 1.0047577810384146e-06,
+      "loss": 0.6406,
+      "step": 7704
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7861292875990769,
+      "learning_rate": 1.003745825047946e-06,
+      "loss": 0.5525,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7860974870586726,
+      "learning_rate": 1.002734322055664e-06,
+      "loss": 0.604,
+      "step": 7706
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9245464161068735,
+      "learning_rate": 1.0017232721762322e-06,
+      "loss": 0.6887,
+      "step": 7707
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.245870208702652,
+      "learning_rate": 1.0007126755242557e-06,
+      "loss": 0.5527,
+      "step": 7708
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9829272032721494,
+      "learning_rate": 9.997025322142934e-07,
+      "loss": 0.5582,
+      "step": 7709
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9923017209495424,
+      "learning_rate": 9.986928423608493e-07,
+      "loss": 0.647,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.027249908466499,
+      "learning_rate": 9.976836060783806e-07,
+      "loss": 0.6092,
+      "step": 7711
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8182521544541843,
+      "learning_rate": 9.966748234812845e-07,
+      "loss": 0.5714,
+      "step": 7712
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1289844169817194,
+      "learning_rate": 9.956664946839173e-07,
+      "loss": 0.5971,
+      "step": 7713
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2228604368044356,
+      "learning_rate": 9.946586198005754e-07,
+      "loss": 0.6114,
+      "step": 7714
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.736074800532946,
+      "learning_rate": 9.9365119894551e-07,
+      "loss": 0.6053,
+      "step": 7715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0883144500735913,
+      "learning_rate": 9.92644232232915e-07,
+      "loss": 0.6102,
+      "step": 7716
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0457525870447477,
+      "learning_rate": 9.91637719776939e-07,
+      "loss": 0.7014,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9544938159774565,
+      "learning_rate": 9.906316616916745e-07,
+      "loss": 0.5763,
+      "step": 7718
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7888980454640862,
+      "learning_rate": 9.89626058091162e-07,
+      "loss": 0.5796,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.138621337538482,
+      "learning_rate": 9.886209090893955e-07,
+      "loss": 0.6714,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8656774932599596,
+      "learning_rate": 9.876162148003121e-07,
+      "loss": 0.5886,
+      "step": 7721
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0270285492211597,
+      "learning_rate": 9.866119753378018e-07,
+      "loss": 0.6527,
+      "step": 7722
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0333059598856047,
+      "learning_rate": 9.856081908156984e-07,
+      "loss": 0.6542,
+      "step": 7723
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.764098539500526,
+      "learning_rate": 9.846048613477894e-07,
+      "loss": 0.5076,
+      "step": 7724
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8576540431009838,
+      "learning_rate": 9.836019870478058e-07,
+      "loss": 0.5937,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8921832430445682,
+      "learning_rate": 9.825995680294298e-07,
+      "loss": 0.5207,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8915147722807397,
+      "learning_rate": 9.815976044062902e-07,
+      "loss": 0.5271,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8331960578185176,
+      "learning_rate": 9.80596096291967e-07,
+      "loss": 0.6478,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7613187657217821,
+      "learning_rate": 9.795950437999852e-07,
+      "loss": 0.5245,
+      "step": 7729
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0307503138525873,
+      "learning_rate": 9.785944470438218e-07,
+      "loss": 0.6444,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7686848346745405,
+      "learning_rate": 9.775943061368982e-07,
+      "loss": 0.5941,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.345642126924032,
+      "learning_rate": 9.765946211925882e-07,
+      "loss": 0.6443,
+      "step": 7732
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9272931213689597,
+      "learning_rate": 9.755953923242102e-07,
+      "loss": 0.5868,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9544735196473395,
+      "learning_rate": 9.74596619645033e-07,
+      "loss": 0.6951,
+      "step": 7734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9460700117324157,
+      "learning_rate": 9.735983032682716e-07,
+      "loss": 0.6072,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1090367377616435,
+      "learning_rate": 9.726004433070935e-07,
+      "loss": 0.7047,
+      "step": 7736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1982409577233133,
+      "learning_rate": 9.716030398746096e-07,
+      "loss": 0.6184,
+      "step": 7737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9379260663865387,
+      "learning_rate": 9.706060930838834e-07,
+      "loss": 0.6641,
+      "step": 7738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3158402348464384,
+      "learning_rate": 9.69609603047922e-07,
+      "loss": 0.6658,
+      "step": 7739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.202051506082379,
+      "learning_rate": 9.686135698796866e-07,
+      "loss": 0.6234,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.946984139913474,
+      "learning_rate": 9.676179936920793e-07,
+      "loss": 0.7025,
+      "step": 7741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9651438174470215,
+      "learning_rate": 9.666228745979571e-07,
+      "loss": 0.6672,
+      "step": 7742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.821834253966292,
+      "learning_rate": 9.656282127101208e-07,
+      "loss": 0.5473,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8646490281696781,
+      "learning_rate": 9.646340081413225e-07,
+      "loss": 0.6667,
+      "step": 7744
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9663132430420378,
+      "learning_rate": 9.636402610042589e-07,
+      "loss": 0.6145,
+      "step": 7745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8591207036266417,
+      "learning_rate": 9.6264697141158e-07,
+      "loss": 0.5758,
+      "step": 7746
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8749424175544835,
+      "learning_rate": 9.61654139475877e-07,
+      "loss": 0.5613,
+      "step": 7747
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8187123900362623,
+      "learning_rate": 9.606617653096967e-07,
+      "loss": 0.5405,
+      "step": 7748
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0773843186976726,
+      "learning_rate": 9.59669849025529e-07,
+      "loss": 0.6071,
+      "step": 7749
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9075752017328036,
+      "learning_rate": 9.586783907358126e-07,
+      "loss": 0.5405,
+      "step": 7750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9368497704701095,
+      "learning_rate": 9.57687390552935e-07,
+      "loss": 0.6222,
+      "step": 7751
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.894016291586957,
+      "learning_rate": 9.566968485892324e-07,
+      "loss": 0.6088,
+      "step": 7752
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8925582435192947,
+      "learning_rate": 9.557067649569873e-07,
+      "loss": 0.5613,
+      "step": 7753
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9135127751222611,
+      "learning_rate": 9.54717139768433e-07,
+      "loss": 0.6596,
+      "step": 7754
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2074368490586944,
+      "learning_rate": 9.537279731357485e-07,
+      "loss": 0.6295,
+      "step": 7755
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9786253685527062,
+      "learning_rate": 9.527392651710598e-07,
+      "loss": 0.6196,
+      "step": 7756
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8644737769919695,
+      "learning_rate": 9.517510159864452e-07,
+      "loss": 0.5505,
+      "step": 7757
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.037734457999557,
+      "learning_rate": 9.507632256939264e-07,
+      "loss": 0.5701,
+      "step": 7758
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8913067735074787,
+      "learning_rate": 9.497758944054769e-07,
+      "loss": 0.5228,
+      "step": 7759
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.961114075237607,
+      "learning_rate": 9.487890222330137e-07,
+      "loss": 0.5865,
+      "step": 7760
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1418281069781733,
+      "learning_rate": 9.478026092884074e-07,
+      "loss": 0.7004,
+      "step": 7761
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.031618259141534,
+      "learning_rate": 9.468166556834724e-07,
+      "loss": 0.6048,
+      "step": 7762
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.983872334557737,
+      "learning_rate": 9.458311615299714e-07,
+      "loss": 0.6104,
+      "step": 7763
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.966621625986287,
+      "learning_rate": 9.448461269396148e-07,
+      "loss": 0.6129,
+      "step": 7764
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8456010848415982,
+      "learning_rate": 9.438615520240651e-07,
+      "loss": 0.6677,
+      "step": 7765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9795646084529916,
+      "learning_rate": 9.428774368949262e-07,
+      "loss": 0.488,
+      "step": 7766
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.778709075372669,
+      "learning_rate": 9.418937816637558e-07,
+      "loss": 0.5943,
+      "step": 7767
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0339698933276407,
+      "learning_rate": 9.409105864420548e-07,
+      "loss": 0.6256,
+      "step": 7768
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1716627541691045,
+      "learning_rate": 9.399278513412757e-07,
+      "loss": 0.6345,
+      "step": 7769
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.099308376029622,
+      "learning_rate": 9.389455764728167e-07,
+      "loss": 0.5985,
+      "step": 7770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.952829993448853,
+      "learning_rate": 9.379637619480236e-07,
+      "loss": 0.6046,
+      "step": 7771
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9217150291138596,
+      "learning_rate": 9.369824078781897e-07,
+      "loss": 0.5854,
+      "step": 7772
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0430771905247034,
+      "learning_rate": 9.360015143745599e-07,
+      "loss": 0.6265,
+      "step": 7773
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9586997014673917,
+      "learning_rate": 9.350210815483207e-07,
+      "loss": 0.7064,
+      "step": 7774
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0477443669373177,
+      "learning_rate": 9.340411095106128e-07,
+      "loss": 0.6626,
+      "step": 7775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8187859005943139,
+      "learning_rate": 9.330615983725194e-07,
+      "loss": 0.6062,
+      "step": 7776
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8676725602236373,
+      "learning_rate": 9.320825482450769e-07,
+      "loss": 0.6194,
+      "step": 7777
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.009081149555021,
+      "learning_rate": 9.311039592392612e-07,
+      "loss": 0.6296,
+      "step": 7778
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9506352943100582,
+      "learning_rate": 9.30125831466005e-07,
+      "loss": 0.6273,
+      "step": 7779
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8634198547031477,
+      "learning_rate": 9.291481650361822e-07,
+      "loss": 0.58,
+      "step": 7780
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2028182575661557,
+      "learning_rate": 9.281709600606193e-07,
+      "loss": 0.6037,
+      "step": 7781
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9124910064744183,
+      "learning_rate": 9.271942166500853e-07,
+      "loss": 0.5988,
+      "step": 7782
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0032674668960504,
+      "learning_rate": 9.262179349153022e-07,
+      "loss": 0.6078,
+      "step": 7783
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0402604629089893,
+      "learning_rate": 9.252421149669349e-07,
+      "loss": 0.5139,
+      "step": 7784
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9476800993967762,
+      "learning_rate": 9.242667569156006e-07,
+      "loss": 0.578,
+      "step": 7785
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.127765015120673,
+      "learning_rate": 9.232918608718599e-07,
+      "loss": 0.6612,
+      "step": 7786
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.963807687238336,
+      "learning_rate": 9.223174269462237e-07,
+      "loss": 0.5882,
+      "step": 7787
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9031394484343827,
+      "learning_rate": 9.213434552491479e-07,
+      "loss": 0.5933,
+      "step": 7788
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8294547712072886,
+      "learning_rate": 9.203699458910397e-07,
+      "loss": 0.6555,
+      "step": 7789
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8309553228658446,
+      "learning_rate": 9.193968989822504e-07,
+      "loss": 0.5514,
+      "step": 7790
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9206616651299844,
+      "learning_rate": 9.184243146330829e-07,
+      "loss": 0.6373,
+      "step": 7791
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1123393182042958,
+      "learning_rate": 9.174521929537827e-07,
+      "loss": 0.6415,
+      "step": 7792
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1317560935213113,
+      "learning_rate": 9.164805340545457e-07,
+      "loss": 0.6525,
+      "step": 7793
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0224002996549735,
+      "learning_rate": 9.15509338045516e-07,
+      "loss": 0.6894,
+      "step": 7794
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.042260851573768,
+      "learning_rate": 9.145386050367827e-07,
+      "loss": 0.5762,
+      "step": 7795
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9891557355289593,
+      "learning_rate": 9.135683351383862e-07,
+      "loss": 0.6294,
+      "step": 7796
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.022029699436418,
+      "learning_rate": 9.125985284603095e-07,
+      "loss": 0.5819,
+      "step": 7797
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1109280193657596,
+      "learning_rate": 9.116291851124887e-07,
+      "loss": 0.6766,
+      "step": 7798
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9502117247555344,
+      "learning_rate": 9.106603052048019e-07,
+      "loss": 0.5748,
+      "step": 7799
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7462106726187883,
+      "learning_rate": 9.096918888470785e-07,
+      "loss": 0.4553,
+      "step": 7800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0378937226681795,
+      "learning_rate": 9.087239361490919e-07,
+      "loss": 0.677,
+      "step": 7801
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8627644962745071,
+      "learning_rate": 9.07756447220568e-07,
+      "loss": 0.5913,
+      "step": 7802
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9507831523660102,
+      "learning_rate": 9.067894221711748e-07,
+      "loss": 0.5372,
+      "step": 7803
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9434714525311196,
+      "learning_rate": 9.058228611105319e-07,
+      "loss": 0.5585,
+      "step": 7804
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.784512235108847,
+      "learning_rate": 9.048567641482031e-07,
+      "loss": 0.6526,
+      "step": 7805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7646771163534876,
+      "learning_rate": 9.038911313937021e-07,
+      "loss": 0.5878,
+      "step": 7806
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9503722707113982,
+      "learning_rate": 9.02925962956489e-07,
+      "loss": 0.6089,
+      "step": 7807
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9933524346721043,
+      "learning_rate": 9.019612589459703e-07,
+      "loss": 0.7098,
+      "step": 7808
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.684179267546423,
+      "learning_rate": 9.009970194714995e-07,
+      "loss": 0.6127,
+      "step": 7809
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.947028893634746,
+      "learning_rate": 9.00033244642382e-07,
+      "loss": 0.571,
+      "step": 7810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8394267571336589,
+      "learning_rate": 8.990699345678633e-07,
+      "loss": 0.5227,
+      "step": 7811
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1484606234393557,
+      "learning_rate": 8.981070893571436e-07,
+      "loss": 0.6421,
+      "step": 7812
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.310340099058539,
+      "learning_rate": 8.971447091193641e-07,
+      "loss": 0.6349,
+      "step": 7813
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2442612121075705,
+      "learning_rate": 8.961827939636198e-07,
+      "loss": 0.6153,
+      "step": 7814
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0223880288341167,
+      "learning_rate": 8.952213439989443e-07,
+      "loss": 0.5675,
+      "step": 7815
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7330228615175547,
+      "learning_rate": 8.942603593343269e-07,
+      "loss": 0.5218,
+      "step": 7816
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9915324011404267,
+      "learning_rate": 8.932998400786985e-07,
+      "loss": 0.6024,
+      "step": 7817
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.098901888535553,
+      "learning_rate": 8.923397863409422e-07,
+      "loss": 0.6626,
+      "step": 7818
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0703368108294784,
+      "learning_rate": 8.913801982298825e-07,
+      "loss": 0.6209,
+      "step": 7819
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0340207412306706,
+      "learning_rate": 8.90421075854297e-07,
+      "loss": 0.744,
+      "step": 7820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0633123671861373,
+      "learning_rate": 8.894624193229051e-07,
+      "loss": 0.6967,
+      "step": 7821
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2559417118692306,
+      "learning_rate": 8.885042287443785e-07,
+      "loss": 0.5755,
+      "step": 7822
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0898486974483483,
+      "learning_rate": 8.875465042273323e-07,
+      "loss": 0.5747,
+      "step": 7823
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8473521886926423,
+      "learning_rate": 8.865892458803288e-07,
+      "loss": 0.5424,
+      "step": 7824
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.00638075054233,
+      "learning_rate": 8.856324538118815e-07,
+      "loss": 0.6036,
+      "step": 7825
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1426291684833663,
+      "learning_rate": 8.846761281304461e-07,
+      "loss": 0.5502,
+      "step": 7826
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9203586483594455,
+      "learning_rate": 8.837202689444274e-07,
+      "loss": 0.5921,
+      "step": 7827
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9063737713261275,
+      "learning_rate": 8.827648763621793e-07,
+      "loss": 0.5702,
+      "step": 7828
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7596477523436753,
+      "learning_rate": 8.818099504919997e-07,
+      "loss": 0.5492,
+      "step": 7829
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.878655835611806,
+      "learning_rate": 8.808554914421341e-07,
+      "loss": 0.5626,
+      "step": 7830
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7387383408855732,
+      "learning_rate": 8.799014993207783e-07,
+      "loss": 0.5791,
+      "step": 7831
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.825422980516178,
+      "learning_rate": 8.789479742360696e-07,
+      "loss": 0.6566,
+      "step": 7832
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.836576312128298,
+      "learning_rate": 8.779949162960988e-07,
+      "loss": 0.4287,
+      "step": 7833
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9235214625697135,
+      "learning_rate": 8.770423256088978e-07,
+      "loss": 0.648,
+      "step": 7834
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8400134406808073,
+      "learning_rate": 8.760902022824502e-07,
+      "loss": 0.5477,
+      "step": 7835
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.095020666362253,
+      "learning_rate": 8.751385464246836e-07,
+      "loss": 0.5633,
+      "step": 7836
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2320814488356078,
+      "learning_rate": 8.74187358143474e-07,
+      "loss": 0.6631,
+      "step": 7837
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.326172772622547,
+      "learning_rate": 8.732366375466422e-07,
+      "loss": 0.6657,
+      "step": 7838
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1461369278054625,
+      "learning_rate": 8.722863847419605e-07,
+      "loss": 0.563,
+      "step": 7839
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.014512726119495,
+      "learning_rate": 8.713365998371431e-07,
+      "loss": 0.674,
+      "step": 7840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.673982067794888,
+      "learning_rate": 8.703872829398563e-07,
+      "loss": 0.4947,
+      "step": 7841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7725031865905028,
+      "learning_rate": 8.694384341577072e-07,
+      "loss": 0.5893,
+      "step": 7842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.84444502749441,
+      "learning_rate": 8.684900535982566e-07,
+      "loss": 0.6629,
+      "step": 7843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7431264743321058,
+      "learning_rate": 8.675421413690072e-07,
+      "loss": 0.6112,
+      "step": 7844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.88630416944583,
+      "learning_rate": 8.665946975774103e-07,
+      "loss": 0.686,
+      "step": 7845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8938332117324725,
+      "learning_rate": 8.656477223308623e-07,
+      "loss": 0.596,
+      "step": 7846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.856990540011655,
+      "learning_rate": 8.647012157367118e-07,
+      "loss": 0.6462,
+      "step": 7847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9573712099806473,
+      "learning_rate": 8.63755177902248e-07,
+      "loss": 0.5727,
+      "step": 7848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7953197325036485,
+      "learning_rate": 8.62809608934711e-07,
+      "loss": 0.5912,
+      "step": 7849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9467543418309012,
+      "learning_rate": 8.618645089412852e-07,
+      "loss": 0.643,
+      "step": 7850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9747829769385832,
+      "learning_rate": 8.609198780291067e-07,
+      "loss": 0.6153,
+      "step": 7851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8468696120697057,
+      "learning_rate": 8.599757163052491e-07,
+      "loss": 0.5527,
+      "step": 7852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0237754704208495,
+      "learning_rate": 8.590320238767425e-07,
+      "loss": 0.6137,
+      "step": 7853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9018135574384365,
+      "learning_rate": 8.580888008505578e-07,
+      "loss": 0.6224,
+      "step": 7854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9945743056461949,
+      "learning_rate": 8.571460473336168e-07,
+      "loss": 0.6345,
+      "step": 7855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8275906820507346,
+      "learning_rate": 8.562037634327836e-07,
+      "loss": 0.6913,
+      "step": 7856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7781481219644808,
+      "learning_rate": 8.552619492548736e-07,
+      "loss": 0.6142,
+      "step": 7857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.101917955513664,
+      "learning_rate": 8.543206049066461e-07,
+      "loss": 0.5878,
+      "step": 7858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7860451310801435,
+      "learning_rate": 8.533797304948066e-07,
+      "loss": 0.6194,
+      "step": 7859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.053297227977161,
+      "learning_rate": 8.524393261260106e-07,
+      "loss": 0.5577,
+      "step": 7860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8999012436315257,
+      "learning_rate": 8.51499391906856e-07,
+      "loss": 0.5324,
+      "step": 7861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9403626400781433,
+      "learning_rate": 8.50559927943892e-07,
+      "loss": 0.6044,
+      "step": 7862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9202525596061424,
+      "learning_rate": 8.496209343436101e-07,
+      "loss": 0.6422,
+      "step": 7863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9258908691490637,
+      "learning_rate": 8.486824112124531e-07,
+      "loss": 0.5877,
+      "step": 7864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.139813896668843,
+      "learning_rate": 8.477443586568068e-07,
+      "loss": 0.6723,
+      "step": 7865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8864183561566554,
+      "learning_rate": 8.46806776783004e-07,
+      "loss": 0.5684,
+      "step": 7866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.086021703795221,
+      "learning_rate": 8.458696656973242e-07,
+      "loss": 0.6466,
+      "step": 7867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7670328138359852,
+      "learning_rate": 8.449330255059974e-07,
+      "loss": 0.5655,
+      "step": 7868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.109669089522648,
+      "learning_rate": 8.439968563151935e-07,
+      "loss": 0.5716,
+      "step": 7869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9731160176083082,
+      "learning_rate": 8.430611582310355e-07,
+      "loss": 0.6779,
+      "step": 7870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6833598941120023,
+      "learning_rate": 8.421259313595881e-07,
+      "loss": 0.5651,
+      "step": 7871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9782640426514457,
+      "learning_rate": 8.411911758068664e-07,
+      "loss": 0.5027,
+      "step": 7872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7638609114287134,
+      "learning_rate": 8.402568916788295e-07,
+      "loss": 0.6116,
+      "step": 7873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0287539947205198,
+      "learning_rate": 8.393230790813834e-07,
+      "loss": 0.6667,
+      "step": 7874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8293808842129964,
+      "learning_rate": 8.383897381203804e-07,
+      "loss": 0.6778,
+      "step": 7875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.889493686769746,
+      "learning_rate": 8.374568689016222e-07,
+      "loss": 0.5777,
+      "step": 7876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0618821665052263,
+      "learning_rate": 8.365244715308524e-07,
+      "loss": 0.6482,
+      "step": 7877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.979835255191372,
+      "learning_rate": 8.355925461137659e-07,
+      "loss": 0.5515,
+      "step": 7878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8234234080540386,
+      "learning_rate": 8.346610927559995e-07,
+      "loss": 0.5685,
+      "step": 7879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.114466124057432,
+      "learning_rate": 8.337301115631408e-07,
+      "loss": 0.6513,
+      "step": 7880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0450819662924973,
+      "learning_rate": 8.327996026407215e-07,
+      "loss": 0.582,
+      "step": 7881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9534975150914224,
+      "learning_rate": 8.318695660942188e-07,
+      "loss": 0.6081,
+      "step": 7882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.766548803762076,
+      "learning_rate": 8.309400020290576e-07,
+      "loss": 0.5923,
+      "step": 7883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8376364935303842,
+      "learning_rate": 8.30010910550611e-07,
+      "loss": 0.5752,
+      "step": 7884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1740334973065294,
+      "learning_rate": 8.29082291764195e-07,
+      "loss": 0.6675,
+      "step": 7885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.955516410752621,
+      "learning_rate": 8.281541457750752e-07,
+      "loss": 0.6615,
+      "step": 7886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0513285670740427,
+      "learning_rate": 8.272264726884611e-07,
+      "loss": 0.6483,
+      "step": 7887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9204631948112747,
+      "learning_rate": 8.262992726095126e-07,
+      "loss": 0.63,
+      "step": 7888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.920799123113018,
+      "learning_rate": 8.253725456433281e-07,
+      "loss": 0.574,
+      "step": 7889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0544728279229307,
+      "learning_rate": 8.244462918949613e-07,
+      "loss": 0.6066,
+      "step": 7890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8312075580282576,
+      "learning_rate": 8.235205114694067e-07,
+      "loss": 0.6475,
+      "step": 7891
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.138681797382035,
+      "learning_rate": 8.225952044716079e-07,
+      "loss": 0.6259,
+      "step": 7892
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.895533332438548,
+      "learning_rate": 8.216703710064516e-07,
+      "loss": 0.5747,
+      "step": 7893
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.904821942247295,
+      "learning_rate": 8.207460111787763e-07,
+      "loss": 0.589,
+      "step": 7894
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9668496729404117,
+      "learning_rate": 8.198221250933613e-07,
+      "loss": 0.6433,
+      "step": 7895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9284587389594263,
+      "learning_rate": 8.188987128549336e-07,
+      "loss": 0.6273,
+      "step": 7896
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1416050872893946,
+      "learning_rate": 8.179757745681693e-07,
+      "loss": 0.6704,
+      "step": 7897
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7866996337890955,
+      "learning_rate": 8.170533103376865e-07,
+      "loss": 0.6691,
+      "step": 7898
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8456024182305906,
+      "learning_rate": 8.161313202680543e-07,
+      "loss": 0.6148,
+      "step": 7899
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7492541132835926,
+      "learning_rate": 8.15209804463783e-07,
+      "loss": 0.5606,
+      "step": 7900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9516406633526484,
+      "learning_rate": 8.142887630293339e-07,
+      "loss": 0.5595,
+      "step": 7901
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7358148718068005,
+      "learning_rate": 8.133681960691098e-07,
+      "loss": 0.4818,
+      "step": 7902
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.963868509775501,
+      "learning_rate": 8.124481036874665e-07,
+      "loss": 0.5443,
+      "step": 7903
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0190989920317866,
+      "learning_rate": 8.115284859886963e-07,
+      "loss": 0.5788,
+      "step": 7904
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9193826384008805,
+      "learning_rate": 8.106093430770473e-07,
+      "loss": 0.7062,
+      "step": 7905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.048467386932486,
+      "learning_rate": 8.096906750567063e-07,
+      "loss": 0.659,
+      "step": 7906
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1066721308629646,
+      "learning_rate": 8.087724820318127e-07,
+      "loss": 0.6521,
+      "step": 7907
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.113774039359718,
+      "learning_rate": 8.07854764106446e-07,
+      "loss": 0.6454,
+      "step": 7908
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9652147888264364,
+      "learning_rate": 8.069375213846381e-07,
+      "loss": 0.5461,
+      "step": 7909
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8955929743963291,
+      "learning_rate": 8.060207539703613e-07,
+      "loss": 0.5916,
+      "step": 7910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.202212850241475,
+      "learning_rate": 8.051044619675368e-07,
+      "loss": 0.7072,
+      "step": 7911
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9825746509636721,
+      "learning_rate": 8.041886454800307e-07,
+      "loss": 0.6028,
+      "step": 7912
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.743284921250136,
+      "learning_rate": 8.032733046116581e-07,
+      "loss": 0.5952,
+      "step": 7913
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9131860315901044,
+      "learning_rate": 8.023584394661754e-07,
+      "loss": 0.5394,
+      "step": 7914
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.08391122810736,
+      "learning_rate": 8.014440501472909e-07,
+      "loss": 0.6065,
+      "step": 7915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9814471916087633,
+      "learning_rate": 8.005301367586532e-07,
+      "loss": 0.5821,
+      "step": 7916
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3000295637313712,
+      "learning_rate": 7.996166994038618e-07,
+      "loss": 0.558,
+      "step": 7917
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0762593164954706,
+      "learning_rate": 7.987037381864587e-07,
+      "loss": 0.7169,
+      "step": 7918
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7514422787011197,
+      "learning_rate": 7.977912532099336e-07,
+      "loss": 0.558,
+      "step": 7919
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.76970855988798,
+      "learning_rate": 7.968792445777207e-07,
+      "loss": 0.5681,
+      "step": 7920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9054210564705762,
+      "learning_rate": 7.95967712393203e-07,
+      "loss": 0.6098,
+      "step": 7921
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9968109035396673,
+      "learning_rate": 7.950566567597067e-07,
+      "loss": 0.5548,
+      "step": 7922
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0679678927457252,
+      "learning_rate": 7.941460777805071e-07,
+      "loss": 0.6346,
+      "step": 7923
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.771436468079132,
+      "learning_rate": 7.932359755588204e-07,
+      "loss": 0.6228,
+      "step": 7924
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.028353877337731,
+      "learning_rate": 7.923263501978151e-07,
+      "loss": 0.6094,
+      "step": 7925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9541780264877802,
+      "learning_rate": 7.914172018006006e-07,
+      "loss": 0.6535,
+      "step": 7926
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.033311941728741,
+      "learning_rate": 7.905085304702348e-07,
+      "loss": 0.6615,
+      "step": 7927
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9664209507744381,
+      "learning_rate": 7.896003363097194e-07,
+      "loss": 0.613,
+      "step": 7928
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.047467628372388,
+      "learning_rate": 7.886926194220051e-07,
+      "loss": 0.6435,
+      "step": 7929
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2322887628362986,
+      "learning_rate": 7.87785379909985e-07,
+      "loss": 0.4776,
+      "step": 7930
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9161552547216953,
+      "learning_rate": 7.86878617876502e-07,
+      "loss": 0.5047,
+      "step": 7931
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9666675466994157,
+      "learning_rate": 7.859723334243414e-07,
+      "loss": 0.6392,
+      "step": 7932
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0673189448510128,
+      "learning_rate": 7.850665266562352e-07,
+      "loss": 0.697,
+      "step": 7933
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.02537722417873,
+      "learning_rate": 7.841611976748637e-07,
+      "loss": 0.6193,
+      "step": 7934
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8026876495600388,
+      "learning_rate": 7.832563465828486e-07,
+      "loss": 0.6503,
+      "step": 7935
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9214738333018813,
+      "learning_rate": 7.823519734827623e-07,
+      "loss": 0.5923,
+      "step": 7936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0886931902273393,
+      "learning_rate": 7.814480784771184e-07,
+      "loss": 0.6328,
+      "step": 7937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7978925952821656,
+      "learning_rate": 7.805446616683815e-07,
+      "loss": 0.6205,
+      "step": 7938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2547069940717863,
+      "learning_rate": 7.796417231589553e-07,
+      "loss": 0.5602,
+      "step": 7939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0421838146360134,
+      "learning_rate": 7.78739263051198e-07,
+      "loss": 0.6162,
+      "step": 7940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8776854409493586,
+      "learning_rate": 7.778372814474028e-07,
+      "loss": 0.5241,
+      "step": 7941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9343979647373022,
+      "learning_rate": 7.769357784498189e-07,
+      "loss": 0.6138,
+      "step": 7942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9126629837911133,
+      "learning_rate": 7.760347541606339e-07,
+      "loss": 0.5349,
+      "step": 7943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.013112677164787,
+      "learning_rate": 7.751342086819864e-07,
+      "loss": 0.5842,
+      "step": 7944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1023021606726475,
+      "learning_rate": 7.742341421159561e-07,
+      "loss": 0.5927,
+      "step": 7945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9172493579642078,
+      "learning_rate": 7.733345545645726e-07,
+      "loss": 0.4841,
+      "step": 7946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8852024848426656,
+      "learning_rate": 7.724354461298089e-07,
+      "loss": 0.6012,
+      "step": 7947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8875527210328427,
+      "learning_rate": 7.71536816913584e-07,
+      "loss": 0.5591,
+      "step": 7948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1986943184940437,
+      "learning_rate": 7.706386670177606e-07,
+      "loss": 0.6477,
+      "step": 7949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9792660395007868,
+      "learning_rate": 7.697409965441527e-07,
+      "loss": 0.5966,
+      "step": 7950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0633054153372186,
+      "learning_rate": 7.68843805594513e-07,
+      "loss": 0.5333,
+      "step": 7951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0106706139040496,
+      "learning_rate": 7.679470942705459e-07,
+      "loss": 0.6114,
+      "step": 7952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9411577381023806,
+      "learning_rate": 7.670508626738959e-07,
+      "loss": 0.6204,
+      "step": 7953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8717425506845027,
+      "learning_rate": 7.661551109061593e-07,
+      "loss": 0.6533,
+      "step": 7954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.957638620055607,
+      "learning_rate": 7.652598390688731e-07,
+      "loss": 0.61,
+      "step": 7955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7361028180084892,
+      "learning_rate": 7.643650472635211e-07,
+      "loss": 0.5245,
+      "step": 7956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.062695795851483,
+      "learning_rate": 7.634707355915321e-07,
+      "loss": 0.5406,
+      "step": 7957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7556940200609303,
+      "learning_rate": 7.625769041542841e-07,
+      "loss": 0.6456,
+      "step": 7958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1244482890856484,
+      "learning_rate": 7.616835530530947e-07,
+      "loss": 0.5838,
+      "step": 7959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.007384668816092,
+      "learning_rate": 7.607906823892341e-07,
+      "loss": 0.5308,
+      "step": 7960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8335296792574647,
+      "learning_rate": 7.598982922639109e-07,
+      "loss": 0.5689,
+      "step": 7961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.89865804594614,
+      "learning_rate": 7.590063827782851e-07,
+      "loss": 0.6264,
+      "step": 7962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.101400701382922,
+      "learning_rate": 7.581149540334587e-07,
+      "loss": 0.6344,
+      "step": 7963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8929227815753287,
+      "learning_rate": 7.572240061304786e-07,
+      "loss": 0.6135,
+      "step": 7964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9995425057991527,
+      "learning_rate": 7.563335391703424e-07,
+      "loss": 0.6244,
+      "step": 7965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8371114068334875,
+      "learning_rate": 7.554435532539872e-07,
+      "loss": 0.6308,
+      "step": 7966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.85970602929272,
+      "learning_rate": 7.545540484822972e-07,
+      "loss": 0.5694,
+      "step": 7967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0786161787069113,
+      "learning_rate": 7.536650249561056e-07,
+      "loss": 0.7314,
+      "step": 7968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7588934796892286,
+      "learning_rate": 7.527764827761863e-07,
+      "loss": 0.5897,
+      "step": 7969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2854940957417815,
+      "learning_rate": 7.518884220432599e-07,
+      "loss": 0.6597,
+      "step": 7970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9764113169266813,
+      "learning_rate": 7.510008428579956e-07,
+      "loss": 0.5809,
+      "step": 7971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.026543602570681,
+      "learning_rate": 7.501137453210027e-07,
+      "loss": 0.7416,
+      "step": 7972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.177075001218227,
+      "learning_rate": 7.492271295328419e-07,
+      "loss": 0.57,
+      "step": 7973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8721218522062755,
+      "learning_rate": 7.483409955940136e-07,
+      "loss": 0.5692,
+      "step": 7974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8713817493079272,
+      "learning_rate": 7.474553436049675e-07,
+      "loss": 0.5742,
+      "step": 7975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8123154657334932,
+      "learning_rate": 7.465701736660963e-07,
+      "loss": 0.5795,
+      "step": 7976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8267308046975017,
+      "learning_rate": 7.456854858777418e-07,
+      "loss": 0.5542,
+      "step": 7977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.124509762504682,
+      "learning_rate": 7.448012803401843e-07,
+      "loss": 0.5719,
+      "step": 7978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7888627753037367,
+      "learning_rate": 7.43917557153656e-07,
+      "loss": 0.6277,
+      "step": 7979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8413228879270656,
+      "learning_rate": 7.430343164183312e-07,
+      "loss": 0.5997,
+      "step": 7980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8432625198692136,
+      "learning_rate": 7.421515582343308e-07,
+      "loss": 0.6055,
+      "step": 7981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0281622706959856,
+      "learning_rate": 7.412692827017193e-07,
+      "loss": 0.5958,
+      "step": 7982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9798861010712308,
+      "learning_rate": 7.4038748992051e-07,
+      "loss": 0.5568,
+      "step": 7983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9779665626638276,
+      "learning_rate": 7.395061799906578e-07,
+      "loss": 0.6735,
+      "step": 7984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.105242899193206,
+      "learning_rate": 7.386253530120635e-07,
+      "loss": 0.6218,
+      "step": 7985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.792392367170328,
+      "learning_rate": 7.377450090845733e-07,
+      "loss": 0.6366,
+      "step": 7986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9121163417362366,
+      "learning_rate": 7.368651483079819e-07,
+      "loss": 0.5815,
+      "step": 7987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.848675634361415,
+      "learning_rate": 7.35985770782024e-07,
+      "loss": 0.6224,
+      "step": 7988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9301143939260674,
+      "learning_rate": 7.35106876606384e-07,
+      "loss": 0.6844,
+      "step": 7989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9834997145749633,
+      "learning_rate": 7.342284658806875e-07,
+      "loss": 0.7157,
+      "step": 7990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8168292987948385,
+      "learning_rate": 7.333505387045108e-07,
+      "loss": 0.5979,
+      "step": 7991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8488952297008012,
+      "learning_rate": 7.324730951773673e-07,
+      "loss": 0.478,
+      "step": 7992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8987998955323417,
+      "learning_rate": 7.315961353987234e-07,
+      "loss": 0.6196,
+      "step": 7993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9061861574645693,
+      "learning_rate": 7.307196594679855e-07,
+      "loss": 0.5492,
+      "step": 7994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.026536733296861,
+      "learning_rate": 7.298436674845099e-07,
+      "loss": 0.6364,
+      "step": 7995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4322053014985077,
+      "learning_rate": 7.289681595475922e-07,
+      "loss": 0.7198,
+      "step": 7996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.851711818464298,
+      "learning_rate": 7.280931357564791e-07,
+      "loss": 0.6029,
+      "step": 7997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.003585262968676,
+      "learning_rate": 7.272185962103567e-07,
+      "loss": 0.5255,
+      "step": 7998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.808289594625113,
+      "learning_rate": 7.263445410083614e-07,
+      "loss": 0.6749,
+      "step": 7999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8692318647300077,
+      "learning_rate": 7.254709702495721e-07,
+      "loss": 0.5514,
+      "step": 8000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.996444861038339,
+      "learning_rate": 7.245978840330103e-07,
+      "loss": 0.575,
+      "step": 8001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0228741290262593,
+      "learning_rate": 7.23725282457649e-07,
+      "loss": 0.6426,
+      "step": 8002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0017135428709207,
+      "learning_rate": 7.228531656223997e-07,
+      "loss": 0.6279,
+      "step": 8003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.894666164788839,
+      "learning_rate": 7.219815336261243e-07,
+      "loss": 0.5261,
+      "step": 8004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1004694991805155,
+      "learning_rate": 7.211103865676255e-07,
+      "loss": 0.5761,
+      "step": 8005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9736439337576523,
+      "learning_rate": 7.202397245456539e-07,
+      "loss": 0.6219,
+      "step": 8006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9906929570826677,
+      "learning_rate": 7.193695476589019e-07,
+      "loss": 0.5685,
+      "step": 8007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7979796734065363,
+      "learning_rate": 7.184998560060114e-07,
+      "loss": 0.5869,
+      "step": 8008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9042889426569207,
+      "learning_rate": 7.176306496855651e-07,
+      "loss": 0.6146,
+      "step": 8009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9254016912472625,
+      "learning_rate": 7.167619287960942e-07,
+      "loss": 0.581,
+      "step": 8010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9512998168719646,
+      "learning_rate": 7.158936934360711e-07,
+      "loss": 0.6168,
+      "step": 8011
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.046608135870485,
+      "learning_rate": 7.150259437039175e-07,
+      "loss": 0.7063,
+      "step": 8012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.878222927702504,
+      "learning_rate": 7.14158679697996e-07,
+      "loss": 0.6154,
+      "step": 8013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.881994367788152,
+      "learning_rate": 7.13291901516619e-07,
+      "loss": 0.5284,
+      "step": 8014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.123034281457884,
+      "learning_rate": 7.124256092580357e-07,
+      "loss": 0.6879,
+      "step": 8015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9677989441827224,
+      "learning_rate": 7.11559803020449e-07,
+      "loss": 0.6354,
+      "step": 8016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0902633305259357,
+      "learning_rate": 7.106944829020013e-07,
+      "loss": 0.5253,
+      "step": 8017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.038512209808121,
+      "learning_rate": 7.098296490007828e-07,
+      "loss": 0.6774,
+      "step": 8018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7981807892646227,
+      "learning_rate": 7.089653014148263e-07,
+      "loss": 0.6563,
+      "step": 8019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.004711187157496,
+      "learning_rate": 7.081014402421115e-07,
+      "loss": 0.6193,
+      "step": 8020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.984447214093464,
+      "learning_rate": 7.072380655805617e-07,
+      "loss": 0.6576,
+      "step": 8021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9593774644521333,
+      "learning_rate": 7.063751775280448e-07,
+      "loss": 0.5683,
+      "step": 8022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.012476005916973,
+      "learning_rate": 7.055127761823732e-07,
+      "loss": 0.6759,
+      "step": 8023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.445817481217792,
+      "learning_rate": 7.046508616413078e-07,
+      "loss": 0.6529,
+      "step": 8024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1997155940021216,
+      "learning_rate": 7.037894340025487e-07,
+      "loss": 0.6053,
+      "step": 8025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5564007641049542,
+      "learning_rate": 7.029284933637454e-07,
+      "loss": 0.5903,
+      "step": 8026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1252616955848116,
+      "learning_rate": 7.020680398224893e-07,
+      "loss": 0.6022,
+      "step": 8027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.877781753370922,
+      "learning_rate": 7.012080734763205e-07,
+      "loss": 0.6165,
+      "step": 8028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.908376242028954,
+      "learning_rate": 7.003485944227162e-07,
+      "loss": 0.516,
+      "step": 8029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8883975626092215,
+      "learning_rate": 6.994896027591074e-07,
+      "loss": 0.5823,
+      "step": 8030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.893765363384913,
+      "learning_rate": 6.986310985828626e-07,
+      "loss": 0.6621,
+      "step": 8031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.939654375998763,
+      "learning_rate": 6.977730819913015e-07,
+      "loss": 0.609,
+      "step": 8032
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9859890115855503,
+      "learning_rate": 6.969155530816824e-07,
+      "loss": 0.6512,
+      "step": 8033
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.036896841652315,
+      "learning_rate": 6.960585119512125e-07,
+      "loss": 0.5567,
+      "step": 8034
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.919130726074716,
+      "learning_rate": 6.952019586970416e-07,
+      "loss": 0.612,
+      "step": 8035
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.928540621573469,
+      "learning_rate": 6.943458934162656e-07,
+      "loss": 0.6911,
+      "step": 8036
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9840568093800386,
+      "learning_rate": 6.934903162059242e-07,
+      "loss": 0.6367,
+      "step": 8037
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.005740659612198,
+      "learning_rate": 6.92635227163001e-07,
+      "loss": 0.6262,
+      "step": 8038
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1474820104829084,
+      "learning_rate": 6.917806263844268e-07,
+      "loss": 0.6138,
+      "step": 8039
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1811307190465885,
+      "learning_rate": 6.909265139670735e-07,
+      "loss": 0.5507,
+      "step": 8040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9714652581473788,
+      "learning_rate": 6.900728900077619e-07,
+      "loss": 0.6353,
+      "step": 8041
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.335366259929589,
+      "learning_rate": 6.89219754603253e-07,
+      "loss": 0.7156,
+      "step": 8042
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0117714750139672,
+      "learning_rate": 6.883671078502574e-07,
+      "loss": 0.5447,
+      "step": 8043
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.117046336034858,
+      "learning_rate": 6.875149498454237e-07,
+      "loss": 0.7257,
+      "step": 8044
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1060454141945257,
+      "learning_rate": 6.866632806853518e-07,
+      "loss": 0.7056,
+      "step": 8045
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8680810716410823,
+      "learning_rate": 6.858121004665813e-07,
+      "loss": 0.5665,
+      "step": 8046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.969897563001284,
+      "learning_rate": 6.849614092856005e-07,
+      "loss": 0.6128,
+      "step": 8047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.038296389027259,
+      "learning_rate": 6.841112072388373e-07,
+      "loss": 0.6069,
+      "step": 8048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.12047751623645,
+      "learning_rate": 6.832614944226695e-07,
+      "loss": 0.6651,
+      "step": 8049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.78106945111783,
+      "learning_rate": 6.824122709334152e-07,
+      "loss": 0.6632,
+      "step": 8050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7715123778299606,
+      "learning_rate": 6.815635368673418e-07,
+      "loss": 0.5322,
+      "step": 8051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9202887091570335,
+      "learning_rate": 6.807152923206528e-07,
+      "loss": 0.6464,
+      "step": 8052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9938870648476452,
+      "learning_rate": 6.798675373895064e-07,
+      "loss": 0.6621,
+      "step": 8053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0986027995623178,
+      "learning_rate": 6.790202721699968e-07,
+      "loss": 0.6684,
+      "step": 8054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0705757135608573,
+      "learning_rate": 6.781734967581699e-07,
+      "loss": 0.5709,
+      "step": 8055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.001524224677035,
+      "learning_rate": 6.77327211250009e-07,
+      "loss": 0.5792,
+      "step": 8056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7953776708987224,
+      "learning_rate": 6.764814157414484e-07,
+      "loss": 0.5446,
+      "step": 8057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9216694263397198,
+      "learning_rate": 6.756361103283626e-07,
+      "loss": 0.5869,
+      "step": 8058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.23119073549652,
+      "learning_rate": 6.747912951065722e-07,
+      "loss": 0.72,
+      "step": 8059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.887099107100426,
+      "learning_rate": 6.739469701718398e-07,
+      "loss": 0.5626,
+      "step": 8060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7724551449703356,
+      "learning_rate": 6.731031356198769e-07,
+      "loss": 0.5565,
+      "step": 8061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8592960000825143,
+      "learning_rate": 6.722597915463352e-07,
+      "loss": 0.6031,
+      "step": 8062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.000586160391033,
+      "learning_rate": 6.714169380468144e-07,
+      "loss": 0.5987,
+      "step": 8063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8188177079291683,
+      "learning_rate": 6.705745752168552e-07,
+      "loss": 0.5919,
+      "step": 8064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8460808208474586,
+      "learning_rate": 6.697327031519452e-07,
+      "loss": 0.5777,
+      "step": 8065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0220325407867525,
+      "learning_rate": 6.688913219475158e-07,
+      "loss": 0.6628,
+      "step": 8066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9653708715581029,
+      "learning_rate": 6.680504316989405e-07,
+      "loss": 0.6537,
+      "step": 8067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1103633643710333,
+      "learning_rate": 6.672100325015396e-07,
+      "loss": 0.7001,
+      "step": 8068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.930693121411337,
+      "learning_rate": 6.663701244505788e-07,
+      "loss": 0.6294,
+      "step": 8069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0081188554424685,
+      "learning_rate": 6.655307076412637e-07,
+      "loss": 0.6435,
+      "step": 8070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9797486027365578,
+      "learning_rate": 6.646917821687504e-07,
+      "loss": 0.6492,
+      "step": 8071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8818227613869194,
+      "learning_rate": 6.638533481281323e-07,
+      "loss": 0.5762,
+      "step": 8072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.05125972802814,
+      "learning_rate": 6.630154056144533e-07,
+      "loss": 0.673,
+      "step": 8073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7769660589729557,
+      "learning_rate": 6.621779547226986e-07,
+      "loss": 0.6233,
+      "step": 8074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.305973486593778,
+      "learning_rate": 6.613409955477962e-07,
+      "loss": 0.6659,
+      "step": 8075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8901669506607903,
+      "learning_rate": 6.605045281846222e-07,
+      "loss": 0.6144,
+      "step": 8076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0978112467258616,
+      "learning_rate": 6.596685527279939e-07,
+      "loss": 0.6885,
+      "step": 8077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.004337979444083,
+      "learning_rate": 6.588330692726747e-07,
+      "loss": 0.6176,
+      "step": 8078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1383294050312145,
+      "learning_rate": 6.579980779133705e-07,
+      "loss": 0.674,
+      "step": 8079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8627858607830665,
+      "learning_rate": 6.571635787447339e-07,
+      "loss": 0.5736,
+      "step": 8080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.059976230713306,
+      "learning_rate": 6.563295718613577e-07,
+      "loss": 0.7069,
+      "step": 8081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0325541362148702,
+      "learning_rate": 6.554960573577834e-07,
+      "loss": 0.586,
+      "step": 8082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9740764701873206,
+      "learning_rate": 6.546630353284927e-07,
+      "loss": 0.5326,
+      "step": 8083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9563646646591624,
+      "learning_rate": 6.538305058679156e-07,
+      "loss": 0.6275,
+      "step": 8084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0898251948049698,
+      "learning_rate": 6.529984690704222e-07,
+      "loss": 0.6423,
+      "step": 8085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2789120695548575,
+      "learning_rate": 6.521669250303303e-07,
+      "loss": 0.6563,
+      "step": 8086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9959617661791869,
+      "learning_rate": 6.51335873841899e-07,
+      "loss": 0.5937,
+      "step": 8087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0252537703122244,
+      "learning_rate": 6.505053155993335e-07,
+      "loss": 0.6138,
+      "step": 8088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9790505498454918,
+      "learning_rate": 6.496752503967801e-07,
+      "loss": 0.635,
+      "step": 8089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.223228305084311,
+      "learning_rate": 6.488456783283343e-07,
+      "loss": 0.6743,
+      "step": 8090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2335882128652336,
+      "learning_rate": 6.480165994880311e-07,
+      "loss": 0.6312,
+      "step": 8091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.153211792356647,
+      "learning_rate": 6.471880139698523e-07,
+      "loss": 0.6157,
+      "step": 8092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1742719609091137,
+      "learning_rate": 6.463599218677214e-07,
+      "loss": 0.6432,
+      "step": 8093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0043189581847036,
+      "learning_rate": 6.455323232755095e-07,
+      "loss": 0.6517,
+      "step": 8094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.135564020927073,
+      "learning_rate": 6.447052182870284e-07,
+      "loss": 0.642,
+      "step": 8095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9084088820683192,
+      "learning_rate": 6.438786069960345e-07,
+      "loss": 0.6292,
+      "step": 8096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9232430703411438,
+      "learning_rate": 6.430524894962292e-07,
+      "loss": 0.6953,
+      "step": 8097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9925604766847647,
+      "learning_rate": 6.422268658812591e-07,
+      "loss": 0.511,
+      "step": 8098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0022550095449083,
+      "learning_rate": 6.414017362447106e-07,
+      "loss": 0.7423,
+      "step": 8099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.761768263429008,
+      "learning_rate": 6.405771006801198e-07,
+      "loss": 0.5593,
+      "step": 8100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8505005255126203,
+      "learning_rate": 6.397529592809615e-07,
+      "loss": 0.6013,
+      "step": 8101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0629170850220375,
+      "learning_rate": 6.389293121406592e-07,
+      "loss": 0.6523,
+      "step": 8102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8557315352311599,
+      "learning_rate": 6.381061593525762e-07,
+      "loss": 0.6192,
+      "step": 8103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8596773536895244,
+      "learning_rate": 6.372835010100215e-07,
+      "loss": 0.5742,
+      "step": 8104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8124890471220407,
+      "learning_rate": 6.364613372062489e-07,
+      "loss": 0.5756,
+      "step": 8105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9876410029659157,
+      "learning_rate": 6.356396680344556e-07,
+      "loss": 0.6143,
+      "step": 8106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8949001243139065,
+      "learning_rate": 6.34818493587781e-07,
+      "loss": 0.6156,
+      "step": 8107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.449001211705069,
+      "learning_rate": 6.339978139593117e-07,
+      "loss": 0.6886,
+      "step": 8108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9631147045408033,
+      "learning_rate": 6.331776292420744e-07,
+      "loss": 0.6128,
+      "step": 8109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9368516611472004,
+      "learning_rate": 6.323579395290435e-07,
+      "loss": 0.5734,
+      "step": 8110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8401589565954515,
+      "learning_rate": 6.315387449131355e-07,
+      "loss": 0.6432,
+      "step": 8111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8297334426235043,
+      "learning_rate": 6.307200454872093e-07,
+      "loss": 0.5988,
+      "step": 8112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.703991110615374,
+      "learning_rate": 6.299018413440705e-07,
+      "loss": 0.5624,
+      "step": 8113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.036173943589799,
+      "learning_rate": 6.290841325764662e-07,
+      "loss": 0.6786,
+      "step": 8114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9534205152629416,
+      "learning_rate": 6.282669192770896e-07,
+      "loss": 0.6305,
+      "step": 8115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.203379643878614,
+      "learning_rate": 6.274502015385747e-07,
+      "loss": 0.61,
+      "step": 8116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0103680975969453,
+      "learning_rate": 6.266339794535043e-07,
+      "loss": 0.6402,
+      "step": 8117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.916711741876715,
+      "learning_rate": 6.258182531143975e-07,
+      "loss": 0.56,
+      "step": 8118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.941936713481425,
+      "learning_rate": 6.250030226137249e-07,
+      "loss": 0.6702,
+      "step": 8119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0225054472897175,
+      "learning_rate": 6.241882880438949e-07,
+      "loss": 0.6005,
+      "step": 8120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8856948884191638,
+      "learning_rate": 6.233740494972651e-07,
+      "loss": 0.5536,
+      "step": 8121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9042258310577747,
+      "learning_rate": 6.225603070661318e-07,
+      "loss": 0.5368,
+      "step": 8122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.004419780870736,
+      "learning_rate": 6.217470608427395e-07,
+      "loss": 0.5644,
+      "step": 8123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9291607571745164,
+      "learning_rate": 6.209343109192728e-07,
+      "loss": 0.5955,
+      "step": 8124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.125008453554657,
+      "learning_rate": 6.201220573878613e-07,
+      "loss": 0.694,
+      "step": 8125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0979216517156836,
+      "learning_rate": 6.193103003405787e-07,
+      "loss": 0.7127,
+      "step": 8126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.134679203517581,
+      "learning_rate": 6.184990398694435e-07,
+      "loss": 0.6655,
+      "step": 8127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8835391406410493,
+      "learning_rate": 6.176882760664149e-07,
+      "loss": 0.545,
+      "step": 8128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9960800018697822,
+      "learning_rate": 6.168780090233994e-07,
+      "loss": 0.6728,
+      "step": 8129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.419923774205455,
+      "learning_rate": 6.160682388322436e-07,
+      "loss": 0.5715,
+      "step": 8130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0042116275814434,
+      "learning_rate": 6.152589655847413e-07,
+      "loss": 0.5711,
+      "step": 8131
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9220772781429358,
+      "learning_rate": 6.14450189372628e-07,
+      "loss": 0.5321,
+      "step": 8132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.015869273943935,
+      "learning_rate": 6.136419102875818e-07,
+      "loss": 0.6413,
+      "step": 8133
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7662095296346332,
+      "learning_rate": 6.128341284212258e-07,
+      "loss": 0.5262,
+      "step": 8134
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.303831477540973,
+      "learning_rate": 6.120268438651283e-07,
+      "loss": 0.7143,
+      "step": 8135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3042755907621832,
+      "learning_rate": 6.112200567107978e-07,
+      "loss": 0.7399,
+      "step": 8136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.071953461037566,
+      "learning_rate": 6.104137670496901e-07,
+      "loss": 0.5517,
+      "step": 8137
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9353300309979724,
+      "learning_rate": 6.096079749732009e-07,
+      "loss": 0.6453,
+      "step": 8138
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.986671989877575,
+      "learning_rate": 6.088026805726727e-07,
+      "loss": 0.6424,
+      "step": 8139
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9911146195693263,
+      "learning_rate": 6.079978839393896e-07,
+      "loss": 0.5474,
+      "step": 8140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0317851662272974,
+      "learning_rate": 6.071935851645794e-07,
+      "loss": 0.5149,
+      "step": 8141
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9665291558339688,
+      "learning_rate": 6.063897843394151e-07,
+      "loss": 0.6124,
+      "step": 8142
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7078468937680698,
+      "learning_rate": 6.055864815550106e-07,
+      "loss": 0.5798,
+      "step": 8143
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5752220742547929,
+      "learning_rate": 6.047836769024268e-07,
+      "loss": 0.5167,
+      "step": 8144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9074047548237727,
+      "learning_rate": 6.03981370472665e-07,
+      "loss": 0.6603,
+      "step": 8145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1307330401924562,
+      "learning_rate": 6.031795623566705e-07,
+      "loss": 0.6601,
+      "step": 8146
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.987581419382653,
+      "learning_rate": 6.023782526453347e-07,
+      "loss": 0.5877,
+      "step": 8147
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9220353371587902,
+      "learning_rate": 6.015774414294894e-07,
+      "loss": 0.5671,
+      "step": 8148
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8531354376553826,
+      "learning_rate": 6.007771287999104e-07,
+      "loss": 0.6311,
+      "step": 8149
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0038979840409037,
+      "learning_rate": 5.999773148473193e-07,
+      "loss": 0.5818,
+      "step": 8150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7687166304424737,
+      "learning_rate": 5.991779996623781e-07,
+      "loss": 0.6035,
+      "step": 8151
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.126672984023645,
+      "learning_rate": 5.983791833356955e-07,
+      "loss": 0.5741,
+      "step": 8152
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7741173112022408,
+      "learning_rate": 5.975808659578197e-07,
+      "loss": 0.5839,
+      "step": 8153
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0262142221276727,
+      "learning_rate": 5.967830476192476e-07,
+      "loss": 0.5856,
+      "step": 8154
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8676247618096367,
+      "learning_rate": 5.959857284104132e-07,
+      "loss": 0.6758,
+      "step": 8155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.059672622044571,
+      "learning_rate": 5.951889084216989e-07,
+      "loss": 0.4721,
+      "step": 8156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1535823631847846,
+      "learning_rate": 5.943925877434276e-07,
+      "loss": 0.6405,
+      "step": 8157
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1292794880898303,
+      "learning_rate": 5.935967664658682e-07,
+      "loss": 0.5985,
+      "step": 8158
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.960817142560252,
+      "learning_rate": 5.928014446792308e-07,
+      "loss": 0.5587,
+      "step": 8159
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8036953737602626,
+      "learning_rate": 5.920066224736703e-07,
+      "loss": 0.5876,
+      "step": 8160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.017494671286078,
+      "learning_rate": 5.912122999392838e-07,
+      "loss": 0.6468,
+      "step": 8161
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9745326076384333,
+      "learning_rate": 5.904184771661126e-07,
+      "loss": 0.6042,
+      "step": 8162
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.115718656081801,
+      "learning_rate": 5.896251542441395e-07,
+      "loss": 0.6424,
+      "step": 8163
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.152383828453354,
+      "learning_rate": 5.888323312632948e-07,
+      "loss": 0.5845,
+      "step": 8164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8618214723348339,
+      "learning_rate": 5.880400083134469e-07,
+      "loss": 0.5711,
+      "step": 8165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9004306913700046,
+      "learning_rate": 5.872481854844126e-07,
+      "loss": 0.6468,
+      "step": 8166
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.079916826005283,
+      "learning_rate": 5.864568628659473e-07,
+      "loss": 0.5972,
+      "step": 8167
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9150425887906342,
+      "learning_rate": 5.856660405477538e-07,
+      "loss": 0.5951,
+      "step": 8168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9768982180845163,
+      "learning_rate": 5.848757186194753e-07,
+      "loss": 0.661,
+      "step": 8169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0685273523691246,
+      "learning_rate": 5.840858971707003e-07,
+      "loss": 0.6187,
+      "step": 8170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.20740535387764,
+      "learning_rate": 5.83296576290957e-07,
+      "loss": 0.61,
+      "step": 8171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.077062544429413,
+      "learning_rate": 5.825077560697224e-07,
+      "loss": 0.6323,
+      "step": 8172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2229980382125083,
+      "learning_rate": 5.817194365964113e-07,
+      "loss": 0.6538,
+      "step": 8173
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0776525637885204,
+      "learning_rate": 5.809316179603863e-07,
+      "loss": 0.6899,
+      "step": 8174
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.060688398838161,
+      "learning_rate": 5.801443002509493e-07,
+      "loss": 0.6166,
+      "step": 8175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9882609174919457,
+      "learning_rate": 5.793574835573495e-07,
+      "loss": 0.6165,
+      "step": 8176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1539749506952197,
+      "learning_rate": 5.785711679687756e-07,
+      "loss": 0.6126,
+      "step": 8177
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.009397380360969,
+      "learning_rate": 5.777853535743605e-07,
+      "loss": 0.539,
+      "step": 8178
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.701887467794953,
+      "learning_rate": 5.770000404631815e-07,
+      "loss": 0.5223,
+      "step": 8179
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8337089423095534,
+      "learning_rate": 5.762152287242578e-07,
+      "loss": 0.523,
+      "step": 8180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.190954607822149,
+      "learning_rate": 5.754309184465534e-07,
+      "loss": 0.5291,
+      "step": 8181
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9339239124719618,
+      "learning_rate": 5.746471097189727e-07,
+      "loss": 0.6148,
+      "step": 8182
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8286911849784362,
+      "learning_rate": 5.738638026303672e-07,
+      "loss": 0.6589,
+      "step": 8183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.763185449223713,
+      "learning_rate": 5.730809972695272e-07,
+      "loss": 0.6251,
+      "step": 8184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.144097951689894,
+      "learning_rate": 5.72298693725189e-07,
+      "loss": 0.6271,
+      "step": 8185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1576327396917168,
+      "learning_rate": 5.715168920860298e-07,
+      "loss": 0.5955,
+      "step": 8186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2777450538296677,
+      "learning_rate": 5.707355924406738e-07,
+      "loss": 0.7751,
+      "step": 8187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2047094872367006,
+      "learning_rate": 5.699547948776829e-07,
+      "loss": 0.6282,
+      "step": 8188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9091741539508933,
+      "learning_rate": 5.691744994855675e-07,
+      "loss": 0.6249,
+      "step": 8189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.021439937726755,
+      "learning_rate": 5.683947063527762e-07,
+      "loss": 0.565,
+      "step": 8190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.729258043483863,
+      "learning_rate": 5.676154155677066e-07,
+      "loss": 0.5358,
+      "step": 8191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.929628451697072,
+      "learning_rate": 5.668366272186915e-07,
+      "loss": 0.5293,
+      "step": 8192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8951688227059889,
+      "learning_rate": 5.660583413940135e-07,
+      "loss": 0.6399,
+      "step": 8193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0138368029617912,
+      "learning_rate": 5.652805581818943e-07,
+      "loss": 0.5991,
+      "step": 8194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9355511215261672,
+      "learning_rate": 5.645032776705023e-07,
+      "loss": 0.5545,
+      "step": 8195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.068165286932841,
+      "learning_rate": 5.637264999479436e-07,
+      "loss": 0.6252,
+      "step": 8196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8040737962816091,
+      "learning_rate": 5.629502251022734e-07,
+      "loss": 0.6947,
+      "step": 8197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1167954493638685,
+      "learning_rate": 5.621744532214856e-07,
+      "loss": 0.6703,
+      "step": 8198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8845230887915438,
+      "learning_rate": 5.613991843935179e-07,
+      "loss": 0.6548,
+      "step": 8199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8029172725787435,
+      "learning_rate": 5.606244187062509e-07,
+      "loss": 0.6001,
+      "step": 8200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9588614842084962,
+      "learning_rate": 5.598501562475111e-07,
+      "loss": 0.5883,
+      "step": 8201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7162881782665964,
+      "learning_rate": 5.590763971050628e-07,
+      "loss": 0.5193,
+      "step": 8202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.048313840022909,
+      "learning_rate": 5.583031413666185e-07,
+      "loss": 0.6895,
+      "step": 8203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9457017560438867,
+      "learning_rate": 5.575303891198286e-07,
+      "loss": 0.5658,
+      "step": 8204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.880442705012502,
+      "learning_rate": 5.567581404522914e-07,
+      "loss": 0.5756,
+      "step": 8205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8335074588378506,
+      "learning_rate": 5.559863954515448e-07,
+      "loss": 0.554,
+      "step": 8206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.957864056813018,
+      "learning_rate": 5.552151542050699e-07,
+      "loss": 0.7151,
+      "step": 8207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0410162015627153,
+      "learning_rate": 5.54444416800291e-07,
+      "loss": 0.6489,
+      "step": 8208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.966686920765713,
+      "learning_rate": 5.536741833245773e-07,
+      "loss": 0.5739,
+      "step": 8209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.097614872666767,
+      "learning_rate": 5.529044538652373e-07,
+      "loss": 0.6079,
+      "step": 8210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8126670428732163,
+      "learning_rate": 5.521352285095261e-07,
+      "loss": 0.5025,
+      "step": 8211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9132465488593897,
+      "learning_rate": 5.513665073446372e-07,
+      "loss": 0.6297,
+      "step": 8212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9350220871821595,
+      "learning_rate": 5.505982904577123e-07,
+      "loss": 0.615,
+      "step": 8213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1347049899077284,
+      "learning_rate": 5.49830577935832e-07,
+      "loss": 0.5208,
+      "step": 8214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8637329068323893,
+      "learning_rate": 5.490633698660197e-07,
+      "loss": 0.5808,
+      "step": 8215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.948838872573589,
+      "learning_rate": 5.482966663352451e-07,
+      "loss": 0.621,
+      "step": 8216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0493579656488325,
+      "learning_rate": 5.47530467430416e-07,
+      "loss": 0.622,
+      "step": 8217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.918223916486134,
+      "learning_rate": 5.467647732383879e-07,
+      "loss": 0.608,
+      "step": 8218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7803045728773015,
+      "learning_rate": 5.459995838459542e-07,
+      "loss": 0.5959,
+      "step": 8219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9266062790106375,
+      "learning_rate": 5.452348993398566e-07,
+      "loss": 0.5599,
+      "step": 8220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7707431361868795,
+      "learning_rate": 5.444707198067722e-07,
+      "loss": 0.6657,
+      "step": 8221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8932789462090363,
+      "learning_rate": 5.437070453333288e-07,
+      "loss": 0.6384,
+      "step": 8222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.236832426733744,
+      "learning_rate": 5.429438760060906e-07,
+      "loss": 0.5938,
+      "step": 8223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9219044021119076,
+      "learning_rate": 5.421812119115699e-07,
+      "loss": 0.6044,
+      "step": 8224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2466433945686757,
+      "learning_rate": 5.414190531362162e-07,
+      "loss": 0.6077,
+      "step": 8225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8069595964244884,
+      "learning_rate": 5.406573997664267e-07,
+      "loss": 0.5446,
+      "step": 8226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8885534423348074,
+      "learning_rate": 5.398962518885375e-07,
+      "loss": 0.5408,
+      "step": 8227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.093745092026466,
+      "learning_rate": 5.391356095888323e-07,
+      "loss": 0.6371,
+      "step": 8228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9475019077484441,
+      "learning_rate": 5.38375472953529e-07,
+      "loss": 0.6218,
+      "step": 8229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.050764295266713,
+      "learning_rate": 5.376158420687977e-07,
+      "loss": 0.7155,
+      "step": 8230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.039172342224332,
+      "learning_rate": 5.368567170207445e-07,
+      "loss": 0.6407,
+      "step": 8231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9356989051958875,
+      "learning_rate": 5.360980978954223e-07,
+      "loss": 0.5751,
+      "step": 8232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.908719869422782,
+      "learning_rate": 5.353399847788233e-07,
+      "loss": 0.6832,
+      "step": 8233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9156768167886171,
+      "learning_rate": 5.345823777568859e-07,
+      "loss": 0.6937,
+      "step": 8234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.094528279582242,
+      "learning_rate": 5.338252769154878e-07,
+      "loss": 0.5975,
+      "step": 8235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9826801963733196,
+      "learning_rate": 5.330686823404507e-07,
+      "loss": 0.5109,
+      "step": 8236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1642191778237656,
+      "learning_rate": 5.323125941175383e-07,
+      "loss": 0.6296,
+      "step": 8237
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1050313683713204,
+      "learning_rate": 5.315570123324593e-07,
+      "loss": 0.5841,
+      "step": 8238
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.86255900999276,
+      "learning_rate": 5.308019370708612e-07,
+      "loss": 0.5258,
+      "step": 8239
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0814563238442214,
+      "learning_rate": 5.300473684183382e-07,
+      "loss": 0.6346,
+      "step": 8240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.004366161593308,
+      "learning_rate": 5.292933064604228e-07,
+      "loss": 0.6448,
+      "step": 8241
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0814251969657196,
+      "learning_rate": 5.28539751282594e-07,
+      "loss": 0.7272,
+      "step": 8242
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8919312569321771,
+      "learning_rate": 5.277867029702716e-07,
+      "loss": 0.5276,
+      "step": 8243
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9044934817666754,
+      "learning_rate": 5.270341616088153e-07,
+      "loss": 0.5631,
+      "step": 8244
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8961844755275268,
+      "learning_rate": 5.262821272835334e-07,
+      "loss": 0.5554,
+      "step": 8245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1041404998731177,
+      "learning_rate": 5.255306000796717e-07,
+      "loss": 0.5704,
+      "step": 8246
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.039862049268198,
+      "learning_rate": 5.24779580082419e-07,
+      "loss": 0.5524,
+      "step": 8247
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8011189836643542,
+      "learning_rate": 5.240290673769099e-07,
+      "loss": 0.5643,
+      "step": 8248
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7674719230033473,
+      "learning_rate": 5.23279062048217e-07,
+      "loss": 0.5861,
+      "step": 8249
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6875423096614102,
+      "learning_rate": 5.225295641813599e-07,
+      "loss": 0.5266,
+      "step": 8250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1684829276074384,
+      "learning_rate": 5.217805738612975e-07,
+      "loss": 0.563,
+      "step": 8251
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.008210226253471,
+      "learning_rate": 5.210320911729311e-07,
+      "loss": 0.5562,
+      "step": 8252
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.146514487983114,
+      "learning_rate": 5.202841162011074e-07,
+      "loss": 0.6285,
+      "step": 8253
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.070368443828562,
+      "learning_rate": 5.195366490306114e-07,
+      "loss": 0.5777,
+      "step": 8254
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9585353530387317,
+      "learning_rate": 5.187896897461752e-07,
+      "loss": 0.5651,
+      "step": 8255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.179156757136343,
+      "learning_rate": 5.180432384324691e-07,
+      "loss": 0.6963,
+      "step": 8256
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9217899295161627,
+      "learning_rate": 5.172972951741096e-07,
+      "loss": 0.5194,
+      "step": 8257
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7949492707964125,
+      "learning_rate": 5.165518600556507e-07,
+      "loss": 0.4705,
+      "step": 8258
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.089497265943696,
+      "learning_rate": 5.158069331615939e-07,
+      "loss": 0.6488,
+      "step": 8259
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.925905784209994,
+      "learning_rate": 5.150625145763794e-07,
+      "loss": 0.523,
+      "step": 8260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8012988786439579,
+      "learning_rate": 5.143186043843934e-07,
+      "loss": 0.5625,
+      "step": 8261
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.073140541720193,
+      "learning_rate": 5.135752026699597e-07,
+      "loss": 0.776,
+      "step": 8262
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0938632113238547,
+      "learning_rate": 5.128323095173498e-07,
+      "loss": 0.6787,
+      "step": 8263
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8505470192539382,
+      "learning_rate": 5.12089925010773e-07,
+      "loss": 0.6027,
+      "step": 8264
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9385142627925132,
+      "learning_rate": 5.113480492343847e-07,
+      "loss": 0.5588,
+      "step": 8265
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.375733866717919,
+      "learning_rate": 5.106066822722782e-07,
+      "loss": 0.66,
+      "step": 8266
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9370422187030516,
+      "learning_rate": 5.098658242084937e-07,
+      "loss": 0.6152,
+      "step": 8267
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9627422041290383,
+      "learning_rate": 5.091254751270097e-07,
+      "loss": 0.6244,
+      "step": 8268
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9673460760767303,
+      "learning_rate": 5.083856351117511e-07,
+      "loss": 0.6486,
+      "step": 8269
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9677413688817145,
+      "learning_rate": 5.076463042465812e-07,
+      "loss": 0.6466,
+      "step": 8270
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9076042255087213,
+      "learning_rate": 5.069074826153097e-07,
+      "loss": 0.7091,
+      "step": 8271
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1522781967814955,
+      "learning_rate": 5.061691703016841e-07,
+      "loss": 0.5958,
+      "step": 8272
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9512598833988453,
+      "learning_rate": 5.054313673893979e-07,
+      "loss": 0.6494,
+      "step": 8273
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8889340369572525,
+      "learning_rate": 5.046940739620826e-07,
+      "loss": 0.5436,
+      "step": 8274
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1089818787586467,
+      "learning_rate": 5.039572901033179e-07,
+      "loss": 0.6802,
+      "step": 8275
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9862877189958716,
+      "learning_rate": 5.0322101589662e-07,
+      "loss": 0.6129,
+      "step": 8276
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8141324404301526,
+      "learning_rate": 5.024852514254513e-07,
+      "loss": 0.5729,
+      "step": 8277
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0791012063858805,
+      "learning_rate": 5.017499967732137e-07,
+      "loss": 0.6168,
+      "step": 8278
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2094091772410938,
+      "learning_rate": 5.010152520232536e-07,
+      "loss": 0.6785,
+      "step": 8279
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0704609733536175,
+      "learning_rate": 5.002810172588584e-07,
+      "loss": 0.6584,
+      "step": 8280
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9853255752421886,
+      "learning_rate": 4.995472925632567e-07,
+      "loss": 0.6358,
+      "step": 8281
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9531326047376758,
+      "learning_rate": 4.988140780196221e-07,
+      "loss": 0.652,
+      "step": 8282
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.057213242161284,
+      "learning_rate": 4.980813737110662e-07,
+      "loss": 0.601,
+      "step": 8283
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.105331445169244,
+      "learning_rate": 4.973491797206481e-07,
+      "loss": 0.5704,
+      "step": 8284
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9364053789589217,
+      "learning_rate": 4.966174961313646e-07,
+      "loss": 0.6009,
+      "step": 8285
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0882062233429544,
+      "learning_rate": 4.958863230261551e-07,
+      "loss": 0.5458,
+      "step": 8286
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9799264378167545,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.582,
+      "step": 8287
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.021495582372774,
+      "learning_rate": 4.94425508599437e-07,
+      "loss": 0.5471,
+      "step": 8288
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8753911264845755,
+      "learning_rate": 4.936958674435178e-07,
+      "loss": 0.6074,
+      "step": 8289
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.066023846671249,
+      "learning_rate": 4.929667371028579e-07,
+      "loss": 0.6597,
+      "step": 8290
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.150677404952387,
+      "learning_rate": 4.922381176601066e-07,
+      "loss": 0.5113,
+      "step": 8291
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.054898643812894,
+      "learning_rate": 4.915100091978591e-07,
+      "loss": 0.6124,
+      "step": 8292
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6742532360960376,
+      "learning_rate": 4.907824117986487e-07,
+      "loss": 0.5931,
+      "step": 8293
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8695315925969587,
+      "learning_rate": 4.900553255449553e-07,
+      "loss": 0.5876,
+      "step": 8294
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8359415649511754,
+      "learning_rate": 4.893287505191946e-07,
+      "loss": 0.5155,
+      "step": 8295
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.06769544820391,
+      "learning_rate": 4.886026868037313e-07,
+      "loss": 0.5753,
+      "step": 8296
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0138686917169335,
+      "learning_rate": 4.878771344808664e-07,
+      "loss": 0.5813,
+      "step": 8297
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9528925103015966,
+      "learning_rate": 4.871520936328478e-07,
+      "loss": 0.5845,
+      "step": 8298
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8439522634772305,
+      "learning_rate": 4.864275643418603e-07,
+      "loss": 0.502,
+      "step": 8299
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2204098320288224,
+      "learning_rate": 4.857035466900361e-07,
+      "loss": 0.6992,
+      "step": 8300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3494412708615866,
+      "learning_rate": 4.849800407594446e-07,
+      "loss": 0.5811,
+      "step": 8301
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3615193836986235,
+      "learning_rate": 4.842570466321023e-07,
+      "loss": 0.6805,
+      "step": 8302
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9959876511710395,
+      "learning_rate": 4.835345643899609e-07,
+      "loss": 0.6178,
+      "step": 8303
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1046472978265878,
+      "learning_rate": 4.828125941149197e-07,
+      "loss": 0.524,
+      "step": 8304
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8352720574795676,
+      "learning_rate": 4.820911358888181e-07,
+      "loss": 0.6001,
+      "step": 8305
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7599800381591997,
+      "learning_rate": 4.813701897934375e-07,
+      "loss": 0.5603,
+      "step": 8306
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.268456220684496,
+      "learning_rate": 4.806497559105011e-07,
+      "loss": 0.5543,
+      "step": 8307
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8005414284190293,
+      "learning_rate": 4.799298343216746e-07,
+      "loss": 0.6011,
+      "step": 8308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.925175771444624,
+      "learning_rate": 4.792104251085655e-07,
+      "loss": 0.5503,
+      "step": 8309
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3450122517710894,
+      "learning_rate": 4.784915283527219e-07,
+      "loss": 0.6776,
+      "step": 8310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1628264585537806,
+      "learning_rate": 4.777731441356342e-07,
+      "loss": 0.5846,
+      "step": 8311
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6450519414376317,
+      "learning_rate": 4.770552725387378e-07,
+      "loss": 0.5043,
+      "step": 8312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7482690149775968,
+      "learning_rate": 4.763379136434054e-07,
+      "loss": 0.6179,
+      "step": 8313
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9801836667225938,
+      "learning_rate": 4.7562106753095527e-07,
+      "loss": 0.5958,
+      "step": 8314
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9733287042306242,
+      "learning_rate": 4.7490473428264406e-07,
+      "loss": 0.7162,
+      "step": 8315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8725031647315058,
+      "learning_rate": 4.741889139796746e-07,
+      "loss": 0.5833,
+      "step": 8316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.799784467033496,
+      "learning_rate": 4.7347360670318756e-07,
+      "loss": 0.6006,
+      "step": 8317
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9366110237287402,
+      "learning_rate": 4.727588125342669e-07,
+      "loss": 0.5255,
+      "step": 8318
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.234824200192645,
+      "learning_rate": 4.7204453155394013e-07,
+      "loss": 0.6816,
+      "step": 8319
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5122506337921116,
+      "learning_rate": 4.7133076384317354e-07,
+      "loss": 0.6069,
+      "step": 8320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.119122687155672,
+      "learning_rate": 4.70617509482878e-07,
+      "loss": 0.6328,
+      "step": 8321
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.004192201331832,
+      "learning_rate": 4.699047685539038e-07,
+      "loss": 0.5006,
+      "step": 8322
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.916590258992686,
+      "learning_rate": 4.6919254113704515e-07,
+      "loss": 0.5811,
+      "step": 8323
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6538686628596007,
+      "learning_rate": 4.68480827313037e-07,
+      "loss": 0.5711,
+      "step": 8324
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6500225555034003,
+      "learning_rate": 4.6776962716255593e-07,
+      "loss": 0.5336,
+      "step": 8325
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9802573141561102,
+      "learning_rate": 4.670589407662196e-07,
+      "loss": 0.5819,
+      "step": 8326
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0088327902438525,
+      "learning_rate": 4.663487682045903e-07,
+      "loss": 0.5182,
+      "step": 8327
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.017583462703393,
+      "learning_rate": 4.656391095581675e-07,
+      "loss": 0.6915,
+      "step": 8328
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1208095680699808,
+      "learning_rate": 4.6492996490739796e-07,
+      "loss": 0.6893,
+      "step": 8329
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9534148806244596,
+      "learning_rate": 4.6422133433266513e-07,
+      "loss": 0.5895,
+      "step": 8330
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8000895668108274,
+      "learning_rate": 4.6351321791429924e-07,
+      "loss": 0.5725,
+      "step": 8331
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9340943599774802,
+      "learning_rate": 4.62805615732565e-07,
+      "loss": 0.6447,
+      "step": 8332
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8956460910814394,
+      "learning_rate": 4.6209852786767593e-07,
+      "loss": 0.5395,
+      "step": 8333
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.709032151810743,
+      "learning_rate": 4.613919543997836e-07,
+      "loss": 0.5488,
+      "step": 8334
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.113498349240345,
+      "learning_rate": 4.606858954089827e-07,
+      "loss": 0.6202,
+      "step": 8335
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.034226927445555,
+      "learning_rate": 4.599803509753081e-07,
+      "loss": 0.6481,
+      "step": 8336
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.853159457928419,
+      "learning_rate": 4.592753211787393e-07,
+      "loss": 0.6445,
+      "step": 8337
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1242559862625416,
+      "learning_rate": 4.585708060991928e-07,
+      "loss": 0.6081,
+      "step": 8338
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2173927759277388,
+      "learning_rate": 4.578668058165325e-07,
+      "loss": 0.6526,
+      "step": 8339
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9183380234550136,
+      "learning_rate": 4.571633204105574e-07,
+      "loss": 0.5716,
+      "step": 8340
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.852559147624212,
+      "learning_rate": 4.564603499610143e-07,
+      "loss": 0.6098,
+      "step": 8341
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9970457306785048,
+      "learning_rate": 4.5575789454758656e-07,
+      "loss": 0.6656,
+      "step": 8342
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.102913775343231,
+      "learning_rate": 4.5505595424990446e-07,
+      "loss": 0.6814,
+      "step": 8343
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1924316214071427,
+      "learning_rate": 4.5435452914753377e-07,
+      "loss": 0.5918,
+      "step": 8344
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.16432717318907,
+      "learning_rate": 4.5365361931998696e-07,
+      "loss": 0.6633,
+      "step": 8345
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0535418697441377,
+      "learning_rate": 4.5295322484671667e-07,
+      "loss": 0.5843,
+      "step": 8346
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9299352635441742,
+      "learning_rate": 4.522533458071149e-07,
+      "loss": 0.6031,
+      "step": 8347
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8036773453246178,
+      "learning_rate": 4.5155398228051707e-07,
+      "loss": 0.5917,
+      "step": 8348
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9397877871030635,
+      "learning_rate": 4.508551343462014e-07,
+      "loss": 0.6233,
+      "step": 8349
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.252675458486445,
+      "learning_rate": 4.501568020833846e-07,
+      "loss": 0.5963,
+      "step": 8350
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.04584672503017,
+      "learning_rate": 4.4945898557122893e-07,
+      "loss": 0.592,
+      "step": 8351
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1792705337695466,
+      "learning_rate": 4.4876168488883267e-07,
+      "loss": 0.6744,
+      "step": 8352
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.852796108199507,
+      "learning_rate": 4.4806490011524205e-07,
+      "loss": 0.5102,
+      "step": 8353
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5734956707152514,
+      "learning_rate": 4.473686313294401e-07,
+      "loss": 0.5494,
+      "step": 8354
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7716365972878156,
+      "learning_rate": 4.466728786103519e-07,
+      "loss": 0.5574,
+      "step": 8355
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9458084993387372,
+      "learning_rate": 4.4597764203684725e-07,
+      "loss": 0.5328,
+      "step": 8356
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.197405580309809,
+      "learning_rate": 4.4528292168773303e-07,
+      "loss": 0.6607,
+      "step": 8357
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0426609447432154,
+      "learning_rate": 4.445887176417613e-07,
+      "loss": 0.6092,
+      "step": 8358
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9637826343530485,
+      "learning_rate": 4.4389502997762236e-07,
+      "loss": 0.5815,
+      "step": 8359
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9580038792141665,
+      "learning_rate": 4.432018587739517e-07,
+      "loss": 0.6124,
+      "step": 8360
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8027490658262888,
+      "learning_rate": 4.425092041093237e-07,
+      "loss": 0.5564,
+      "step": 8361
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.890109404627932,
+      "learning_rate": 4.418170660622539e-07,
+      "loss": 0.6641,
+      "step": 8362
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9496169997143404,
+      "learning_rate": 4.4112544471119954e-07,
+      "loss": 0.6565,
+      "step": 8363
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1410985482384,
+      "learning_rate": 4.404343401345612e-07,
+      "loss": 0.7114,
+      "step": 8364
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7444177282452187,
+      "learning_rate": 4.39743752410679e-07,
+      "loss": 0.6074,
+      "step": 8365
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9722413049276115,
+      "learning_rate": 4.390536816178353e-07,
+      "loss": 0.6795,
+      "step": 8366
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8931068079905056,
+      "learning_rate": 4.3836412783425265e-07,
+      "loss": 0.6789,
+      "step": 8367
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.956276084763643,
+      "learning_rate": 4.3767509113809836e-07,
+      "loss": 0.6538,
+      "step": 8368
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7819181415754366,
+      "learning_rate": 4.3698657160747504e-07,
+      "loss": 0.6105,
+      "step": 8369
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.852985794251653,
+      "learning_rate": 4.36298569320433e-07,
+      "loss": 0.4922,
+      "step": 8370
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.074863975049127,
+      "learning_rate": 4.3561108435495936e-07,
+      "loss": 0.5791,
+      "step": 8371
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8898011683002591,
+      "learning_rate": 4.349241167889867e-07,
+      "loss": 0.6205,
+      "step": 8372
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.859913410648297,
+      "learning_rate": 4.342376667003845e-07,
+      "loss": 0.6641,
+      "step": 8373
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.835582843864355,
+      "learning_rate": 4.335517341669676e-07,
+      "loss": 0.5306,
+      "step": 8374
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1789774219199276,
+      "learning_rate": 4.3286631926648834e-07,
+      "loss": 0.6437,
+      "step": 8375
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1690881296258766,
+      "learning_rate": 4.321814220766457e-07,
+      "loss": 0.5847,
+      "step": 8376
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9311374676014503,
+      "learning_rate": 4.3149704267507254e-07,
+      "loss": 0.5837,
+      "step": 8377
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8601406230446484,
+      "learning_rate": 4.3081318113935013e-07,
+      "loss": 0.5966,
+      "step": 8378
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9117613858119065,
+      "learning_rate": 4.3012983754699645e-07,
+      "loss": 0.5844,
+      "step": 8379
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1005104659206735,
+      "learning_rate": 4.29447011975474e-07,
+      "loss": 0.5632,
+      "step": 8380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8160285873217523,
+      "learning_rate": 4.2876470450218254e-07,
+      "loss": 0.5465,
+      "step": 8381
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8764970657380244,
+      "learning_rate": 4.2808291520446856e-07,
+      "loss": 0.6718,
+      "step": 8382
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9302392899798297,
+      "learning_rate": 4.274016441596146e-07,
+      "loss": 0.6263,
+      "step": 8383
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.380086794220247,
+      "learning_rate": 4.267208914448467e-07,
+      "loss": 0.6717,
+      "step": 8384
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8467666239151335,
+      "learning_rate": 4.2604065713733376e-07,
+      "loss": 0.4995,
+      "step": 8385
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1309058059197357,
+      "learning_rate": 4.253609413141824e-07,
+      "loss": 0.6071,
+      "step": 8386
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.056063519146599,
+      "learning_rate": 4.2468174405244255e-07,
+      "loss": 0.6373,
+      "step": 8387
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.997005731740509,
+      "learning_rate": 4.240030654291061e-07,
+      "loss": 0.6215,
+      "step": 8388
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8805146093287572,
+      "learning_rate": 4.2332490552110363e-07,
+      "loss": 0.6959,
+      "step": 8389
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9311754319444367,
+      "learning_rate": 4.2264726440531036e-07,
+      "loss": 0.6109,
+      "step": 8390
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9639047760658772,
+      "learning_rate": 4.2197014215853926e-07,
+      "loss": 0.6419,
+      "step": 8391
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9795175845484239,
+      "learning_rate": 4.2129353885754564e-07,
+      "loss": 0.6291,
+      "step": 8392
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.249186746122543,
+      "learning_rate": 4.206174545790281e-07,
+      "loss": 0.5673,
+      "step": 8393
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.253664003461183,
+      "learning_rate": 4.199418893996232e-07,
+      "loss": 0.6916,
+      "step": 8394
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1217256531704485,
+      "learning_rate": 4.192668433959113e-07,
+      "loss": 0.6001,
+      "step": 8395
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9901519915823938,
+      "learning_rate": 4.1859231664441115e-07,
+      "loss": 0.5294,
+      "step": 8396
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9997277010717542,
+      "learning_rate": 4.179183092215855e-07,
+      "loss": 0.7411,
+      "step": 8397
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.165446184188909,
+      "learning_rate": 4.172448212038371e-07,
+      "loss": 0.6355,
+      "step": 8398
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.11255812027217,
+      "learning_rate": 4.165718526675083e-07,
+      "loss": 0.5731,
+      "step": 8399
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.919796366514845,
+      "learning_rate": 4.158994036888847e-07,
+      "loss": 0.5309,
+      "step": 8400
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.390639655248388,
+      "learning_rate": 4.152274743441925e-07,
+      "loss": 0.5966,
+      "step": 8401
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.738681794830981,
+      "learning_rate": 4.1455606470959755e-07,
+      "loss": 0.4943,
+      "step": 8402
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2208990337370897,
+      "learning_rate": 4.138851748612099e-07,
+      "loss": 0.6778,
+      "step": 8403
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0630590778482705,
+      "learning_rate": 4.132148048750767e-07,
+      "loss": 0.5795,
+      "step": 8404
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9018986833626161,
+      "learning_rate": 4.125449548271909e-07,
+      "loss": 0.5599,
+      "step": 8405
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0210569711520066,
+      "learning_rate": 4.118756247934802e-07,
+      "loss": 0.6431,
+      "step": 8406
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0602854396292396,
+      "learning_rate": 4.112068148498199e-07,
+      "loss": 0.7268,
+      "step": 8407
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9460298623096413,
+      "learning_rate": 4.1053852507202117e-07,
+      "loss": 0.6629,
+      "step": 8408
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7595563936601968,
+      "learning_rate": 4.098707555358411e-07,
+      "loss": 0.5799,
+      "step": 8409
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0271997118305483,
+      "learning_rate": 4.092035063169725e-07,
+      "loss": 0.6046,
+      "step": 8410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0289865415330928,
+      "learning_rate": 4.0853677749105426e-07,
+      "loss": 0.5695,
+      "step": 8411
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9559717154468896,
+      "learning_rate": 4.078705691336621e-07,
+      "loss": 0.6128,
+      "step": 8412
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8769271149503086,
+      "learning_rate": 4.072048813203161e-07,
+      "loss": 0.6361,
+      "step": 8413
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0028628194390103,
+      "learning_rate": 4.065397141264737e-07,
+      "loss": 0.6202,
+      "step": 8414
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9916985701401626,
+      "learning_rate": 4.0587506762753747e-07,
+      "loss": 0.6754,
+      "step": 8415
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9127873478665742,
+      "learning_rate": 4.05210941898847e-07,
+      "loss": 0.5592,
+      "step": 8416
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0282432551440217,
+      "learning_rate": 4.045473370156866e-07,
+      "loss": 0.7055,
+      "step": 8417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8459979565416385,
+      "learning_rate": 4.038842530532777e-07,
+      "loss": 0.5877,
+      "step": 8418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8860989137846904,
+      "learning_rate": 4.032216900867869e-07,
+      "loss": 0.5879,
+      "step": 8419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.303077578313458,
+      "learning_rate": 4.02559648191318e-07,
+      "loss": 0.6251,
+      "step": 8420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0308831657626274,
+      "learning_rate": 4.0189812744191647e-07,
+      "loss": 0.6051,
+      "step": 8421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.784559200096564,
+      "learning_rate": 4.0123712791357185e-07,
+      "loss": 0.5577,
+      "step": 8422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0297559541852874,
+      "learning_rate": 4.005766496812097e-07,
+      "loss": 0.6529,
+      "step": 8423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9111708728025478,
+      "learning_rate": 3.999166928197007e-07,
+      "loss": 0.5998,
+      "step": 8424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.132095921466122,
+      "learning_rate": 3.9925725740385503e-07,
+      "loss": 0.6852,
+      "step": 8425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8697129123540708,
+      "learning_rate": 3.985983435084212e-07,
+      "loss": 0.5694,
+      "step": 8426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9320332116706223,
+      "learning_rate": 3.979399512080928e-07,
+      "loss": 0.6194,
+      "step": 8427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7865258361708811,
+      "learning_rate": 3.972820805775024e-07,
+      "loss": 0.5196,
+      "step": 8428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9635188121091862,
+      "learning_rate": 3.966247316912214e-07,
+      "loss": 0.6628,
+      "step": 8429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.052755454266565,
+      "learning_rate": 3.959679046237663e-07,
+      "loss": 0.5529,
+      "step": 8430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2660073623651082,
+      "learning_rate": 3.9531159944959094e-07,
+      "loss": 0.5532,
+      "step": 8431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1619277389271128,
+      "learning_rate": 3.9465581624309245e-07,
+      "loss": 0.6274,
+      "step": 8432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9488850141266936,
+      "learning_rate": 3.9400055507860525e-07,
+      "loss": 0.5921,
+      "step": 8433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9310958922073735,
+      "learning_rate": 3.933458160304099e-07,
+      "loss": 0.5115,
+      "step": 8434
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7377699202843322,
+      "learning_rate": 3.9269159917272313e-07,
+      "loss": 0.5034,
+      "step": 8435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.020931667657547,
+      "learning_rate": 3.920379045797046e-07,
+      "loss": 0.6726,
+      "step": 8436
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.045919126231745,
+      "learning_rate": 3.9138473232545326e-07,
+      "loss": 0.6392,
+      "step": 8437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9556100117349695,
+      "learning_rate": 3.907320824840111e-07,
+      "loss": 0.6564,
+      "step": 8438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.830463491913251,
+      "learning_rate": 3.900799551293588e-07,
+      "loss": 0.5916,
+      "step": 8439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.117119952335453,
+      "learning_rate": 3.8942835033542013e-07,
+      "loss": 0.6425,
+      "step": 8440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.351242053992765,
+      "learning_rate": 3.8877726817605655e-07,
+      "loss": 0.6671,
+      "step": 8441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9606052679245647,
+      "learning_rate": 3.8812670872507454e-07,
+      "loss": 0.572,
+      "step": 8442
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.001344741894956,
+      "learning_rate": 3.874766720562151e-07,
+      "loss": 0.6716,
+      "step": 8443
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8404479443713242,
+      "learning_rate": 3.8682715824316594e-07,
+      "loss": 0.588,
+      "step": 8444
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9912689065492977,
+      "learning_rate": 3.861781673595516e-07,
+      "loss": 0.5291,
+      "step": 8445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8709075400500266,
+      "learning_rate": 3.855296994789415e-07,
+      "loss": 0.6012,
+      "step": 8446
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1625558589548106,
+      "learning_rate": 3.8488175467484015e-07,
+      "loss": 0.5877,
+      "step": 8447
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0313618701656835,
+      "learning_rate": 3.8423433302069824e-07,
+      "loss": 0.6104,
+      "step": 8448
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.76595693209394,
+      "learning_rate": 3.8358743458990267e-07,
+      "loss": 0.5309,
+      "step": 8449
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9556132619540931,
+      "learning_rate": 3.829410594557859e-07,
+      "loss": 0.5177,
+      "step": 8450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9424701856527413,
+      "learning_rate": 3.8229520769161474e-07,
+      "loss": 0.5511,
+      "step": 8451
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0982651348249606,
+      "learning_rate": 3.816498793706025e-07,
+      "loss": 0.6457,
+      "step": 8452
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1999610052684497,
+      "learning_rate": 3.810050745658994e-07,
+      "loss": 0.7338,
+      "step": 8453
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9720801346355596,
+      "learning_rate": 3.803607933505993e-07,
+      "loss": 0.5504,
+      "step": 8454
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9568723610962957,
+      "learning_rate": 3.797170357977337e-07,
+      "loss": 0.633,
+      "step": 8455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9100529824005883,
+      "learning_rate": 3.790738019802775e-07,
+      "loss": 0.6134,
+      "step": 8456
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.997788576607131,
+      "learning_rate": 3.7843109197114426e-07,
+      "loss": 0.7004,
+      "step": 8457
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0031467913611256,
+      "learning_rate": 3.7778890584318773e-07,
+      "loss": 0.6788,
+      "step": 8458
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9267261626502268,
+      "learning_rate": 3.771472436692053e-07,
+      "loss": 0.6607,
+      "step": 8459
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0280693159384757,
+      "learning_rate": 3.7650610552193157e-07,
+      "loss": 0.6408,
+      "step": 8460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.047634744310128,
+      "learning_rate": 3.758654914740445e-07,
+      "loss": 0.5969,
+      "step": 8461
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1014846487101164,
+      "learning_rate": 3.752254015981599e-07,
+      "loss": 0.6298,
+      "step": 8462
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9895447225484795,
+      "learning_rate": 3.745858359668375e-07,
+      "loss": 0.6054,
+      "step": 8463
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.90717225882143,
+      "learning_rate": 3.7394679465257435e-07,
+      "loss": 0.5842,
+      "step": 8464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.051788494418991,
+      "learning_rate": 3.7330827772780967e-07,
+      "loss": 0.5711,
+      "step": 8465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9343377474924703,
+      "learning_rate": 3.726702852649222e-07,
+      "loss": 0.5839,
+      "step": 8466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3496780116095097,
+      "learning_rate": 3.720328173362331e-07,
+      "loss": 0.6441,
+      "step": 8467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.050871079546467,
+      "learning_rate": 3.713958740140022e-07,
+      "loss": 0.6083,
+      "step": 8468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9364655434741551,
+      "learning_rate": 3.7075945537043244e-07,
+      "loss": 0.5007,
+      "step": 8469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1012266768594476,
+      "learning_rate": 3.701235614776627e-07,
+      "loss": 0.7012,
+      "step": 8470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0691007203153036,
+      "learning_rate": 3.6948819240777756e-07,
+      "loss": 0.5571,
+      "step": 8471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7160709095761875,
+      "learning_rate": 3.688533482327994e-07,
+      "loss": 0.5777,
+      "step": 8472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8206494989587712,
+      "learning_rate": 3.6821902902469066e-07,
+      "loss": 0.525,
+      "step": 8473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.711893505353368,
+      "learning_rate": 3.675852348553538e-07,
+      "loss": 0.5737,
+      "step": 8474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.033288025220472,
+      "learning_rate": 3.6695196579663583e-07,
+      "loss": 0.534,
+      "step": 8475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0240144326187415,
+      "learning_rate": 3.6631922192031934e-07,
+      "loss": 0.7105,
+      "step": 8476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2155007112046166,
+      "learning_rate": 3.656870032981308e-07,
+      "loss": 0.627,
+      "step": 8477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.021651389737251,
+      "learning_rate": 3.65055310001734e-07,
+      "loss": 0.6434,
+      "step": 8478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.350132802250893,
+      "learning_rate": 3.6442414210273834e-07,
+      "loss": 0.6545,
+      "step": 8479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.043985363928126,
+      "learning_rate": 3.637934996726861e-07,
+      "loss": 0.6513,
+      "step": 8480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.817926490943509,
+      "learning_rate": 3.6316338278306715e-07,
+      "loss": 0.621,
+      "step": 8481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1668932428840284,
+      "learning_rate": 3.6253379150530676e-07,
+      "loss": 0.6419,
+      "step": 8482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0924027965702496,
+      "learning_rate": 3.619047259107744e-07,
+      "loss": 0.7059,
+      "step": 8483
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.973897319602242,
+      "learning_rate": 3.6127618607077754e-07,
+      "loss": 0.6232,
+      "step": 8484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1019528532461513,
+      "learning_rate": 3.606481720565652e-07,
+      "loss": 0.6515,
+      "step": 8485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8078476936519057,
+      "learning_rate": 3.600206839393261e-07,
+      "loss": 0.5487,
+      "step": 8486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6102160702115826,
+      "learning_rate": 3.593937217901894e-07,
+      "loss": 0.6213,
+      "step": 8487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7697681408557449,
+      "learning_rate": 3.587672856802238e-07,
+      "loss": 0.6497,
+      "step": 8488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9259481594858332,
+      "learning_rate": 3.581413756804414e-07,
+      "loss": 0.6138,
+      "step": 8489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.078149202383723,
+      "learning_rate": 3.5751599186179155e-07,
+      "loss": 0.604,
+      "step": 8490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.014467331458678,
+      "learning_rate": 3.568911342951653e-07,
+      "loss": 0.7161,
+      "step": 8491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9429447928577988,
+      "learning_rate": 3.5626680305139383e-07,
+      "loss": 0.5646,
+      "step": 8492
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8978526166540983,
+      "learning_rate": 3.5564299820124883e-07,
+      "loss": 0.6737,
+      "step": 8493
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.78340609310645,
+      "learning_rate": 3.5501971981544206e-07,
+      "loss": 0.6035,
+      "step": 8494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1841185692701757,
+      "learning_rate": 3.5439696796462474e-07,
+      "loss": 0.687,
+      "step": 8495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9069053525425088,
+      "learning_rate": 3.5377474271939103e-07,
+      "loss": 0.5794,
+      "step": 8496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9781368239694577,
+      "learning_rate": 3.531530441502723e-07,
+      "loss": 0.6754,
+      "step": 8497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.787260799174934,
+      "learning_rate": 3.525318723277427e-07,
+      "loss": 0.621,
+      "step": 8498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8797198389956506,
+      "learning_rate": 3.519112273222142e-07,
+      "loss": 0.5571,
+      "step": 8499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0387277170307674,
+      "learning_rate": 3.512911092040422e-07,
+      "loss": 0.6346,
+      "step": 8500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9635088913195,
+      "learning_rate": 3.506715180435194e-07,
+      "loss": 0.6559,
+      "step": 8501
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8491276195881368,
+      "learning_rate": 3.500524539108807e-07,
+      "loss": 0.5446,
+      "step": 8502
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9939506343422873,
+      "learning_rate": 3.4943391687629946e-07,
+      "loss": 0.4936,
+      "step": 8503
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9671648301858564,
+      "learning_rate": 3.4881590700989175e-07,
+      "loss": 0.5706,
+      "step": 8504
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.044394803943145,
+      "learning_rate": 3.48198424381711e-07,
+      "loss": 0.625,
+      "step": 8505
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9877868545973922,
+      "learning_rate": 3.475814690617541e-07,
+      "loss": 0.5645,
+      "step": 8506
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.052299829237452,
+      "learning_rate": 3.469650411199543e-07,
+      "loss": 0.6363,
+      "step": 8507
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.07731152157618,
+      "learning_rate": 3.4634914062618984e-07,
+      "loss": 0.5551,
+      "step": 8508
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8411429732595614,
+      "learning_rate": 3.457337676502753e-07,
+      "loss": 0.6741,
+      "step": 8509
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7952870625461101,
+      "learning_rate": 3.451189222619661e-07,
+      "loss": 0.5644,
+      "step": 8510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.024486086455206,
+      "learning_rate": 3.445046045309586e-07,
+      "loss": 0.6004,
+      "step": 8511
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.812082428978913,
+      "learning_rate": 3.438908145268904e-07,
+      "loss": 0.5468,
+      "step": 8512
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9415016178715299,
+      "learning_rate": 3.4327755231933603e-07,
+      "loss": 0.6659,
+      "step": 8513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4030823748882164,
+      "learning_rate": 3.426648179778147e-07,
+      "loss": 0.6063,
+      "step": 8514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8938865020813795,
+      "learning_rate": 3.420526115717815e-07,
+      "loss": 0.5623,
+      "step": 8515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.851080556637924,
+      "learning_rate": 3.4144093317063586e-07,
+      "loss": 0.5629,
+      "step": 8516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8727554350128697,
+      "learning_rate": 3.4082978284371127e-07,
+      "loss": 0.5925,
+      "step": 8517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9815146666422891,
+      "learning_rate": 3.4021916066028837e-07,
+      "loss": 0.5842,
+      "step": 8518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7584842117350237,
+      "learning_rate": 3.396090666895829e-07,
+      "loss": 0.7568,
+      "step": 8519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9915690900587948,
+      "learning_rate": 3.3899950100075354e-07,
+      "loss": 0.6972,
+      "step": 8520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7809472800894863,
+      "learning_rate": 3.383904636628965e-07,
+      "loss": 0.5753,
+      "step": 8521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8229838348029193,
+      "learning_rate": 3.377819547450517e-07,
+      "loss": 0.5927,
+      "step": 8522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8899181595589158,
+      "learning_rate": 3.3717397431619614e-07,
+      "loss": 0.6775,
+      "step": 8523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.111746304571928,
+      "learning_rate": 3.365665224452469e-07,
+      "loss": 0.7002,
+      "step": 8524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7960693428634793,
+      "learning_rate": 3.359595992010639e-07,
+      "loss": 0.5953,
+      "step": 8525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0102795188304614,
+      "learning_rate": 3.3535320465244494e-07,
+      "loss": 0.6059,
+      "step": 8526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8455053001836659,
+      "learning_rate": 3.3474733886812606e-07,
+      "loss": 0.6622,
+      "step": 8527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9846483394387668,
+      "learning_rate": 3.3414200191678903e-07,
+      "loss": 0.6186,
+      "step": 8528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.864288965670221,
+      "learning_rate": 3.3353719386704954e-07,
+      "loss": 0.54,
+      "step": 8529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8117014185107374,
+      "learning_rate": 3.329329147874677e-07,
+      "loss": 0.5355,
+      "step": 8530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.050544467198889,
+      "learning_rate": 3.3232916474654154e-07,
+      "loss": 0.6408,
+      "step": 8531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9753592793290733,
+      "learning_rate": 3.317259438127085e-07,
+      "loss": 0.6121,
+      "step": 8532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9685029438749013,
+      "learning_rate": 3.3112325205434834e-07,
+      "loss": 0.5375,
+      "step": 8533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0390676005965145,
+      "learning_rate": 3.305210895397792e-07,
+      "loss": 0.6119,
+      "step": 8534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.147082033891267,
+      "learning_rate": 3.299194563372604e-07,
+      "loss": 0.6216,
+      "step": 8535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9294865710173905,
+      "learning_rate": 3.2931835251498845e-07,
+      "loss": 0.5659,
+      "step": 8536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.091471816118534,
+      "learning_rate": 3.287177781411044e-07,
+      "loss": 0.7131,
+      "step": 8537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9174669992475244,
+      "learning_rate": 3.2811773328368604e-07,
+      "loss": 0.6418,
+      "step": 8538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.00937920247632,
+      "learning_rate": 3.2751821801075055e-07,
+      "loss": 0.5425,
+      "step": 8539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9361390403260308,
+      "learning_rate": 3.26919232390257e-07,
+      "loss": 0.6752,
+      "step": 8540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.976383613072552,
+      "learning_rate": 3.2632077649010495e-07,
+      "loss": 0.6174,
+      "step": 8541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0442738816418555,
+      "learning_rate": 3.2572285037813123e-07,
+      "loss": 0.6375,
+      "step": 8542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9529903619318187,
+      "learning_rate": 3.251254541221155e-07,
+      "loss": 0.6939,
+      "step": 8543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8756515993312053,
+      "learning_rate": 3.245285877897747e-07,
+      "loss": 0.6838,
+      "step": 8544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0880957169144088,
+      "learning_rate": 3.239322514487686e-07,
+      "loss": 0.5843,
+      "step": 8545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9669044853905082,
+      "learning_rate": 3.233364451666948e-07,
+      "loss": 0.633,
+      "step": 8546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1809013639310924,
+      "learning_rate": 3.227411690110904e-07,
+      "loss": 0.7172,
+      "step": 8547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7695203042424918,
+      "learning_rate": 3.2214642304943364e-07,
+      "loss": 0.5523,
+      "step": 8548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.427593226307166,
+      "learning_rate": 3.215522073491434e-07,
+      "loss": 0.5712,
+      "step": 8549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.816262167078391,
+      "learning_rate": 3.2095852197757625e-07,
+      "loss": 0.5903,
+      "step": 8550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0096160380666808,
+      "learning_rate": 3.203653670020307e-07,
+      "loss": 0.6423,
+      "step": 8551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9964846253698296,
+      "learning_rate": 3.1977274248974286e-07,
+      "loss": 0.5813,
+      "step": 8552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.057566069204673,
+      "learning_rate": 3.1918064850789297e-07,
+      "loss": 0.5512,
+      "step": 8553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9850613079984498,
+      "learning_rate": 3.1858908512359456e-07,
+      "loss": 0.5835,
+      "step": 8554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.847173933051401,
+      "learning_rate": 3.1799805240390723e-07,
+      "loss": 0.6499,
+      "step": 8555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8814036275877406,
+      "learning_rate": 3.1740755041582694e-07,
+      "loss": 0.5895,
+      "step": 8556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.072777684061676,
+      "learning_rate": 3.1681757922629063e-07,
+      "loss": 0.6682,
+      "step": 8557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8957016203434267,
+      "learning_rate": 3.1622813890217483e-07,
+      "loss": 0.5977,
+      "step": 8558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9201797550472164,
+      "learning_rate": 3.156392295102967e-07,
+      "loss": 0.4838,
+      "step": 8559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8415451231635673,
+      "learning_rate": 3.1505085111741165e-07,
+      "loss": 0.6179,
+      "step": 8560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7549802487786756,
+      "learning_rate": 3.144630037902152e-07,
+      "loss": 0.5285,
+      "step": 8561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.982659812561418,
+      "learning_rate": 3.1387568759534523e-07,
+      "loss": 0.6163,
+      "step": 8562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.875249224583908,
+      "learning_rate": 3.132889025993746e-07,
+      "loss": 0.6487,
+      "step": 8563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.117308403382591,
+      "learning_rate": 3.127026488688217e-07,
+      "loss": 0.6205,
+      "step": 8564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1811566929908777,
+      "learning_rate": 3.121169264701396e-07,
+      "loss": 0.5575,
+      "step": 8565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7310740844214478,
+      "learning_rate": 3.1153173546972395e-07,
+      "loss": 0.5113,
+      "step": 8566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0363528766549575,
+      "learning_rate": 3.109470759339095e-07,
+      "loss": 0.6399,
+      "step": 8567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.808682458516396,
+      "learning_rate": 3.1036294792897103e-07,
+      "loss": 0.5567,
+      "step": 8568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0304155594704945,
+      "learning_rate": 3.097793515211211e-07,
+      "loss": 0.638,
+      "step": 8569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8611948646613983,
+      "learning_rate": 3.0919628677651636e-07,
+      "loss": 0.5643,
+      "step": 8570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8861149189282882,
+      "learning_rate": 3.086137537612488e-07,
+      "loss": 0.5412,
+      "step": 8571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9998343579666824,
+      "learning_rate": 3.080317525413523e-07,
+      "loss": 0.6044,
+      "step": 8572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9620610866969879,
+      "learning_rate": 3.074502831827997e-07,
+      "loss": 0.658,
+      "step": 8573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9372850625188789,
+      "learning_rate": 3.0686934575150484e-07,
+      "loss": 0.6131,
+      "step": 8574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.010086692177107,
+      "learning_rate": 3.06288940313319e-07,
+      "loss": 0.5936,
+      "step": 8575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7050273978062502,
+      "learning_rate": 3.057090669340357e-07,
+      "loss": 0.4815,
+      "step": 8576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.011687551457418,
+      "learning_rate": 3.0512972567938505e-07,
+      "loss": 0.6161,
+      "step": 8577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9730483163807309,
+      "learning_rate": 3.045509166150412e-07,
+      "loss": 0.5797,
+      "step": 8578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9574359302513729,
+      "learning_rate": 3.0397263980661283e-07,
+      "loss": 0.6348,
+      "step": 8579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8261271087547564,
+      "learning_rate": 3.0339489531965307e-07,
+      "loss": 0.5946,
+      "step": 8580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.01638250418493,
+      "learning_rate": 3.028176832196511e-07,
+      "loss": 0.6325,
+      "step": 8581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0561733418751396,
+      "learning_rate": 3.022410035720391e-07,
+      "loss": 0.657,
+      "step": 8582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8801117833119756,
+      "learning_rate": 3.0166485644218423e-07,
+      "loss": 0.579,
+      "step": 8583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.17562854791616,
+      "learning_rate": 3.010892418953981e-07,
+      "loss": 0.627,
+      "step": 8584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8988445529001572,
+      "learning_rate": 3.005141599969286e-07,
+      "loss": 0.559,
+      "step": 8585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0209929465985708,
+      "learning_rate": 2.999396108119662e-07,
+      "loss": 0.5639,
+      "step": 8586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7706616500453063,
+      "learning_rate": 2.9936559440563727e-07,
+      "loss": 0.577,
+      "step": 8587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7695880019783534,
+      "learning_rate": 2.9879211084301194e-07,
+      "loss": 0.4723,
+      "step": 8588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8372393183117093,
+      "learning_rate": 2.9821916018909603e-07,
+      "loss": 0.5891,
+      "step": 8589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.323982778463408,
+      "learning_rate": 2.9764674250883917e-07,
+      "loss": 0.6652,
+      "step": 8590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.850804961341406,
+      "learning_rate": 2.970748578671251e-07,
+      "loss": 0.6241,
+      "step": 8591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.947215875031301,
+      "learning_rate": 2.9650350632878246e-07,
+      "loss": 0.6403,
+      "step": 8592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0757901778312218,
+      "learning_rate": 2.959326879585755e-07,
+      "loss": 0.6949,
+      "step": 8593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3184331772097204,
+      "learning_rate": 2.953624028212115e-07,
+      "loss": 0.6928,
+      "step": 8594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3251534514180134,
+      "learning_rate": 2.947926509813337e-07,
+      "loss": 0.6444,
+      "step": 8595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9132798847854695,
+      "learning_rate": 2.942234325035287e-07,
+      "loss": 0.584,
+      "step": 8596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9965630415954567,
+      "learning_rate": 2.9365474745231935e-07,
+      "loss": 0.6394,
+      "step": 8597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.044198645987738,
+      "learning_rate": 2.9308659589216913e-07,
+      "loss": 0.5541,
+      "step": 8598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.897414457067917,
+      "learning_rate": 2.925189778874826e-07,
+      "loss": 0.5109,
+      "step": 8599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.956046702740908,
+      "learning_rate": 2.9195189350260055e-07,
+      "loss": 0.6672,
+      "step": 8600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9022595348835627,
+      "learning_rate": 2.9138534280180706e-07,
+      "loss": 0.6785,
+      "step": 8601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.026765034791194,
+      "learning_rate": 2.908193258493236e-07,
+      "loss": 0.6166,
+      "step": 8602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.127506565979339,
+      "learning_rate": 2.9025384270931043e-07,
+      "loss": 0.5703,
+      "step": 8603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.213452785659447,
+      "learning_rate": 2.8968889344586915e-07,
+      "loss": 0.6445,
+      "step": 8604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9799657762514415,
+      "learning_rate": 2.8912447812303956e-07,
+      "loss": 0.6423,
+      "step": 8605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1495360937337464,
+      "learning_rate": 2.885605968048011e-07,
+      "loss": 0.6764,
+      "step": 8606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.715241435422298,
+      "learning_rate": 2.8799724955507367e-07,
+      "loss": 0.555,
+      "step": 8607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.00052872630302,
+      "learning_rate": 2.874344364377152e-07,
+      "loss": 0.643,
+      "step": 8608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9284952108877513,
+      "learning_rate": 2.868721575165245e-07,
+      "loss": 0.6311,
+      "step": 8609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.690339382498523,
+      "learning_rate": 2.86310412855238e-07,
+      "loss": 0.5235,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9565147307410093,
+      "learning_rate": 2.85749202517534e-07,
+      "loss": 0.6184,
+      "step": 8611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0904675537845843,
+      "learning_rate": 2.8518852656702845e-07,
+      "loss": 0.6863,
+      "step": 8612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.106472750498347,
+      "learning_rate": 2.8462838506727707e-07,
+      "loss": 0.697,
+      "step": 8613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.961157989135241,
+      "learning_rate": 2.8406877808177414e-07,
+      "loss": 0.5415,
+      "step": 8614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.954339439444753,
+      "learning_rate": 2.8350970567395555e-07,
+      "loss": 0.6069,
+      "step": 8615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.020476422284541,
+      "learning_rate": 2.8295116790719444e-07,
+      "loss": 0.6706,
+      "step": 8616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.092819502393807,
+      "learning_rate": 2.8239316484480527e-07,
+      "loss": 0.5515,
+      "step": 8617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.835575899833264,
+      "learning_rate": 2.8183569655003963e-07,
+      "loss": 0.4899,
+      "step": 8618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0289917891127454,
+      "learning_rate": 2.812787630860919e-07,
+      "loss": 0.5401,
+      "step": 8619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.811975427170239,
+      "learning_rate": 2.807223645160906e-07,
+      "loss": 0.5122,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.200065793760229,
+      "learning_rate": 2.801665009031096e-07,
+      "loss": 0.536,
+      "step": 8621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8723693676317084,
+      "learning_rate": 2.796111723101563e-07,
+      "loss": 0.5162,
+      "step": 8622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9167058977018545,
+      "learning_rate": 2.7905637880018324e-07,
+      "loss": 0.6457,
+      "step": 8623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9368506827247765,
+      "learning_rate": 2.7850212043607715e-07,
+      "loss": 0.6408,
+      "step": 8624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8101557694487689,
+      "learning_rate": 2.779483972806685e-07,
+      "loss": 0.6135,
+      "step": 8625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1315207329438612,
+      "learning_rate": 2.773952093967225e-07,
+      "loss": 0.5771,
+      "step": 8626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0848835082063646,
+      "learning_rate": 2.7684255684694903e-07,
+      "loss": 0.6573,
+      "step": 8627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.91807737163709,
+      "learning_rate": 2.7629043969399193e-07,
+      "loss": 0.6022,
+      "step": 8628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.118468230505594,
+      "learning_rate": 2.7573885800043775e-07,
+      "loss": 0.6365,
+      "step": 8629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9790619469804935,
+      "learning_rate": 2.7518781182881096e-07,
+      "loss": 0.6301,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.022098894804944,
+      "learning_rate": 2.7463730124157706e-07,
+      "loss": 0.6485,
+      "step": 8631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.056757846364307,
+      "learning_rate": 2.7408732630113787e-07,
+      "loss": 0.565,
+      "step": 8632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9947893769894807,
+      "learning_rate": 2.73537887069838e-07,
+      "loss": 0.5146,
+      "step": 8633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.231051088541484,
+      "learning_rate": 2.729889836099581e-07,
+      "loss": 0.5901,
+      "step": 8634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9660039530430966,
+      "learning_rate": 2.7244061598371953e-07,
+      "loss": 0.7301,
+      "step": 8635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9431122518352824,
+      "learning_rate": 2.7189278425328426e-07,
+      "loss": 0.6849,
+      "step": 8636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8396886613553558,
+      "learning_rate": 2.713454884807504e-07,
+      "loss": 0.5825,
+      "step": 8637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8078269413999202,
+      "learning_rate": 2.707987287281583e-07,
+      "loss": 0.6539,
+      "step": 8638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.862585778610424,
+      "learning_rate": 2.70252505057485e-07,
+      "loss": 0.5778,
+      "step": 8639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.894036420630567,
+      "learning_rate": 2.6970681753065e-07,
+      "loss": 0.5502,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8819699326290948,
+      "learning_rate": 2.691616662095092e-07,
+      "loss": 0.6226,
+      "step": 8641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9204845961184525,
+      "learning_rate": 2.686170511558578e-07,
+      "loss": 0.637,
+      "step": 8642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8825697272527415,
+      "learning_rate": 2.680729724314313e-07,
+      "loss": 0.6391,
+      "step": 8643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8943724619603919,
+      "learning_rate": 2.675294300979053e-07,
+      "loss": 0.5944,
+      "step": 8644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.259795468206444,
+      "learning_rate": 2.6698642421689124e-07,
+      "loss": 0.6657,
+      "step": 8645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9778845403528227,
+      "learning_rate": 2.664439548499448e-07,
+      "loss": 0.6692,
+      "step": 8646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0445734899171097,
+      "learning_rate": 2.6590202205855506e-07,
+      "loss": 0.6772,
+      "step": 8647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3171034519987823,
+      "learning_rate": 2.6536062590415577e-07,
+      "loss": 0.5824,
+      "step": 8648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8081871008672612,
+      "learning_rate": 2.648197664481156e-07,
+      "loss": 0.6396,
+      "step": 8649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1307670329224937,
+      "learning_rate": 2.6427944375174484e-07,
+      "loss": 0.6678,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.253854108832021,
+      "learning_rate": 2.637396578762913e-07,
+      "loss": 0.5559,
+      "step": 8651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.094905790658024,
+      "learning_rate": 2.6320040888294373e-07,
+      "loss": 0.6969,
+      "step": 8652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8667934315030559,
+      "learning_rate": 2.6266169683282827e-07,
+      "loss": 0.5351,
+      "step": 8653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9530819540310358,
+      "learning_rate": 2.621235217870116e-07,
+      "loss": 0.6533,
+      "step": 8654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9977273869765968,
+      "learning_rate": 2.615858838064983e-07,
+      "loss": 0.6239,
+      "step": 8655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9377945345738095,
+      "learning_rate": 2.6104878295223455e-07,
+      "loss": 0.6392,
+      "step": 8656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0652175471062133,
+      "learning_rate": 2.605122192851012e-07,
+      "loss": 0.6128,
+      "step": 8657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0175796302069418,
+      "learning_rate": 2.5997619286592224e-07,
+      "loss": 0.6744,
+      "step": 8658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.747368958393763,
+      "learning_rate": 2.594407037554586e-07,
+      "loss": 0.6059,
+      "step": 8659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2493432755306895,
+      "learning_rate": 2.5890575201441224e-07,
+      "loss": 0.6867,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0016522845047566,
+      "learning_rate": 2.5837133770342135e-07,
+      "loss": 0.6622,
+      "step": 8661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2223145338114767,
+      "learning_rate": 2.578374608830664e-07,
+      "loss": 0.6139,
+      "step": 8662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.004690950521899,
+      "learning_rate": 2.5730412161386386e-07,
+      "loss": 0.6133,
+      "step": 8663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9682711738064742,
+      "learning_rate": 2.567713199562727e-07,
+      "loss": 0.6053,
+      "step": 8664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9739531150421892,
+      "learning_rate": 2.562390559706879e-07,
+      "loss": 0.6154,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1119152599912105,
+      "learning_rate": 2.5570732971744395e-07,
+      "loss": 0.6286,
+      "step": 8666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.893301025135361,
+      "learning_rate": 2.55176141256816e-07,
+      "loss": 0.6225,
+      "step": 8667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1130422748129,
+      "learning_rate": 2.54645490649017e-07,
+      "loss": 0.6161,
+      "step": 8668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1581512462565833,
+      "learning_rate": 2.5411537795419883e-07,
+      "loss": 0.7077,
+      "step": 8669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9273252313199596,
+      "learning_rate": 2.5358580323245396e-07,
+      "loss": 0.6358,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9499885610352081,
+      "learning_rate": 2.530567665438116e-07,
+      "loss": 0.5547,
+      "step": 8671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.111913389915149,
+      "learning_rate": 2.52528267948241e-07,
+      "loss": 0.54,
+      "step": 8672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0260864755391164,
+      "learning_rate": 2.520003075056521e-07,
+      "loss": 0.5447,
+      "step": 8673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8671832739571868,
+      "learning_rate": 2.5147288527588964e-07,
+      "loss": 0.6723,
+      "step": 8674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7405827829670664,
+      "learning_rate": 2.5094600131874205e-07,
+      "loss": 0.5051,
+      "step": 8675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0584072380487015,
+      "learning_rate": 2.5041965569393366e-07,
+      "loss": 0.59,
+      "step": 8676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0569408420447046,
+      "learning_rate": 2.498938484611296e-07,
+      "loss": 0.6142,
+      "step": 8677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8546671023261774,
+      "learning_rate": 2.4936857967993166e-07,
+      "loss": 0.6434,
+      "step": 8678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.012713259045472,
+      "learning_rate": 2.4884384940988436e-07,
+      "loss": 0.6429,
+      "step": 8679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1338257948479336,
+      "learning_rate": 2.4831965771046574e-07,
+      "loss": 0.6217,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.978217762866347,
+      "learning_rate": 2.4779600464109874e-07,
+      "loss": 0.6405,
+      "step": 8681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3616307774259555,
+      "learning_rate": 2.4727289026114043e-07,
+      "loss": 0.6803,
+      "step": 8682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1877070740639097,
+      "learning_rate": 2.4675031462988995e-07,
+      "loss": 0.6009,
+      "step": 8683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9569526648473758,
+      "learning_rate": 2.462282778065839e-07,
+      "loss": 0.6932,
+      "step": 8684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8683437289001,
+      "learning_rate": 2.4570677985039817e-07,
+      "loss": 0.5918,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.89796611020582,
+      "learning_rate": 2.451858208204477e-07,
+      "loss": 0.61,
+      "step": 8686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9025712390079088,
+      "learning_rate": 2.446654007757865e-07,
+      "loss": 0.5998,
+      "step": 8687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9360903945505326,
+      "learning_rate": 2.441455197754056e-07,
+      "loss": 0.5562,
+      "step": 8688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8305256442447795,
+      "learning_rate": 2.436261778782378e-07,
+      "loss": 0.6337,
+      "step": 8689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9413740891789482,
+      "learning_rate": 2.431073751431529e-07,
+      "loss": 0.6631,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8320514002798973,
+      "learning_rate": 2.4258911162896083e-07,
+      "loss": 0.6427,
+      "step": 8691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6997980036645046,
+      "learning_rate": 2.4207138739440914e-07,
+      "loss": 0.5699,
+      "step": 8692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.768129171675751,
+      "learning_rate": 2.4155420249818596e-07,
+      "loss": 0.5332,
+      "step": 8693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.866393250711023,
+      "learning_rate": 2.4103755699891487e-07,
+      "loss": 0.5497,
+      "step": 8694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1352325058003974,
+      "learning_rate": 2.405214509551623e-07,
+      "loss": 0.5664,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9594544194885912,
+      "learning_rate": 2.40005884425431e-07,
+      "loss": 0.5781,
+      "step": 8696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7619113591533766,
+      "learning_rate": 2.3949085746816424e-07,
+      "loss": 0.5066,
+      "step": 8697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1093914969198413,
+      "learning_rate": 2.389763701417419e-07,
+      "loss": 0.7102,
+      "step": 8698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.10399786419913,
+      "learning_rate": 2.3846242250448624e-07,
+      "loss": 0.7246,
+      "step": 8699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0803726938672593,
+      "learning_rate": 2.3794901461465402e-07,
+      "loss": 0.7128,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8698045186841377,
+      "learning_rate": 2.3743614653044423e-07,
+      "loss": 0.5957,
+      "step": 8701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6970683847366228,
+      "learning_rate": 2.3692381830999255e-07,
+      "loss": 0.5625,
+      "step": 8702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.096324173910222,
+      "learning_rate": 2.364120300113748e-07,
+      "loss": 0.667,
+      "step": 8703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.975751309515821,
+      "learning_rate": 2.3590078169260512e-07,
+      "loss": 0.5752,
+      "step": 8704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.934365429629889,
+      "learning_rate": 2.3539007341163656e-07,
+      "loss": 0.6003,
+      "step": 8705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.049087859270842,
+      "learning_rate": 2.3487990522636005e-07,
+      "loss": 0.6966,
+      "step": 8706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1228999756456597,
+      "learning_rate": 2.3437027719460659e-07,
+      "loss": 0.5632,
+      "step": 8707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.988718489834557,
+      "learning_rate": 2.3386118937414602e-07,
+      "loss": 0.6215,
+      "step": 8708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9566653933319493,
+      "learning_rate": 2.3335264182268502e-07,
+      "loss": 0.4998,
+      "step": 8709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8116059731020802,
+      "learning_rate": 2.328446345978713e-07,
+      "loss": 0.505,
+      "step": 8710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9242008346185047,
+      "learning_rate": 2.3233716775728943e-07,
+      "loss": 0.5418,
+      "step": 8711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8515564888972094,
+      "learning_rate": 2.3183024135846554e-07,
+      "loss": 0.6389,
+      "step": 8712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8985578066102073,
+      "learning_rate": 2.3132385545886037e-07,
+      "loss": 0.5512,
+      "step": 8713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9046098103072053,
+      "learning_rate": 2.3081801011587745e-07,
+      "loss": 0.5628,
+      "step": 8714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1917228952182097,
+      "learning_rate": 2.3031270538685701e-07,
+      "loss": 0.664,
+      "step": 8715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5641131401356265,
+      "learning_rate": 2.2980794132907713e-07,
+      "loss": 0.7537,
+      "step": 8716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8083432847429621,
+      "learning_rate": 2.2930371799975593e-07,
+      "loss": 0.5948,
+      "step": 8717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.157666039891891,
+      "learning_rate": 2.2880003545605101e-07,
+      "loss": 0.6228,
+      "step": 8718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9598888616221026,
+      "learning_rate": 2.2829689375505616e-07,
+      "loss": 0.6904,
+      "step": 8719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0529866203222418,
+      "learning_rate": 2.2779429295380683e-07,
+      "loss": 0.7202,
+      "step": 8720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.025400351121733,
+      "learning_rate": 2.2729223310927473e-07,
+      "loss": 0.6518,
+      "step": 8721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.643070030489791,
+      "learning_rate": 2.267907142783715e-07,
+      "loss": 0.5422,
+      "step": 8722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.168941535240748,
+      "learning_rate": 2.2628973651794783e-07,
+      "loss": 0.6902,
+      "step": 8723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.025938493633533,
+      "learning_rate": 2.257892998847916e-07,
+      "loss": 0.6059,
+      "step": 8724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9086079703312495,
+      "learning_rate": 2.2528940443562964e-07,
+      "loss": 0.5982,
+      "step": 8725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8506109120515981,
+      "learning_rate": 2.247900502271294e-07,
+      "loss": 0.5936,
+      "step": 8726
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.129293598828574,
+      "learning_rate": 2.2429123731589397e-07,
+      "loss": 0.4808,
+      "step": 8727
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.199598134027323,
+      "learning_rate": 2.2379296575846809e-07,
+      "loss": 0.6588,
+      "step": 8728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9581770705902077,
+      "learning_rate": 2.2329523561133215e-07,
+      "loss": 0.5429,
+      "step": 8729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.070975877443173,
+      "learning_rate": 2.2279804693090934e-07,
+      "loss": 0.6249,
+      "step": 8730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8912896810784905,
+      "learning_rate": 2.223013997735557e-07,
+      "loss": 0.6175,
+      "step": 8731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.104502168982993,
+      "learning_rate": 2.2180529419557062e-07,
+      "loss": 0.6044,
+      "step": 8732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.868832405954279,
+      "learning_rate": 2.213097302531897e-07,
+      "loss": 0.5568,
+      "step": 8733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8040318964006492,
+      "learning_rate": 2.208147080025891e-07,
+      "loss": 0.5683,
+      "step": 8734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9255176394764801,
+      "learning_rate": 2.2032022749988113e-07,
+      "loss": 0.6989,
+      "step": 8735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8424660754915125,
+      "learning_rate": 2.1982628880111988e-07,
+      "loss": 0.578,
+      "step": 8736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9013342427890227,
+      "learning_rate": 2.1933289196229335e-07,
+      "loss": 0.6747,
+      "step": 8737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.947218569687289,
+      "learning_rate": 2.1884003703933343e-07,
+      "loss": 0.5696,
+      "step": 8738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.124368808695498,
+      "learning_rate": 2.1834772408810712e-07,
+      "loss": 0.6159,
+      "step": 8739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.140229776900523,
+      "learning_rate": 2.1785595316442032e-07,
+      "loss": 0.6732,
+      "step": 8740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9381452957689793,
+      "learning_rate": 2.17364724324019e-07,
+      "loss": 0.5212,
+      "step": 8741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1126927729884417,
+      "learning_rate": 2.1687403762258585e-07,
+      "loss": 0.6332,
+      "step": 8742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0856561623513987,
+      "learning_rate": 2.1638389311574358e-07,
+      "loss": 0.7432,
+      "step": 8743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.05971362182267,
+      "learning_rate": 2.1589429085905278e-07,
+      "loss": 0.6012,
+      "step": 8744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0239460998799848,
+      "learning_rate": 2.1540523090801292e-07,
+      "loss": 0.5419,
+      "step": 8745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9997879159734424,
+      "learning_rate": 2.149167133180602e-07,
+      "loss": 0.5633,
+      "step": 8746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.059015466218665,
+      "learning_rate": 2.1442873814457364e-07,
+      "loss": 0.6194,
+      "step": 8747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.525897595654169,
+      "learning_rate": 2.1394130544286519e-07,
+      "loss": 0.6233,
+      "step": 8748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9910958324419423,
+      "learning_rate": 2.1345441526819e-07,
+      "loss": 0.5732,
+      "step": 8749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9316768520667007,
+      "learning_rate": 2.1296806767573897e-07,
+      "loss": 0.667,
+      "step": 8750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1742785881902753,
+      "learning_rate": 2.1248226272064299e-07,
+      "loss": 0.6773,
+      "step": 8751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.129579988218092,
+      "learning_rate": 2.1199700045797077e-07,
+      "loss": 0.5753,
+      "step": 8752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.188883794073873,
+      "learning_rate": 2.115122809427289e-07,
+      "loss": 0.6846,
+      "step": 8753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8777793148808293,
+      "learning_rate": 2.1102810422986286e-07,
+      "loss": 0.5493,
+      "step": 8754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9743047341702247,
+      "learning_rate": 2.105444703742582e-07,
+      "loss": 0.608,
+      "step": 8755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.95744377141077,
+      "learning_rate": 2.1006137943073602e-07,
+      "loss": 0.6769,
+      "step": 8756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8324319226472476,
+      "learning_rate": 2.095788314540592e-07,
+      "loss": 0.5666,
+      "step": 8757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6567842512967919,
+      "learning_rate": 2.0909682649892516e-07,
+      "loss": 0.5137,
+      "step": 8758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7813909272995272,
+      "learning_rate": 2.08615364619974e-07,
+      "loss": 0.6392,
+      "step": 8759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9961485822604668,
+      "learning_rate": 2.0813444587178156e-07,
+      "loss": 0.6946,
+      "step": 8760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9287070838365437,
+      "learning_rate": 2.0765407030886197e-07,
+      "loss": 0.6178,
+      "step": 8761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0168575757502505,
+      "learning_rate": 2.0717423798566838e-07,
+      "loss": 0.5699,
+      "step": 8762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.940310038251417,
+      "learning_rate": 2.0669494895659391e-07,
+      "loss": 0.5678,
+      "step": 8763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8342339019571705,
+      "learning_rate": 2.0621620327596735e-07,
+      "loss": 0.6365,
+      "step": 8764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9270699483483476,
+      "learning_rate": 2.057380009980581e-07,
+      "loss": 0.5998,
+      "step": 8765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1596691648652593,
+      "learning_rate": 2.0526034217707213e-07,
+      "loss": 0.654,
+      "step": 8766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0471564825995765,
+      "learning_rate": 2.0478322686715735e-07,
+      "loss": 0.6248,
+      "step": 8767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.009135979527017,
+      "learning_rate": 2.0430665512239377e-07,
+      "loss": 0.6103,
+      "step": 8768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0477920499162057,
+      "learning_rate": 2.0383062699680601e-07,
+      "loss": 0.7414,
+      "step": 8769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8535142344799098,
+      "learning_rate": 2.0335514254435363e-07,
+      "loss": 0.6597,
+      "step": 8770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7529105422442415,
+      "learning_rate": 2.028802018189363e-07,
+      "loss": 0.5557,
+      "step": 8771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.064542466324043,
+      "learning_rate": 2.0240580487438988e-07,
+      "loss": 0.5888,
+      "step": 8772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1972196941591506,
+      "learning_rate": 2.0193195176449188e-07,
+      "loss": 0.7178,
+      "step": 8773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9128197640557918,
+      "learning_rate": 2.0145864254295434e-07,
+      "loss": 0.6155,
+      "step": 8774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9423327842910736,
+      "learning_rate": 2.0098587726343156e-07,
+      "loss": 0.5647,
+      "step": 8775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8146626640382046,
+      "learning_rate": 2.0051365597951233e-07,
+      "loss": 0.5116,
+      "step": 8776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1038513954027906,
+      "learning_rate": 2.000419787447261e-07,
+      "loss": 0.6498,
+      "step": 8777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9053349463179003,
+      "learning_rate": 1.9957084561254114e-07,
+      "loss": 0.6542,
+      "step": 8778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0831840015435117,
+      "learning_rate": 1.9910025663636146e-07,
+      "loss": 0.5748,
+      "step": 8779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.997692900719902,
+      "learning_rate": 1.9863021186953268e-07,
+      "loss": 0.6376,
+      "step": 8780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1968215047753756,
+      "learning_rate": 1.9816071136533665e-07,
+      "loss": 0.6415,
+      "step": 8781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.068923142726295,
+      "learning_rate": 1.9769175517699302e-07,
+      "loss": 0.6246,
+      "step": 8782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.047076265368446,
+      "learning_rate": 1.9722334335766092e-07,
+      "loss": 0.5755,
+      "step": 8783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9159180146333123,
+      "learning_rate": 1.9675547596043787e-07,
+      "loss": 0.5303,
+      "step": 8784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.054746023991903,
+      "learning_rate": 1.962881530383587e-07,
+      "loss": 0.6917,
+      "step": 8785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8723658262962497,
+      "learning_rate": 1.9582137464439876e-07,
+      "loss": 0.5934,
+      "step": 8786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9256110172198415,
+      "learning_rate": 1.95355140831468e-07,
+      "loss": 0.5799,
+      "step": 8787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0613496981654134,
+      "learning_rate": 1.94889451652418e-07,
+      "loss": 0.64,
+      "step": 8788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8928277560645934,
+      "learning_rate": 1.9442430716003713e-07,
+      "loss": 0.5309,
+      "step": 8789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.211268371166142,
+      "learning_rate": 1.9395970740705205e-07,
+      "loss": 0.6969,
+      "step": 8790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0242471939323763,
+      "learning_rate": 1.9349565244612678e-07,
+      "loss": 0.6129,
+      "step": 8791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.055751552611855,
+      "learning_rate": 1.9303214232986588e-07,
+      "loss": 0.6545,
+      "step": 8792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.936122063029454,
+      "learning_rate": 1.9256917711081014e-07,
+      "loss": 0.5843,
+      "step": 8793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.995976353696018,
+      "learning_rate": 1.921067568414403e-07,
+      "loss": 0.5792,
+      "step": 8794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9787829732592972,
+      "learning_rate": 1.9164488157417284e-07,
+      "loss": 0.6806,
+      "step": 8795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.812060229047015,
+      "learning_rate": 1.9118355136136523e-07,
+      "loss": 0.587,
+      "step": 8796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0037208498765238,
+      "learning_rate": 1.9072276625531127e-07,
+      "loss": 0.5948,
+      "step": 8797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9635031676753756,
+      "learning_rate": 1.9026252630824415e-07,
+      "loss": 0.5545,
+      "step": 8798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.910679902637944,
+      "learning_rate": 1.898028315723338e-07,
+      "loss": 0.5776,
+      "step": 8799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9739607133635115,
+      "learning_rate": 1.8934368209969023e-07,
+      "loss": 0.5948,
+      "step": 8800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9046769737170033,
+      "learning_rate": 1.88885077942359e-07,
+      "loss": 0.6231,
+      "step": 8801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9551105878545139,
+      "learning_rate": 1.8842701915232743e-07,
+      "loss": 0.52,
+      "step": 8802
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8166528029787048,
+      "learning_rate": 1.8796950578151785e-07,
+      "loss": 0.4986,
+      "step": 8803
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.794656916003373,
+      "learning_rate": 1.8751253788179325e-07,
+      "loss": 0.6422,
+      "step": 8804
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.173722308812529,
+      "learning_rate": 1.8705611550495218e-07,
+      "loss": 0.6271,
+      "step": 8805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.730807391557464,
+      "learning_rate": 1.866002387027338e-07,
+      "loss": 0.509,
+      "step": 8806
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7958190844792186,
+      "learning_rate": 1.8614490752681292e-07,
+      "loss": 0.5214,
+      "step": 8807
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9894503413864093,
+      "learning_rate": 1.8569012202880599e-07,
+      "loss": 0.5939,
+      "step": 8808
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1598352658993623,
+      "learning_rate": 1.852358822602629e-07,
+      "loss": 0.5809,
+      "step": 8809
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0071051604662937,
+      "learning_rate": 1.847821882726769e-07,
+      "loss": 0.605,
+      "step": 8810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0832891855842615,
+      "learning_rate": 1.843290401174752e-07,
+      "loss": 0.6761,
+      "step": 8811
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0663532941669702,
+      "learning_rate": 1.838764378460256e-07,
+      "loss": 0.6676,
+      "step": 8812
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0405747575233977,
+      "learning_rate": 1.834243815096326e-07,
+      "loss": 0.5634,
+      "step": 8813
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.086360893875116,
+      "learning_rate": 1.829728711595391e-07,
+      "loss": 0.6271,
+      "step": 8814
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9649083555496234,
+      "learning_rate": 1.825219068469275e-07,
+      "loss": 0.6017,
+      "step": 8815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.006301857023903,
+      "learning_rate": 1.820714886229158e-07,
+      "loss": 0.6402,
+      "step": 8816
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0959766376734814,
+      "learning_rate": 1.8162161653856257e-07,
+      "loss": 0.5846,
+      "step": 8817
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9703517347002897,
+      "learning_rate": 1.8117229064486264e-07,
+      "loss": 0.6658,
+      "step": 8818
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9458890586613302,
+      "learning_rate": 1.8072351099275077e-07,
+      "loss": 0.5241,
+      "step": 8819
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.96184039789254,
+      "learning_rate": 1.8027527763309682e-07,
+      "loss": 0.5714,
+      "step": 8820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8509877835265842,
+      "learning_rate": 1.798275906167124e-07,
+      "loss": 0.621,
+      "step": 8821
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8661444783986674,
+      "learning_rate": 1.7938044999434412e-07,
+      "loss": 0.6553,
+      "step": 8822
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8170939072394554,
+      "learning_rate": 1.7893385581667866e-07,
+      "loss": 0.6345,
+      "step": 8823
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9101906587424282,
+      "learning_rate": 1.7848780813433942e-07,
+      "loss": 0.604,
+      "step": 8824
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2378093679248,
+      "learning_rate": 1.7804230699788983e-07,
+      "loss": 0.6194,
+      "step": 8825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.28535490117752,
+      "learning_rate": 1.7759735245782838e-07,
+      "loss": 0.6506,
+      "step": 8826
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.969139478413972,
+      "learning_rate": 1.7715294456459419e-07,
+      "loss": 0.5687,
+      "step": 8827
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9704670196931522,
+      "learning_rate": 1.7670908336856252e-07,
+      "loss": 0.6122,
+      "step": 8828
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.046800006905025,
+      "learning_rate": 1.762657689200481e-07,
+      "loss": 0.6601,
+      "step": 8829
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0533236305561267,
+      "learning_rate": 1.7582300126930296e-07,
+      "loss": 0.6144,
+      "step": 8830
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8968834934718648,
+      "learning_rate": 1.7538078046651808e-07,
+      "loss": 0.6129,
+      "step": 8831
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8440975962291353,
+      "learning_rate": 1.7493910656182056e-07,
+      "loss": 0.5748,
+      "step": 8832
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.054011675246738,
+      "learning_rate": 1.744979796052776e-07,
+      "loss": 0.6494,
+      "step": 8833
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.350762103488185,
+      "learning_rate": 1.7405739964689362e-07,
+      "loss": 0.6574,
+      "step": 8834
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.194253819345266,
+      "learning_rate": 1.7361736673660978e-07,
+      "loss": 0.582,
+      "step": 8835
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9923276521726647,
+      "learning_rate": 1.7317788092430676e-07,
+      "loss": 0.5446,
+      "step": 8836
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9545991453195062,
+      "learning_rate": 1.7273894225980303e-07,
+      "loss": 0.5959,
+      "step": 8837
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8931944180409803,
+      "learning_rate": 1.7230055079285435e-07,
+      "loss": 0.6128,
+      "step": 8838
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.064916087058621,
+      "learning_rate": 1.7186270657315596e-07,
+      "loss": 0.5658,
+      "step": 8839
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.014461259471736,
+      "learning_rate": 1.7142540965033815e-07,
+      "loss": 0.6944,
+      "step": 8840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0937362350654034,
+      "learning_rate": 1.7098866007397296e-07,
+      "loss": 0.6447,
+      "step": 8841
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6339582774602281,
+      "learning_rate": 1.70552457893568e-07,
+      "loss": 0.5278,
+      "step": 8842
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9130656985961463,
+      "learning_rate": 1.701168031585676e-07,
+      "loss": 0.594,
+      "step": 8843
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8312506205126324,
+      "learning_rate": 1.696816959183578e-07,
+      "loss": 0.6467,
+      "step": 8844
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.241570074288019,
+      "learning_rate": 1.6924713622225975e-07,
+      "loss": 0.6756,
+      "step": 8845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1124221106992396,
+      "learning_rate": 1.6881312411953288e-07,
+      "loss": 0.6671,
+      "step": 8846
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9213107738949424,
+      "learning_rate": 1.6837965965937564e-07,
+      "loss": 0.6519,
+      "step": 8847
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8382074321131783,
+      "learning_rate": 1.6794674289092316e-07,
+      "loss": 0.5749,
+      "step": 8848
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.860478511195977,
+      "learning_rate": 1.6751437386324842e-07,
+      "loss": 0.5574,
+      "step": 8849
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0995492889812497,
+      "learning_rate": 1.6708255262536443e-07,
+      "loss": 0.5592,
+      "step": 8850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0108036150741375,
+      "learning_rate": 1.6665127922621927e-07,
+      "loss": 0.5698,
+      "step": 8851
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.029768719174924,
+      "learning_rate": 1.6622055371470103e-07,
+      "loss": 0.5616,
+      "step": 8852
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.069671016263712,
+      "learning_rate": 1.6579037613963456e-07,
+      "loss": 0.7072,
+      "step": 8853
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0304812240208134,
+      "learning_rate": 1.6536074654978307e-07,
+      "loss": 0.568,
+      "step": 8854
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9320514761894227,
+      "learning_rate": 1.6493166499384762e-07,
+      "loss": 0.6442,
+      "step": 8855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0939593574383397,
+      "learning_rate": 1.645031315204676e-07,
+      "loss": 0.6313,
+      "step": 8856
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0131860014511966,
+      "learning_rate": 1.6407514617821752e-07,
+      "loss": 0.7299,
+      "step": 8857
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9602967302745016,
+      "learning_rate": 1.6364770901561467e-07,
+      "loss": 0.5883,
+      "step": 8858
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.029561620284226,
+      "learning_rate": 1.6322082008110974e-07,
+      "loss": 0.58,
+      "step": 8859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.016812776940906,
+      "learning_rate": 1.6279447942309345e-07,
+      "loss": 0.687,
+      "step": 8860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.138409567531603,
+      "learning_rate": 1.6236868708989438e-07,
+      "loss": 0.686,
+      "step": 8861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7290146664194534,
+      "learning_rate": 1.619434431297784e-07,
+      "loss": 0.559,
+      "step": 8862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9890454463913607,
+      "learning_rate": 1.615187475909491e-07,
+      "loss": 0.6277,
+      "step": 8863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9194698017431477,
+      "learning_rate": 1.6109460052154802e-07,
+      "loss": 0.5385,
+      "step": 8864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.873346670966746,
+      "learning_rate": 1.6067100196965447e-07,
+      "loss": 0.548,
+      "step": 8865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7853281470209055,
+      "learning_rate": 1.602479519832867e-07,
+      "loss": 0.5921,
+      "step": 8866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.007330702321665,
+      "learning_rate": 1.5982545061039855e-07,
+      "loss": 0.6518,
+      "step": 8867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9895802955142883,
+      "learning_rate": 1.5940349789888398e-07,
+      "loss": 0.6133,
+      "step": 8868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.977575299921113,
+      "learning_rate": 1.5898209389657305e-07,
+      "loss": 0.6876,
+      "step": 8869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9175713363661888,
+      "learning_rate": 1.585612386512353e-07,
+      "loss": 0.594,
+      "step": 8870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0749683134033026,
+      "learning_rate": 1.5814093221057647e-07,
+      "loss": 0.5403,
+      "step": 8871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1233122307163885,
+      "learning_rate": 1.5772117462224068e-07,
+      "loss": 0.6821,
+      "step": 8872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0961791591392056,
+      "learning_rate": 1.5730196593380877e-07,
+      "loss": 0.644,
+      "step": 8873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.944886823301487,
+      "learning_rate": 1.5688330619280269e-07,
+      "loss": 0.6329,
+      "step": 8874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8193650429809556,
+      "learning_rate": 1.5646519544667783e-07,
+      "loss": 0.5126,
+      "step": 8875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9800998991702814,
+      "learning_rate": 1.5604763374283073e-07,
+      "loss": 0.5899,
+      "step": 8876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0403033797812267,
+      "learning_rate": 1.556306211285935e-07,
+      "loss": 0.623,
+      "step": 8877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.035385871609747,
+      "learning_rate": 1.5521415765123783e-07,
+      "loss": 0.544,
+      "step": 8878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8780946782463006,
+      "learning_rate": 1.54798243357972e-07,
+      "loss": 0.6026,
+      "step": 8879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4224720214407203,
+      "learning_rate": 1.5438287829594113e-07,
+      "loss": 0.6607,
+      "step": 8880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0498507932770895,
+      "learning_rate": 1.5396806251223085e-07,
+      "loss": 0.6207,
+      "step": 8881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.011661827763579,
+      "learning_rate": 1.535537960538619e-07,
+      "loss": 0.6032,
+      "step": 8882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1615800677651666,
+      "learning_rate": 1.5314007896779393e-07,
+      "loss": 0.6291,
+      "step": 8883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2134352807827535,
+      "learning_rate": 1.5272691130092443e-07,
+      "loss": 0.5684,
+      "step": 8884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.113655773007037,
+      "learning_rate": 1.5231429310008817e-07,
+      "loss": 0.6024,
+      "step": 8885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7751520094120168,
+      "learning_rate": 1.5190222441205715e-07,
+      "loss": 0.5622,
+      "step": 8886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9421602467307941,
+      "learning_rate": 1.5149070528354238e-07,
+      "loss": 0.6058,
+      "step": 8887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6698253787569741,
+      "learning_rate": 1.510797357611915e-07,
+      "loss": 0.6074,
+      "step": 8888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9009707059362406,
+      "learning_rate": 1.5066931589159118e-07,
+      "loss": 0.6354,
+      "step": 8889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7195770528720253,
+      "learning_rate": 1.502594457212636e-07,
+      "loss": 0.6722,
+      "step": 8890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9207288432623015,
+      "learning_rate": 1.4985012529667052e-07,
+      "loss": 0.6761,
+      "step": 8891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.139605040626212,
+      "learning_rate": 1.4944135466421095e-07,
+      "loss": 0.6148,
+      "step": 8892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0452258517628152,
+      "learning_rate": 1.4903313387022168e-07,
+      "loss": 0.6248,
+      "step": 8893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.107910493475442,
+      "learning_rate": 1.4862546296097514e-07,
+      "loss": 0.6257,
+      "step": 8894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9371679169071094,
+      "learning_rate": 1.4821834198268493e-07,
+      "loss": 0.6685,
+      "step": 8895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.269015494506768,
+      "learning_rate": 1.4781177098149912e-07,
+      "loss": 0.6222,
+      "step": 8896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0322257763626053,
+      "learning_rate": 1.4740575000350643e-07,
+      "loss": 0.6172,
+      "step": 8897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9689096944036948,
+      "learning_rate": 1.4700027909473003e-07,
+      "loss": 0.5686,
+      "step": 8898
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8577584797794915,
+      "learning_rate": 1.4659535830113368e-07,
+      "loss": 0.7123,
+      "step": 8899
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.087469919331919,
+      "learning_rate": 1.461909876686174e-07,
+      "loss": 0.6695,
+      "step": 8900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0831701171131347,
+      "learning_rate": 1.4578716724301834e-07,
+      "loss": 0.5916,
+      "step": 8901
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.765591749554606,
+      "learning_rate": 1.4538389707011103e-07,
+      "loss": 0.5457,
+      "step": 8902
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0101458451306575,
+      "learning_rate": 1.4498117719561056e-07,
+      "loss": 0.655,
+      "step": 8903
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.057138304301439,
+      "learning_rate": 1.445790076651654e-07,
+      "loss": 0.6631,
+      "step": 8904
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.01208335411512,
+      "learning_rate": 1.4417738852436523e-07,
+      "loss": 0.6347,
+      "step": 8905
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.233841840702759,
+      "learning_rate": 1.4377631981873474e-07,
+      "loss": 0.6104,
+      "step": 8906
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.992020434510179,
+      "learning_rate": 1.4337580159373864e-07,
+      "loss": 0.5969,
+      "step": 8907
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9569295294426226,
+      "learning_rate": 1.4297583389477675e-07,
+      "loss": 0.5636,
+      "step": 8908
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0519576156254904,
+      "learning_rate": 1.4257641676718891e-07,
+      "loss": 0.6166,
+      "step": 8909
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9110514363828413,
+      "learning_rate": 1.4217755025624946e-07,
+      "loss": 0.6078,
+      "step": 8910
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8948115316189942,
+      "learning_rate": 1.4177923440717445e-07,
+      "loss": 0.6172,
+      "step": 8911
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.910510986217153,
+      "learning_rate": 1.413814692651133e-07,
+      "loss": 0.5804,
+      "step": 8912
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8590318355294009,
+      "learning_rate": 1.4098425487515665e-07,
+      "loss": 0.622,
+      "step": 8913
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9410471171303303,
+      "learning_rate": 1.4058759128232957e-07,
+      "loss": 0.5916,
+      "step": 8914
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.866173074226064,
+      "learning_rate": 1.4019147853159663e-07,
+      "loss": 0.5734,
+      "step": 8915
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9339751206748805,
+      "learning_rate": 1.397959166678603e-07,
+      "loss": 0.5967,
+      "step": 8916
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0814425508882373,
+      "learning_rate": 1.3940090573595854e-07,
+      "loss": 0.6032,
+      "step": 8917
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7468867551652556,
+      "learning_rate": 1.3900644578066892e-07,
+      "loss": 0.5055,
+      "step": 8918
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9505230853370727,
+      "learning_rate": 1.386125368467045e-07,
+      "loss": 0.5505,
+      "step": 8919
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8138209073254565,
+      "learning_rate": 1.3821917897871905e-07,
+      "loss": 0.5381,
+      "step": 8920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8811701807030197,
+      "learning_rate": 1.378263722213008e-07,
+      "loss": 0.5717,
+      "step": 8921
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6894404622210073,
+      "learning_rate": 1.3743411661897688e-07,
+      "loss": 0.5267,
+      "step": 8922
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.045490954150234,
+      "learning_rate": 1.3704241221621062e-07,
+      "loss": 0.6505,
+      "step": 8923
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8548965079774178,
+      "learning_rate": 1.366512590574054e-07,
+      "loss": 0.5188,
+      "step": 8924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1367292127451414,
+      "learning_rate": 1.3626065718689962e-07,
+      "loss": 0.6023,
+      "step": 8925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9485421579702578,
+      "learning_rate": 1.3587060664897178e-07,
+      "loss": 0.5993,
+      "step": 8926
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9473491176314763,
+      "learning_rate": 1.3548110748783426e-07,
+      "loss": 0.6362,
+      "step": 8927
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.150426581736567,
+      "learning_rate": 1.3509215974764067e-07,
+      "loss": 0.5799,
+      "step": 8928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.145747658974006,
+      "learning_rate": 1.347037634724796e-07,
+      "loss": 0.5796,
+      "step": 8929
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.955864982526995,
+      "learning_rate": 1.3431591870637862e-07,
+      "loss": 0.7073,
+      "step": 8930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0360528975248453,
+      "learning_rate": 1.3392862549330087e-07,
+      "loss": 0.6163,
+      "step": 8931
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.066226650062881,
+      "learning_rate": 1.3354188387715017e-07,
+      "loss": 0.6395,
+      "step": 8932
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7992072253913418,
+      "learning_rate": 1.3315569390176364e-07,
+      "loss": 0.4591,
+      "step": 8933
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9422148778935013,
+      "learning_rate": 1.3277005561092016e-07,
+      "loss": 0.6084,
+      "step": 8934
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.099312537538956,
+      "learning_rate": 1.3238496904833254e-07,
+      "loss": 0.6624,
+      "step": 8935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.106560416500555,
+      "learning_rate": 1.3200043425765364e-07,
+      "loss": 0.5163,
+      "step": 8936
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0539286083044805,
+      "learning_rate": 1.3161645128247247e-07,
+      "loss": 0.6933,
+      "step": 8937
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0500688798555298,
+      "learning_rate": 1.3123302016631477e-07,
+      "loss": 0.6283,
+      "step": 8938
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0385543350456485,
+      "learning_rate": 1.3085014095264527e-07,
+      "loss": 0.5445,
+      "step": 8939
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.279005385123845,
+      "learning_rate": 1.3046781368486638e-07,
+      "loss": 0.6009,
+      "step": 8940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.050341700548169,
+      "learning_rate": 1.3008603840631516e-07,
+      "loss": 0.6678,
+      "step": 8941
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8506971858384356,
+      "learning_rate": 1.2970481516026922e-07,
+      "loss": 0.4761,
+      "step": 8942
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.009727635049314,
+      "learning_rate": 1.293241439899423e-07,
+      "loss": 0.5926,
+      "step": 8943
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0818680696814704,
+      "learning_rate": 1.28944024938486e-07,
+      "loss": 0.6389,
+      "step": 8944
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.903971138639191,
+      "learning_rate": 1.2856445804898866e-07,
+      "loss": 0.5683,
+      "step": 8945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.254377222554142,
+      "learning_rate": 1.2818544336447637e-07,
+      "loss": 0.5999,
+      "step": 8946
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0107754811059886,
+      "learning_rate": 1.2780698092791145e-07,
+      "loss": 0.5586,
+      "step": 8947
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.811450767600721,
+      "learning_rate": 1.274290707821968e-07,
+      "loss": 0.5123,
+      "step": 8948
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2487747274358436,
+      "learning_rate": 1.2705171297016873e-07,
+      "loss": 0.6004,
+      "step": 8949
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0659723146256956,
+      "learning_rate": 1.2667490753460465e-07,
+      "loss": 0.6025,
+      "step": 8950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.857448937961545,
+      "learning_rate": 1.2629865451821656e-07,
+      "loss": 0.6056,
+      "step": 8951
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.179553087454338,
+      "learning_rate": 1.2592295396365527e-07,
+      "loss": 0.668,
+      "step": 8952
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.002595953566603,
+      "learning_rate": 1.2554780591350845e-07,
+      "loss": 0.5864,
+      "step": 8953
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0266575419282606,
+      "learning_rate": 1.2517321041030096e-07,
+      "loss": 0.6822,
+      "step": 8954
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9381132086572017,
+      "learning_rate": 1.2479916749649657e-07,
+      "loss": 0.5354,
+      "step": 8955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.877289303138945,
+      "learning_rate": 1.2442567721449307e-07,
+      "loss": 0.5831,
+      "step": 8956
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0137513658224586,
+      "learning_rate": 1.240527396066299e-07,
+      "loss": 0.5456,
+      "step": 8957
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8438678243217668,
+      "learning_rate": 1.2368035471517993e-07,
+      "loss": 0.578,
+      "step": 8958
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0226666721004958,
+      "learning_rate": 1.233085225823566e-07,
+      "loss": 0.6474,
+      "step": 8959
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9125971670943722,
+      "learning_rate": 1.2293724325030786e-07,
+      "loss": 0.6467,
+      "step": 8960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0296936030038264,
+      "learning_rate": 1.2256651676112118e-07,
+      "loss": 0.6714,
+      "step": 8961
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9650618330501934,
+      "learning_rate": 1.2219634315681962e-07,
+      "loss": 0.6185,
+      "step": 8962
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9626286181880022,
+      "learning_rate": 1.2182672247936566e-07,
+      "loss": 0.574,
+      "step": 8963
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8374055121902448,
+      "learning_rate": 1.214576547706575e-07,
+      "loss": 0.5862,
+      "step": 8964
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0329421955743108,
+      "learning_rate": 1.210891400725306e-07,
+      "loss": 0.5647,
+      "step": 8965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9931490923121717,
+      "learning_rate": 1.2072117842675867e-07,
+      "loss": 0.5806,
+      "step": 8966
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.829680075590195,
+      "learning_rate": 1.2035376987505288e-07,
+      "loss": 0.6675,
+      "step": 8967
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.847421420466505,
+      "learning_rate": 1.199869144590593e-07,
+      "loss": 0.5583,
+      "step": 8968
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8374941888305347,
+      "learning_rate": 1.196206122203647e-07,
+      "loss": 0.5854,
+      "step": 8969
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1554222663190084,
+      "learning_rate": 1.192548632004903e-07,
+      "loss": 0.7002,
+      "step": 8970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9239746198381393,
+      "learning_rate": 1.188896674408968e-07,
+      "loss": 0.6828,
+      "step": 8971
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.995032207458123,
+      "learning_rate": 1.1852502498298058e-07,
+      "loss": 0.5992,
+      "step": 8972
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8669745553621828,
+      "learning_rate": 1.1816093586807687e-07,
+      "loss": 0.5442,
+      "step": 8973
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1107807844678823,
+      "learning_rate": 1.1779740013745655e-07,
+      "loss": 0.5963,
+      "step": 8974
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8560952940963382,
+      "learning_rate": 1.174344178323289e-07,
+      "loss": 0.6525,
+      "step": 8975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.752115937083064,
+      "learning_rate": 1.1707198899383875e-07,
+      "loss": 0.6869,
+      "step": 8976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7421395713844312,
+      "learning_rate": 1.1671011366307105e-07,
+      "loss": 0.556,
+      "step": 8977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8270357499897079,
+      "learning_rate": 1.163487918810452e-07,
+      "loss": 0.5845,
+      "step": 8978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.740650769846731,
+      "learning_rate": 1.1598802368872009e-07,
+      "loss": 0.5365,
+      "step": 8979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8365042275780272,
+      "learning_rate": 1.1562780912699023e-07,
+      "loss": 0.5251,
+      "step": 8980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.886097167360647,
+      "learning_rate": 1.1526814823668853e-07,
+      "loss": 0.6388,
+      "step": 8981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8793682490212578,
+      "learning_rate": 1.1490904105858403e-07,
+      "loss": 0.6169,
+      "step": 8982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7536349174266053,
+      "learning_rate": 1.1455048763338361e-07,
+      "loss": 0.5711,
+      "step": 8983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.898534836280984,
+      "learning_rate": 1.1419248800173199e-07,
+      "loss": 0.5357,
+      "step": 8984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8602519762704828,
+      "learning_rate": 1.1383504220421004e-07,
+      "loss": 0.7386,
+      "step": 8985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.193413971204252,
+      "learning_rate": 1.1347815028133536e-07,
+      "loss": 0.6072,
+      "step": 8986
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.885655084013139,
+      "learning_rate": 1.1312181227356556e-07,
+      "loss": 0.7008,
+      "step": 8987
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.145617108224491,
+      "learning_rate": 1.1276602822129223e-07,
+      "loss": 0.6087,
+      "step": 8988
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1748127911102175,
+      "learning_rate": 1.1241079816484591e-07,
+      "loss": 0.6246,
+      "step": 8989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0305142220354355,
+      "learning_rate": 1.1205612214449434e-07,
+      "loss": 0.6088,
+      "step": 8990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8258785832904052,
+      "learning_rate": 1.1170200020044098e-07,
+      "loss": 0.549,
+      "step": 8991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9969958511662391,
+      "learning_rate": 1.1134843237282922e-07,
+      "loss": 0.5544,
+      "step": 8992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9296466850652296,
+      "learning_rate": 1.1099541870173591e-07,
+      "loss": 0.6445,
+      "step": 8993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8055438751017052,
+      "learning_rate": 1.1064295922717904e-07,
+      "loss": 0.6468,
+      "step": 8994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.133732513612682,
+      "learning_rate": 1.1029105398911055e-07,
+      "loss": 0.5653,
+      "step": 8995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8705602462539526,
+      "learning_rate": 1.0993970302742297e-07,
+      "loss": 0.6245,
+      "step": 8996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8390147969929742,
+      "learning_rate": 1.0958890638194108e-07,
+      "loss": 0.6049,
+      "step": 8997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9877021610759997,
+      "learning_rate": 1.0923866409243144e-07,
+      "loss": 0.5138,
+      "step": 8998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9376071697368407,
+      "learning_rate": 1.0888897619859506e-07,
+      "loss": 0.5826,
+      "step": 8999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8774347240206823,
+      "learning_rate": 1.0853984274007246e-07,
+      "loss": 0.64,
+      "step": 9000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.083899548076166,
+      "learning_rate": 1.0819126375643862e-07,
+      "loss": 0.6244,
+      "step": 9001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.158124305561164,
+      "learning_rate": 1.0784323928720753e-07,
+      "loss": 0.6079,
+      "step": 9002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.932121628840871,
+      "learning_rate": 1.0749576937182926e-07,
+      "loss": 0.5824,
+      "step": 9003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.052197137452352,
+      "learning_rate": 1.0714885404969288e-07,
+      "loss": 0.6556,
+      "step": 9004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9618385932007807,
+      "learning_rate": 1.0680249336012139e-07,
+      "loss": 0.5538,
+      "step": 9005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1643380059315005,
+      "learning_rate": 1.064566873423778e-07,
+      "loss": 0.5097,
+      "step": 9006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0854664855258283,
+      "learning_rate": 1.0611143603566076e-07,
+      "loss": 0.6295,
+      "step": 9007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9952119931589922,
+      "learning_rate": 1.0576673947910732e-07,
+      "loss": 0.599,
+      "step": 9008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9799496601050393,
+      "learning_rate": 1.0542259771178898e-07,
+      "loss": 0.6321,
+      "step": 9009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9816211810449393,
+      "learning_rate": 1.0507901077271843e-07,
+      "loss": 0.5686,
+      "step": 9010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8540076930182927,
+      "learning_rate": 1.0473597870084174e-07,
+      "loss": 0.5351,
+      "step": 9011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.741984167133165,
+      "learning_rate": 1.0439350153504446e-07,
+      "loss": 0.5656,
+      "step": 9012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9845451433362102,
+      "learning_rate": 1.0405157931414667e-07,
+      "loss": 0.596,
+      "step": 9013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6155972863453967,
+      "learning_rate": 1.0371021207690957e-07,
+      "loss": 0.4353,
+      "step": 9014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8418415679031528,
+      "learning_rate": 1.0336939986202666e-07,
+      "loss": 0.6544,
+      "step": 9015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.886523860804278,
+      "learning_rate": 1.0302914270813314e-07,
+      "loss": 0.6301,
+      "step": 9016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8822214646480715,
+      "learning_rate": 1.0268944065379704e-07,
+      "loss": 0.5913,
+      "step": 9017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0695162665439426,
+      "learning_rate": 1.0235029373752758e-07,
+      "loss": 0.629,
+      "step": 9018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.001389219375752,
+      "learning_rate": 1.0201170199776789e-07,
+      "loss": 0.5956,
+      "step": 9019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1057510275279165,
+      "learning_rate": 1.0167366547289892e-07,
+      "loss": 0.6267,
+      "step": 9020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9837252540821202,
+      "learning_rate": 1.0133618420123947e-07,
+      "loss": 0.6496,
+      "step": 9021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.04320201586192,
+      "learning_rate": 1.0099925822104562e-07,
+      "loss": 0.6135,
+      "step": 9022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9760632612491782,
+      "learning_rate": 1.0066288757050846e-07,
+      "loss": 0.6513,
+      "step": 9023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.037409421200088,
+      "learning_rate": 1.0032707228775917e-07,
+      "loss": 0.5312,
+      "step": 9024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0731308535241197,
+      "learning_rate": 9.999181241086231e-08,
+      "loss": 0.6405,
+      "step": 9025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0034900044017414,
+      "learning_rate": 9.965710797782358e-08,
+      "loss": 0.6196,
+      "step": 9026
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0664636409457473,
+      "learning_rate": 9.932295902658263e-08,
+      "loss": 0.7305,
+      "step": 9027
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0057196873279506,
+      "learning_rate": 9.898936559501637e-08,
+      "loss": 0.6518,
+      "step": 9028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.920647903527355,
+      "learning_rate": 9.865632772094124e-08,
+      "loss": 0.568,
+      "step": 9029
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.061404388610512,
+      "learning_rate": 9.832384544210704e-08,
+      "loss": 0.6644,
+      "step": 9030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.051063631357103,
+      "learning_rate": 9.799191879620474e-08,
+      "loss": 0.6747,
+      "step": 9031
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.11086077130743,
+      "learning_rate": 9.76605478208581e-08,
+      "loss": 0.5537,
+      "step": 9032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8661021852558275,
+      "learning_rate": 9.732973255363154e-08,
+      "loss": 0.5615,
+      "step": 9033
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8129931355838271,
+      "learning_rate": 9.699947303202339e-08,
+      "loss": 0.6432,
+      "step": 9034
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8858498295677275,
+      "learning_rate": 9.666976929347205e-08,
+      "loss": 0.5187,
+      "step": 9035
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.167973700167487,
+      "learning_rate": 9.634062137534927e-08,
+      "loss": 0.6469,
+      "step": 9036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8363965553392205,
+      "learning_rate": 9.601202931496745e-08,
+      "loss": 0.5113,
+      "step": 9037
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8634840550978502,
+      "learning_rate": 9.5683993149574e-08,
+      "loss": 0.6035,
+      "step": 9038
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2143595974193717,
+      "learning_rate": 9.535651291635362e-08,
+      "loss": 0.6819,
+      "step": 9039
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6812826245515171,
+      "learning_rate": 9.502958865242774e-08,
+      "loss": 0.5876,
+      "step": 9040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8248003304779605,
+      "learning_rate": 9.470322039485614e-08,
+      "loss": 0.5248,
+      "step": 9041
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9421864031203093,
+      "learning_rate": 9.437740818063312e-08,
+      "loss": 0.6556,
+      "step": 9042
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.906943046443007,
+      "learning_rate": 9.405215204669193e-08,
+      "loss": 0.5409,
+      "step": 9043
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0185391093753142,
+      "learning_rate": 9.372745202990253e-08,
+      "loss": 0.6874,
+      "step": 9044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9843021074005947,
+      "learning_rate": 9.3403308167071e-08,
+      "loss": 0.5017,
+      "step": 9045
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9854311881929885,
+      "learning_rate": 9.30797204949413e-08,
+      "loss": 0.6676,
+      "step": 9046
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0826537011454134,
+      "learning_rate": 9.275668905019409e-08,
+      "loss": 0.6396,
+      "step": 9047
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.039120363907294,
+      "learning_rate": 9.243421386944618e-08,
+      "loss": 0.65,
+      "step": 9048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8978766367979325,
+      "learning_rate": 9.211229498925278e-08,
+      "loss": 0.5569,
+      "step": 9049
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7595071125989983,
+      "learning_rate": 9.179093244610415e-08,
+      "loss": 0.486,
+      "step": 9050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.890778877362623,
+      "learning_rate": 9.147012627642949e-08,
+      "loss": 0.6964,
+      "step": 9051
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.005268335224658,
+      "learning_rate": 9.114987651659357e-08,
+      "loss": 0.5917,
+      "step": 9052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1730014002086895,
+      "learning_rate": 9.083018320289849e-08,
+      "loss": 0.6614,
+      "step": 9053
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9735903938030195,
+      "learning_rate": 9.051104637158359e-08,
+      "loss": 0.5689,
+      "step": 9054
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8456661586038106,
+      "learning_rate": 9.019246605882492e-08,
+      "loss": 0.5551,
+      "step": 9055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.052240286399753,
+      "learning_rate": 8.987444230073528e-08,
+      "loss": 0.6064,
+      "step": 9056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8535713753595429,
+      "learning_rate": 8.955697513336415e-08,
+      "loss": 0.5835,
+      "step": 9057
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8303132592401052,
+      "learning_rate": 8.924006459269886e-08,
+      "loss": 0.5045,
+      "step": 9058
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.901318346829754,
+      "learning_rate": 8.892371071466176e-08,
+      "loss": 0.6645,
+      "step": 9059
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9257932527237893,
+      "learning_rate": 8.860791353511532e-08,
+      "loss": 0.6335,
+      "step": 9060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9043707207031637,
+      "learning_rate": 8.829267308985535e-08,
+      "loss": 0.5067,
+      "step": 9061
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9917505790714765,
+      "learning_rate": 8.797798941461655e-08,
+      "loss": 0.6557,
+      "step": 9062
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8937462209908957,
+      "learning_rate": 8.766386254507043e-08,
+      "loss": 0.5871,
+      "step": 9063
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9406431576153542,
+      "learning_rate": 8.735029251682515e-08,
+      "loss": 0.5824,
+      "step": 9064
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.969695305422505,
+      "learning_rate": 8.703727936542505e-08,
+      "loss": 0.6058,
+      "step": 9065
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9684624208295343,
+      "learning_rate": 8.672482312635233e-08,
+      "loss": 0.7302,
+      "step": 9066
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9063600516524655,
+      "learning_rate": 8.641292383502531e-08,
+      "loss": 0.5817,
+      "step": 9067
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.913210044504492,
+      "learning_rate": 8.610158152680071e-08,
+      "loss": 0.6381,
+      "step": 9068
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9742629875445776,
+      "learning_rate": 8.579079623696917e-08,
+      "loss": 0.7128,
+      "step": 9069
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8004319724207307,
+      "learning_rate": 8.548056800076199e-08,
+      "loss": 0.6184,
+      "step": 9070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8260617741245229,
+      "learning_rate": 8.517089685334323e-08,
+      "loss": 0.665,
+      "step": 9071
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0768267807130627,
+      "learning_rate": 8.486178282981761e-08,
+      "loss": 0.701,
+      "step": 9072
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8819509447944622,
+      "learning_rate": 8.455322596522375e-08,
+      "loss": 0.6165,
+      "step": 9073
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9499123138063066,
+      "learning_rate": 8.424522629453924e-08,
+      "loss": 0.6395,
+      "step": 9074
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.978374390635286,
+      "learning_rate": 8.39377838526767e-08,
+      "loss": 0.597,
+      "step": 9075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8940455664553872,
+      "learning_rate": 8.363089867448715e-08,
+      "loss": 0.5436,
+      "step": 9076
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9531936063259534,
+      "learning_rate": 8.332457079475831e-08,
+      "loss": 0.5746,
+      "step": 9077
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.823839062817681,
+      "learning_rate": 8.301880024821296e-08,
+      "loss": 0.5773,
+      "step": 9078
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0376195888887345,
+      "learning_rate": 8.271358706951227e-08,
+      "loss": 0.5653,
+      "step": 9079
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.999992907214282,
+      "learning_rate": 8.240893129325412e-08,
+      "loss": 0.6454,
+      "step": 9080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8778664959112257,
+      "learning_rate": 8.210483295397309e-08,
+      "loss": 0.6124,
+      "step": 9081
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.173549074135219,
+      "learning_rate": 8.180129208613996e-08,
+      "loss": 0.6335,
+      "step": 9082
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9313462323770607,
+      "learning_rate": 8.149830872416331e-08,
+      "loss": 0.6208,
+      "step": 9083
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0185542339124622,
+      "learning_rate": 8.119588290238845e-08,
+      "loss": 0.6189,
+      "step": 9084
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9778143235997587,
+      "learning_rate": 8.089401465509628e-08,
+      "loss": 0.6107,
+      "step": 9085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8124396471473323,
+      "learning_rate": 8.059270401650555e-08,
+      "loss": 0.6052,
+      "step": 9086
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8136504927323756,
+      "learning_rate": 8.029195102077114e-08,
+      "loss": 0.6718,
+      "step": 9087
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9373249127008352,
+      "learning_rate": 7.999175570198526e-08,
+      "loss": 0.5822,
+      "step": 9088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.048051219588817,
+      "learning_rate": 7.969211809417732e-08,
+      "loss": 0.6618,
+      "step": 9089
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9640947013669483,
+      "learning_rate": 7.939303823131295e-08,
+      "loss": 0.6105,
+      "step": 9090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8903042757750237,
+      "learning_rate": 7.909451614729335e-08,
+      "loss": 0.5093,
+      "step": 9091
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9700603234134746,
+      "learning_rate": 7.87965518759598e-08,
+      "loss": 0.6289,
+      "step": 9092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.068173773432267,
+      "learning_rate": 7.84991454510864e-08,
+      "loss": 0.6136,
+      "step": 9093
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9719085980940332,
+      "learning_rate": 7.820229690638615e-08,
+      "loss": 0.5706,
+      "step": 9094
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8669821696404074,
+      "learning_rate": 7.790600627550937e-08,
+      "loss": 0.6763,
+      "step": 9095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0351969275458415,
+      "learning_rate": 7.761027359204088e-08,
+      "loss": 0.6411,
+      "step": 9096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.034610079179534,
+      "learning_rate": 7.731509888950551e-08,
+      "loss": 0.5367,
+      "step": 9097
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9903590314113853,
+      "learning_rate": 7.702048220136149e-08,
+      "loss": 0.6549,
+      "step": 9098
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.040096970601361,
+      "learning_rate": 7.672642356100656e-08,
+      "loss": 0.5158,
+      "step": 9099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8819140831818857,
+      "learning_rate": 7.643292300177296e-08,
+      "loss": 0.4779,
+      "step": 9100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.684999361885513,
+      "learning_rate": 7.613998055693073e-08,
+      "loss": 0.6065,
+      "step": 9101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0139571541265853,
+      "learning_rate": 7.584759625968663e-08,
+      "loss": 0.554,
+      "step": 9102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8729824867562832,
+      "learning_rate": 7.555577014318472e-08,
+      "loss": 0.5385,
+      "step": 9103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.080253251347515,
+      "learning_rate": 7.526450224050407e-08,
+      "loss": 0.5282,
+      "step": 9104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8347182779960427,
+      "learning_rate": 7.497379258466275e-08,
+      "loss": 0.5965,
+      "step": 9105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.164534043831013,
+      "learning_rate": 7.468364120861272e-08,
+      "loss": 0.6141,
+      "step": 9106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2212477876578496,
+      "learning_rate": 7.439404814524654e-08,
+      "loss": 0.5259,
+      "step": 9107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.87666751211576,
+      "learning_rate": 7.410501342738963e-08,
+      "loss": 0.5406,
+      "step": 9108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.042540734017541,
+      "learning_rate": 7.381653708780578e-08,
+      "loss": 0.6218,
+      "step": 9109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9767662921202052,
+      "learning_rate": 7.352861915919607e-08,
+      "loss": 0.5197,
+      "step": 9110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9563546830435936,
+      "learning_rate": 7.324125967419715e-08,
+      "loss": 0.6026,
+      "step": 9111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0233014232164117,
+      "learning_rate": 7.295445866538297e-08,
+      "loss": 0.5401,
+      "step": 9112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.059552075508856,
+      "learning_rate": 7.26682161652642e-08,
+      "loss": 0.5435,
+      "step": 9113
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.13299560125978,
+      "learning_rate": 7.238253220628822e-08,
+      "loss": 0.7473,
+      "step": 9114
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2557692843805333,
+      "learning_rate": 7.20974068208391e-08,
+      "loss": 0.5964,
+      "step": 9115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9903729640281569,
+      "learning_rate": 7.181284004123601e-08,
+      "loss": 0.6408,
+      "step": 9116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8199528667987253,
+      "learning_rate": 7.152883189973759e-08,
+      "loss": 0.5784,
+      "step": 9117
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.35897713970099,
+      "learning_rate": 7.124538242853751e-08,
+      "loss": 0.6388,
+      "step": 9118
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1030461531404296,
+      "learning_rate": 7.096249165976621e-08,
+      "loss": 0.6247,
+      "step": 9119
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8662951848090812,
+      "learning_rate": 7.06801596254908e-08,
+      "loss": 0.6072,
+      "step": 9120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9138440363291855,
+      "learning_rate": 7.039838635771623e-08,
+      "loss": 0.5645,
+      "step": 9121
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.734258665755537,
+      "learning_rate": 7.011717188838196e-08,
+      "loss": 0.604,
+      "step": 9122
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.196151154914006,
+      "learning_rate": 6.983651624936527e-08,
+      "loss": 0.5313,
+      "step": 9123
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7132829172978423,
+      "learning_rate": 6.955641947248127e-08,
+      "loss": 0.5929,
+      "step": 9124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9418755624331163,
+      "learning_rate": 6.927688158947954e-08,
+      "loss": 0.5897,
+      "step": 9125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9872540559183436,
+      "learning_rate": 6.899790263204643e-08,
+      "loss": 0.5944,
+      "step": 9126
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1639015797194205,
+      "learning_rate": 6.871948263180772e-08,
+      "loss": 0.6074,
+      "step": 9127
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9873225783662412,
+      "learning_rate": 6.844162162032265e-08,
+      "loss": 0.5684,
+      "step": 9128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0519891397743266,
+      "learning_rate": 6.816431962908931e-08,
+      "loss": 0.6091,
+      "step": 9129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9955079387546661,
+      "learning_rate": 6.788757668954038e-08,
+      "loss": 0.5752,
+      "step": 9130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0818890242182357,
+      "learning_rate": 6.761139283304685e-08,
+      "loss": 0.6259,
+      "step": 9131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9861722922728569,
+      "learning_rate": 6.73357680909159e-08,
+      "loss": 0.6536,
+      "step": 9132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.955473365977711,
+      "learning_rate": 6.706070249439034e-08,
+      "loss": 0.5794,
+      "step": 9133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8569417282687593,
+      "learning_rate": 6.678619607465131e-08,
+      "loss": 0.5563,
+      "step": 9134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9271564060152484,
+      "learning_rate": 6.651224886281504e-08,
+      "loss": 0.6414,
+      "step": 9135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0471491330858442,
+      "learning_rate": 6.623886088993559e-08,
+      "loss": 0.5188,
+      "step": 9136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9551415429882093,
+      "learning_rate": 6.59660321870026e-08,
+      "loss": 0.5605,
+      "step": 9137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.083480071147602,
+      "learning_rate": 6.5693762784943e-08,
+      "loss": 0.5904,
+      "step": 9138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.828002306812625,
+      "learning_rate": 6.542205271461988e-08,
+      "loss": 0.5919,
+      "step": 9139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0114884598779534,
+      "learning_rate": 6.515090200683361e-08,
+      "loss": 0.5696,
+      "step": 9140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8440201739093747,
+      "learning_rate": 6.48803106923196e-08,
+      "loss": 0.6082,
+      "step": 9141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7543344260471143,
+      "learning_rate": 6.461027880175219e-08,
+      "loss": 0.6036,
+      "step": 9142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7472069441257652,
+      "learning_rate": 6.434080636574025e-08,
+      "loss": 0.6077,
+      "step": 9143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8148441768694168,
+      "learning_rate": 6.407189341483044e-08,
+      "loss": 0.5992,
+      "step": 9144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8986202705276924,
+      "learning_rate": 6.380353997950506e-08,
+      "loss": 0.5562,
+      "step": 9145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0313020160531634,
+      "learning_rate": 6.35357460901842e-08,
+      "loss": 0.5758,
+      "step": 9146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9249252266461854,
+      "learning_rate": 6.326851177722304e-08,
+      "loss": 0.5855,
+      "step": 9147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.302195837407535,
+      "learning_rate": 6.300183707091457e-08,
+      "loss": 0.5979,
+      "step": 9148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.942121352149436,
+      "learning_rate": 6.273572200148792e-08,
+      "loss": 0.5166,
+      "step": 9149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.821356962288708,
+      "learning_rate": 6.247016659910842e-08,
+      "loss": 0.6556,
+      "step": 9150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0069073914638875,
+      "learning_rate": 6.220517089387867e-08,
+      "loss": 0.6701,
+      "step": 9151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7514971240436819,
+      "learning_rate": 6.194073491583796e-08,
+      "loss": 0.5036,
+      "step": 9152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0429327714393395,
+      "learning_rate": 6.167685869495954e-08,
+      "loss": 0.6446,
+      "step": 9153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.136167526614985,
+      "learning_rate": 6.141354226115781e-08,
+      "loss": 0.6722,
+      "step": 9154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.036591943231998,
+      "learning_rate": 6.115078564427946e-08,
+      "loss": 0.6177,
+      "step": 9155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.957728748548829,
+      "learning_rate": 6.088858887411064e-08,
+      "loss": 0.5896,
+      "step": 9156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8733656466152195,
+      "learning_rate": 6.06269519803715e-08,
+      "loss": 0.6692,
+      "step": 9157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9703078943600634,
+      "learning_rate": 6.036587499272161e-08,
+      "loss": 0.5802,
+      "step": 9158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6708428863614988,
+      "learning_rate": 6.010535794075455e-08,
+      "loss": 0.4799,
+      "step": 9159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.063875697905319,
+      "learning_rate": 5.984540085400114e-08,
+      "loss": 0.57,
+      "step": 9160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.85335140179406,
+      "learning_rate": 5.9586003761930024e-08,
+      "loss": 0.5223,
+      "step": 9161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7270860909482966,
+      "learning_rate": 5.932716669394489e-08,
+      "loss": 0.5401,
+      "step": 9162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0866778648399165,
+      "learning_rate": 5.9068889679385624e-08,
+      "loss": 0.563,
+      "step": 9163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.047048714162409,
+      "learning_rate": 5.8811172747530457e-08,
+      "loss": 0.5375,
+      "step": 9164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8605628615428378,
+      "learning_rate": 5.855401592759269e-08,
+      "loss": 0.5913,
+      "step": 9165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6883951918162357,
+      "learning_rate": 5.8297419248722345e-08,
+      "loss": 0.5417,
+      "step": 9166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9874976605092245,
+      "learning_rate": 5.80413827400067e-08,
+      "loss": 0.5758,
+      "step": 9167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.095976679677836,
+      "learning_rate": 5.778590643046811e-08,
+      "loss": 0.5551,
+      "step": 9168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.368713155416316,
+      "learning_rate": 5.7530990349067285e-08,
+      "loss": 0.5212,
+      "step": 9169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9132853451864427,
+      "learning_rate": 5.727663452469945e-08,
+      "loss": 0.6589,
+      "step": 9170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8104655617619578,
+      "learning_rate": 5.702283898619765e-08,
+      "loss": 0.5446,
+      "step": 9171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.895549005171521,
+      "learning_rate": 5.6769603762331096e-08,
+      "loss": 0.6207,
+      "step": 9172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.010640938641115,
+      "learning_rate": 5.651692888180516e-08,
+      "loss": 0.5883,
+      "step": 9173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9263129999779227,
+      "learning_rate": 5.626481437326303e-08,
+      "loss": 0.6767,
+      "step": 9174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9656014437831926,
+      "learning_rate": 5.601326026528187e-08,
+      "loss": 0.6445,
+      "step": 9175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0843236537066003,
+      "learning_rate": 5.5762266586377734e-08,
+      "loss": 0.6064,
+      "step": 9176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2182950535700776,
+      "learning_rate": 5.551183336500177e-08,
+      "loss": 0.597,
+      "step": 9177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0720744469238306,
+      "learning_rate": 5.526196062954181e-08,
+      "loss": 0.5568,
+      "step": 9178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9350111511493413,
+      "learning_rate": 5.501264840832299e-08,
+      "loss": 0.6638,
+      "step": 9179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9323446366770365,
+      "learning_rate": 5.4763896729606023e-08,
+      "loss": 0.578,
+      "step": 9180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.09711440356542,
+      "learning_rate": 5.451570562158892e-08,
+      "loss": 0.6727,
+      "step": 9181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8486259014947013,
+      "learning_rate": 5.4268075112403615e-08,
+      "loss": 0.5326,
+      "step": 9182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.002990590014281,
+      "learning_rate": 5.402100523012266e-08,
+      "loss": 0.6319,
+      "step": 9183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.762628792862233,
+      "learning_rate": 5.377449600275142e-08,
+      "loss": 0.5471,
+      "step": 9184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.80757756492264,
+      "learning_rate": 5.352854745823366e-08,
+      "loss": 0.5654,
+      "step": 9185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.111766800819143,
+      "learning_rate": 5.3283159624448745e-08,
+      "loss": 0.6046,
+      "step": 9186
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.136818353546279,
+      "learning_rate": 5.3038332529213865e-08,
+      "loss": 0.633,
+      "step": 9187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7831004493513567,
+      "learning_rate": 5.279406620028016e-08,
+      "loss": 0.5601,
+      "step": 9188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9671860829871322,
+      "learning_rate": 5.255036066533714e-08,
+      "loss": 0.6037,
+      "step": 9189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2695919619215563,
+      "learning_rate": 5.230721595201049e-08,
+      "loss": 0.6201,
+      "step": 9190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9143495780323916,
+      "learning_rate": 5.20646320878615e-08,
+      "loss": 0.6209,
+      "step": 9191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7335234797845942,
+      "learning_rate": 5.182260910038928e-08,
+      "loss": 0.5238,
+      "step": 9192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1520062494936623,
+      "learning_rate": 5.1581147017027434e-08,
+      "loss": 0.5854,
+      "step": 9193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8251696136209343,
+      "learning_rate": 5.134024586514796e-08,
+      "loss": 0.5771,
+      "step": 9194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8451479021419956,
+      "learning_rate": 5.109990567205792e-08,
+      "loss": 0.5387,
+      "step": 9195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9829306324790406,
+      "learning_rate": 5.08601264650016e-08,
+      "loss": 0.5184,
+      "step": 9196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0004402204722864,
+      "learning_rate": 5.06209082711584e-08,
+      "loss": 0.5989,
+      "step": 9197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9331331565930703,
+      "learning_rate": 5.038225111764605e-08,
+      "loss": 0.6236,
+      "step": 9198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1842830535619893,
+      "learning_rate": 5.0144155031517926e-08,
+      "loss": 0.5366,
+      "step": 9199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0003594845759616,
+      "learning_rate": 4.990662003976243e-08,
+      "loss": 0.6162,
+      "step": 9200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9969106571586461,
+      "learning_rate": 4.966964616930692e-08,
+      "loss": 0.6471,
+      "step": 9201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.971192026882797,
+      "learning_rate": 4.943323344701212e-08,
+      "loss": 0.6109,
+      "step": 9202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.158762303401497,
+      "learning_rate": 4.919738189967771e-08,
+      "loss": 0.6428,
+      "step": 9203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6185547120279675,
+      "learning_rate": 4.8962091554039524e-08,
+      "loss": 0.6368,
+      "step": 9204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.015253292430434,
+      "learning_rate": 4.8727362436767344e-08,
+      "loss": 0.6126,
+      "step": 9205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.116270646369772,
+      "learning_rate": 4.8493194574470436e-08,
+      "loss": 0.6873,
+      "step": 9206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7183519515168044,
+      "learning_rate": 4.825958799369201e-08,
+      "loss": 0.508,
+      "step": 9207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7666831560398584,
+      "learning_rate": 4.802654272091367e-08,
+      "loss": 0.5841,
+      "step": 9208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8230875968232338,
+      "learning_rate": 4.779405878255206e-08,
+      "loss": 0.5549,
+      "step": 9209
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.152194890452504,
+      "learning_rate": 4.756213620496053e-08,
+      "loss": 0.6126,
+      "step": 9210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.015006634328195,
+      "learning_rate": 4.733077501442862e-08,
+      "loss": 0.6799,
+      "step": 9211
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.256874151082544,
+      "learning_rate": 4.709997523718257e-08,
+      "loss": 0.5797,
+      "step": 9212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9338870307485518,
+      "learning_rate": 4.6869736899385345e-08,
+      "loss": 0.5389,
+      "step": 9213
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.866170129258445,
+      "learning_rate": 4.664006002713495e-08,
+      "loss": 0.6559,
+      "step": 9214
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1576898953891415,
+      "learning_rate": 4.641094464646667e-08,
+      "loss": 0.6158,
+      "step": 9215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8488446862845445,
+      "learning_rate": 4.618239078335307e-08,
+      "loss": 0.5886,
+      "step": 9216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.046528350714155,
+      "learning_rate": 4.5954398463700647e-08,
+      "loss": 0.6249,
+      "step": 9217
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.860971563552356,
+      "learning_rate": 4.572696771335483e-08,
+      "loss": 0.6386,
+      "step": 9218
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9383068774147776,
+      "learning_rate": 4.5500098558095565e-08,
+      "loss": 0.5341,
+      "step": 9219
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1917997000300655,
+      "learning_rate": 4.5273791023639494e-08,
+      "loss": 0.5904,
+      "step": 9220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0815462160143805,
+      "learning_rate": 4.504804513564054e-08,
+      "loss": 0.6328,
+      "step": 9221
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8893378428676797,
+      "learning_rate": 4.482286091968768e-08,
+      "loss": 0.5827,
+      "step": 9222
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8432453644442008,
+      "learning_rate": 4.459823840130717e-08,
+      "loss": 0.5942,
+      "step": 9223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0539820466658094,
+      "learning_rate": 4.437417760596141e-08,
+      "loss": 0.5402,
+      "step": 9224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0722178972880947,
+      "learning_rate": 4.415067855904842e-08,
+      "loss": 0.6071,
+      "step": 9225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2177273773031763,
+      "learning_rate": 4.392774128590349e-08,
+      "loss": 0.6256,
+      "step": 9226
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0694785751628038,
+      "learning_rate": 4.3705365811797515e-08,
+      "loss": 0.668,
+      "step": 9227
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8951352083608757,
+      "learning_rate": 4.348355216193867e-08,
+      "loss": 0.6042,
+      "step": 9228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.028419678421276,
+      "learning_rate": 4.326230036147017e-08,
+      "loss": 0.5914,
+      "step": 9229
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0823053199915336,
+      "learning_rate": 4.30416104354725e-08,
+      "loss": 0.5709,
+      "step": 9230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.823570228290745,
+      "learning_rate": 4.282148240896178e-08,
+      "loss": 0.599,
+      "step": 9231
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.74260159641539,
+      "learning_rate": 4.2601916306891365e-08,
+      "loss": 0.5198,
+      "step": 9232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7015132786393319,
+      "learning_rate": 4.2382912154150244e-08,
+      "loss": 0.5233,
+      "step": 9233
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.961974043992251,
+      "learning_rate": 4.216446997556245e-08,
+      "loss": 0.5958,
+      "step": 9234
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.013194471656937,
+      "learning_rate": 4.19465897958915e-08,
+      "loss": 0.541,
+      "step": 9235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.210012908565479,
+      "learning_rate": 4.1729271639834315e-08,
+      "loss": 0.6799,
+      "step": 9236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0115717834978177,
+      "learning_rate": 4.151251553202562e-08,
+      "loss": 0.5222,
+      "step": 9237
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9749451130038853,
+      "learning_rate": 4.129632149703522e-08,
+      "loss": 0.6103,
+      "step": 9238
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8871785221244468,
+      "learning_rate": 4.1080689559370724e-08,
+      "loss": 0.6575,
+      "step": 9239
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.039353039205422,
+      "learning_rate": 4.086561974347536e-08,
+      "loss": 0.6047,
+      "step": 9240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.191591399613144,
+      "learning_rate": 4.065111207372796e-08,
+      "loss": 0.6459,
+      "step": 9241
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.071197060104028,
+      "learning_rate": 4.043716657444407e-08,
+      "loss": 0.6256,
+      "step": 9242
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.920300538975992,
+      "learning_rate": 4.022378326987597e-08,
+      "loss": 0.6548,
+      "step": 9243
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0560108290708023,
+      "learning_rate": 4.001096218421152e-08,
+      "loss": 0.5192,
+      "step": 9244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.002316195882517,
+      "learning_rate": 3.9798703341575875e-08,
+      "loss": 0.5916,
+      "step": 9245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.191893706906724,
+      "learning_rate": 3.9587006766029225e-08,
+      "loss": 0.7022,
+      "step": 9246
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.960519681813049,
+      "learning_rate": 3.937587248156904e-08,
+      "loss": 0.5285,
+      "step": 9247
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1307289887200067,
+      "learning_rate": 3.916530051212841e-08,
+      "loss": 0.7106,
+      "step": 9248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0097373182537246,
+      "learning_rate": 3.8955290881576566e-08,
+      "loss": 0.6415,
+      "step": 9249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.088973178619039,
+      "learning_rate": 3.874584361371947e-08,
+      "loss": 0.619,
+      "step": 9250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3453820853091965,
+      "learning_rate": 3.8536958732299234e-08,
+      "loss": 0.7008,
+      "step": 9251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8925152599216863,
+      "learning_rate": 3.832863626099359e-08,
+      "loss": 0.6295,
+      "step": 9252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9730246462677488,
+      "learning_rate": 3.8120876223418646e-08,
+      "loss": 0.6821,
+      "step": 9253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.995162008625557,
+      "learning_rate": 3.791367864312334e-08,
+      "loss": 0.5596,
+      "step": 9254
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9608437281347626,
+      "learning_rate": 3.770704354359611e-08,
+      "loss": 0.6098,
+      "step": 9255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9119220279863318,
+      "learning_rate": 3.750097094825933e-08,
+      "loss": 0.5848,
+      "step": 9256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8167225502830717,
+      "learning_rate": 3.729546088047264e-08,
+      "loss": 0.5353,
+      "step": 9257
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8657223958933558,
+      "learning_rate": 3.709051336353187e-08,
+      "loss": 0.5794,
+      "step": 9258
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3673185647016375,
+      "learning_rate": 3.688612842066952e-08,
+      "loss": 0.6088,
+      "step": 9259
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.908061434806496,
+      "learning_rate": 3.6682306075052634e-08,
+      "loss": 0.519,
+      "step": 9260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9989992305528361,
+      "learning_rate": 3.64790463497866e-08,
+      "loss": 0.5655,
+      "step": 9261
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.077972598614312,
+      "learning_rate": 3.627634926791246e-08,
+      "loss": 0.6019,
+      "step": 9262
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9263719734688765,
+      "learning_rate": 3.6074214852405695e-08,
+      "loss": 0.6626,
+      "step": 9263
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8445050717486768,
+      "learning_rate": 3.587264312618022e-08,
+      "loss": 0.5336,
+      "step": 9264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.085353318157501,
+      "learning_rate": 3.567163411208552e-08,
+      "loss": 0.5532,
+      "step": 9265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8421560673376067,
+      "learning_rate": 3.547118783290615e-08,
+      "loss": 0.6325,
+      "step": 9266
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9754557962477426,
+      "learning_rate": 3.527130431136505e-08,
+      "loss": 0.5936,
+      "step": 9267
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.199110288587306,
+      "learning_rate": 3.507198357011909e-08,
+      "loss": 0.6782,
+      "step": 9268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9254296087735812,
+      "learning_rate": 3.487322563176354e-08,
+      "loss": 0.5547,
+      "step": 9269
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0537058557901986,
+      "learning_rate": 3.467503051882815e-08,
+      "loss": 0.5562,
+      "step": 9270
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8764811043642,
+      "learning_rate": 3.4477398253778826e-08,
+      "loss": 0.6119,
+      "step": 9271
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9573268571815328,
+      "learning_rate": 3.4280328859019885e-08,
+      "loss": 0.6284,
+      "step": 9272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2651084124987184,
+      "learning_rate": 3.408382235688845e-08,
+      "loss": 0.5998,
+      "step": 9273
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7190886471188342,
+      "learning_rate": 3.388787876966115e-08,
+      "loss": 0.5464,
+      "step": 9274
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.898764604876328,
+      "learning_rate": 3.3692498119548e-08,
+      "loss": 0.5864,
+      "step": 9275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9545866295420675,
+      "learning_rate": 3.3497680428697943e-08,
+      "loss": 0.6006,
+      "step": 9276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9882529876550887,
+      "learning_rate": 3.330342571919332e-08,
+      "loss": 0.6126,
+      "step": 9277
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1666195490031344,
+      "learning_rate": 3.3109734013055396e-08,
+      "loss": 0.7462,
+      "step": 9278
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8273279290401259,
+      "learning_rate": 3.2916605332238284e-08,
+      "loss": 0.6519,
+      "step": 9279
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.25238653257137,
+      "learning_rate": 3.2724039698636135e-08,
+      "loss": 0.6955,
+      "step": 9280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1016705536415943,
+      "learning_rate": 3.2532037134076486e-08,
+      "loss": 0.6079,
+      "step": 9281
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0163167940465256,
+      "learning_rate": 3.234059766032416e-08,
+      "loss": 0.56,
+      "step": 9282
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9318858697010097,
+      "learning_rate": 3.214972129907956e-08,
+      "loss": 0.6073,
+      "step": 9283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7757658787840196,
+      "learning_rate": 3.195940807198039e-08,
+      "loss": 0.5379,
+      "step": 9284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8460998635082828,
+      "learning_rate": 3.1769658000598835e-08,
+      "loss": 0.5544,
+      "step": 9285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0236445760390427,
+      "learning_rate": 3.158047110644436e-08,
+      "loss": 0.5358,
+      "step": 9286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9270061530827873,
+      "learning_rate": 3.1391847410962573e-08,
+      "loss": 0.6229,
+      "step": 9287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9604700327535756,
+      "learning_rate": 3.1203786935535275e-08,
+      "loss": 0.6332,
+      "step": 9288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9975031878953207,
+      "learning_rate": 3.1016289701479296e-08,
+      "loss": 0.552,
+      "step": 9289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8939971445756167,
+      "learning_rate": 3.082935573004986e-08,
+      "loss": 0.5689,
+      "step": 9290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.804739917206502,
+      "learning_rate": 3.064298504243612e-08,
+      "loss": 0.6185,
+      "step": 9291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9170101195891174,
+      "learning_rate": 3.0457177659764524e-08,
+      "loss": 0.6434,
+      "step": 9292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.934719583858853,
+      "learning_rate": 3.02719336030971e-08,
+      "loss": 0.6144,
+      "step": 9293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.929080277217315,
+      "learning_rate": 3.008725289343206e-08,
+      "loss": 0.6113,
+      "step": 9294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0128393803689226,
+      "learning_rate": 2.990313555170488e-08,
+      "loss": 0.6372,
+      "step": 9295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9044953243775646,
+      "learning_rate": 2.9719581598786072e-08,
+      "loss": 0.6269,
+      "step": 9296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.987130561174613,
+      "learning_rate": 2.953659105548179e-08,
+      "loss": 0.5736,
+      "step": 9297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9601612132663782,
+      "learning_rate": 2.9354163942535983e-08,
+      "loss": 0.5485,
+      "step": 9298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8276899215854248,
+      "learning_rate": 2.9172300280627674e-08,
+      "loss": 0.5822,
+      "step": 9299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9585863100809016,
+      "learning_rate": 2.899100009037148e-08,
+      "loss": 0.6306,
+      "step": 9300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8717322423573106,
+      "learning_rate": 2.8810263392319293e-08,
+      "loss": 0.5468,
+      "step": 9301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8433778506189058,
+      "learning_rate": 2.863009020695917e-08,
+      "loss": 0.6548,
+      "step": 9302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8375022349395036,
+      "learning_rate": 2.845048055471311e-08,
+      "loss": 0.635,
+      "step": 9303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.986408985770117,
+      "learning_rate": 2.8271434455942604e-08,
+      "loss": 0.6994,
+      "step": 9304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7962105054371804,
+      "learning_rate": 2.809295193094308e-08,
+      "loss": 0.6022,
+      "step": 9305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.129112709564285,
+      "learning_rate": 2.7915032999946133e-08,
+      "loss": 0.6974,
+      "step": 9306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2015618631548834,
+      "learning_rate": 2.7737677683120077e-08,
+      "loss": 0.5899,
+      "step": 9307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.032080111074871,
+      "learning_rate": 2.7560886000569386e-08,
+      "loss": 0.6206,
+      "step": 9308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9551756649376424,
+      "learning_rate": 2.7384657972334137e-08,
+      "loss": 0.5469,
+      "step": 9309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.957396990331527,
+      "learning_rate": 2.7208993618390578e-08,
+      "loss": 0.6458,
+      "step": 9310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9599537127158781,
+      "learning_rate": 2.7033892958651665e-08,
+      "loss": 0.6791,
+      "step": 9311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7920889911825337,
+      "learning_rate": 2.6859356012965964e-08,
+      "loss": 0.4747,
+      "step": 9312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.944794784351862,
+      "learning_rate": 2.6685382801118765e-08,
+      "loss": 0.5882,
+      "step": 9313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8238895229633358,
+      "learning_rate": 2.6511973342829843e-08,
+      "loss": 0.5857,
+      "step": 9314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.772948928337245,
+      "learning_rate": 2.6339127657756814e-08,
+      "loss": 0.5756,
+      "step": 9315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1942165661966677,
+      "learning_rate": 2.6166845765492333e-08,
+      "loss": 0.5701,
+      "step": 9316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.980825317874205,
+      "learning_rate": 2.5995127685566335e-08,
+      "loss": 0.579,
+      "step": 9317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9101493425224827,
+      "learning_rate": 2.5823973437442696e-08,
+      "loss": 0.6031,
+      "step": 9318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8369726379682219,
+      "learning_rate": 2.5653383040524228e-08,
+      "loss": 0.6034,
+      "step": 9319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8791588215949244,
+      "learning_rate": 2.5483356514147128e-08,
+      "loss": 0.6301,
+      "step": 9320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9328706087321703,
+      "learning_rate": 2.531389387758598e-08,
+      "loss": 0.6878,
+      "step": 9321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0484986499553894,
+      "learning_rate": 2.5144995150049312e-08,
+      "loss": 0.6561,
+      "step": 9322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8432600727248396,
+      "learning_rate": 2.497666035068347e-08,
+      "loss": 0.5751,
+      "step": 9323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.221330921501575,
+      "learning_rate": 2.48088894985693e-08,
+      "loss": 0.6359,
+      "step": 9324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0728270839250973,
+      "learning_rate": 2.464168261272548e-08,
+      "loss": 0.5583,
+      "step": 9325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0001901715742494,
+      "learning_rate": 2.4475039712105742e-08,
+      "loss": 0.6303,
+      "step": 9326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9665561149483282,
+      "learning_rate": 2.4308960815599412e-08,
+      "loss": 0.6741,
+      "step": 9327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.008847781228865,
+      "learning_rate": 2.4143445942033105e-08,
+      "loss": 0.5627,
+      "step": 9328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9899789683173035,
+      "learning_rate": 2.3978495110168477e-08,
+      "loss": 0.6163,
+      "step": 9329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9583106013630442,
+      "learning_rate": 2.38141083387039e-08,
+      "loss": 0.6204,
+      "step": 9330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1674235835611495,
+      "learning_rate": 2.3650285646273362e-08,
+      "loss": 0.5559,
+      "step": 9331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8975429563101518,
+      "learning_rate": 2.348702705144701e-08,
+      "loss": 0.615,
+      "step": 9332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.090178727546794,
+      "learning_rate": 2.33243325727317e-08,
+      "loss": 0.7192,
+      "step": 9333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.881027148542671,
+      "learning_rate": 2.3162202228569353e-08,
+      "loss": 0.5326,
+      "step": 9334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9837370957943103,
+      "learning_rate": 2.300063603733804e-08,
+      "loss": 0.6979,
+      "step": 9335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7547614413184414,
+      "learning_rate": 2.2839634017353118e-08,
+      "loss": 0.5466,
+      "step": 9336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0185544324841307,
+      "learning_rate": 2.267919618686443e-08,
+      "loss": 0.6056,
+      "step": 9337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0292071357538766,
+      "learning_rate": 2.251932256405853e-08,
+      "loss": 0.6469,
+      "step": 9338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.804027643043366,
+      "learning_rate": 2.2360013167057602e-08,
+      "loss": 0.5343,
+      "step": 9339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9766515344332558,
+      "learning_rate": 2.220126801392164e-08,
+      "loss": 0.5715,
+      "step": 9340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9002895224444987,
+      "learning_rate": 2.2043087122644023e-08,
+      "loss": 0.7093,
+      "step": 9341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1821936473764985,
+      "learning_rate": 2.188547051115597e-08,
+      "loss": 0.5981,
+      "step": 9342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9245969157170688,
+      "learning_rate": 2.1728418197323742e-08,
+      "loss": 0.5908,
+      "step": 9343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.12228430228684,
+      "learning_rate": 2.1571930198950874e-08,
+      "loss": 0.5723,
+      "step": 9344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8732070599355801,
+      "learning_rate": 2.1416006533775957e-08,
+      "loss": 0.5704,
+      "step": 9345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.897499138291424,
+      "learning_rate": 2.1260647219473742e-08,
+      "loss": 0.5922,
+      "step": 9346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1515440733884086,
+      "learning_rate": 2.110585227365458e-08,
+      "loss": 0.6734,
+      "step": 9347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0743929440733906,
+      "learning_rate": 2.095162171386611e-08,
+      "loss": 0.6869,
+      "step": 9348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9802962077153008,
+      "learning_rate": 2.0797955557590454e-08,
+      "loss": 0.5603,
+      "step": 9349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.902902250037822,
+      "learning_rate": 2.064485382224757e-08,
+      "loss": 0.6033,
+      "step": 9350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9135159924715113,
+      "learning_rate": 2.04923165251919e-08,
+      "loss": 0.6096,
+      "step": 9351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.001111887938867,
+      "learning_rate": 2.0340343683714624e-08,
+      "loss": 0.66,
+      "step": 9352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2845616405345233,
+      "learning_rate": 2.018893531504196e-08,
+      "loss": 0.7283,
+      "step": 9353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0236378220572275,
+      "learning_rate": 2.0038091436337392e-08,
+      "loss": 0.5974,
+      "step": 9354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7188227459501022,
+      "learning_rate": 1.9887812064700028e-08,
+      "loss": 0.5671,
+      "step": 9355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9222401255258874,
+      "learning_rate": 1.973809721716513e-08,
+      "loss": 0.5832,
+      "step": 9356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0028373714720367,
+      "learning_rate": 1.9588946910703567e-08,
+      "loss": 0.5924,
+      "step": 9357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8852559821091195,
+      "learning_rate": 1.9440361162222367e-08,
+      "loss": 0.6009,
+      "step": 9358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6645381563062454,
+      "learning_rate": 1.929233998856417e-08,
+      "loss": 0.5973,
+      "step": 9359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8784683299762617,
+      "learning_rate": 1.914488340650833e-08,
+      "loss": 0.6166,
+      "step": 9360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9423516733751873,
+      "learning_rate": 1.8997991432769812e-08,
+      "loss": 0.5375,
+      "step": 9361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.896972762313844,
+      "learning_rate": 1.8851664083999742e-08,
+      "loss": 0.5419,
+      "step": 9362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9254153543066406,
+      "learning_rate": 1.8705901376784852e-08,
+      "loss": 0.5956,
+      "step": 9363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.837387913964868,
+      "learning_rate": 1.8560703327649144e-08,
+      "loss": 0.5569,
+      "step": 9364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1808835722204445,
+      "learning_rate": 1.8416069953050565e-08,
+      "loss": 0.6779,
+      "step": 9365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1182269084331415,
+      "learning_rate": 1.8272001269384886e-08,
+      "loss": 0.6208,
+      "step": 9366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8505257750618553,
+      "learning_rate": 1.812849729298238e-08,
+      "loss": 0.5383,
+      "step": 9367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.85357421039629,
+      "learning_rate": 1.7985558040110594e-08,
+      "loss": 0.657,
+      "step": 9368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0404214691662395,
+      "learning_rate": 1.784318352697212e-08,
+      "loss": 0.6548,
+      "step": 9369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0413228816834392,
+      "learning_rate": 1.7701373769706265e-08,
+      "loss": 0.6042,
+      "step": 9370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.192428086651773,
+      "learning_rate": 1.7560128784387953e-08,
+      "loss": 0.6355,
+      "step": 9371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.693491337066328,
+      "learning_rate": 1.741944858702771e-08,
+      "loss": 0.5473,
+      "step": 9372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9447757214859827,
+      "learning_rate": 1.7279333193573332e-08,
+      "loss": 0.5628,
+      "step": 9373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9521877617740033,
+      "learning_rate": 1.7139782619906565e-08,
+      "loss": 0.5326,
+      "step": 9374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1318514544476956,
+      "learning_rate": 1.700079688184697e-08,
+      "loss": 0.5991,
+      "step": 9375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.111380016447978,
+      "learning_rate": 1.686237599514917e-08,
+      "loss": 0.5793,
+      "step": 9376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0061463420649823,
+      "learning_rate": 1.672451997550395e-08,
+      "loss": 0.5981,
+      "step": 9377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9289965935272066,
+      "learning_rate": 1.658722883853825e-08,
+      "loss": 0.6348,
+      "step": 9378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.759822966934597,
+      "learning_rate": 1.6450502599814622e-08,
+      "loss": 0.5902,
+      "step": 9379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.830609332103879,
+      "learning_rate": 1.631434127483178e-08,
+      "loss": 0.6012,
+      "step": 9380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.903223961744007,
+      "learning_rate": 1.6178744879024045e-08,
+      "loss": 0.5514,
+      "step": 9381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9105186920277744,
+      "learning_rate": 1.604371342776301e-08,
+      "loss": 0.528,
+      "step": 9382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7738895886516455,
+      "learning_rate": 1.5909246936354205e-08,
+      "loss": 0.5702,
+      "step": 9383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9203976490248245,
+      "learning_rate": 1.5775345420041e-08,
+      "loss": 0.509,
+      "step": 9384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.981620797130497,
+      "learning_rate": 1.564200889400125e-08,
+      "loss": 0.6442,
+      "step": 9385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.686120661097064,
+      "learning_rate": 1.5509237373349527e-08,
+      "loss": 0.5022,
+      "step": 9386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.011187623061088,
+      "learning_rate": 1.537703087313658e-08,
+      "loss": 0.6468,
+      "step": 9387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7837135872441217,
+      "learning_rate": 1.5245389408348744e-08,
+      "loss": 0.5189,
+      "step": 9388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9655077598433057,
+      "learning_rate": 1.5114312993908532e-08,
+      "loss": 0.5999,
+      "step": 9389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.052384135182507,
+      "learning_rate": 1.4983801644672948e-08,
+      "loss": 0.5486,
+      "step": 9390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1277424515467844,
+      "learning_rate": 1.4853855375437377e-08,
+      "loss": 0.6428,
+      "step": 9391
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.977814419207284,
+      "learning_rate": 1.4724474200931704e-08,
+      "loss": 0.6306,
+      "step": 9392
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.138743859165059,
+      "learning_rate": 1.4595658135822533e-08,
+      "loss": 0.7269,
+      "step": 9393
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7904427754103556,
+      "learning_rate": 1.4467407194710958e-08,
+      "loss": 0.5781,
+      "step": 9394
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9993046254554308,
+      "learning_rate": 1.4339721392135352e-08,
+      "loss": 0.6969,
+      "step": 9395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0825943027963394,
+      "learning_rate": 1.4212600742569694e-08,
+      "loss": 0.6154,
+      "step": 9396
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9421694999551333,
+      "learning_rate": 1.4086045260423564e-08,
+      "loss": 0.6173,
+      "step": 9397
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0324173008481954,
+      "learning_rate": 1.3960054960043267e-08,
+      "loss": 0.6327,
+      "step": 9398
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3108147574118845,
+      "learning_rate": 1.3834629855710158e-08,
+      "loss": 0.7125,
+      "step": 9399
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.973482473276303,
+      "learning_rate": 1.370976996164175e-08,
+      "loss": 0.6279,
+      "step": 9400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.086224009235821,
+      "learning_rate": 1.3585475291991724e-08,
+      "loss": 0.5979,
+      "step": 9401
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9357123351662384,
+      "learning_rate": 1.3461745860849917e-08,
+      "loss": 0.6264,
+      "step": 9402
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7861483587494011,
+      "learning_rate": 1.333858168224178e-08,
+      "loss": 0.561,
+      "step": 9403
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0311118954318155,
+      "learning_rate": 1.3215982770128366e-08,
+      "loss": 0.7493,
+      "step": 9404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.045942396521214,
+      "learning_rate": 1.3093949138406892e-08,
+      "loss": 0.661,
+      "step": 9405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.07625840929344,
+      "learning_rate": 1.2972480800910181e-08,
+      "loss": 0.6479,
+      "step": 9406
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.067265450826643,
+      "learning_rate": 1.285157777140833e-08,
+      "loss": 0.587,
+      "step": 9407
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.119763936274861,
+      "learning_rate": 1.2731240063605931e-08,
+      "loss": 0.5975,
+      "step": 9408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.024916147212216,
+      "learning_rate": 1.2611467691144297e-08,
+      "loss": 0.6936,
+      "step": 9409
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.983127501150537,
+      "learning_rate": 1.2492260667599232e-08,
+      "loss": 0.6452,
+      "step": 9410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8649494084823937,
+      "learning_rate": 1.2373619006484927e-08,
+      "loss": 0.5512,
+      "step": 9411
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0159592835650364,
+      "learning_rate": 1.2255542721248959e-08,
+      "loss": 0.4776,
+      "step": 9412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1385407875276283,
+      "learning_rate": 1.213803182527673e-08,
+      "loss": 0.617,
+      "step": 9413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.834974861679934,
+      "learning_rate": 1.2021086331888143e-08,
+      "loss": 0.6695,
+      "step": 9414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0853828764496343,
+      "learning_rate": 1.190470625434037e-08,
+      "loss": 0.6612,
+      "step": 9415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.98313825019437,
+      "learning_rate": 1.1788891605825081e-08,
+      "loss": 0.6369,
+      "step": 9416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.117767892785416,
+      "learning_rate": 1.1673642399470663e-08,
+      "loss": 0.5904,
+      "step": 9417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.003554226087581,
+      "learning_rate": 1.1558958648341667e-08,
+      "loss": 0.6737,
+      "step": 9418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8760117400728373,
+      "learning_rate": 1.1444840365437692e-08,
+      "loss": 0.5358,
+      "step": 9419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.826652775318849,
+      "learning_rate": 1.1331287563695059e-08,
+      "loss": 0.5802,
+      "step": 9420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.210283365637128,
+      "learning_rate": 1.1218300255985137e-08,
+      "loss": 0.6695,
+      "step": 9421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9114471453067057,
+      "learning_rate": 1.1105878455116015e-08,
+      "loss": 0.6807,
+      "step": 9422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.308622820891408,
+      "learning_rate": 1.0994022173831386e-08,
+      "loss": 0.5271,
+      "step": 9423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.154207219627266,
+      "learning_rate": 1.088273142481111e-08,
+      "loss": 0.6708,
+      "step": 9424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.062686057137944,
+      "learning_rate": 1.0772006220670094e-08,
+      "loss": 0.6164,
+      "step": 9425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9871469414224412,
+      "learning_rate": 1.0661846573959412e-08,
+      "loss": 0.5549,
+      "step": 9426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9153912700098923,
+      "learning_rate": 1.0552252497166849e-08,
+      "loss": 0.5732,
+      "step": 9427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0642917207760196,
+      "learning_rate": 1.0443224002715801e-08,
+      "loss": 0.6291,
+      "step": 9428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8687845348076741,
+      "learning_rate": 1.0334761102964163e-08,
+      "loss": 0.5605,
+      "step": 9429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0119112889139665,
+      "learning_rate": 1.022686381020821e-08,
+      "loss": 0.6568,
+      "step": 9430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.142024130237965,
+      "learning_rate": 1.01195321366776e-08,
+      "loss": 0.6009,
+      "step": 9431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9652097501051895,
+      "learning_rate": 1.0012766094539273e-08,
+      "loss": 0.5757,
+      "step": 9432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.11686381269212,
+      "learning_rate": 9.90656569589632e-09,
+      "loss": 0.6156,
+      "step": 9433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8540572366176127,
+      "learning_rate": 9.800930952786336e-09,
+      "loss": 0.5143,
+      "step": 9434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8884066141241278,
+      "learning_rate": 9.695861877184187e-09,
+      "loss": 0.6136,
+      "step": 9435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8442181601552678,
+      "learning_rate": 9.59135848099979e-09,
+      "loss": 0.556,
+      "step": 9436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.010704790123921,
+      "learning_rate": 9.487420776079226e-09,
+      "loss": 0.6489,
+      "step": 9437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1347997954400997,
+      "learning_rate": 9.38404877420418e-09,
+      "loss": 0.6559,
+      "step": 9438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0640248046696636,
+      "learning_rate": 9.281242487093612e-09,
+      "loss": 0.622,
+      "step": 9439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0903299573728393,
+      "learning_rate": 9.179001926399866e-09,
+      "loss": 0.5661,
+      "step": 9440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.011961237897929,
+      "learning_rate": 9.077327103713119e-09,
+      "loss": 0.6185,
+      "step": 9441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9101374524807846,
+      "learning_rate": 8.976218030558592e-09,
+      "loss": 0.5676,
+      "step": 9442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.997803767273175,
+      "learning_rate": 8.87567471839712e-09,
+      "loss": 0.5351,
+      "step": 9443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8370994226150568,
+      "learning_rate": 8.775697178626807e-09,
+      "loss": 0.5526,
+      "step": 9444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8661873226938193,
+      "learning_rate": 8.676285422580255e-09,
+      "loss": 0.6151,
+      "step": 9445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0505710073168952,
+      "learning_rate": 8.577439461526782e-09,
+      "loss": 0.593,
+      "step": 9446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9162215230441964,
+      "learning_rate": 8.479159306670203e-09,
+      "loss": 0.6494,
+      "step": 9447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9933094696955849,
+      "learning_rate": 8.381444969151608e-09,
+      "loss": 0.5787,
+      "step": 9448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8322768596191124,
+      "learning_rate": 8.284296460047691e-09,
+      "loss": 0.6197,
+      "step": 9449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9649119341867372,
+      "learning_rate": 8.1877137903702e-09,
+      "loss": 0.5414,
+      "step": 9450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8174793553111006,
+      "learning_rate": 8.091696971068152e-09,
+      "loss": 0.5382,
+      "step": 9451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.024317306699993,
+      "learning_rate": 7.996246013025067e-09,
+      "loss": 0.5741,
+      "step": 9452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.293287635439528,
+      "learning_rate": 7.901360927061174e-09,
+      "loss": 0.6518,
+      "step": 9453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.086365621299315,
+      "learning_rate": 7.807041723931763e-09,
+      "loss": 0.5751,
+      "step": 9454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.936273640010896,
+      "learning_rate": 7.713288414328835e-09,
+      "loss": 0.7171,
+      "step": 9455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.768772970005881,
+      "learning_rate": 7.620101008879444e-09,
+      "loss": 0.498,
+      "step": 9456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8866484555298333,
+      "learning_rate": 7.527479518147918e-09,
+      "loss": 0.5336,
+      "step": 9457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9101721466577788,
+      "learning_rate": 7.435423952631971e-09,
+      "loss": 0.5666,
+      "step": 9458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1247138345157066,
+      "learning_rate": 7.343934322767699e-09,
+      "loss": 0.636,
+      "step": 9459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0203673144733316,
+      "learning_rate": 7.253010638925695e-09,
+      "loss": 0.4786,
+      "step": 9460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0489168644317868,
+      "learning_rate": 7.1626529114127155e-09,
+      "loss": 0.6674,
+      "step": 9461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.014479744809945,
+      "learning_rate": 7.072861150471122e-09,
+      "loss": 0.6233,
+      "step": 9462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1318020390134382,
+      "learning_rate": 6.98363536627944e-09,
+      "loss": 0.6574,
+      "step": 9463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.126617683494603,
+      "learning_rate": 6.894975568951801e-09,
+      "loss": 0.6201,
+      "step": 9464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1193407018704273,
+      "learning_rate": 6.806881768539053e-09,
+      "loss": 0.5861,
+      "step": 9465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1218760503414384,
+      "learning_rate": 6.719353975025989e-09,
+      "loss": 0.7565,
+      "step": 9466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9224166553741697,
+      "learning_rate": 6.63239219833467e-09,
+      "loss": 0.6918,
+      "step": 9467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.90476914149826,
+      "learning_rate": 6.5459964483233215e-09,
+      "loss": 0.5775,
+      "step": 9468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0805059353113027,
+      "learning_rate": 6.460166734785223e-09,
+      "loss": 0.6907,
+      "step": 9469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.039038658401796,
+      "learning_rate": 6.374903067448701e-09,
+      "loss": 0.5588,
+      "step": 9470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9051077555679738,
+      "learning_rate": 6.290205455980469e-09,
+      "loss": 0.6101,
+      "step": 9471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9114694925966835,
+      "learning_rate": 6.2060739099800704e-09,
+      "loss": 0.6417,
+      "step": 9472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0034549196304288,
+      "learning_rate": 6.122508438984875e-09,
+      "loss": 0.6337,
+      "step": 9473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.929605593178843,
+      "learning_rate": 6.039509052467862e-09,
+      "loss": 0.6391,
+      "step": 9474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0565609104314695,
+      "learning_rate": 5.957075759837061e-09,
+      "loss": 0.6141,
+      "step": 9475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8833486655812495,
+      "learning_rate": 5.875208570436663e-09,
+      "loss": 0.5839,
+      "step": 9476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.023643705570554,
+      "learning_rate": 5.793907493546469e-09,
+      "loss": 0.5759,
+      "step": 9477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8797156245993185,
+      "learning_rate": 5.713172538383549e-09,
+      "loss": 0.5925,
+      "step": 9478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.029520402534025,
+      "learning_rate": 5.6330037140989166e-09,
+      "loss": 0.5766,
+      "step": 9479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1692397510491266,
+      "learning_rate": 5.5534010297803034e-09,
+      "loss": 0.5764,
+      "step": 9480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.101204297729523,
+      "learning_rate": 5.474364494451045e-09,
+      "loss": 0.5781,
+      "step": 9481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9301730906644166,
+      "learning_rate": 5.395894117070089e-09,
+      "loss": 0.7224,
+      "step": 9482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7290399911078331,
+      "learning_rate": 5.317989906533094e-09,
+      "loss": 0.5331,
+      "step": 9483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7841568775539753,
+      "learning_rate": 5.2406518716707766e-09,
+      "loss": 0.63,
+      "step": 9484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.154539636467135,
+      "learning_rate": 5.1638800212494566e-09,
+      "loss": 0.6452,
+      "step": 9485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0702545934630017,
+      "learning_rate": 5.087674363972173e-09,
+      "loss": 0.5551,
+      "step": 9486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9771019143634083,
+      "learning_rate": 5.01203490847646e-09,
+      "loss": 0.5728,
+      "step": 9487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8392551379936382,
+      "learning_rate": 4.93696166333768e-09,
+      "loss": 0.6022,
+      "step": 9488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9594213971787073,
+      "learning_rate": 4.8624546370651395e-09,
+      "loss": 0.6141,
+      "step": 9489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8586510143718253,
+      "learning_rate": 4.788513838104858e-09,
+      "loss": 0.6084,
+      "step": 9490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6622364581523985,
+      "learning_rate": 4.7151392748379095e-09,
+      "loss": 0.5573,
+      "step": 9491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.07679473110362,
+      "learning_rate": 4.642330955582641e-09,
+      "loss": 0.6039,
+      "step": 9492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8678922922176417,
+      "learning_rate": 4.57008888859134e-09,
+      "loss": 0.5118,
+      "step": 9493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9392398900547896,
+      "learning_rate": 4.498413082053566e-09,
+      "loss": 0.6151,
+      "step": 9494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9618678399300282,
+      "learning_rate": 4.42730354409393e-09,
+      "loss": 0.6178,
+      "step": 9495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8062383269521096,
+      "learning_rate": 4.356760282773209e-09,
+      "loss": 0.5622,
+      "step": 9496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8782190565478545,
+      "learning_rate": 4.286783306087783e-09,
+      "loss": 0.5159,
+      "step": 9497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.976593913395866,
+      "learning_rate": 4.21737262197075e-09,
+      "loss": 0.6743,
+      "step": 9498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9818379691434536,
+      "learning_rate": 4.148528238289151e-09,
+      "loss": 0.6773,
+      "step": 9499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0152612123522133,
+      "learning_rate": 4.080250162847299e-09,
+      "loss": 0.6145,
+      "step": 9500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1683488804338475,
+      "learning_rate": 4.0125384033845586e-09,
+      "loss": 0.5787,
+      "step": 9501
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8175768439932203,
+      "learning_rate": 3.945392967577011e-09,
+      "loss": 0.555,
+      "step": 9502
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1157456006926907,
+      "learning_rate": 3.8788138630357905e-09,
+      "loss": 0.5811,
+      "step": 9503
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1523135188211087,
+      "learning_rate": 3.812801097308194e-09,
+      "loss": 0.6436,
+      "step": 9504
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9801354903550479,
+      "learning_rate": 3.747354677876569e-09,
+      "loss": 0.7153,
+      "step": 9505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9816879347353105,
+      "learning_rate": 3.682474612159981e-09,
+      "loss": 0.561,
+      "step": 9506
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.036183290464764,
+      "learning_rate": 3.6181609075131017e-09,
+      "loss": 0.7561,
+      "step": 9507
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1541774964882485,
+      "learning_rate": 3.5544135712262116e-09,
+      "loss": 0.642,
+      "step": 9508
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2831346674446213,
+      "learning_rate": 3.4912326105246418e-09,
+      "loss": 0.6434,
+      "step": 9509
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.958292617985834,
+      "learning_rate": 3.4286180325715513e-09,
+      "loss": 0.6362,
+      "step": 9510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9314782542001665,
+      "learning_rate": 3.3665698444640406e-09,
+      "loss": 0.5673,
+      "step": 9511
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.991690621452874,
+      "learning_rate": 3.3050880532359277e-09,
+      "loss": 0.5816,
+      "step": 9512
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5630124336650852,
+      "learning_rate": 3.2441726658560825e-09,
+      "loss": 0.591,
+      "step": 9513
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.98736244563508,
+      "learning_rate": 3.183823689230092e-09,
+      "loss": 0.661,
+      "step": 9514
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.976417843136004,
+      "learning_rate": 3.1240411301980413e-09,
+      "loss": 0.6741,
+      "step": 9515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0882239803944302,
+      "learning_rate": 3.0648249955378405e-09,
+      "loss": 0.6652,
+      "step": 9516
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.163565371948662,
+      "learning_rate": 3.006175291960789e-09,
+      "loss": 0.5793,
+      "step": 9517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.00691462554776,
+      "learning_rate": 2.9480920261154565e-09,
+      "loss": 0.6054,
+      "step": 9518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.076838889221059,
+      "learning_rate": 2.8905752045865766e-09,
+      "loss": 0.6402,
+      "step": 9519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.115754866675787,
+      "learning_rate": 2.833624833893933e-09,
+      "loss": 0.6234,
+      "step": 9520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0742777041776113,
+      "learning_rate": 2.7772409204923633e-09,
+      "loss": 0.529,
+      "step": 9521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.033700985807466,
+      "learning_rate": 2.721423470773421e-09,
+      "loss": 0.5545,
+      "step": 9522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8799003192685757,
+      "learning_rate": 2.6661724910653774e-09,
+      "loss": 0.626,
+      "step": 9523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.276364282368124,
+      "learning_rate": 2.6114879876298905e-09,
+      "loss": 0.7034,
+      "step": 9524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9794806333649237,
+      "learning_rate": 2.5573699666664455e-09,
+      "loss": 0.5946,
+      "step": 9525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.849392363068563,
+      "learning_rate": 2.5038184343101346e-09,
+      "loss": 0.6102,
+      "step": 9526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9217115956324478,
+      "learning_rate": 2.4508333966305473e-09,
+      "loss": 0.5229,
+      "step": 9527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.019210354811399,
+      "learning_rate": 2.3984148596339907e-09,
+      "loss": 0.5405,
+      "step": 9528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3043229195920816,
+      "learning_rate": 2.3465628292623776e-09,
+      "loss": 0.6815,
+      "step": 9529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.080083068675512,
+      "learning_rate": 2.295277311393784e-09,
+      "loss": 0.579,
+      "step": 9530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1199920401449552,
+      "learning_rate": 2.2445583118413384e-09,
+      "loss": 0.6651,
+      "step": 9531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0842626790362053,
+      "learning_rate": 2.19440583635433e-09,
+      "loss": 0.6451,
+      "step": 9532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0915589293813173,
+      "learning_rate": 2.1448198906182106e-09,
+      "loss": 0.6338,
+      "step": 9533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0306946472868996,
+      "learning_rate": 2.0958004802529297e-09,
+      "loss": 0.5401,
+      "step": 9534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.055855540710789,
+      "learning_rate": 2.047347610816819e-09,
+      "loss": 0.586,
+      "step": 9535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9878400684546984,
+      "learning_rate": 1.999461287800486e-09,
+      "loss": 0.5352,
+      "step": 9536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.905336891439375,
+      "learning_rate": 1.9521415166329216e-09,
+      "loss": 0.484,
+      "step": 9537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8815579402867484,
+      "learning_rate": 1.9053883026781685e-09,
+      "loss": 0.6296,
+      "step": 9538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.984575489338281,
+      "learning_rate": 1.8592016512358757e-09,
+      "loss": 0.5511,
+      "step": 9539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.046470877718721,
+      "learning_rate": 1.8135815675418556e-09,
+      "loss": 0.6024,
+      "step": 9540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9722402840427116,
+      "learning_rate": 1.7685280567664165e-09,
+      "loss": 0.5257,
+      "step": 9541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.000880220808784,
+      "learning_rate": 1.7240411240176946e-09,
+      "loss": 0.5769,
+      "step": 9542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1063839838590925,
+      "learning_rate": 1.680120774338323e-09,
+      "loss": 0.6462,
+      "step": 9543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0056915245462923,
+      "learning_rate": 1.6367670127059864e-09,
+      "loss": 0.63,
+      "step": 9544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8659277173450295,
+      "learning_rate": 1.5939798440367527e-09,
+      "loss": 0.5298,
+      "step": 9545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.864106975642594,
+      "learning_rate": 1.5517592731789654e-09,
+      "loss": 0.5465,
+      "step": 9546
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9575734799120101,
+      "learning_rate": 1.5101053049199065e-09,
+      "loss": 0.5879,
+      "step": 9547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9740114393181525,
+      "learning_rate": 1.4690179439807993e-09,
+      "loss": 0.64,
+      "step": 9548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9215329174434153,
+      "learning_rate": 1.4284971950195847e-09,
+      "loss": 0.652,
+      "step": 9549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8879530552664527,
+      "learning_rate": 1.3885430626287e-09,
+      "loss": 0.5143,
+      "step": 9550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0021158704501776,
+      "learning_rate": 1.3491555513378552e-09,
+      "loss": 0.6374,
+      "step": 9551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.750227327277223,
+      "learning_rate": 1.310334665611257e-09,
+      "loss": 0.6026,
+      "step": 9552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.924162397266902,
+      "learning_rate": 1.2720804098498297e-09,
+      "loss": 0.6646,
+      "step": 9553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9347971214125042,
+      "learning_rate": 1.2343927883901042e-09,
+      "loss": 0.6712,
+      "step": 9554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.006885502405168,
+      "learning_rate": 1.1972718055036637e-09,
+      "loss": 0.5202,
+      "step": 9555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.689423303629976,
+      "learning_rate": 1.1607174653988085e-09,
+      "loss": 0.5011,
+      "step": 9556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9127280989270514,
+      "learning_rate": 1.124729772219446e-09,
+      "loss": 0.4981,
+      "step": 9557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0010664585250657,
+      "learning_rate": 1.089308730043981e-09,
+      "loss": 0.6541,
+      "step": 9558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.07992521114073,
+      "learning_rate": 1.0544543428886445e-09,
+      "loss": 0.6736,
+      "step": 9559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9283583857469802,
+      "learning_rate": 1.0201666147041656e-09,
+      "loss": 0.6337,
+      "step": 9560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9700243707520468,
+      "learning_rate": 9.864455493763247e-10,
+      "loss": 0.6305,
+      "step": 9561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.147094232957025,
+      "learning_rate": 9.5329115072873e-10,
+      "loss": 0.6686,
+      "step": 9562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.079316149178158,
+      "learning_rate": 9.207034225189315e-10,
+      "loss": 0.5308,
+      "step": 9563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.354067111579948,
+      "learning_rate": 8.886823684417512e-10,
+      "loss": 0.6647,
+      "step": 9564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7465751308270212,
+      "learning_rate": 8.57227992125953e-10,
+      "loss": 0.5295,
+      "step": 9565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9868861820598962,
+      "learning_rate": 8.263402971375734e-10,
+      "loss": 0.5493,
+      "step": 9566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.22374318466106,
+      "learning_rate": 7.960192869782557e-10,
+      "loss": 0.6371,
+      "step": 9567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9923933780189522,
+      "learning_rate": 7.662649650841403e-10,
+      "loss": 0.5635,
+      "step": 9568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.038120541474471,
+      "learning_rate": 7.370773348286397e-10,
+      "loss": 0.5779,
+      "step": 9569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9771844157809515,
+      "learning_rate": 7.084563995202187e-10,
+      "loss": 0.526,
+      "step": 9570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.812053455374471,
+      "learning_rate": 6.80402162403504e-10,
+      "loss": 0.5634,
+      "step": 9571
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9921272648147166,
+      "learning_rate": 6.529146266587294e-10,
+      "loss": 0.5553,
+      "step": 9572
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.859834751177411,
+      "learning_rate": 6.259937954006257e-10,
+      "loss": 0.6282,
+      "step": 9573
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.040795878230898,
+      "learning_rate": 5.996396716823061e-10,
+      "loss": 0.641,
+      "step": 9574
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9718501791261143,
+      "learning_rate": 5.738522584897155e-10,
+      "loss": 0.5137,
+      "step": 9575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.019272879708207,
+      "learning_rate": 5.486315587471813e-10,
+      "loss": 0.613,
+      "step": 9576
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0140937375266206,
+      "learning_rate": 5.239775753129728e-10,
+      "loss": 0.5492,
+      "step": 9577
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7876855888321224,
+      "learning_rate": 4.998903109826314e-10,
+      "loss": 0.5968,
+      "step": 9578
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.937692711387017,
+      "learning_rate": 4.763697684850854e-10,
+      "loss": 0.5326,
+      "step": 9579
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8871781497039766,
+      "learning_rate": 4.5341595048764564e-10,
+      "loss": 0.5952,
+      "step": 9580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.657956554866962,
+      "learning_rate": 4.310288595921197e-10,
+      "loss": 0.4911,
+      "step": 9581
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.953405180081528,
+      "learning_rate": 4.0920849833592236e-10,
+      "loss": 0.6131,
+      "step": 9582
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.091422857268565,
+      "learning_rate": 3.879548691926305e-10,
+      "loss": 0.6022,
+      "step": 9583
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9263886894181483,
+      "learning_rate": 3.6726797457198314e-10,
+      "loss": 0.5952,
+      "step": 9584
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.146763105003528,
+      "learning_rate": 3.471478168176612e-10,
+      "loss": 0.6512,
+      "step": 9585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.931113208646632,
+      "learning_rate": 3.27594398211728e-10,
+      "loss": 0.5383,
+      "step": 9586
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9413320484945569,
+      "learning_rate": 3.086077209701888e-10,
+      "loss": 0.6305,
+      "step": 9587
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.917465576733899,
+      "learning_rate": 2.901877872452108e-10,
+      "loss": 0.5411,
+      "step": 9588
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9595757628143469,
+      "learning_rate": 2.723345991245685e-10,
+      "loss": 0.5646,
+      "step": 9589
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.196310127527789,
+      "learning_rate": 2.550481586321985e-10,
+      "loss": 0.5639,
+      "step": 9590
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.900419785333137,
+      "learning_rate": 2.3832846772819937e-10,
+      "loss": 0.7065,
+      "step": 9591
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9730652890498885,
+      "learning_rate": 2.2217552830716693e-10,
+      "loss": 0.6706,
+      "step": 9592
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9593091425150002,
+      "learning_rate": 2.0658934219985883e-10,
+      "loss": 0.5465,
+      "step": 9593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.135942361894095,
+      "learning_rate": 1.9156991117430523e-10,
+      "loss": 0.758,
+      "step": 9594
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.930526843929985,
+      "learning_rate": 1.7711723693192296e-10,
+      "loss": 0.5605,
+      "step": 9595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9256736889575217,
+      "learning_rate": 1.6323132111084605e-10,
+      "loss": 0.6454,
+      "step": 9596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.064044604856687,
+      "learning_rate": 1.49912165286481e-10,
+      "loss": 0.6168,
+      "step": 9597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8522204350674754,
+      "learning_rate": 1.3715977096706578e-10,
+      "loss": 0.626,
+      "step": 9598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0195678223686127,
+      "learning_rate": 1.24974139599221e-10,
+      "loss": 0.7152,
+      "step": 9599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.974776889449131,
+      "learning_rate": 1.1335527256350898e-10,
+      "loss": 0.5263,
+      "step": 9600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9816482265824762,
+      "learning_rate": 1.0230317117776445e-10,
+      "loss": 0.5823,
+      "step": 9601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9456288642203419,
+      "learning_rate": 9.181783669431898e-11,
+      "loss": 0.5181,
+      "step": 9602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8505141834481915,
+      "learning_rate": 8.189927030222145e-11,
+      "loss": 0.5149,
+      "step": 9603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9702244129196227,
+      "learning_rate": 7.254747312501754e-11,
+      "loss": 0.6123,
+      "step": 9604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9383150259419573,
+      "learning_rate": 6.376244622297023e-11,
+      "loss": 0.6127,
+      "step": 9605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9864553209022995,
+      "learning_rate": 5.554419059250471e-11,
+      "loss": 0.6042,
+      "step": 9606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9772866782374987,
+      "learning_rate": 4.7892707164542975e-11,
+      "loss": 0.5077,
+      "step": 9607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0014007844865565,
+      "learning_rate": 4.080799680727943e-11,
+      "loss": 0.6155,
+      "step": 9608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2486787110200264,
+      "learning_rate": 3.42900603228502e-11,
+      "loss": 0.5979,
+      "step": 9609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.064265143022498,
+      "learning_rate": 2.833889845010873e-11,
+      "loss": 0.6211,
+      "step": 9610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.081375211775521,
+      "learning_rate": 2.2954511864625717e-11,
+      "loss": 0.5994,
+      "step": 9611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7557657446910315,
+      "learning_rate": 1.8136901175913602e-11,
+      "loss": 0.6341,
+      "step": 9612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8772616582779724,
+      "learning_rate": 1.3886066930202113e-11,
+      "loss": 0.6788,
+      "step": 9613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6975943413231218,
+      "learning_rate": 1.0202009609883157e-11,
+      "loss": 0.4914,
+      "step": 9614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9089824368308848,
+      "learning_rate": 7.084729631845477e-12,
+      "loss": 0.5957,
+      "step": 9615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8543367974168876,
+      "learning_rate": 4.534227349695108e-12,
+      "loss": 0.6776,
+      "step": 9616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.842851875350606,
+      "learning_rate": 2.5505030532002594e-12,
+      "loss": 0.5382,
+      "step": 9617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8926880273749038,
+      "learning_rate": 1.13355696607087e-12,
+      "loss": 0.6009,
+      "step": 9618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9155764490752814,
+      "learning_rate": 2.8338924928927867e-13,
+      "loss": 0.5445,
+      "step": 9619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8883391558284497,
+      "learning_rate": 0.0,
+      "loss": 0.6125,
+      "step": 9620
+    },
+    {
+      "epoch": 1.0,
+      "step": 9620,
+      "total_flos": 2995729119199232.0,
+      "train_loss": 0.6451899706387966,
+      "train_runtime": 324149.63,
+      "train_samples_per_second": 3.799,
+      "train_steps_per_second": 0.03
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9620,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 2995729119199232.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}