diff --git "a/finetune-1.1b-llava-tiny-llama-eva640/trainer_state.json" "b/finetune-1.1b-llava-tiny-llama-eva640/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/finetune-1.1b-llava-tiny-llama-eva640/trainer_state.json"
@@ -0,0 +1,50904 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 7266,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00013762730525736306,
+      "grad_norm": 125.19473920812882,
+      "learning_rate": 9.174311926605506e-08,
+      "loss": 9.1743,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002752546105147261,
+      "grad_norm": 122.30798929948465,
+      "learning_rate": 1.8348623853211012e-07,
+      "loss": 9.1505,
+      "step": 2
+    },
+    {
+      "epoch": 0.00041288191577208916,
+      "grad_norm": 127.96765390189857,
+      "learning_rate": 2.752293577981652e-07,
+      "loss": 9.1093,
+      "step": 3
+    },
+    {
+      "epoch": 0.0005505092210294523,
+      "grad_norm": 113.35082853293781,
+      "learning_rate": 3.6697247706422023e-07,
+      "loss": 9.1053,
+      "step": 4
+    },
+    {
+      "epoch": 0.0006881365262868153,
+      "grad_norm": 109.99694948301452,
+      "learning_rate": 4.587155963302753e-07,
+      "loss": 9.1692,
+      "step": 5
+    },
+    {
+      "epoch": 0.0008257638315441783,
+      "grad_norm": 69.78117222147398,
+      "learning_rate": 5.504587155963304e-07,
+      "loss": 8.6909,
+      "step": 6
+    },
+    {
+      "epoch": 0.0009633911368015414,
+      "grad_norm": 57.80940552084882,
+      "learning_rate": 6.422018348623854e-07,
+      "loss": 8.5063,
+      "step": 7
+    },
+    {
+      "epoch": 0.0011010184420589045,
+      "grad_norm": 67.77576466371379,
+      "learning_rate": 7.339449541284405e-07,
+      "loss": 8.5205,
+      "step": 8
+    },
+    {
+      "epoch": 0.0012386457473162675,
+      "grad_norm": 70.46469666692839,
+      "learning_rate": 8.256880733944956e-07,
+      "loss": 8.2066,
+      "step": 9
+    },
+    {
+      "epoch": 0.0013762730525736307,
+      "grad_norm": 57.423476779940195,
+      "learning_rate": 9.174311926605506e-07,
+      "loss": 7.907,
+      "step": 10
+    },
+    {
+      "epoch": 0.0015139003578309937,
+      "grad_norm": 46.52489442826151,
+      "learning_rate": 1.0091743119266057e-06,
+      "loss": 7.3497,
+      "step": 11
+    },
+    {
+      "epoch": 0.0016515276630883566,
+      "grad_norm": 42.39966056483745,
+      "learning_rate": 1.1009174311926608e-06,
+      "loss": 6.988,
+      "step": 12
+    },
+    {
+      "epoch": 0.0017891549683457198,
+      "grad_norm": 33.67260907893102,
+      "learning_rate": 1.1926605504587159e-06,
+      "loss": 6.9455,
+      "step": 13
+    },
+    {
+      "epoch": 0.0019267822736030828,
+      "grad_norm": 41.146175167699894,
+      "learning_rate": 1.2844036697247707e-06,
+      "loss": 6.297,
+      "step": 14
+    },
+    {
+      "epoch": 0.002064409578860446,
+      "grad_norm": 28.177944155862917,
+      "learning_rate": 1.3761467889908258e-06,
+      "loss": 6.4059,
+      "step": 15
+    },
+    {
+      "epoch": 0.002202036884117809,
+      "grad_norm": 26.610618628606783,
+      "learning_rate": 1.467889908256881e-06,
+      "loss": 5.8543,
+      "step": 16
+    },
+    {
+      "epoch": 0.002339664189375172,
+      "grad_norm": 31.837106689327197,
+      "learning_rate": 1.559633027522936e-06,
+      "loss": 6.1293,
+      "step": 17
+    },
+    {
+      "epoch": 0.002477291494632535,
+      "grad_norm": 29.090116611244447,
+      "learning_rate": 1.6513761467889911e-06,
+      "loss": 5.7072,
+      "step": 18
+    },
+    {
+      "epoch": 0.002614918799889898,
+      "grad_norm": 26.22572277646329,
+      "learning_rate": 1.743119266055046e-06,
+      "loss": 5.5007,
+      "step": 19
+    },
+    {
+      "epoch": 0.0027525461051472614,
+      "grad_norm": 26.77430720735284,
+      "learning_rate": 1.8348623853211011e-06,
+      "loss": 5.5339,
+      "step": 20
+    },
+    {
+      "epoch": 0.002890173410404624,
+      "grad_norm": 24.538818580717084,
+      "learning_rate": 1.9266055045871564e-06,
+      "loss": 5.1773,
+      "step": 21
+    },
+    {
+      "epoch": 0.0030278007156619873,
+      "grad_norm": 25.789682554194393,
+      "learning_rate": 2.0183486238532113e-06,
+      "loss": 4.9759,
+      "step": 22
+    },
+    {
+      "epoch": 0.0031654280209193505,
+      "grad_norm": 26.77784988960033,
+      "learning_rate": 2.110091743119266e-06,
+      "loss": 4.7808,
+      "step": 23
+    },
+    {
+      "epoch": 0.0033030553261767133,
+      "grad_norm": 30.783704917578614,
+      "learning_rate": 2.2018348623853215e-06,
+      "loss": 4.5444,
+      "step": 24
+    },
+    {
+      "epoch": 0.0034406826314340765,
+      "grad_norm": 35.40576005126993,
+      "learning_rate": 2.2935779816513764e-06,
+      "loss": 4.3848,
+      "step": 25
+    },
+    {
+      "epoch": 0.0035783099366914397,
+      "grad_norm": 29.629770483842115,
+      "learning_rate": 2.3853211009174317e-06,
+      "loss": 4.3358,
+      "step": 26
+    },
+    {
+      "epoch": 0.0037159372419488025,
+      "grad_norm": 19.958437560482693,
+      "learning_rate": 2.4770642201834866e-06,
+      "loss": 3.9298,
+      "step": 27
+    },
+    {
+      "epoch": 0.0038535645472061657,
+      "grad_norm": 18.113189543914967,
+      "learning_rate": 2.5688073394495415e-06,
+      "loss": 3.9833,
+      "step": 28
+    },
+    {
+      "epoch": 0.003991191852463528,
+      "grad_norm": 19.629459014814753,
+      "learning_rate": 2.6605504587155968e-06,
+      "loss": 3.9207,
+      "step": 29
+    },
+    {
+      "epoch": 0.004128819157720892,
+      "grad_norm": 17.91583591372297,
+      "learning_rate": 2.7522935779816517e-06,
+      "loss": 3.8832,
+      "step": 30
+    },
+    {
+      "epoch": 0.004266446462978255,
+      "grad_norm": 18.46401475377401,
+      "learning_rate": 2.844036697247707e-06,
+      "loss": 3.6923,
+      "step": 31
+    },
+    {
+      "epoch": 0.004404073768235618,
+      "grad_norm": 18.72285351538884,
+      "learning_rate": 2.935779816513762e-06,
+      "loss": 3.4699,
+      "step": 32
+    },
+    {
+      "epoch": 0.004541701073492981,
+      "grad_norm": 16.259640150416583,
+      "learning_rate": 3.0275229357798168e-06,
+      "loss": 3.3905,
+      "step": 33
+    },
+    {
+      "epoch": 0.004679328378750344,
+      "grad_norm": 14.570187923394842,
+      "learning_rate": 3.119266055045872e-06,
+      "loss": 3.3477,
+      "step": 34
+    },
+    {
+      "epoch": 0.004816955684007707,
+      "grad_norm": 16.939322171144347,
+      "learning_rate": 3.211009174311927e-06,
+      "loss": 3.2639,
+      "step": 35
+    },
+    {
+      "epoch": 0.00495458298926507,
+      "grad_norm": 16.966681420183825,
+      "learning_rate": 3.3027522935779823e-06,
+      "loss": 3.5091,
+      "step": 36
+    },
+    {
+      "epoch": 0.005092210294522433,
+      "grad_norm": 13.209059958726078,
+      "learning_rate": 3.394495412844037e-06,
+      "loss": 3.0526,
+      "step": 37
+    },
+    {
+      "epoch": 0.005229837599779796,
+      "grad_norm": 15.651744192367802,
+      "learning_rate": 3.486238532110092e-06,
+      "loss": 2.9199,
+      "step": 38
+    },
+    {
+      "epoch": 0.0053674649050371595,
+      "grad_norm": 14.311846143454474,
+      "learning_rate": 3.5779816513761473e-06,
+      "loss": 2.9452,
+      "step": 39
+    },
+    {
+      "epoch": 0.005505092210294523,
+      "grad_norm": 12.613915749020137,
+      "learning_rate": 3.6697247706422022e-06,
+      "loss": 2.7292,
+      "step": 40
+    },
+    {
+      "epoch": 0.005642719515551885,
+      "grad_norm": 13.223492123273966,
+      "learning_rate": 3.7614678899082575e-06,
+      "loss": 2.4475,
+      "step": 41
+    },
+    {
+      "epoch": 0.005780346820809248,
+      "grad_norm": 14.336826364729564,
+      "learning_rate": 3.853211009174313e-06,
+      "loss": 2.6571,
+      "step": 42
+    },
+    {
+      "epoch": 0.0059179741260666115,
+      "grad_norm": 10.383910530550867,
+      "learning_rate": 3.944954128440367e-06,
+      "loss": 2.4292,
+      "step": 43
+    },
+    {
+      "epoch": 0.006055601431323975,
+      "grad_norm": 8.99565021368346,
+      "learning_rate": 4.036697247706423e-06,
+      "loss": 2.3244,
+      "step": 44
+    },
+    {
+      "epoch": 0.006193228736581338,
+      "grad_norm": 8.817102013735807,
+      "learning_rate": 4.128440366972478e-06,
+      "loss": 2.1943,
+      "step": 45
+    },
+    {
+      "epoch": 0.006330856041838701,
+      "grad_norm": 9.702298784362961,
+      "learning_rate": 4.220183486238532e-06,
+      "loss": 2.1952,
+      "step": 46
+    },
+    {
+      "epoch": 0.006468483347096064,
+      "grad_norm": 8.88701992027617,
+      "learning_rate": 4.311926605504588e-06,
+      "loss": 2.0957,
+      "step": 47
+    },
+    {
+      "epoch": 0.006606110652353427,
+      "grad_norm": 9.05653946287146,
+      "learning_rate": 4.403669724770643e-06,
+      "loss": 1.9548,
+      "step": 48
+    },
+    {
+      "epoch": 0.00674373795761079,
+      "grad_norm": 9.945685017558295,
+      "learning_rate": 4.4954128440366975e-06,
+      "loss": 2.1238,
+      "step": 49
+    },
+    {
+      "epoch": 0.006881365262868153,
+      "grad_norm": 9.86954527309786,
+      "learning_rate": 4.587155963302753e-06,
+      "loss": 1.8804,
+      "step": 50
+    },
+    {
+      "epoch": 0.007018992568125516,
+      "grad_norm": 11.147524766075072,
+      "learning_rate": 4.678899082568808e-06,
+      "loss": 1.7962,
+      "step": 51
+    },
+    {
+      "epoch": 0.007156619873382879,
+      "grad_norm": 12.103687673579824,
+      "learning_rate": 4.770642201834863e-06,
+      "loss": 1.7479,
+      "step": 52
+    },
+    {
+      "epoch": 0.007294247178640243,
+      "grad_norm": 11.28196444108274,
+      "learning_rate": 4.862385321100918e-06,
+      "loss": 1.7689,
+      "step": 53
+    },
+    {
+      "epoch": 0.007431874483897605,
+      "grad_norm": 8.318397086615121,
+      "learning_rate": 4.954128440366973e-06,
+      "loss": 1.6661,
+      "step": 54
+    },
+    {
+      "epoch": 0.007569501789154968,
+      "grad_norm": 7.8950793538117665,
+      "learning_rate": 5.045871559633028e-06,
+      "loss": 1.7077,
+      "step": 55
+    },
+    {
+      "epoch": 0.007707129094412331,
+      "grad_norm": 5.5054801936451545,
+      "learning_rate": 5.137614678899083e-06,
+      "loss": 1.5939,
+      "step": 56
+    },
+    {
+      "epoch": 0.007844756399669695,
+      "grad_norm": 4.169025037965987,
+      "learning_rate": 5.229357798165137e-06,
+      "loss": 1.5904,
+      "step": 57
+    },
+    {
+      "epoch": 0.007982383704927057,
+      "grad_norm": 4.897484074363938,
+      "learning_rate": 5.3211009174311936e-06,
+      "loss": 1.6256,
+      "step": 58
+    },
+    {
+      "epoch": 0.008120011010184421,
+      "grad_norm": 6.909386381109011,
+      "learning_rate": 5.412844036697248e-06,
+      "loss": 1.5304,
+      "step": 59
+    },
+    {
+      "epoch": 0.008257638315441783,
+      "grad_norm": 3.8966671981572865,
+      "learning_rate": 5.504587155963303e-06,
+      "loss": 1.6455,
+      "step": 60
+    },
+    {
+      "epoch": 0.008395265620699147,
+      "grad_norm": 4.3631520694277395,
+      "learning_rate": 5.596330275229358e-06,
+      "loss": 1.6049,
+      "step": 61
+    },
+    {
+      "epoch": 0.00853289292595651,
+      "grad_norm": 3.7637060529275987,
+      "learning_rate": 5.688073394495414e-06,
+      "loss": 1.6034,
+      "step": 62
+    },
+    {
+      "epoch": 0.008670520231213872,
+      "grad_norm": 5.80413139969054,
+      "learning_rate": 5.7798165137614684e-06,
+      "loss": 1.5927,
+      "step": 63
+    },
+    {
+      "epoch": 0.008808147536471236,
+      "grad_norm": 6.399309128340181,
+      "learning_rate": 5.871559633027524e-06,
+      "loss": 1.6421,
+      "step": 64
+    },
+    {
+      "epoch": 0.008945774841728598,
+      "grad_norm": 3.5967425904318255,
+      "learning_rate": 5.963302752293578e-06,
+      "loss": 1.6728,
+      "step": 65
+    },
+    {
+      "epoch": 0.009083402146985962,
+      "grad_norm": 4.164940529668355,
+      "learning_rate": 6.0550458715596335e-06,
+      "loss": 1.5774,
+      "step": 66
+    },
+    {
+      "epoch": 0.009221029452243325,
+      "grad_norm": 3.942303098856934,
+      "learning_rate": 6.146788990825688e-06,
+      "loss": 1.4309,
+      "step": 67
+    },
+    {
+      "epoch": 0.009358656757500689,
+      "grad_norm": 3.6098390288887434,
+      "learning_rate": 6.238532110091744e-06,
+      "loss": 1.5521,
+      "step": 68
+    },
+    {
+      "epoch": 0.009496284062758051,
+      "grad_norm": 3.613790576808798,
+      "learning_rate": 6.330275229357799e-06,
+      "loss": 1.4476,
+      "step": 69
+    },
+    {
+      "epoch": 0.009633911368015413,
+      "grad_norm": 3.6054707610689527,
+      "learning_rate": 6.422018348623854e-06,
+      "loss": 1.5639,
+      "step": 70
+    },
+    {
+      "epoch": 0.009771538673272778,
+      "grad_norm": 3.7183900147459004,
+      "learning_rate": 6.513761467889908e-06,
+      "loss": 1.5219,
+      "step": 71
+    },
+    {
+      "epoch": 0.00990916597853014,
+      "grad_norm": 3.1318978731203426,
+      "learning_rate": 6.6055045871559645e-06,
+      "loss": 1.5748,
+      "step": 72
+    },
+    {
+      "epoch": 0.010046793283787504,
+      "grad_norm": 2.980705402594747,
+      "learning_rate": 6.697247706422019e-06,
+      "loss": 1.5212,
+      "step": 73
+    },
+    {
+      "epoch": 0.010184420589044866,
+      "grad_norm": 2.5719692786623063,
+      "learning_rate": 6.788990825688074e-06,
+      "loss": 1.6141,
+      "step": 74
+    },
+    {
+      "epoch": 0.01032204789430223,
+      "grad_norm": 2.9602160150446353,
+      "learning_rate": 6.880733944954129e-06,
+      "loss": 1.5481,
+      "step": 75
+    },
+    {
+      "epoch": 0.010459675199559593,
+      "grad_norm": 4.6484036208468895,
+      "learning_rate": 6.972477064220184e-06,
+      "loss": 1.5382,
+      "step": 76
+    },
+    {
+      "epoch": 0.010597302504816955,
+      "grad_norm": 2.6422519125904738,
+      "learning_rate": 7.0642201834862385e-06,
+      "loss": 1.5199,
+      "step": 77
+    },
+    {
+      "epoch": 0.010734929810074319,
+      "grad_norm": 2.2462609269807463,
+      "learning_rate": 7.155963302752295e-06,
+      "loss": 1.4589,
+      "step": 78
+    },
+    {
+      "epoch": 0.010872557115331681,
+      "grad_norm": 2.5958416602354526,
+      "learning_rate": 7.247706422018349e-06,
+      "loss": 1.4196,
+      "step": 79
+    },
+    {
+      "epoch": 0.011010184420589045,
+      "grad_norm": 2.4322557882552434,
+      "learning_rate": 7.3394495412844045e-06,
+      "loss": 1.5169,
+      "step": 80
+    },
+    {
+      "epoch": 0.011147811725846408,
+      "grad_norm": 2.3721454375389763,
+      "learning_rate": 7.431192660550459e-06,
+      "loss": 1.4884,
+      "step": 81
+    },
+    {
+      "epoch": 0.01128543903110377,
+      "grad_norm": 2.517517646959331,
+      "learning_rate": 7.522935779816515e-06,
+      "loss": 1.5131,
+      "step": 82
+    },
+    {
+      "epoch": 0.011423066336361134,
+      "grad_norm": 2.1111513768983947,
+      "learning_rate": 7.6146788990825695e-06,
+      "loss": 1.5024,
+      "step": 83
+    },
+    {
+      "epoch": 0.011560693641618497,
+      "grad_norm": 2.19144633613799,
+      "learning_rate": 7.706422018348626e-06,
+      "loss": 1.5059,
+      "step": 84
+    },
+    {
+      "epoch": 0.01169832094687586,
+      "grad_norm": 2.649489817808233,
+      "learning_rate": 7.79816513761468e-06,
+      "loss": 1.437,
+      "step": 85
+    },
+    {
+      "epoch": 0.011835948252133223,
+      "grad_norm": 2.215611260494801,
+      "learning_rate": 7.889908256880735e-06,
+      "loss": 1.4768,
+      "step": 86
+    },
+    {
+      "epoch": 0.011973575557390587,
+      "grad_norm": 2.223805042274753,
+      "learning_rate": 7.981651376146789e-06,
+      "loss": 1.4697,
+      "step": 87
+    },
+    {
+      "epoch": 0.01211120286264795,
+      "grad_norm": 3.0018648238797803,
+      "learning_rate": 8.073394495412845e-06,
+      "loss": 1.6319,
+      "step": 88
+    },
+    {
+      "epoch": 0.012248830167905312,
+      "grad_norm": 2.0243528943519937,
+      "learning_rate": 8.1651376146789e-06,
+      "loss": 1.4016,
+      "step": 89
+    },
+    {
+      "epoch": 0.012386457473162676,
+      "grad_norm": 2.386818589224275,
+      "learning_rate": 8.256880733944956e-06,
+      "loss": 1.4194,
+      "step": 90
+    },
+    {
+      "epoch": 0.012524084778420038,
+      "grad_norm": 2.0395702062949073,
+      "learning_rate": 8.34862385321101e-06,
+      "loss": 1.5033,
+      "step": 91
+    },
+    {
+      "epoch": 0.012661712083677402,
+      "grad_norm": 2.633921139273025,
+      "learning_rate": 8.440366972477065e-06,
+      "loss": 1.5427,
+      "step": 92
+    },
+    {
+      "epoch": 0.012799339388934764,
+      "grad_norm": 2.276591928540183,
+      "learning_rate": 8.53211009174312e-06,
+      "loss": 1.4927,
+      "step": 93
+    },
+    {
+      "epoch": 0.012936966694192129,
+      "grad_norm": 2.288218912762329,
+      "learning_rate": 8.623853211009175e-06,
+      "loss": 1.4559,
+      "step": 94
+    },
+    {
+      "epoch": 0.01307459399944949,
+      "grad_norm": 2.1128719242838665,
+      "learning_rate": 8.71559633027523e-06,
+      "loss": 1.4829,
+      "step": 95
+    },
+    {
+      "epoch": 0.013212221304706853,
+      "grad_norm": 2.2711246581766913,
+      "learning_rate": 8.807339449541286e-06,
+      "loss": 1.4727,
+      "step": 96
+    },
+    {
+      "epoch": 0.013349848609964217,
+      "grad_norm": 2.1130231906451664,
+      "learning_rate": 8.89908256880734e-06,
+      "loss": 1.4609,
+      "step": 97
+    },
+    {
+      "epoch": 0.01348747591522158,
+      "grad_norm": 2.220821852164481,
+      "learning_rate": 8.990825688073395e-06,
+      "loss": 1.3396,
+      "step": 98
+    },
+    {
+      "epoch": 0.013625103220478944,
+      "grad_norm": 2.3633420704730743,
+      "learning_rate": 9.08256880733945e-06,
+      "loss": 1.5284,
+      "step": 99
+    },
+    {
+      "epoch": 0.013762730525736306,
+      "grad_norm": 2.7357866827984103,
+      "learning_rate": 9.174311926605506e-06,
+      "loss": 1.5814,
+      "step": 100
+    },
+    {
+      "epoch": 0.013900357830993668,
+      "grad_norm": 2.5183026028685274,
+      "learning_rate": 9.26605504587156e-06,
+      "loss": 1.4183,
+      "step": 101
+    },
+    {
+      "epoch": 0.014037985136251032,
+      "grad_norm": 2.220648136823632,
+      "learning_rate": 9.357798165137616e-06,
+      "loss": 1.3791,
+      "step": 102
+    },
+    {
+      "epoch": 0.014175612441508395,
+      "grad_norm": 2.620225939456079,
+      "learning_rate": 9.44954128440367e-06,
+      "loss": 1.5901,
+      "step": 103
+    },
+    {
+      "epoch": 0.014313239746765759,
+      "grad_norm": 2.373693059130836,
+      "learning_rate": 9.541284403669727e-06,
+      "loss": 1.3812,
+      "step": 104
+    },
+    {
+      "epoch": 0.014450867052023121,
+      "grad_norm": 2.290956215731559,
+      "learning_rate": 9.633027522935781e-06,
+      "loss": 1.5668,
+      "step": 105
+    },
+    {
+      "epoch": 0.014588494357280485,
+      "grad_norm": 2.151827990042943,
+      "learning_rate": 9.724770642201836e-06,
+      "loss": 1.4094,
+      "step": 106
+    },
+    {
+      "epoch": 0.014726121662537848,
+      "grad_norm": 2.25935072143266,
+      "learning_rate": 9.81651376146789e-06,
+      "loss": 1.389,
+      "step": 107
+    },
+    {
+      "epoch": 0.01486374896779521,
+      "grad_norm": 1.9427811247759121,
+      "learning_rate": 9.908256880733946e-06,
+      "loss": 1.3752,
+      "step": 108
+    },
+    {
+      "epoch": 0.015001376273052574,
+      "grad_norm": 2.670354810474395,
+      "learning_rate": 1e-05,
+      "loss": 1.3999,
+      "step": 109
+    },
+    {
+      "epoch": 0.015139003578309936,
+      "grad_norm": 2.6207542392337064,
+      "learning_rate": 1.0091743119266055e-05,
+      "loss": 1.4274,
+      "step": 110
+    },
+    {
+      "epoch": 0.0152766308835673,
+      "grad_norm": 2.032814746469394,
+      "learning_rate": 1.018348623853211e-05,
+      "loss": 1.4037,
+      "step": 111
+    },
+    {
+      "epoch": 0.015414258188824663,
+      "grad_norm": 2.3553013514581376,
+      "learning_rate": 1.0275229357798166e-05,
+      "loss": 1.3847,
+      "step": 112
+    },
+    {
+      "epoch": 0.015551885494082027,
+      "grad_norm": 2.3604510534366185,
+      "learning_rate": 1.036697247706422e-05,
+      "loss": 1.4721,
+      "step": 113
+    },
+    {
+      "epoch": 0.01568951279933939,
+      "grad_norm": 1.990776274775859,
+      "learning_rate": 1.0458715596330275e-05,
+      "loss": 1.4307,
+      "step": 114
+    },
+    {
+      "epoch": 0.01582714010459675,
+      "grad_norm": 2.4366239781037335,
+      "learning_rate": 1.055045871559633e-05,
+      "loss": 1.3784,
+      "step": 115
+    },
+    {
+      "epoch": 0.015964767409854114,
+      "grad_norm": 2.1530688662663025,
+      "learning_rate": 1.0642201834862387e-05,
+      "loss": 1.5301,
+      "step": 116
+    },
+    {
+      "epoch": 0.01610239471511148,
+      "grad_norm": 2.4856890182030273,
+      "learning_rate": 1.0733944954128442e-05,
+      "loss": 1.5627,
+      "step": 117
+    },
+    {
+      "epoch": 0.016240022020368842,
+      "grad_norm": 2.4715849958308707,
+      "learning_rate": 1.0825688073394496e-05,
+      "loss": 1.5468,
+      "step": 118
+    },
+    {
+      "epoch": 0.016377649325626204,
+      "grad_norm": 2.535683782971731,
+      "learning_rate": 1.091743119266055e-05,
+      "loss": 1.4756,
+      "step": 119
+    },
+    {
+      "epoch": 0.016515276630883566,
+      "grad_norm": 2.4541634820421754,
+      "learning_rate": 1.1009174311926607e-05,
+      "loss": 1.5382,
+      "step": 120
+    },
+    {
+      "epoch": 0.01665290393614093,
+      "grad_norm": 2.1539947467318554,
+      "learning_rate": 1.1100917431192661e-05,
+      "loss": 1.358,
+      "step": 121
+    },
+    {
+      "epoch": 0.016790531241398295,
+      "grad_norm": 1.9474224783077931,
+      "learning_rate": 1.1192660550458716e-05,
+      "loss": 1.3763,
+      "step": 122
+    },
+    {
+      "epoch": 0.016928158546655657,
+      "grad_norm": 2.078875556516513,
+      "learning_rate": 1.128440366972477e-05,
+      "loss": 1.3807,
+      "step": 123
+    },
+    {
+      "epoch": 0.01706578585191302,
+      "grad_norm": 2.4055668082267108,
+      "learning_rate": 1.1376146788990828e-05,
+      "loss": 1.3872,
+      "step": 124
+    },
+    {
+      "epoch": 0.01720341315717038,
+      "grad_norm": 2.3132588920774912,
+      "learning_rate": 1.1467889908256882e-05,
+      "loss": 1.4327,
+      "step": 125
+    },
+    {
+      "epoch": 0.017341040462427744,
+      "grad_norm": 2.4527793924936936,
+      "learning_rate": 1.1559633027522937e-05,
+      "loss": 1.5077,
+      "step": 126
+    },
+    {
+      "epoch": 0.01747866776768511,
+      "grad_norm": 2.0978600503608504,
+      "learning_rate": 1.1651376146788991e-05,
+      "loss": 1.3669,
+      "step": 127
+    },
+    {
+      "epoch": 0.017616295072942472,
+      "grad_norm": 2.1239701609371626,
+      "learning_rate": 1.1743119266055047e-05,
+      "loss": 1.401,
+      "step": 128
+    },
+    {
+      "epoch": 0.017753922378199834,
+      "grad_norm": 2.5224577529066083,
+      "learning_rate": 1.1834862385321102e-05,
+      "loss": 1.4933,
+      "step": 129
+    },
+    {
+      "epoch": 0.017891549683457197,
+      "grad_norm": 1.9091441052093827,
+      "learning_rate": 1.1926605504587156e-05,
+      "loss": 1.47,
+      "step": 130
+    },
+    {
+      "epoch": 0.018029176988714563,
+      "grad_norm": 2.117237280408027,
+      "learning_rate": 1.2018348623853211e-05,
+      "loss": 1.5108,
+      "step": 131
+    },
+    {
+      "epoch": 0.018166804293971925,
+      "grad_norm": 2.2082410149915574,
+      "learning_rate": 1.2110091743119267e-05,
+      "loss": 1.4637,
+      "step": 132
+    },
+    {
+      "epoch": 0.018304431599229287,
+      "grad_norm": 2.6122172487718007,
+      "learning_rate": 1.2201834862385321e-05,
+      "loss": 1.5086,
+      "step": 133
+    },
+    {
+      "epoch": 0.01844205890448665,
+      "grad_norm": 2.008839943937337,
+      "learning_rate": 1.2293577981651376e-05,
+      "loss": 1.2705,
+      "step": 134
+    },
+    {
+      "epoch": 0.018579686209744012,
+      "grad_norm": 2.3695876697774083,
+      "learning_rate": 1.238532110091743e-05,
+      "loss": 1.4219,
+      "step": 135
+    },
+    {
+      "epoch": 0.018717313515001378,
+      "grad_norm": 2.2118887087092185,
+      "learning_rate": 1.2477064220183488e-05,
+      "loss": 1.4856,
+      "step": 136
+    },
+    {
+      "epoch": 0.01885494082025874,
+      "grad_norm": 2.189617479512794,
+      "learning_rate": 1.2568807339449543e-05,
+      "loss": 1.4182,
+      "step": 137
+    },
+    {
+      "epoch": 0.018992568125516102,
+      "grad_norm": 1.9338747414395177,
+      "learning_rate": 1.2660550458715597e-05,
+      "loss": 1.3458,
+      "step": 138
+    },
+    {
+      "epoch": 0.019130195430773465,
+      "grad_norm": 2.5686958772824613,
+      "learning_rate": 1.2752293577981652e-05,
+      "loss": 1.3838,
+      "step": 139
+    },
+    {
+      "epoch": 0.019267822736030827,
+      "grad_norm": 2.030881709271741,
+      "learning_rate": 1.2844036697247708e-05,
+      "loss": 1.3908,
+      "step": 140
+    },
+    {
+      "epoch": 0.019405450041288193,
+      "grad_norm": 2.563599227740635,
+      "learning_rate": 1.2935779816513762e-05,
+      "loss": 1.4337,
+      "step": 141
+    },
+    {
+      "epoch": 0.019543077346545555,
+      "grad_norm": 1.8176109793242443,
+      "learning_rate": 1.3027522935779817e-05,
+      "loss": 1.3158,
+      "step": 142
+    },
+    {
+      "epoch": 0.019680704651802917,
+      "grad_norm": 1.8019193031728145,
+      "learning_rate": 1.3119266055045871e-05,
+      "loss": 1.3311,
+      "step": 143
+    },
+    {
+      "epoch": 0.01981833195706028,
+      "grad_norm": 2.698646522727441,
+      "learning_rate": 1.3211009174311929e-05,
+      "loss": 1.5612,
+      "step": 144
+    },
+    {
+      "epoch": 0.019955959262317642,
+      "grad_norm": 1.9171636115670228,
+      "learning_rate": 1.3302752293577984e-05,
+      "loss": 1.305,
+      "step": 145
+    },
+    {
+      "epoch": 0.020093586567575008,
+      "grad_norm": 1.7563775578267051,
+      "learning_rate": 1.3394495412844038e-05,
+      "loss": 1.346,
+      "step": 146
+    },
+    {
+      "epoch": 0.02023121387283237,
+      "grad_norm": 1.683948752339103,
+      "learning_rate": 1.3486238532110092e-05,
+      "loss": 1.3838,
+      "step": 147
+    },
+    {
+      "epoch": 0.020368841178089733,
+      "grad_norm": 1.9218370036130723,
+      "learning_rate": 1.3577981651376149e-05,
+      "loss": 1.3623,
+      "step": 148
+    },
+    {
+      "epoch": 0.020506468483347095,
+      "grad_norm": 1.9968951763915517,
+      "learning_rate": 1.3669724770642203e-05,
+      "loss": 1.3466,
+      "step": 149
+    },
+    {
+      "epoch": 0.02064409578860446,
+      "grad_norm": 1.7657209630970832,
+      "learning_rate": 1.3761467889908258e-05,
+      "loss": 1.335,
+      "step": 150
+    },
+    {
+      "epoch": 0.020781723093861823,
+      "grad_norm": 2.275067067517579,
+      "learning_rate": 1.3853211009174312e-05,
+      "loss": 1.3795,
+      "step": 151
+    },
+    {
+      "epoch": 0.020919350399119185,
+      "grad_norm": 2.0946969906583455,
+      "learning_rate": 1.3944954128440368e-05,
+      "loss": 1.4477,
+      "step": 152
+    },
+    {
+      "epoch": 0.021056977704376548,
+      "grad_norm": 1.94844802962327,
+      "learning_rate": 1.4036697247706423e-05,
+      "loss": 1.2773,
+      "step": 153
+    },
+    {
+      "epoch": 0.02119460500963391,
+      "grad_norm": 1.895306116339568,
+      "learning_rate": 1.4128440366972477e-05,
+      "loss": 1.3181,
+      "step": 154
+    },
+    {
+      "epoch": 0.021332232314891276,
+      "grad_norm": 2.367957867670617,
+      "learning_rate": 1.4220183486238533e-05,
+      "loss": 1.2996,
+      "step": 155
+    },
+    {
+      "epoch": 0.021469859620148638,
+      "grad_norm": 2.275921389073251,
+      "learning_rate": 1.431192660550459e-05,
+      "loss": 1.3412,
+      "step": 156
+    },
+    {
+      "epoch": 0.021607486925406,
+      "grad_norm": 2.311174722296487,
+      "learning_rate": 1.4403669724770644e-05,
+      "loss": 1.4006,
+      "step": 157
+    },
+    {
+      "epoch": 0.021745114230663363,
+      "grad_norm": 1.9058240934259199,
+      "learning_rate": 1.4495412844036698e-05,
+      "loss": 1.3754,
+      "step": 158
+    },
+    {
+      "epoch": 0.021882741535920725,
+      "grad_norm": 2.0591383358940933,
+      "learning_rate": 1.4587155963302753e-05,
+      "loss": 1.3629,
+      "step": 159
+    },
+    {
+      "epoch": 0.02202036884117809,
+      "grad_norm": 2.2042816428987124,
+      "learning_rate": 1.4678899082568809e-05,
+      "loss": 1.4356,
+      "step": 160
+    },
+    {
+      "epoch": 0.022157996146435453,
+      "grad_norm": 2.324548086443444,
+      "learning_rate": 1.4770642201834863e-05,
+      "loss": 1.3237,
+      "step": 161
+    },
+    {
+      "epoch": 0.022295623451692816,
+      "grad_norm": 2.297329627871107,
+      "learning_rate": 1.4862385321100918e-05,
+      "loss": 1.3938,
+      "step": 162
+    },
+    {
+      "epoch": 0.022433250756950178,
+      "grad_norm": 1.845311250165973,
+      "learning_rate": 1.4954128440366972e-05,
+      "loss": 1.3465,
+      "step": 163
+    },
+    {
+      "epoch": 0.02257087806220754,
+      "grad_norm": 2.3928422276791292,
+      "learning_rate": 1.504587155963303e-05,
+      "loss": 1.2567,
+      "step": 164
+    },
+    {
+      "epoch": 0.022708505367464906,
+      "grad_norm": 2.0869370512078005,
+      "learning_rate": 1.5137614678899085e-05,
+      "loss": 1.354,
+      "step": 165
+    },
+    {
+      "epoch": 0.02284613267272227,
+      "grad_norm": 2.2948529551834866,
+      "learning_rate": 1.5229357798165139e-05,
+      "loss": 1.3956,
+      "step": 166
+    },
+    {
+      "epoch": 0.02298375997797963,
+      "grad_norm": 2.1356579266269633,
+      "learning_rate": 1.5321100917431192e-05,
+      "loss": 1.3029,
+      "step": 167
+    },
+    {
+      "epoch": 0.023121387283236993,
+      "grad_norm": 1.879315448535232,
+      "learning_rate": 1.541284403669725e-05,
+      "loss": 1.4359,
+      "step": 168
+    },
+    {
+      "epoch": 0.02325901458849436,
+      "grad_norm": 2.2985390452183383,
+      "learning_rate": 1.5504587155963304e-05,
+      "loss": 1.428,
+      "step": 169
+    },
+    {
+      "epoch": 0.02339664189375172,
+      "grad_norm": 2.1614017756521218,
+      "learning_rate": 1.559633027522936e-05,
+      "loss": 1.362,
+      "step": 170
+    },
+    {
+      "epoch": 0.023534269199009084,
+      "grad_norm": 1.9571233839509083,
+      "learning_rate": 1.5688073394495413e-05,
+      "loss": 1.3408,
+      "step": 171
+    },
+    {
+      "epoch": 0.023671896504266446,
+      "grad_norm": 2.204507265231331,
+      "learning_rate": 1.577981651376147e-05,
+      "loss": 1.3756,
+      "step": 172
+    },
+    {
+      "epoch": 0.023809523809523808,
+      "grad_norm": 1.8598561696357072,
+      "learning_rate": 1.5871559633027525e-05,
+      "loss": 1.3254,
+      "step": 173
+    },
+    {
+      "epoch": 0.023947151114781174,
+      "grad_norm": 2.028593408652933,
+      "learning_rate": 1.5963302752293578e-05,
+      "loss": 1.3713,
+      "step": 174
+    },
+    {
+      "epoch": 0.024084778420038536,
+      "grad_norm": 2.068922340732782,
+      "learning_rate": 1.6055045871559634e-05,
+      "loss": 1.4301,
+      "step": 175
+    },
+    {
+      "epoch": 0.0242224057252959,
+      "grad_norm": 2.6177132341150635,
+      "learning_rate": 1.614678899082569e-05,
+      "loss": 1.3865,
+      "step": 176
+    },
+    {
+      "epoch": 0.02436003303055326,
+      "grad_norm": 1.9257037964088355,
+      "learning_rate": 1.6238532110091743e-05,
+      "loss": 1.3321,
+      "step": 177
+    },
+    {
+      "epoch": 0.024497660335810623,
+      "grad_norm": 2.089865470637852,
+      "learning_rate": 1.63302752293578e-05,
+      "loss": 1.3191,
+      "step": 178
+    },
+    {
+      "epoch": 0.02463528764106799,
+      "grad_norm": 2.2931036240406115,
+      "learning_rate": 1.6422018348623852e-05,
+      "loss": 1.4369,
+      "step": 179
+    },
+    {
+      "epoch": 0.02477291494632535,
+      "grad_norm": 2.0069970933947836,
+      "learning_rate": 1.6513761467889912e-05,
+      "loss": 1.3773,
+      "step": 180
+    },
+    {
+      "epoch": 0.024910542251582714,
+      "grad_norm": 2.17213886306559,
+      "learning_rate": 1.6605504587155964e-05,
+      "loss": 1.434,
+      "step": 181
+    },
+    {
+      "epoch": 0.025048169556840076,
+      "grad_norm": 1.8345577373199526,
+      "learning_rate": 1.669724770642202e-05,
+      "loss": 1.292,
+      "step": 182
+    },
+    {
+      "epoch": 0.02518579686209744,
+      "grad_norm": 2.2382844643998583,
+      "learning_rate": 1.6788990825688073e-05,
+      "loss": 1.3798,
+      "step": 183
+    },
+    {
+      "epoch": 0.025323424167354804,
+      "grad_norm": 2.3380349507564033,
+      "learning_rate": 1.688073394495413e-05,
+      "loss": 1.329,
+      "step": 184
+    },
+    {
+      "epoch": 0.025461051472612167,
+      "grad_norm": 1.802959440174109,
+      "learning_rate": 1.6972477064220186e-05,
+      "loss": 1.308,
+      "step": 185
+    },
+    {
+      "epoch": 0.02559867877786953,
+      "grad_norm": 2.152666761656666,
+      "learning_rate": 1.706422018348624e-05,
+      "loss": 1.2755,
+      "step": 186
+    },
+    {
+      "epoch": 0.02573630608312689,
+      "grad_norm": 1.8448538889980963,
+      "learning_rate": 1.7155963302752295e-05,
+      "loss": 1.3412,
+      "step": 187
+    },
+    {
+      "epoch": 0.025873933388384257,
+      "grad_norm": 1.8126166741608833,
+      "learning_rate": 1.724770642201835e-05,
+      "loss": 1.3208,
+      "step": 188
+    },
+    {
+      "epoch": 0.02601156069364162,
+      "grad_norm": 2.468183706025089,
+      "learning_rate": 1.7339449541284407e-05,
+      "loss": 1.3522,
+      "step": 189
+    },
+    {
+      "epoch": 0.02614918799889898,
+      "grad_norm": 2.0889631328324145,
+      "learning_rate": 1.743119266055046e-05,
+      "loss": 1.303,
+      "step": 190
+    },
+    {
+      "epoch": 0.026286815304156344,
+      "grad_norm": 1.8802549683018899,
+      "learning_rate": 1.7522935779816516e-05,
+      "loss": 1.3033,
+      "step": 191
+    },
+    {
+      "epoch": 0.026424442609413706,
+      "grad_norm": 2.102205226581823,
+      "learning_rate": 1.7614678899082572e-05,
+      "loss": 1.3699,
+      "step": 192
+    },
+    {
+      "epoch": 0.026562069914671072,
+      "grad_norm": 1.9127372886810032,
+      "learning_rate": 1.7706422018348625e-05,
+      "loss": 1.3024,
+      "step": 193
+    },
+    {
+      "epoch": 0.026699697219928435,
+      "grad_norm": 2.263119460489956,
+      "learning_rate": 1.779816513761468e-05,
+      "loss": 1.3812,
+      "step": 194
+    },
+    {
+      "epoch": 0.026837324525185797,
+      "grad_norm": 3.2316683630938847,
+      "learning_rate": 1.7889908256880734e-05,
+      "loss": 1.449,
+      "step": 195
+    },
+    {
+      "epoch": 0.02697495183044316,
+      "grad_norm": 2.2282560974129475,
+      "learning_rate": 1.798165137614679e-05,
+      "loss": 1.3174,
+      "step": 196
+    },
+    {
+      "epoch": 0.02711257913570052,
+      "grad_norm": 2.0650109257707827,
+      "learning_rate": 1.8073394495412846e-05,
+      "loss": 1.401,
+      "step": 197
+    },
+    {
+      "epoch": 0.027250206440957887,
+      "grad_norm": 1.9541818117446426,
+      "learning_rate": 1.81651376146789e-05,
+      "loss": 1.3052,
+      "step": 198
+    },
+    {
+      "epoch": 0.02738783374621525,
+      "grad_norm": 2.29522795249014,
+      "learning_rate": 1.8256880733944955e-05,
+      "loss": 1.4411,
+      "step": 199
+    },
+    {
+      "epoch": 0.027525461051472612,
+      "grad_norm": 2.0593967064852485,
+      "learning_rate": 1.834862385321101e-05,
+      "loss": 1.3792,
+      "step": 200
+    },
+    {
+      "epoch": 0.027663088356729974,
+      "grad_norm": 1.9422640470971306,
+      "learning_rate": 1.8440366972477067e-05,
+      "loss": 1.2357,
+      "step": 201
+    },
+    {
+      "epoch": 0.027800715661987337,
+      "grad_norm": 1.8175882378871986,
+      "learning_rate": 1.853211009174312e-05,
+      "loss": 1.3464,
+      "step": 202
+    },
+    {
+      "epoch": 0.027938342967244702,
+      "grad_norm": 1.8652650357485816,
+      "learning_rate": 1.8623853211009176e-05,
+      "loss": 1.3388,
+      "step": 203
+    },
+    {
+      "epoch": 0.028075970272502065,
+      "grad_norm": 1.9321739266272018,
+      "learning_rate": 1.8715596330275232e-05,
+      "loss": 1.4048,
+      "step": 204
+    },
+    {
+      "epoch": 0.028213597577759427,
+      "grad_norm": 1.9395894514827925,
+      "learning_rate": 1.8807339449541285e-05,
+      "loss": 1.3342,
+      "step": 205
+    },
+    {
+      "epoch": 0.02835122488301679,
+      "grad_norm": 2.1963289957192593,
+      "learning_rate": 1.889908256880734e-05,
+      "loss": 1.3347,
+      "step": 206
+    },
+    {
+      "epoch": 0.028488852188274155,
+      "grad_norm": 2.3476467008195225,
+      "learning_rate": 1.8990825688073394e-05,
+      "loss": 1.3726,
+      "step": 207
+    },
+    {
+      "epoch": 0.028626479493531518,
+      "grad_norm": 1.7420923036134865,
+      "learning_rate": 1.9082568807339454e-05,
+      "loss": 1.3369,
+      "step": 208
+    },
+    {
+      "epoch": 0.02876410679878888,
+      "grad_norm": 1.939565410980372,
+      "learning_rate": 1.9174311926605506e-05,
+      "loss": 1.31,
+      "step": 209
+    },
+    {
+      "epoch": 0.028901734104046242,
+      "grad_norm": 2.1394287453368537,
+      "learning_rate": 1.9266055045871563e-05,
+      "loss": 1.3769,
+      "step": 210
+    },
+    {
+      "epoch": 0.029039361409303605,
+      "grad_norm": 2.071505816789772,
+      "learning_rate": 1.9357798165137615e-05,
+      "loss": 1.339,
+      "step": 211
+    },
+    {
+      "epoch": 0.02917698871456097,
+      "grad_norm": 2.4395853297564636,
+      "learning_rate": 1.944954128440367e-05,
+      "loss": 1.3921,
+      "step": 212
+    },
+    {
+      "epoch": 0.029314616019818333,
+      "grad_norm": 1.9367232795084275,
+      "learning_rate": 1.9541284403669728e-05,
+      "loss": 1.3692,
+      "step": 213
+    },
+    {
+      "epoch": 0.029452243325075695,
+      "grad_norm": 2.1528438766735403,
+      "learning_rate": 1.963302752293578e-05,
+      "loss": 1.4347,
+      "step": 214
+    },
+    {
+      "epoch": 0.029589870630333057,
+      "grad_norm": 2.0232019543566793,
+      "learning_rate": 1.9724770642201837e-05,
+      "loss": 1.3108,
+      "step": 215
+    },
+    {
+      "epoch": 0.02972749793559042,
+      "grad_norm": 2.048476375562197,
+      "learning_rate": 1.9816513761467893e-05,
+      "loss": 1.3485,
+      "step": 216
+    },
+    {
+      "epoch": 0.029865125240847785,
+      "grad_norm": 1.8724633867335398,
+      "learning_rate": 1.9908256880733945e-05,
+      "loss": 1.3775,
+      "step": 217
+    },
+    {
+      "epoch": 0.030002752546105148,
+      "grad_norm": 1.9466022384594104,
+      "learning_rate": 2e-05,
+      "loss": 1.3636,
+      "step": 218
+    },
+    {
+      "epoch": 0.03014037985136251,
+      "grad_norm": 1.7548298294927687,
+      "learning_rate": 1.9999999006568445e-05,
+      "loss": 1.3478,
+      "step": 219
+    },
+    {
+      "epoch": 0.030278007156619872,
+      "grad_norm": 1.8347380541503355,
+      "learning_rate": 1.999999602627397e-05,
+      "loss": 1.3415,
+      "step": 220
+    },
+    {
+      "epoch": 0.030415634461877235,
+      "grad_norm": 1.9335671410050412,
+      "learning_rate": 1.9999991059117167e-05,
+      "loss": 1.2799,
+      "step": 221
+    },
+    {
+      "epoch": 0.0305532617671346,
+      "grad_norm": 2.00152236637167,
+      "learning_rate": 1.999998410509903e-05,
+      "loss": 1.279,
+      "step": 222
+    },
+    {
+      "epoch": 0.030690889072391963,
+      "grad_norm": 1.8593056101521477,
+      "learning_rate": 1.999997516422093e-05,
+      "loss": 1.355,
+      "step": 223
+    },
+    {
+      "epoch": 0.030828516377649325,
+      "grad_norm": 2.0407496049002467,
+      "learning_rate": 1.9999964236484656e-05,
+      "loss": 1.297,
+      "step": 224
+    },
+    {
+      "epoch": 0.030966143682906688,
+      "grad_norm": 2.198753152129754,
+      "learning_rate": 1.9999951321892372e-05,
+      "loss": 1.3341,
+      "step": 225
+    },
+    {
+      "epoch": 0.031103770988164053,
+      "grad_norm": 1.9251066337928342,
+      "learning_rate": 1.9999936420446643e-05,
+      "loss": 1.2439,
+      "step": 226
+    },
+    {
+      "epoch": 0.031241398293421416,
+      "grad_norm": 2.0392613455325375,
+      "learning_rate": 1.9999919532150435e-05,
+      "loss": 1.362,
+      "step": 227
+    },
+    {
+      "epoch": 0.03137902559867878,
+      "grad_norm": 2.0012281863675416,
+      "learning_rate": 1.9999900657007094e-05,
+      "loss": 1.3278,
+      "step": 228
+    },
+    {
+      "epoch": 0.03151665290393614,
+      "grad_norm": 2.091729274288602,
+      "learning_rate": 1.999987979502038e-05,
+      "loss": 1.2744,
+      "step": 229
+    },
+    {
+      "epoch": 0.0316542802091935,
+      "grad_norm": 1.9583046678445861,
+      "learning_rate": 1.9999856946194435e-05,
+      "loss": 1.3289,
+      "step": 230
+    },
+    {
+      "epoch": 0.031791907514450865,
+      "grad_norm": 2.2321874709738494,
+      "learning_rate": 1.9999832110533794e-05,
+      "loss": 1.4639,
+      "step": 231
+    },
+    {
+      "epoch": 0.03192953481970823,
+      "grad_norm": 1.9568634380870933,
+      "learning_rate": 1.9999805288043398e-05,
+      "loss": 1.404,
+      "step": 232
+    },
+    {
+      "epoch": 0.03206716212496559,
+      "grad_norm": 2.0794142752238804,
+      "learning_rate": 1.9999776478728574e-05,
+      "loss": 1.2842,
+      "step": 233
+    },
+    {
+      "epoch": 0.03220478943022296,
+      "grad_norm": 2.03025781353407,
+      "learning_rate": 1.9999745682595044e-05,
+      "loss": 1.2752,
+      "step": 234
+    },
+    {
+      "epoch": 0.03234241673548032,
+      "grad_norm": 1.8111188733288912,
+      "learning_rate": 1.999971289964893e-05,
+      "loss": 1.3175,
+      "step": 235
+    },
+    {
+      "epoch": 0.032480044040737684,
+      "grad_norm": 1.9569011952305575,
+      "learning_rate": 1.9999678129896742e-05,
+      "loss": 1.3261,
+      "step": 236
+    },
+    {
+      "epoch": 0.032617671345995046,
+      "grad_norm": 1.9990194066635878,
+      "learning_rate": 1.9999641373345394e-05,
+      "loss": 1.336,
+      "step": 237
+    },
+    {
+      "epoch": 0.03275529865125241,
+      "grad_norm": 1.8364956102738534,
+      "learning_rate": 1.999960263000218e-05,
+      "loss": 1.255,
+      "step": 238
+    },
+    {
+      "epoch": 0.03289292595650977,
+      "grad_norm": 1.929709663665978,
+      "learning_rate": 1.9999561899874807e-05,
+      "loss": 1.3415,
+      "step": 239
+    },
+    {
+      "epoch": 0.03303055326176713,
+      "grad_norm": 1.8032861518293837,
+      "learning_rate": 1.9999519182971364e-05,
+      "loss": 1.3253,
+      "step": 240
+    },
+    {
+      "epoch": 0.033168180567024495,
+      "grad_norm": 2.0279852481671714,
+      "learning_rate": 1.9999474479300335e-05,
+      "loss": 1.2381,
+      "step": 241
+    },
+    {
+      "epoch": 0.03330580787228186,
+      "grad_norm": 1.792431506777118,
+      "learning_rate": 1.9999427788870607e-05,
+      "loss": 1.2443,
+      "step": 242
+    },
+    {
+      "epoch": 0.03344343517753923,
+      "grad_norm": 2.3241923034830894,
+      "learning_rate": 1.9999379111691455e-05,
+      "loss": 1.3475,
+      "step": 243
+    },
+    {
+      "epoch": 0.03358106248279659,
+      "grad_norm": 2.0510620159686885,
+      "learning_rate": 1.999932844777255e-05,
+      "loss": 1.339,
+      "step": 244
+    },
+    {
+      "epoch": 0.03371868978805395,
+      "grad_norm": 2.1677207215823655,
+      "learning_rate": 1.9999275797123955e-05,
+      "loss": 1.3016,
+      "step": 245
+    },
+    {
+      "epoch": 0.033856317093311314,
+      "grad_norm": 1.7919750093120932,
+      "learning_rate": 1.9999221159756137e-05,
+      "loss": 1.3405,
+      "step": 246
+    },
+    {
+      "epoch": 0.033993944398568676,
+      "grad_norm": 1.9042317968595934,
+      "learning_rate": 1.9999164535679948e-05,
+      "loss": 1.2821,
+      "step": 247
+    },
+    {
+      "epoch": 0.03413157170382604,
+      "grad_norm": 3.6777603979460642,
+      "learning_rate": 1.9999105924906642e-05,
+      "loss": 1.3485,
+      "step": 248
+    },
+    {
+      "epoch": 0.0342691990090834,
+      "grad_norm": 1.9428729713164685,
+      "learning_rate": 1.999904532744786e-05,
+      "loss": 1.3255,
+      "step": 249
+    },
+    {
+      "epoch": 0.03440682631434076,
+      "grad_norm": 2.383411066942213,
+      "learning_rate": 1.9998982743315643e-05,
+      "loss": 1.35,
+      "step": 250
+    },
+    {
+      "epoch": 0.034544453619598126,
+      "grad_norm": 2.2379833061766097,
+      "learning_rate": 1.9998918172522422e-05,
+      "loss": 1.4816,
+      "step": 251
+    },
+    {
+      "epoch": 0.03468208092485549,
+      "grad_norm": 1.923733923959379,
+      "learning_rate": 1.9998851615081037e-05,
+      "loss": 1.3208,
+      "step": 252
+    },
+    {
+      "epoch": 0.03481970823011286,
+      "grad_norm": 2.0879502477647005,
+      "learning_rate": 1.9998783071004704e-05,
+      "loss": 1.4089,
+      "step": 253
+    },
+    {
+      "epoch": 0.03495733553537022,
+      "grad_norm": 1.965937217365374,
+      "learning_rate": 1.999871254030704e-05,
+      "loss": 1.3721,
+      "step": 254
+    },
+    {
+      "epoch": 0.03509496284062758,
+      "grad_norm": 2.1222299354082796,
+      "learning_rate": 1.9998640023002065e-05,
+      "loss": 1.3264,
+      "step": 255
+    },
+    {
+      "epoch": 0.035232590145884944,
+      "grad_norm": 1.9253363012862292,
+      "learning_rate": 1.9998565519104182e-05,
+      "loss": 1.3766,
+      "step": 256
+    },
+    {
+      "epoch": 0.035370217451142306,
+      "grad_norm": 2.1348805530495536,
+      "learning_rate": 1.9998489028628197e-05,
+      "loss": 1.3228,
+      "step": 257
+    },
+    {
+      "epoch": 0.03550784475639967,
+      "grad_norm": 2.044301356776676,
+      "learning_rate": 1.9998410551589308e-05,
+      "loss": 1.3664,
+      "step": 258
+    },
+    {
+      "epoch": 0.03564547206165703,
+      "grad_norm": 1.9081403517070847,
+      "learning_rate": 1.99983300880031e-05,
+      "loss": 1.3123,
+      "step": 259
+    },
+    {
+      "epoch": 0.03578309936691439,
+      "grad_norm": 2.3317066324035514,
+      "learning_rate": 1.9998247637885574e-05,
+      "loss": 1.4008,
+      "step": 260
+    },
+    {
+      "epoch": 0.035920726672171756,
+      "grad_norm": 1.8314659800631017,
+      "learning_rate": 1.9998163201253097e-05,
+      "loss": 1.2814,
+      "step": 261
+    },
+    {
+      "epoch": 0.036058353977429125,
+      "grad_norm": 2.0508282899685923,
+      "learning_rate": 1.999807677812245e-05,
+      "loss": 1.3333,
+      "step": 262
+    },
+    {
+      "epoch": 0.03619598128268649,
+      "grad_norm": 2.299867489692615,
+      "learning_rate": 1.9997988368510812e-05,
+      "loss": 1.2614,
+      "step": 263
+    },
+    {
+      "epoch": 0.03633360858794385,
+      "grad_norm": 1.8252336045175042,
+      "learning_rate": 1.9997897972435744e-05,
+      "loss": 1.3035,
+      "step": 264
+    },
+    {
+      "epoch": 0.03647123589320121,
+      "grad_norm": 2.48587944963577,
+      "learning_rate": 1.99978055899152e-05,
+      "loss": 1.4024,
+      "step": 265
+    },
+    {
+      "epoch": 0.036608863198458574,
+      "grad_norm": 2.2016066964884375,
+      "learning_rate": 1.999771122096754e-05,
+      "loss": 1.3368,
+      "step": 266
+    },
+    {
+      "epoch": 0.03674649050371594,
+      "grad_norm": 2.257269395715839,
+      "learning_rate": 1.9997614865611515e-05,
+      "loss": 1.3245,
+      "step": 267
+    },
+    {
+      "epoch": 0.0368841178089733,
+      "grad_norm": 1.9392772728395122,
+      "learning_rate": 1.9997516523866273e-05,
+      "loss": 1.3736,
+      "step": 268
+    },
+    {
+      "epoch": 0.03702174511423066,
+      "grad_norm": 1.800868859241569,
+      "learning_rate": 1.9997416195751343e-05,
+      "loss": 1.2669,
+      "step": 269
+    },
+    {
+      "epoch": 0.037159372419488024,
+      "grad_norm": 1.8835866861637278,
+      "learning_rate": 1.999731388128667e-05,
+      "loss": 1.3274,
+      "step": 270
+    },
+    {
+      "epoch": 0.037296999724745386,
+      "grad_norm": 1.86509997187314,
+      "learning_rate": 1.999720958049257e-05,
+      "loss": 1.2804,
+      "step": 271
+    },
+    {
+      "epoch": 0.037434627030002755,
+      "grad_norm": 1.9535121874766557,
+      "learning_rate": 1.999710329338978e-05,
+      "loss": 1.1958,
+      "step": 272
+    },
+    {
+      "epoch": 0.03757225433526012,
+      "grad_norm": 1.8226933365592513,
+      "learning_rate": 1.9996995019999408e-05,
+      "loss": 1.2574,
+      "step": 273
+    },
+    {
+      "epoch": 0.03770988164051748,
+      "grad_norm": 2.1225172773983725,
+      "learning_rate": 1.999688476034297e-05,
+      "loss": 1.2712,
+      "step": 274
+    },
+    {
+      "epoch": 0.03784750894577484,
+      "grad_norm": 1.8854002368178748,
+      "learning_rate": 1.9996772514442372e-05,
+      "loss": 1.3378,
+      "step": 275
+    },
+    {
+      "epoch": 0.037985136251032205,
+      "grad_norm": 1.9575746838816686,
+      "learning_rate": 1.999665828231992e-05,
+      "loss": 1.326,
+      "step": 276
+    },
+    {
+      "epoch": 0.03812276355628957,
+      "grad_norm": 1.992758874242281,
+      "learning_rate": 1.9996542063998302e-05,
+      "loss": 1.3026,
+      "step": 277
+    },
+    {
+      "epoch": 0.03826039086154693,
+      "grad_norm": 2.0898539554785738,
+      "learning_rate": 1.999642385950061e-05,
+      "loss": 1.243,
+      "step": 278
+    },
+    {
+      "epoch": 0.03839801816680429,
+      "grad_norm": 2.0479644406210724,
+      "learning_rate": 1.9996303668850343e-05,
+      "loss": 1.2691,
+      "step": 279
+    },
+    {
+      "epoch": 0.038535645472061654,
+      "grad_norm": 1.7496656826647126,
+      "learning_rate": 1.9996181492071365e-05,
+      "loss": 1.2616,
+      "step": 280
+    },
+    {
+      "epoch": 0.03867327277731902,
+      "grad_norm": 1.9329945252568672,
+      "learning_rate": 1.999605732918796e-05,
+      "loss": 1.2337,
+      "step": 281
+    },
+    {
+      "epoch": 0.038810900082576386,
+      "grad_norm": 1.897288648846803,
+      "learning_rate": 1.9995931180224794e-05,
+      "loss": 1.3319,
+      "step": 282
+    },
+    {
+      "epoch": 0.03894852738783375,
+      "grad_norm": 1.971862006752804,
+      "learning_rate": 1.9995803045206932e-05,
+      "loss": 1.2937,
+      "step": 283
+    },
+    {
+      "epoch": 0.03908615469309111,
+      "grad_norm": 1.97870082633005,
+      "learning_rate": 1.9995672924159833e-05,
+      "loss": 1.3127,
+      "step": 284
+    },
+    {
+      "epoch": 0.03922378199834847,
+      "grad_norm": 1.9065868730995048,
+      "learning_rate": 1.9995540817109348e-05,
+      "loss": 1.377,
+      "step": 285
+    },
+    {
+      "epoch": 0.039361409303605835,
+      "grad_norm": 1.9755311239559321,
+      "learning_rate": 1.999540672408173e-05,
+      "loss": 1.2757,
+      "step": 286
+    },
+    {
+      "epoch": 0.0394990366088632,
+      "grad_norm": 1.8140810813211568,
+      "learning_rate": 1.9995270645103612e-05,
+      "loss": 1.2681,
+      "step": 287
+    },
+    {
+      "epoch": 0.03963666391412056,
+      "grad_norm": 2.0146856861227542,
+      "learning_rate": 1.9995132580202043e-05,
+      "loss": 1.323,
+      "step": 288
+    },
+    {
+      "epoch": 0.03977429121937792,
+      "grad_norm": 2.01051832955701,
+      "learning_rate": 1.9994992529404448e-05,
+      "loss": 1.2374,
+      "step": 289
+    },
+    {
+      "epoch": 0.039911918524635284,
+      "grad_norm": 1.8492282150780337,
+      "learning_rate": 1.9994850492738653e-05,
+      "loss": 1.2018,
+      "step": 290
+    },
+    {
+      "epoch": 0.040049545829892653,
+      "grad_norm": 1.9389421531955668,
+      "learning_rate": 1.9994706470232878e-05,
+      "loss": 1.3533,
+      "step": 291
+    },
+    {
+      "epoch": 0.040187173135150016,
+      "grad_norm": 2.172373919946752,
+      "learning_rate": 1.999456046191574e-05,
+      "loss": 1.3005,
+      "step": 292
+    },
+    {
+      "epoch": 0.04032480044040738,
+      "grad_norm": 1.9029291943778814,
+      "learning_rate": 1.9994412467816252e-05,
+      "loss": 1.3316,
+      "step": 293
+    },
+    {
+      "epoch": 0.04046242774566474,
+      "grad_norm": 2.036563530414646,
+      "learning_rate": 1.9994262487963814e-05,
+      "loss": 1.3751,
+      "step": 294
+    },
+    {
+      "epoch": 0.0406000550509221,
+      "grad_norm": 2.0099481046515573,
+      "learning_rate": 1.9994110522388227e-05,
+      "loss": 1.3179,
+      "step": 295
+    },
+    {
+      "epoch": 0.040737682356179465,
+      "grad_norm": 1.9347221479381318,
+      "learning_rate": 1.999395657111968e-05,
+      "loss": 1.3262,
+      "step": 296
+    },
+    {
+      "epoch": 0.04087530966143683,
+      "grad_norm": 1.8746567728621637,
+      "learning_rate": 1.999380063418877e-05,
+      "loss": 1.2625,
+      "step": 297
+    },
+    {
+      "epoch": 0.04101293696669419,
+      "grad_norm": 2.000673507533951,
+      "learning_rate": 1.999364271162647e-05,
+      "loss": 1.2587,
+      "step": 298
+    },
+    {
+      "epoch": 0.04115056427195155,
+      "grad_norm": 2.1470349843525955,
+      "learning_rate": 1.9993482803464163e-05,
+      "loss": 1.3166,
+      "step": 299
+    },
+    {
+      "epoch": 0.04128819157720892,
+      "grad_norm": 2.082712481226252,
+      "learning_rate": 1.9993320909733618e-05,
+      "loss": 1.2266,
+      "step": 300
+    },
+    {
+      "epoch": 0.041425818882466284,
+      "grad_norm": 1.8518462224302648,
+      "learning_rate": 1.9993157030467002e-05,
+      "loss": 1.2869,
+      "step": 301
+    },
+    {
+      "epoch": 0.041563446187723646,
+      "grad_norm": 1.9955519906665373,
+      "learning_rate": 1.9992991165696875e-05,
+      "loss": 1.1822,
+      "step": 302
+    },
+    {
+      "epoch": 0.04170107349298101,
+      "grad_norm": 2.2311341822797583,
+      "learning_rate": 1.9992823315456197e-05,
+      "loss": 1.3345,
+      "step": 303
+    },
+    {
+      "epoch": 0.04183870079823837,
+      "grad_norm": 2.0750759989995444,
+      "learning_rate": 1.9992653479778307e-05,
+      "loss": 1.2559,
+      "step": 304
+    },
+    {
+      "epoch": 0.04197632810349573,
+      "grad_norm": 1.9139735838239553,
+      "learning_rate": 1.9992481658696957e-05,
+      "loss": 1.2901,
+      "step": 305
+    },
+    {
+      "epoch": 0.042113955408753095,
+      "grad_norm": 1.9785785414243133,
+      "learning_rate": 1.9992307852246283e-05,
+      "loss": 1.3835,
+      "step": 306
+    },
+    {
+      "epoch": 0.04225158271401046,
+      "grad_norm": 1.950811371080704,
+      "learning_rate": 1.9992132060460824e-05,
+      "loss": 1.2565,
+      "step": 307
+    },
+    {
+      "epoch": 0.04238921001926782,
+      "grad_norm": 1.8806893274339707,
+      "learning_rate": 1.9991954283375494e-05,
+      "loss": 1.2616,
+      "step": 308
+    },
+    {
+      "epoch": 0.04252683732452518,
+      "grad_norm": 1.9627813991714094,
+      "learning_rate": 1.9991774521025625e-05,
+      "loss": 1.285,
+      "step": 309
+    },
+    {
+      "epoch": 0.04266446462978255,
+      "grad_norm": 2.1000262070869478,
+      "learning_rate": 1.9991592773446934e-05,
+      "loss": 1.2997,
+      "step": 310
+    },
+    {
+      "epoch": 0.042802091935039914,
+      "grad_norm": 2.2443392965099385,
+      "learning_rate": 1.999140904067553e-05,
+      "loss": 1.3041,
+      "step": 311
+    },
+    {
+      "epoch": 0.042939719240297276,
+      "grad_norm": 1.904198006761273,
+      "learning_rate": 1.9991223322747914e-05,
+      "loss": 1.3205,
+      "step": 312
+    },
+    {
+      "epoch": 0.04307734654555464,
+      "grad_norm": 2.037733430757215,
+      "learning_rate": 1.9991035619700988e-05,
+      "loss": 1.3453,
+      "step": 313
+    },
+    {
+      "epoch": 0.043214973850812,
+      "grad_norm": 2.0271634868244717,
+      "learning_rate": 1.9990845931572048e-05,
+      "loss": 1.3254,
+      "step": 314
+    },
+    {
+      "epoch": 0.04335260115606936,
+      "grad_norm": 2.056518991454578,
+      "learning_rate": 1.999065425839878e-05,
+      "loss": 1.2907,
+      "step": 315
+    },
+    {
+      "epoch": 0.043490228461326726,
+      "grad_norm": 1.9450411910085335,
+      "learning_rate": 1.999046060021927e-05,
+      "loss": 1.246,
+      "step": 316
+    },
+    {
+      "epoch": 0.04362785576658409,
+      "grad_norm": 2.3013272256160477,
+      "learning_rate": 1.9990264957071996e-05,
+      "loss": 1.3439,
+      "step": 317
+    },
+    {
+      "epoch": 0.04376548307184145,
+      "grad_norm": 1.7469933983299086,
+      "learning_rate": 1.999006732899582e-05,
+      "loss": 1.4132,
+      "step": 318
+    },
+    {
+      "epoch": 0.04390311037709882,
+      "grad_norm": 1.9843759897211115,
+      "learning_rate": 1.998986771603002e-05,
+      "loss": 1.2616,
+      "step": 319
+    },
+    {
+      "epoch": 0.04404073768235618,
+      "grad_norm": 2.0912297698851816,
+      "learning_rate": 1.998966611821425e-05,
+      "loss": 1.2584,
+      "step": 320
+    },
+    {
+      "epoch": 0.044178364987613544,
+      "grad_norm": 1.9979224115725074,
+      "learning_rate": 1.9989462535588563e-05,
+      "loss": 1.2447,
+      "step": 321
+    },
+    {
+      "epoch": 0.04431599229287091,
+      "grad_norm": 1.8046442671301806,
+      "learning_rate": 1.9989256968193412e-05,
+      "loss": 1.305,
+      "step": 322
+    },
+    {
+      "epoch": 0.04445361959812827,
+      "grad_norm": 1.8304603195229936,
+      "learning_rate": 1.9989049416069643e-05,
+      "loss": 1.2999,
+      "step": 323
+    },
+    {
+      "epoch": 0.04459124690338563,
+      "grad_norm": 1.6554488060689538,
+      "learning_rate": 1.998883987925849e-05,
+      "loss": 1.1887,
+      "step": 324
+    },
+    {
+      "epoch": 0.044728874208642994,
+      "grad_norm": 2.2301382251330946,
+      "learning_rate": 1.998862835780158e-05,
+      "loss": 1.3089,
+      "step": 325
+    },
+    {
+      "epoch": 0.044866501513900356,
+      "grad_norm": 2.012467783875781,
+      "learning_rate": 1.9988414851740948e-05,
+      "loss": 1.336,
+      "step": 326
+    },
+    {
+      "epoch": 0.04500412881915772,
+      "grad_norm": 2.039803083423073,
+      "learning_rate": 1.9988199361119012e-05,
+      "loss": 1.2423,
+      "step": 327
+    },
+    {
+      "epoch": 0.04514175612441508,
+      "grad_norm": 1.7328938639393288,
+      "learning_rate": 1.9987981885978587e-05,
+      "loss": 1.1952,
+      "step": 328
+    },
+    {
+      "epoch": 0.04527938342967245,
+      "grad_norm": 1.8719596794351048,
+      "learning_rate": 1.998776242636288e-05,
+      "loss": 1.3441,
+      "step": 329
+    },
+    {
+      "epoch": 0.04541701073492981,
+      "grad_norm": 2.1170910662433444,
+      "learning_rate": 1.9987540982315497e-05,
+      "loss": 1.3259,
+      "step": 330
+    },
+    {
+      "epoch": 0.045554638040187174,
+      "grad_norm": 1.8939646339648109,
+      "learning_rate": 1.9987317553880435e-05,
+      "loss": 1.2743,
+      "step": 331
+    },
+    {
+      "epoch": 0.04569226534544454,
+      "grad_norm": 2.0732442964571827,
+      "learning_rate": 1.9987092141102088e-05,
+      "loss": 1.3028,
+      "step": 332
+    },
+    {
+      "epoch": 0.0458298926507019,
+      "grad_norm": 1.9756422826790319,
+      "learning_rate": 1.998686474402524e-05,
+      "loss": 1.2877,
+      "step": 333
+    },
+    {
+      "epoch": 0.04596751995595926,
+      "grad_norm": 2.081035273139465,
+      "learning_rate": 1.998663536269507e-05,
+      "loss": 1.3367,
+      "step": 334
+    },
+    {
+      "epoch": 0.046105147261216624,
+      "grad_norm": 1.9192226676341722,
+      "learning_rate": 1.998640399715716e-05,
+      "loss": 1.3373,
+      "step": 335
+    },
+    {
+      "epoch": 0.046242774566473986,
+      "grad_norm": 1.9271005799805245,
+      "learning_rate": 1.9986170647457473e-05,
+      "loss": 1.3318,
+      "step": 336
+    },
+    {
+      "epoch": 0.04638040187173135,
+      "grad_norm": 2.19044500789315,
+      "learning_rate": 1.998593531364238e-05,
+      "loss": 1.2437,
+      "step": 337
+    },
+    {
+      "epoch": 0.04651802917698872,
+      "grad_norm": 1.9494024594105408,
+      "learning_rate": 1.9985697995758623e-05,
+      "loss": 1.2268,
+      "step": 338
+    },
+    {
+      "epoch": 0.04665565648224608,
+      "grad_norm": 1.8829341282836296,
+      "learning_rate": 1.9985458693853367e-05,
+      "loss": 1.2664,
+      "step": 339
+    },
+    {
+      "epoch": 0.04679328378750344,
+      "grad_norm": 1.8870065990956553,
+      "learning_rate": 1.9985217407974157e-05,
+      "loss": 1.218,
+      "step": 340
+    },
+    {
+      "epoch": 0.046930911092760805,
+      "grad_norm": 2.060396980479203,
+      "learning_rate": 1.9984974138168927e-05,
+      "loss": 1.2898,
+      "step": 341
+    },
+    {
+      "epoch": 0.04706853839801817,
+      "grad_norm": 1.9419054971590057,
+      "learning_rate": 1.998472888448602e-05,
+      "loss": 1.3088,
+      "step": 342
+    },
+    {
+      "epoch": 0.04720616570327553,
+      "grad_norm": 1.6749626290520405,
+      "learning_rate": 1.9984481646974156e-05,
+      "loss": 1.1737,
+      "step": 343
+    },
+    {
+      "epoch": 0.04734379300853289,
+      "grad_norm": 2.07319264303262,
+      "learning_rate": 1.998423242568246e-05,
+      "loss": 1.3571,
+      "step": 344
+    },
+    {
+      "epoch": 0.047481420313790254,
+      "grad_norm": 2.049637865300209,
+      "learning_rate": 1.9983981220660453e-05,
+      "loss": 1.2898,
+      "step": 345
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 1.7279163765246852,
+      "learning_rate": 1.9983728031958045e-05,
+      "loss": 1.3704,
+      "step": 346
+    },
+    {
+      "epoch": 0.04775667492430498,
+      "grad_norm": 2.19329447436634,
+      "learning_rate": 1.9983472859625538e-05,
+      "loss": 1.2793,
+      "step": 347
+    },
+    {
+      "epoch": 0.04789430222956235,
+      "grad_norm": 2.0000656809954114,
+      "learning_rate": 1.998321570371363e-05,
+      "loss": 1.3677,
+      "step": 348
+    },
+    {
+      "epoch": 0.04803192953481971,
+      "grad_norm": 2.1140808385113994,
+      "learning_rate": 1.998295656427342e-05,
+      "loss": 1.2437,
+      "step": 349
+    },
+    {
+      "epoch": 0.04816955684007707,
+      "grad_norm": 2.2620320368775313,
+      "learning_rate": 1.998269544135639e-05,
+      "loss": 1.1969,
+      "step": 350
+    },
+    {
+      "epoch": 0.048307184145334435,
+      "grad_norm": 2.035491051074993,
+      "learning_rate": 1.9982432335014424e-05,
+      "loss": 1.2543,
+      "step": 351
+    },
+    {
+      "epoch": 0.0484448114505918,
+      "grad_norm": 1.9188178767874544,
+      "learning_rate": 1.99821672452998e-05,
+      "loss": 1.3715,
+      "step": 352
+    },
+    {
+      "epoch": 0.04858243875584916,
+      "grad_norm": 1.773103606908032,
+      "learning_rate": 1.9981900172265183e-05,
+      "loss": 1.2937,
+      "step": 353
+    },
+    {
+      "epoch": 0.04872006606110652,
+      "grad_norm": 1.8255602342492092,
+      "learning_rate": 1.998163111596364e-05,
+      "loss": 1.2485,
+      "step": 354
+    },
+    {
+      "epoch": 0.048857693366363884,
+      "grad_norm": 1.8177790018006474,
+      "learning_rate": 1.998136007644863e-05,
+      "loss": 1.2067,
+      "step": 355
+    },
+    {
+      "epoch": 0.04899532067162125,
+      "grad_norm": 1.8523525631451863,
+      "learning_rate": 1.9981087053774e-05,
+      "loss": 1.2702,
+      "step": 356
+    },
+    {
+      "epoch": 0.049132947976878616,
+      "grad_norm": 1.9180926138060057,
+      "learning_rate": 1.9980812047993998e-05,
+      "loss": 1.2702,
+      "step": 357
+    },
+    {
+      "epoch": 0.04927057528213598,
+      "grad_norm": 1.8398547657295286,
+      "learning_rate": 1.9980535059163268e-05,
+      "loss": 1.3007,
+      "step": 358
+    },
+    {
+      "epoch": 0.04940820258739334,
+      "grad_norm": 2.248073413784736,
+      "learning_rate": 1.9980256087336837e-05,
+      "loss": 1.374,
+      "step": 359
+    },
+    {
+      "epoch": 0.0495458298926507,
+      "grad_norm": 2.1565660311421055,
+      "learning_rate": 1.9979975132570136e-05,
+      "loss": 1.3018,
+      "step": 360
+    },
+    {
+      "epoch": 0.049683457197908065,
+      "grad_norm": 2.1246831299298248,
+      "learning_rate": 1.997969219491899e-05,
+      "loss": 1.264,
+      "step": 361
+    },
+    {
+      "epoch": 0.04982108450316543,
+      "grad_norm": 1.914329865097868,
+      "learning_rate": 1.997940727443961e-05,
+      "loss": 1.2822,
+      "step": 362
+    },
+    {
+      "epoch": 0.04995871180842279,
+      "grad_norm": 1.8992103495887487,
+      "learning_rate": 1.997912037118861e-05,
+      "loss": 1.3134,
+      "step": 363
+    },
+    {
+      "epoch": 0.05009633911368015,
+      "grad_norm": 1.9877935298192693,
+      "learning_rate": 1.997883148522299e-05,
+      "loss": 1.3024,
+      "step": 364
+    },
+    {
+      "epoch": 0.050233966418937515,
+      "grad_norm": 2.0183664107663164,
+      "learning_rate": 1.997854061660015e-05,
+      "loss": 1.3124,
+      "step": 365
+    },
+    {
+      "epoch": 0.05037159372419488,
+      "grad_norm": 2.100563135989665,
+      "learning_rate": 1.997824776537788e-05,
+      "loss": 1.3131,
+      "step": 366
+    },
+    {
+      "epoch": 0.050509221029452246,
+      "grad_norm": 1.9323042709870288,
+      "learning_rate": 1.9977952931614367e-05,
+      "loss": 1.3546,
+      "step": 367
+    },
+    {
+      "epoch": 0.05064684833470961,
+      "grad_norm": 2.0135447926682417,
+      "learning_rate": 1.9977656115368188e-05,
+      "loss": 1.2968,
+      "step": 368
+    },
+    {
+      "epoch": 0.05078447563996697,
+      "grad_norm": 2.0423576289706395,
+      "learning_rate": 1.997735731669832e-05,
+      "loss": 1.3582,
+      "step": 369
+    },
+    {
+      "epoch": 0.05092210294522433,
+      "grad_norm": 1.93104028666782,
+      "learning_rate": 1.9977056535664126e-05,
+      "loss": 1.2859,
+      "step": 370
+    },
+    {
+      "epoch": 0.051059730250481696,
+      "grad_norm": 2.0062015604170735,
+      "learning_rate": 1.997675377232537e-05,
+      "loss": 1.338,
+      "step": 371
+    },
+    {
+      "epoch": 0.05119735755573906,
+      "grad_norm": 1.7190018801986924,
+      "learning_rate": 1.9976449026742207e-05,
+      "loss": 1.1972,
+      "step": 372
+    },
+    {
+      "epoch": 0.05133498486099642,
+      "grad_norm": 2.0383342953000696,
+      "learning_rate": 1.9976142298975185e-05,
+      "loss": 1.3037,
+      "step": 373
+    },
+    {
+      "epoch": 0.05147261216625378,
+      "grad_norm": 1.9237636324072152,
+      "learning_rate": 1.9975833589085248e-05,
+      "loss": 1.2522,
+      "step": 374
+    },
+    {
+      "epoch": 0.051610239471511145,
+      "grad_norm": 2.019093419845557,
+      "learning_rate": 1.997552289713373e-05,
+      "loss": 1.3249,
+      "step": 375
+    },
+    {
+      "epoch": 0.051747866776768514,
+      "grad_norm": 2.0185035479057434,
+      "learning_rate": 1.9975210223182362e-05,
+      "loss": 1.274,
+      "step": 376
+    },
+    {
+      "epoch": 0.051885494082025876,
+      "grad_norm": 2.1183820265903948,
+      "learning_rate": 1.997489556729327e-05,
+      "loss": 1.2689,
+      "step": 377
+    },
+    {
+      "epoch": 0.05202312138728324,
+      "grad_norm": 1.9445286126627803,
+      "learning_rate": 1.997457892952897e-05,
+      "loss": 1.2651,
+      "step": 378
+    },
+    {
+      "epoch": 0.0521607486925406,
+      "grad_norm": 1.8633943800092068,
+      "learning_rate": 1.997426030995237e-05,
+      "loss": 1.1868,
+      "step": 379
+    },
+    {
+      "epoch": 0.05229837599779796,
+      "grad_norm": 1.8001205037987675,
+      "learning_rate": 1.9973939708626783e-05,
+      "loss": 1.2185,
+      "step": 380
+    },
+    {
+      "epoch": 0.052436003303055326,
+      "grad_norm": 1.691984854979059,
+      "learning_rate": 1.9973617125615905e-05,
+      "loss": 1.2676,
+      "step": 381
+    },
+    {
+      "epoch": 0.05257363060831269,
+      "grad_norm": 1.9956581709026606,
+      "learning_rate": 1.9973292560983824e-05,
+      "loss": 1.3274,
+      "step": 382
+    },
+    {
+      "epoch": 0.05271125791357005,
+      "grad_norm": 1.9127609189649284,
+      "learning_rate": 1.9972966014795037e-05,
+      "loss": 1.2833,
+      "step": 383
+    },
+    {
+      "epoch": 0.05284888521882741,
+      "grad_norm": 2.36956331830658,
+      "learning_rate": 1.997263748711441e-05,
+      "loss": 1.3796,
+      "step": 384
+    },
+    {
+      "epoch": 0.052986512524084775,
+      "grad_norm": 1.8591358110423712,
+      "learning_rate": 1.997230697800723e-05,
+      "loss": 1.206,
+      "step": 385
+    },
+    {
+      "epoch": 0.053124139829342144,
+      "grad_norm": 1.7971418097910214,
+      "learning_rate": 1.9971974487539157e-05,
+      "loss": 1.2734,
+      "step": 386
+    },
+    {
+      "epoch": 0.05326176713459951,
+      "grad_norm": 2.162818982720296,
+      "learning_rate": 1.997164001577626e-05,
+      "loss": 1.3393,
+      "step": 387
+    },
+    {
+      "epoch": 0.05339939443985687,
+      "grad_norm": 1.9010844236248636,
+      "learning_rate": 1.997130356278498e-05,
+      "loss": 1.2105,
+      "step": 388
+    },
+    {
+      "epoch": 0.05353702174511423,
+      "grad_norm": 1.8862243986294585,
+      "learning_rate": 1.9970965128632182e-05,
+      "loss": 1.299,
+      "step": 389
+    },
+    {
+      "epoch": 0.053674649050371594,
+      "grad_norm": 1.8811403580279635,
+      "learning_rate": 1.9970624713385098e-05,
+      "loss": 1.1756,
+      "step": 390
+    },
+    {
+      "epoch": 0.053812276355628956,
+      "grad_norm": 1.934476324240299,
+      "learning_rate": 1.9970282317111366e-05,
+      "loss": 1.3337,
+      "step": 391
+    },
+    {
+      "epoch": 0.05394990366088632,
+      "grad_norm": 1.9113824359188867,
+      "learning_rate": 1.9969937939879015e-05,
+      "loss": 1.2909,
+      "step": 392
+    },
+    {
+      "epoch": 0.05408753096614368,
+      "grad_norm": 2.17083176906069,
+      "learning_rate": 1.996959158175647e-05,
+      "loss": 1.2969,
+      "step": 393
+    },
+    {
+      "epoch": 0.05422515827140104,
+      "grad_norm": 1.863157109046354,
+      "learning_rate": 1.9969243242812548e-05,
+      "loss": 1.3732,
+      "step": 394
+    },
+    {
+      "epoch": 0.05436278557665841,
+      "grad_norm": 1.918559551273086,
+      "learning_rate": 1.9968892923116458e-05,
+      "loss": 1.2147,
+      "step": 395
+    },
+    {
+      "epoch": 0.054500412881915775,
+      "grad_norm": 1.8970874447218284,
+      "learning_rate": 1.99685406227378e-05,
+      "loss": 1.2886,
+      "step": 396
+    },
+    {
+      "epoch": 0.05463804018717314,
+      "grad_norm": 1.7106394486847707,
+      "learning_rate": 1.9968186341746577e-05,
+      "loss": 1.2718,
+      "step": 397
+    },
+    {
+      "epoch": 0.0547756674924305,
+      "grad_norm": 2.194502873483654,
+      "learning_rate": 1.9967830080213177e-05,
+      "loss": 1.3638,
+      "step": 398
+    },
+    {
+      "epoch": 0.05491329479768786,
+      "grad_norm": 1.8064799309541488,
+      "learning_rate": 1.9967471838208384e-05,
+      "loss": 1.2916,
+      "step": 399
+    },
+    {
+      "epoch": 0.055050922102945224,
+      "grad_norm": 1.8271244373348412,
+      "learning_rate": 1.9967111615803375e-05,
+      "loss": 1.1705,
+      "step": 400
+    },
+    {
+      "epoch": 0.055188549408202586,
+      "grad_norm": 1.6791195988031211,
+      "learning_rate": 1.9966749413069725e-05,
+      "loss": 1.2967,
+      "step": 401
+    },
+    {
+      "epoch": 0.05532617671345995,
+      "grad_norm": 2.2396720681627222,
+      "learning_rate": 1.9966385230079398e-05,
+      "loss": 1.2927,
+      "step": 402
+    },
+    {
+      "epoch": 0.05546380401871731,
+      "grad_norm": 2.0996442757188944,
+      "learning_rate": 1.996601906690475e-05,
+      "loss": 1.4073,
+      "step": 403
+    },
+    {
+      "epoch": 0.05560143132397467,
+      "grad_norm": 1.8849503810783812,
+      "learning_rate": 1.9965650923618533e-05,
+      "loss": 1.2971,
+      "step": 404
+    },
+    {
+      "epoch": 0.05573905862923204,
+      "grad_norm": 1.7862736226937408,
+      "learning_rate": 1.996528080029389e-05,
+      "loss": 1.1044,
+      "step": 405
+    },
+    {
+      "epoch": 0.055876685934489405,
+      "grad_norm": 1.7587660367767275,
+      "learning_rate": 1.9964908697004365e-05,
+      "loss": 1.2746,
+      "step": 406
+    },
+    {
+      "epoch": 0.05601431323974677,
+      "grad_norm": 1.9919955565245953,
+      "learning_rate": 1.9964534613823887e-05,
+      "loss": 1.2618,
+      "step": 407
+    },
+    {
+      "epoch": 0.05615194054500413,
+      "grad_norm": 2.2063945668263165,
+      "learning_rate": 1.996415855082678e-05,
+      "loss": 1.2899,
+      "step": 408
+    },
+    {
+      "epoch": 0.05628956785026149,
+      "grad_norm": 1.8569914802032457,
+      "learning_rate": 1.9963780508087762e-05,
+      "loss": 1.2451,
+      "step": 409
+    },
+    {
+      "epoch": 0.056427195155518854,
+      "grad_norm": 1.9653190955473436,
+      "learning_rate": 1.996340048568195e-05,
+      "loss": 1.2956,
+      "step": 410
+    },
+    {
+      "epoch": 0.056564822460776217,
+      "grad_norm": 1.8139149050748382,
+      "learning_rate": 1.996301848368484e-05,
+      "loss": 1.2045,
+      "step": 411
+    },
+    {
+      "epoch": 0.05670244976603358,
+      "grad_norm": 1.564938337794532,
+      "learning_rate": 1.9962634502172337e-05,
+      "loss": 1.237,
+      "step": 412
+    },
+    {
+      "epoch": 0.05684007707129094,
+      "grad_norm": 1.934555588251456,
+      "learning_rate": 1.9962248541220736e-05,
+      "loss": 1.2473,
+      "step": 413
+    },
+    {
+      "epoch": 0.05697770437654831,
+      "grad_norm": 1.6613692931490067,
+      "learning_rate": 1.9961860600906715e-05,
+      "loss": 1.209,
+      "step": 414
+    },
+    {
+      "epoch": 0.05711533168180567,
+      "grad_norm": 1.7323167090809082,
+      "learning_rate": 1.9961470681307356e-05,
+      "loss": 1.1927,
+      "step": 415
+    },
+    {
+      "epoch": 0.057252958987063035,
+      "grad_norm": 1.9121359938175644,
+      "learning_rate": 1.996107878250013e-05,
+      "loss": 1.2966,
+      "step": 416
+    },
+    {
+      "epoch": 0.0573905862923204,
+      "grad_norm": 2.276850724118106,
+      "learning_rate": 1.9960684904562906e-05,
+      "loss": 1.3582,
+      "step": 417
+    },
+    {
+      "epoch": 0.05752821359757776,
+      "grad_norm": 1.8421745621957972,
+      "learning_rate": 1.9960289047573937e-05,
+      "loss": 1.3685,
+      "step": 418
+    },
+    {
+      "epoch": 0.05766584090283512,
+      "grad_norm": 1.9317876497639175,
+      "learning_rate": 1.9959891211611873e-05,
+      "loss": 1.2558,
+      "step": 419
+    },
+    {
+      "epoch": 0.057803468208092484,
+      "grad_norm": 1.7312446850719045,
+      "learning_rate": 1.995949139675576e-05,
+      "loss": 1.2059,
+      "step": 420
+    },
+    {
+      "epoch": 0.05794109551334985,
+      "grad_norm": 1.876212419182311,
+      "learning_rate": 1.9959089603085043e-05,
+      "loss": 1.2748,
+      "step": 421
+    },
+    {
+      "epoch": 0.05807872281860721,
+      "grad_norm": 1.9622129284643401,
+      "learning_rate": 1.995868583067954e-05,
+      "loss": 1.2435,
+      "step": 422
+    },
+    {
+      "epoch": 0.05821635012386457,
+      "grad_norm": 1.9174901875998451,
+      "learning_rate": 1.995828007961949e-05,
+      "loss": 1.3094,
+      "step": 423
+    },
+    {
+      "epoch": 0.05835397742912194,
+      "grad_norm": 2.110436852907964,
+      "learning_rate": 1.9957872349985497e-05,
+      "loss": 1.1854,
+      "step": 424
+    },
+    {
+      "epoch": 0.0584916047343793,
+      "grad_norm": 1.8515117355286672,
+      "learning_rate": 1.9957462641858578e-05,
+      "loss": 1.3051,
+      "step": 425
+    },
+    {
+      "epoch": 0.058629232039636665,
+      "grad_norm": 2.0012891728170987,
+      "learning_rate": 1.9957050955320132e-05,
+      "loss": 1.3385,
+      "step": 426
+    },
+    {
+      "epoch": 0.05876685934489403,
+      "grad_norm": 1.84818568117921,
+      "learning_rate": 1.995663729045196e-05,
+      "loss": 1.2645,
+      "step": 427
+    },
+    {
+      "epoch": 0.05890448665015139,
+      "grad_norm": 1.6990181462479568,
+      "learning_rate": 1.995622164733625e-05,
+      "loss": 1.302,
+      "step": 428
+    },
+    {
+      "epoch": 0.05904211395540875,
+      "grad_norm": 2.0195448260835747,
+      "learning_rate": 1.9955804026055588e-05,
+      "loss": 1.357,
+      "step": 429
+    },
+    {
+      "epoch": 0.059179741260666115,
+      "grad_norm": 2.2826671523874134,
+      "learning_rate": 1.995538442669294e-05,
+      "loss": 1.2609,
+      "step": 430
+    },
+    {
+      "epoch": 0.05931736856592348,
+      "grad_norm": 1.855375433612593,
+      "learning_rate": 1.9954962849331683e-05,
+      "loss": 1.3405,
+      "step": 431
+    },
+    {
+      "epoch": 0.05945499587118084,
+      "grad_norm": 1.8944896973759686,
+      "learning_rate": 1.995453929405558e-05,
+      "loss": 1.2659,
+      "step": 432
+    },
+    {
+      "epoch": 0.05959262317643821,
+      "grad_norm": 2.0665824320121193,
+      "learning_rate": 1.995411376094878e-05,
+      "loss": 1.3329,
+      "step": 433
+    },
+    {
+      "epoch": 0.05973025048169557,
+      "grad_norm": 1.890664622189023,
+      "learning_rate": 1.9953686250095832e-05,
+      "loss": 1.2831,
+      "step": 434
+    },
+    {
+      "epoch": 0.05986787778695293,
+      "grad_norm": 1.9210316852899385,
+      "learning_rate": 1.9953256761581676e-05,
+      "loss": 1.2864,
+      "step": 435
+    },
+    {
+      "epoch": 0.060005505092210296,
+      "grad_norm": 2.0807789068763256,
+      "learning_rate": 1.9952825295491646e-05,
+      "loss": 1.2978,
+      "step": 436
+    },
+    {
+      "epoch": 0.06014313239746766,
+      "grad_norm": 1.9289002736800502,
+      "learning_rate": 1.9952391851911472e-05,
+      "loss": 1.2796,
+      "step": 437
+    },
+    {
+      "epoch": 0.06028075970272502,
+      "grad_norm": 1.7939765493175006,
+      "learning_rate": 1.995195643092727e-05,
+      "loss": 1.1806,
+      "step": 438
+    },
+    {
+      "epoch": 0.06041838700798238,
+      "grad_norm": 1.929164866322231,
+      "learning_rate": 1.9951519032625553e-05,
+      "loss": 1.233,
+      "step": 439
+    },
+    {
+      "epoch": 0.060556014313239745,
+      "grad_norm": 1.743479810393273,
+      "learning_rate": 1.9951079657093224e-05,
+      "loss": 1.2521,
+      "step": 440
+    },
+    {
+      "epoch": 0.06069364161849711,
+      "grad_norm": 2.0613972824025306,
+      "learning_rate": 1.9950638304417586e-05,
+      "loss": 1.2734,
+      "step": 441
+    },
+    {
+      "epoch": 0.06083126892375447,
+      "grad_norm": 1.8358539949381785,
+      "learning_rate": 1.9950194974686323e-05,
+      "loss": 1.2735,
+      "step": 442
+    },
+    {
+      "epoch": 0.06096889622901184,
+      "grad_norm": 1.8588496519748607,
+      "learning_rate": 1.994974966798752e-05,
+      "loss": 1.2849,
+      "step": 443
+    },
+    {
+      "epoch": 0.0611065235342692,
+      "grad_norm": 1.9323398411409636,
+      "learning_rate": 1.9949302384409658e-05,
+      "loss": 1.207,
+      "step": 444
+    },
+    {
+      "epoch": 0.061244150839526564,
+      "grad_norm": 1.8837049212364023,
+      "learning_rate": 1.9948853124041605e-05,
+      "loss": 1.2625,
+      "step": 445
+    },
+    {
+      "epoch": 0.061381778144783926,
+      "grad_norm": 2.083906517694679,
+      "learning_rate": 1.994840188697262e-05,
+      "loss": 1.3255,
+      "step": 446
+    },
+    {
+      "epoch": 0.06151940545004129,
+      "grad_norm": 2.118115994699614,
+      "learning_rate": 1.9947948673292355e-05,
+      "loss": 1.226,
+      "step": 447
+    },
+    {
+      "epoch": 0.06165703275529865,
+      "grad_norm": 1.9495937007307207,
+      "learning_rate": 1.9947493483090868e-05,
+      "loss": 1.275,
+      "step": 448
+    },
+    {
+      "epoch": 0.06179466006055601,
+      "grad_norm": 1.8613733424514662,
+      "learning_rate": 1.9947036316458586e-05,
+      "loss": 1.2404,
+      "step": 449
+    },
+    {
+      "epoch": 0.061932287365813375,
+      "grad_norm": 2.00336335993132,
+      "learning_rate": 1.994657717348635e-05,
+      "loss": 1.2119,
+      "step": 450
+    },
+    {
+      "epoch": 0.06206991467107074,
+      "grad_norm": 1.7603164089522412,
+      "learning_rate": 1.994611605426538e-05,
+      "loss": 1.2477,
+      "step": 451
+    },
+    {
+      "epoch": 0.06220754197632811,
+      "grad_norm": 1.9460811460971439,
+      "learning_rate": 1.9945652958887307e-05,
+      "loss": 1.2729,
+      "step": 452
+    },
+    {
+      "epoch": 0.06234516928158547,
+      "grad_norm": 2.158714337051103,
+      "learning_rate": 1.9945187887444125e-05,
+      "loss": 1.2744,
+      "step": 453
+    },
+    {
+      "epoch": 0.06248279658684283,
+      "grad_norm": 2.0849629537107606,
+      "learning_rate": 1.994472084002825e-05,
+      "loss": 1.2764,
+      "step": 454
+    },
+    {
+      "epoch": 0.0626204238921002,
+      "grad_norm": 1.8921686259368096,
+      "learning_rate": 1.9944251816732465e-05,
+      "loss": 1.2631,
+      "step": 455
+    },
+    {
+      "epoch": 0.06275805119735756,
+      "grad_norm": 1.8139476210515333,
+      "learning_rate": 1.994378081764997e-05,
+      "loss": 1.2492,
+      "step": 456
+    },
+    {
+      "epoch": 0.06289567850261492,
+      "grad_norm": 1.658948647418169,
+      "learning_rate": 1.9943307842874343e-05,
+      "loss": 1.265,
+      "step": 457
+    },
+    {
+      "epoch": 0.06303330580787228,
+      "grad_norm": 1.957374565040711,
+      "learning_rate": 1.9942832892499554e-05,
+      "loss": 1.3416,
+      "step": 458
+    },
+    {
+      "epoch": 0.06317093311312964,
+      "grad_norm": 1.784477911383927,
+      "learning_rate": 1.9942355966619973e-05,
+      "loss": 1.1849,
+      "step": 459
+    },
+    {
+      "epoch": 0.063308560418387,
+      "grad_norm": 1.7841123823226424,
+      "learning_rate": 1.994187706533036e-05,
+      "loss": 1.2017,
+      "step": 460
+    },
+    {
+      "epoch": 0.06344618772364437,
+      "grad_norm": 2.023149833569769,
+      "learning_rate": 1.994139618872586e-05,
+      "loss": 1.2725,
+      "step": 461
+    },
+    {
+      "epoch": 0.06358381502890173,
+      "grad_norm": 1.6274094999384983,
+      "learning_rate": 1.994091333690202e-05,
+      "loss": 1.2145,
+      "step": 462
+    },
+    {
+      "epoch": 0.06372144233415909,
+      "grad_norm": 1.9408764622805104,
+      "learning_rate": 1.994042850995478e-05,
+      "loss": 1.2045,
+      "step": 463
+    },
+    {
+      "epoch": 0.06385906963941645,
+      "grad_norm": 1.8871658537140907,
+      "learning_rate": 1.9939941707980462e-05,
+      "loss": 1.2864,
+      "step": 464
+    },
+    {
+      "epoch": 0.06399669694467382,
+      "grad_norm": 2.1552974071740807,
+      "learning_rate": 1.993945293107579e-05,
+      "loss": 1.3313,
+      "step": 465
+    },
+    {
+      "epoch": 0.06413432424993118,
+      "grad_norm": 1.8076542039299266,
+      "learning_rate": 1.9938962179337877e-05,
+      "loss": 1.2683,
+      "step": 466
+    },
+    {
+      "epoch": 0.06427195155518856,
+      "grad_norm": 2.0413606219325544,
+      "learning_rate": 1.9938469452864228e-05,
+      "loss": 1.1935,
+      "step": 467
+    },
+    {
+      "epoch": 0.06440957886044592,
+      "grad_norm": 1.8034038379822015,
+      "learning_rate": 1.9937974751752742e-05,
+      "loss": 1.2596,
+      "step": 468
+    },
+    {
+      "epoch": 0.06454720616570328,
+      "grad_norm": 2.192878316769907,
+      "learning_rate": 1.993747807610171e-05,
+      "loss": 1.2511,
+      "step": 469
+    },
+    {
+      "epoch": 0.06468483347096064,
+      "grad_norm": 1.9280308180215493,
+      "learning_rate": 1.993697942600981e-05,
+      "loss": 1.2295,
+      "step": 470
+    },
+    {
+      "epoch": 0.064822460776218,
+      "grad_norm": 2.0030590524892764,
+      "learning_rate": 1.9936478801576126e-05,
+      "loss": 1.2074,
+      "step": 471
+    },
+    {
+      "epoch": 0.06496008808147537,
+      "grad_norm": 1.9156370956774154,
+      "learning_rate": 1.9935976202900113e-05,
+      "loss": 1.3269,
+      "step": 472
+    },
+    {
+      "epoch": 0.06509771538673273,
+      "grad_norm": 1.8945829849145701,
+      "learning_rate": 1.9935471630081643e-05,
+      "loss": 1.1089,
+      "step": 473
+    },
+    {
+      "epoch": 0.06523534269199009,
+      "grad_norm": 2.0557742392810447,
+      "learning_rate": 1.9934965083220962e-05,
+      "loss": 1.2919,
+      "step": 474
+    },
+    {
+      "epoch": 0.06537296999724745,
+      "grad_norm": 1.895026334766847,
+      "learning_rate": 1.993445656241871e-05,
+      "loss": 1.2658,
+      "step": 475
+    },
+    {
+      "epoch": 0.06551059730250482,
+      "grad_norm": 2.085726631951332,
+      "learning_rate": 1.9933946067775926e-05,
+      "loss": 1.3042,
+      "step": 476
+    },
+    {
+      "epoch": 0.06564822460776218,
+      "grad_norm": 2.1898164350676605,
+      "learning_rate": 1.9933433599394044e-05,
+      "loss": 1.2352,
+      "step": 477
+    },
+    {
+      "epoch": 0.06578585191301954,
+      "grad_norm": 1.7912662915920923,
+      "learning_rate": 1.9932919157374875e-05,
+      "loss": 1.2668,
+      "step": 478
+    },
+    {
+      "epoch": 0.0659234792182769,
+      "grad_norm": 1.751634242571438,
+      "learning_rate": 1.993240274182064e-05,
+      "loss": 1.2506,
+      "step": 479
+    },
+    {
+      "epoch": 0.06606110652353427,
+      "grad_norm": 1.7499311679189105,
+      "learning_rate": 1.993188435283394e-05,
+      "loss": 1.2484,
+      "step": 480
+    },
+    {
+      "epoch": 0.06619873382879163,
+      "grad_norm": 1.8954821340679753,
+      "learning_rate": 1.993136399051777e-05,
+      "loss": 1.3322,
+      "step": 481
+    },
+    {
+      "epoch": 0.06633636113404899,
+      "grad_norm": 1.99824946164007,
+      "learning_rate": 1.993084165497552e-05,
+      "loss": 1.2665,
+      "step": 482
+    },
+    {
+      "epoch": 0.06647398843930635,
+      "grad_norm": 1.9107533598514703,
+      "learning_rate": 1.9930317346310973e-05,
+      "loss": 1.2524,
+      "step": 483
+    },
+    {
+      "epoch": 0.06661161574456372,
+      "grad_norm": 1.8697004034047635,
+      "learning_rate": 1.9929791064628298e-05,
+      "loss": 1.2589,
+      "step": 484
+    },
+    {
+      "epoch": 0.06674924304982108,
+      "grad_norm": 1.8787892591244844,
+      "learning_rate": 1.9929262810032066e-05,
+      "loss": 1.1634,
+      "step": 485
+    },
+    {
+      "epoch": 0.06688687035507845,
+      "grad_norm": 1.7249769684886362,
+      "learning_rate": 1.9928732582627227e-05,
+      "loss": 1.2129,
+      "step": 486
+    },
+    {
+      "epoch": 0.06702449766033582,
+      "grad_norm": 1.9434460654529309,
+      "learning_rate": 1.9928200382519137e-05,
+      "loss": 1.3121,
+      "step": 487
+    },
+    {
+      "epoch": 0.06716212496559318,
+      "grad_norm": 1.8087670749345492,
+      "learning_rate": 1.9927666209813533e-05,
+      "loss": 1.2442,
+      "step": 488
+    },
+    {
+      "epoch": 0.06729975227085054,
+      "grad_norm": 1.9667890540167725,
+      "learning_rate": 1.9927130064616545e-05,
+      "loss": 1.2028,
+      "step": 489
+    },
+    {
+      "epoch": 0.0674373795761079,
+      "grad_norm": 1.9770404104985562,
+      "learning_rate": 1.9926591947034703e-05,
+      "loss": 1.296,
+      "step": 490
+    },
+    {
+      "epoch": 0.06757500688136527,
+      "grad_norm": 1.9943108838405652,
+      "learning_rate": 1.9926051857174918e-05,
+      "loss": 1.1825,
+      "step": 491
+    },
+    {
+      "epoch": 0.06771263418662263,
+      "grad_norm": 1.8889648957124738,
+      "learning_rate": 1.9925509795144507e-05,
+      "loss": 1.2721,
+      "step": 492
+    },
+    {
+      "epoch": 0.06785026149187999,
+      "grad_norm": 1.7832558671605818,
+      "learning_rate": 1.9924965761051162e-05,
+      "loss": 1.3489,
+      "step": 493
+    },
+    {
+      "epoch": 0.06798788879713735,
+      "grad_norm": 1.9964227718191707,
+      "learning_rate": 1.992441975500298e-05,
+      "loss": 1.1739,
+      "step": 494
+    },
+    {
+      "epoch": 0.06812551610239471,
+      "grad_norm": 1.7785734577726646,
+      "learning_rate": 1.9923871777108444e-05,
+      "loss": 1.3104,
+      "step": 495
+    },
+    {
+      "epoch": 0.06826314340765208,
+      "grad_norm": 1.854080157068782,
+      "learning_rate": 1.9923321827476424e-05,
+      "loss": 1.2536,
+      "step": 496
+    },
+    {
+      "epoch": 0.06840077071290944,
+      "grad_norm": 1.8296990592019273,
+      "learning_rate": 1.9922769906216197e-05,
+      "loss": 1.2576,
+      "step": 497
+    },
+    {
+      "epoch": 0.0685383980181668,
+      "grad_norm": 1.9168431893427422,
+      "learning_rate": 1.9922216013437416e-05,
+      "loss": 1.2094,
+      "step": 498
+    },
+    {
+      "epoch": 0.06867602532342416,
+      "grad_norm": 1.884493633168247,
+      "learning_rate": 1.9921660149250136e-05,
+      "loss": 1.2731,
+      "step": 499
+    },
+    {
+      "epoch": 0.06881365262868153,
+      "grad_norm": 1.7962972392112313,
+      "learning_rate": 1.9921102313764792e-05,
+      "loss": 1.2245,
+      "step": 500
+    },
+    {
+      "epoch": 0.06895127993393889,
+      "grad_norm": 1.913754930363172,
+      "learning_rate": 1.9920542507092224e-05,
+      "loss": 1.1765,
+      "step": 501
+    },
+    {
+      "epoch": 0.06908890723919625,
+      "grad_norm": 2.058309558958913,
+      "learning_rate": 1.991998072934366e-05,
+      "loss": 1.3113,
+      "step": 502
+    },
+    {
+      "epoch": 0.06922653454445361,
+      "grad_norm": 1.8918350737789915,
+      "learning_rate": 1.9919416980630716e-05,
+      "loss": 1.2009,
+      "step": 503
+    },
+    {
+      "epoch": 0.06936416184971098,
+      "grad_norm": 1.6294278475860247,
+      "learning_rate": 1.9918851261065396e-05,
+      "loss": 1.1632,
+      "step": 504
+    },
+    {
+      "epoch": 0.06950178915496835,
+      "grad_norm": 2.3358958648407957,
+      "learning_rate": 1.9918283570760106e-05,
+      "loss": 1.2233,
+      "step": 505
+    },
+    {
+      "epoch": 0.06963941646022571,
+      "grad_norm": 1.9964956885839813,
+      "learning_rate": 1.991771390982764e-05,
+      "loss": 1.1868,
+      "step": 506
+    },
+    {
+      "epoch": 0.06977704376548308,
+      "grad_norm": 2.1455556527906774,
+      "learning_rate": 1.9917142278381174e-05,
+      "loss": 1.3121,
+      "step": 507
+    },
+    {
+      "epoch": 0.06991467107074044,
+      "grad_norm": 1.8536691004070176,
+      "learning_rate": 1.9916568676534293e-05,
+      "loss": 1.3062,
+      "step": 508
+    },
+    {
+      "epoch": 0.0700522983759978,
+      "grad_norm": 1.8421357833064747,
+      "learning_rate": 1.9915993104400956e-05,
+      "loss": 1.264,
+      "step": 509
+    },
+    {
+      "epoch": 0.07018992568125516,
+      "grad_norm": 1.704082348208926,
+      "learning_rate": 1.9915415562095526e-05,
+      "loss": 1.0597,
+      "step": 510
+    },
+    {
+      "epoch": 0.07032755298651253,
+      "grad_norm": 1.7081094380669708,
+      "learning_rate": 1.991483604973275e-05,
+      "loss": 1.2118,
+      "step": 511
+    },
+    {
+      "epoch": 0.07046518029176989,
+      "grad_norm": 2.137867699063183,
+      "learning_rate": 1.991425456742777e-05,
+      "loss": 1.2689,
+      "step": 512
+    },
+    {
+      "epoch": 0.07060280759702725,
+      "grad_norm": 2.1243982146673672,
+      "learning_rate": 1.991367111529612e-05,
+      "loss": 1.2886,
+      "step": 513
+    },
+    {
+      "epoch": 0.07074043490228461,
+      "grad_norm": 1.8434831560047957,
+      "learning_rate": 1.991308569345372e-05,
+      "loss": 1.1534,
+      "step": 514
+    },
+    {
+      "epoch": 0.07087806220754198,
+      "grad_norm": 2.0962906673915844,
+      "learning_rate": 1.991249830201689e-05,
+      "loss": 1.2563,
+      "step": 515
+    },
+    {
+      "epoch": 0.07101568951279934,
+      "grad_norm": 1.7898084007142196,
+      "learning_rate": 1.9911908941102336e-05,
+      "loss": 1.2567,
+      "step": 516
+    },
+    {
+      "epoch": 0.0711533168180567,
+      "grad_norm": 1.8977713607965296,
+      "learning_rate": 1.991131761082715e-05,
+      "loss": 1.1845,
+      "step": 517
+    },
+    {
+      "epoch": 0.07129094412331406,
+      "grad_norm": 2.016993814550598,
+      "learning_rate": 1.991072431130883e-05,
+      "loss": 1.2505,
+      "step": 518
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 1.9529632818952092,
+      "learning_rate": 1.9910129042665253e-05,
+      "loss": 1.2945,
+      "step": 519
+    },
+    {
+      "epoch": 0.07156619873382879,
+      "grad_norm": 2.0760690048294643,
+      "learning_rate": 1.9909531805014686e-05,
+      "loss": 1.2622,
+      "step": 520
+    },
+    {
+      "epoch": 0.07170382603908615,
+      "grad_norm": 2.091813347100997,
+      "learning_rate": 1.99089325984758e-05,
+      "loss": 1.3105,
+      "step": 521
+    },
+    {
+      "epoch": 0.07184145334434351,
+      "grad_norm": 1.9985380056672712,
+      "learning_rate": 1.9908331423167644e-05,
+      "loss": 1.343,
+      "step": 522
+    },
+    {
+      "epoch": 0.07197908064960087,
+      "grad_norm": 1.9039271946062237,
+      "learning_rate": 1.9907728279209665e-05,
+      "loss": 1.2964,
+      "step": 523
+    },
+    {
+      "epoch": 0.07211670795485825,
+      "grad_norm": 1.925355312411513,
+      "learning_rate": 1.99071231667217e-05,
+      "loss": 1.1624,
+      "step": 524
+    },
+    {
+      "epoch": 0.07225433526011561,
+      "grad_norm": 1.9208508557748165,
+      "learning_rate": 1.9906516085823974e-05,
+      "loss": 1.18,
+      "step": 525
+    },
+    {
+      "epoch": 0.07239196256537297,
+      "grad_norm": 1.8401921667012449,
+      "learning_rate": 1.9905907036637103e-05,
+      "loss": 1.2058,
+      "step": 526
+    },
+    {
+      "epoch": 0.07252958987063034,
+      "grad_norm": 1.8185225338885393,
+      "learning_rate": 1.9905296019282105e-05,
+      "loss": 1.2032,
+      "step": 527
+    },
+    {
+      "epoch": 0.0726672171758877,
+      "grad_norm": 2.0907999865265614,
+      "learning_rate": 1.990468303388038e-05,
+      "loss": 1.2156,
+      "step": 528
+    },
+    {
+      "epoch": 0.07280484448114506,
+      "grad_norm": 1.9473419406852845,
+      "learning_rate": 1.990406808055371e-05,
+      "loss": 1.2697,
+      "step": 529
+    },
+    {
+      "epoch": 0.07294247178640242,
+      "grad_norm": 1.8795191621729275,
+      "learning_rate": 1.990345115942429e-05,
+      "loss": 1.2713,
+      "step": 530
+    },
+    {
+      "epoch": 0.07308009909165979,
+      "grad_norm": 1.7089408163592847,
+      "learning_rate": 1.9902832270614683e-05,
+      "loss": 1.2537,
+      "step": 531
+    },
+    {
+      "epoch": 0.07321772639691715,
+      "grad_norm": 1.938789752281339,
+      "learning_rate": 1.9902211414247864e-05,
+      "loss": 1.273,
+      "step": 532
+    },
+    {
+      "epoch": 0.07335535370217451,
+      "grad_norm": 1.9324931804949115,
+      "learning_rate": 1.990158859044718e-05,
+      "loss": 1.3781,
+      "step": 533
+    },
+    {
+      "epoch": 0.07349298100743187,
+      "grad_norm": 1.864103670510535,
+      "learning_rate": 1.990096379933638e-05,
+      "loss": 1.2843,
+      "step": 534
+    },
+    {
+      "epoch": 0.07363060831268924,
+      "grad_norm": 1.8999361286453935,
+      "learning_rate": 1.9900337041039605e-05,
+      "loss": 1.2507,
+      "step": 535
+    },
+    {
+      "epoch": 0.0737682356179466,
+      "grad_norm": 1.6643939919544746,
+      "learning_rate": 1.989970831568138e-05,
+      "loss": 1.237,
+      "step": 536
+    },
+    {
+      "epoch": 0.07390586292320396,
+      "grad_norm": 1.6263359925684833,
+      "learning_rate": 1.9899077623386628e-05,
+      "loss": 1.1924,
+      "step": 537
+    },
+    {
+      "epoch": 0.07404349022846132,
+      "grad_norm": 1.7678926810869806,
+      "learning_rate": 1.989844496428065e-05,
+      "loss": 1.245,
+      "step": 538
+    },
+    {
+      "epoch": 0.07418111753371869,
+      "grad_norm": 1.854960029020722,
+      "learning_rate": 1.9897810338489157e-05,
+      "loss": 1.1983,
+      "step": 539
+    },
+    {
+      "epoch": 0.07431874483897605,
+      "grad_norm": 1.7955959752445425,
+      "learning_rate": 1.9897173746138234e-05,
+      "loss": 1.3088,
+      "step": 540
+    },
+    {
+      "epoch": 0.07445637214423341,
+      "grad_norm": 1.7420834283480413,
+      "learning_rate": 1.9896535187354365e-05,
+      "loss": 1.2503,
+      "step": 541
+    },
+    {
+      "epoch": 0.07459399944949077,
+      "grad_norm": 1.7211647391789828,
+      "learning_rate": 1.9895894662264423e-05,
+      "loss": 1.2186,
+      "step": 542
+    },
+    {
+      "epoch": 0.07473162675474815,
+      "grad_norm": 1.8325303616295099,
+      "learning_rate": 1.9895252170995675e-05,
+      "loss": 1.2486,
+      "step": 543
+    },
+    {
+      "epoch": 0.07486925406000551,
+      "grad_norm": 1.8518483511369566,
+      "learning_rate": 1.9894607713675765e-05,
+      "loss": 1.2479,
+      "step": 544
+    },
+    {
+      "epoch": 0.07500688136526287,
+      "grad_norm": 1.7080146845786037,
+      "learning_rate": 1.989396129043275e-05,
+      "loss": 1.241,
+      "step": 545
+    },
+    {
+      "epoch": 0.07514450867052024,
+      "grad_norm": 1.787776166443195,
+      "learning_rate": 1.989331290139506e-05,
+      "loss": 1.2329,
+      "step": 546
+    },
+    {
+      "epoch": 0.0752821359757776,
+      "grad_norm": 1.9675816082582034,
+      "learning_rate": 1.989266254669152e-05,
+      "loss": 1.2143,
+      "step": 547
+    },
+    {
+      "epoch": 0.07541976328103496,
+      "grad_norm": 2.0264806515707265,
+      "learning_rate": 1.9892010226451347e-05,
+      "loss": 1.2665,
+      "step": 548
+    },
+    {
+      "epoch": 0.07555739058629232,
+      "grad_norm": 1.8518089805026696,
+      "learning_rate": 1.989135594080415e-05,
+      "loss": 1.267,
+      "step": 549
+    },
+    {
+      "epoch": 0.07569501789154968,
+      "grad_norm": 1.9321483022536319,
+      "learning_rate": 1.9890699689879926e-05,
+      "loss": 1.2446,
+      "step": 550
+    },
+    {
+      "epoch": 0.07583264519680705,
+      "grad_norm": 1.8052040599963532,
+      "learning_rate": 1.989004147380906e-05,
+      "loss": 1.1975,
+      "step": 551
+    },
+    {
+      "epoch": 0.07597027250206441,
+      "grad_norm": 1.7004916155362677,
+      "learning_rate": 1.9889381292722336e-05,
+      "loss": 1.1435,
+      "step": 552
+    },
+    {
+      "epoch": 0.07610789980732177,
+      "grad_norm": 1.8804445637446154,
+      "learning_rate": 1.9888719146750918e-05,
+      "loss": 1.2393,
+      "step": 553
+    },
+    {
+      "epoch": 0.07624552711257913,
+      "grad_norm": 1.967804620888914,
+      "learning_rate": 1.988805503602637e-05,
+      "loss": 1.2418,
+      "step": 554
+    },
+    {
+      "epoch": 0.0763831544178365,
+      "grad_norm": 1.7138058240798248,
+      "learning_rate": 1.9887388960680638e-05,
+      "loss": 1.2663,
+      "step": 555
+    },
+    {
+      "epoch": 0.07652078172309386,
+      "grad_norm": 2.113844612507439,
+      "learning_rate": 1.9886720920846063e-05,
+      "loss": 1.3577,
+      "step": 556
+    },
+    {
+      "epoch": 0.07665840902835122,
+      "grad_norm": 1.8325053943133653,
+      "learning_rate": 1.9886050916655375e-05,
+      "loss": 1.2186,
+      "step": 557
+    },
+    {
+      "epoch": 0.07679603633360858,
+      "grad_norm": 1.8909385711887048,
+      "learning_rate": 1.9885378948241696e-05,
+      "loss": 1.275,
+      "step": 558
+    },
+    {
+      "epoch": 0.07693366363886595,
+      "grad_norm": 1.883956694389189,
+      "learning_rate": 1.9884705015738536e-05,
+      "loss": 1.2713,
+      "step": 559
+    },
+    {
+      "epoch": 0.07707129094412331,
+      "grad_norm": 1.8226106789412526,
+      "learning_rate": 1.9884029119279794e-05,
+      "loss": 1.2087,
+      "step": 560
+    },
+    {
+      "epoch": 0.07720891824938067,
+      "grad_norm": 2.003347655718789,
+      "learning_rate": 1.9883351258999764e-05,
+      "loss": 1.2244,
+      "step": 561
+    },
+    {
+      "epoch": 0.07734654555463805,
+      "grad_norm": 1.8504277193618903,
+      "learning_rate": 1.988267143503313e-05,
+      "loss": 1.234,
+      "step": 562
+    },
+    {
+      "epoch": 0.07748417285989541,
+      "grad_norm": 1.9751309965127748,
+      "learning_rate": 1.988198964751496e-05,
+      "loss": 1.2329,
+      "step": 563
+    },
+    {
+      "epoch": 0.07762180016515277,
+      "grad_norm": 1.7120810852198987,
+      "learning_rate": 1.9881305896580715e-05,
+      "loss": 1.2759,
+      "step": 564
+    },
+    {
+      "epoch": 0.07775942747041013,
+      "grad_norm": 1.7660846849413923,
+      "learning_rate": 1.988062018236625e-05,
+      "loss": 1.2106,
+      "step": 565
+    },
+    {
+      "epoch": 0.0778970547756675,
+      "grad_norm": 2.1765335761884814,
+      "learning_rate": 1.9879932505007803e-05,
+      "loss": 1.3522,
+      "step": 566
+    },
+    {
+      "epoch": 0.07803468208092486,
+      "grad_norm": 1.8089315794528318,
+      "learning_rate": 1.9879242864642013e-05,
+      "loss": 1.2475,
+      "step": 567
+    },
+    {
+      "epoch": 0.07817230938618222,
+      "grad_norm": 2.1708236031325034,
+      "learning_rate": 1.9878551261405893e-05,
+      "loss": 1.4144,
+      "step": 568
+    },
+    {
+      "epoch": 0.07830993669143958,
+      "grad_norm": 2.001722612405579,
+      "learning_rate": 1.9877857695436863e-05,
+      "loss": 1.3721,
+      "step": 569
+    },
+    {
+      "epoch": 0.07844756399669695,
+      "grad_norm": 1.79750491041061,
+      "learning_rate": 1.9877162166872727e-05,
+      "loss": 1.2368,
+      "step": 570
+    },
+    {
+      "epoch": 0.07858519130195431,
+      "grad_norm": 1.8581748444348096,
+      "learning_rate": 1.9876464675851665e-05,
+      "loss": 1.2423,
+      "step": 571
+    },
+    {
+      "epoch": 0.07872281860721167,
+      "grad_norm": 2.018617395741517,
+      "learning_rate": 1.9875765222512267e-05,
+      "loss": 1.1859,
+      "step": 572
+    },
+    {
+      "epoch": 0.07886044591246903,
+      "grad_norm": 1.9573830078171723,
+      "learning_rate": 1.9875063806993508e-05,
+      "loss": 1.2288,
+      "step": 573
+    },
+    {
+      "epoch": 0.0789980732177264,
+      "grad_norm": 1.780591420553638,
+      "learning_rate": 1.987436042943474e-05,
+      "loss": 1.2473,
+      "step": 574
+    },
+    {
+      "epoch": 0.07913570052298376,
+      "grad_norm": 1.7854249175916004,
+      "learning_rate": 1.9873655089975722e-05,
+      "loss": 1.2346,
+      "step": 575
+    },
+    {
+      "epoch": 0.07927332782824112,
+      "grad_norm": 1.7471304396664455,
+      "learning_rate": 1.9872947788756592e-05,
+      "loss": 1.2789,
+      "step": 576
+    },
+    {
+      "epoch": 0.07941095513349848,
+      "grad_norm": 1.9086477608673977,
+      "learning_rate": 1.9872238525917885e-05,
+      "loss": 1.3156,
+      "step": 577
+    },
+    {
+      "epoch": 0.07954858243875584,
+      "grad_norm": 1.74526336497638,
+      "learning_rate": 1.9871527301600516e-05,
+      "loss": 1.3196,
+      "step": 578
+    },
+    {
+      "epoch": 0.0796862097440132,
+      "grad_norm": 1.7796989086839412,
+      "learning_rate": 1.98708141159458e-05,
+      "loss": 1.2096,
+      "step": 579
+    },
+    {
+      "epoch": 0.07982383704927057,
+      "grad_norm": 1.7409836389829947,
+      "learning_rate": 1.987009896909544e-05,
+      "loss": 1.1783,
+      "step": 580
+    },
+    {
+      "epoch": 0.07996146435452794,
+      "grad_norm": 2.0288717992342704,
+      "learning_rate": 1.9869381861191518e-05,
+      "loss": 1.362,
+      "step": 581
+    },
+    {
+      "epoch": 0.08009909165978531,
+      "grad_norm": 2.22104670935202,
+      "learning_rate": 1.9868662792376518e-05,
+      "loss": 1.1934,
+      "step": 582
+    },
+    {
+      "epoch": 0.08023671896504267,
+      "grad_norm": 1.6938462514595205,
+      "learning_rate": 1.9867941762793307e-05,
+      "loss": 1.1924,
+      "step": 583
+    },
+    {
+      "epoch": 0.08037434627030003,
+      "grad_norm": 1.7501484004589887,
+      "learning_rate": 1.9867218772585146e-05,
+      "loss": 1.1451,
+      "step": 584
+    },
+    {
+      "epoch": 0.0805119735755574,
+      "grad_norm": 1.8104157588693333,
+      "learning_rate": 1.9866493821895684e-05,
+      "loss": 1.2997,
+      "step": 585
+    },
+    {
+      "epoch": 0.08064960088081476,
+      "grad_norm": 1.881624799711022,
+      "learning_rate": 1.9865766910868954e-05,
+      "loss": 1.3111,
+      "step": 586
+    },
+    {
+      "epoch": 0.08078722818607212,
+      "grad_norm": 1.9013486114759162,
+      "learning_rate": 1.9865038039649386e-05,
+      "loss": 1.1747,
+      "step": 587
+    },
+    {
+      "epoch": 0.08092485549132948,
+      "grad_norm": 1.8162016199870816,
+      "learning_rate": 1.9864307208381802e-05,
+      "loss": 1.3138,
+      "step": 588
+    },
+    {
+      "epoch": 0.08106248279658684,
+      "grad_norm": 1.8299051952470389,
+      "learning_rate": 1.98635744172114e-05,
+      "loss": 1.2967,
+      "step": 589
+    },
+    {
+      "epoch": 0.0812001101018442,
+      "grad_norm": 1.6327750679709911,
+      "learning_rate": 1.986283966628378e-05,
+      "loss": 1.1681,
+      "step": 590
+    },
+    {
+      "epoch": 0.08133773740710157,
+      "grad_norm": 1.7536784820951652,
+      "learning_rate": 1.9862102955744923e-05,
+      "loss": 1.2247,
+      "step": 591
+    },
+    {
+      "epoch": 0.08147536471235893,
+      "grad_norm": 1.7359732299869188,
+      "learning_rate": 1.986136428574121e-05,
+      "loss": 1.0886,
+      "step": 592
+    },
+    {
+      "epoch": 0.08161299201761629,
+      "grad_norm": 1.7604144586515782,
+      "learning_rate": 1.98606236564194e-05,
+      "loss": 1.162,
+      "step": 593
+    },
+    {
+      "epoch": 0.08175061932287365,
+      "grad_norm": 1.8099089635305472,
+      "learning_rate": 1.9859881067926646e-05,
+      "loss": 1.2731,
+      "step": 594
+    },
+    {
+      "epoch": 0.08188824662813102,
+      "grad_norm": 1.7795524106857699,
+      "learning_rate": 1.985913652041049e-05,
+      "loss": 1.1829,
+      "step": 595
+    },
+    {
+      "epoch": 0.08202587393338838,
+      "grad_norm": 1.9820053136758584,
+      "learning_rate": 1.985839001401887e-05,
+      "loss": 1.2816,
+      "step": 596
+    },
+    {
+      "epoch": 0.08216350123864574,
+      "grad_norm": 1.7325962212105188,
+      "learning_rate": 1.985764154890009e-05,
+      "loss": 1.2378,
+      "step": 597
+    },
+    {
+      "epoch": 0.0823011285439031,
+      "grad_norm": 1.951390479975755,
+      "learning_rate": 1.9856891125202876e-05,
+      "loss": 1.2897,
+      "step": 598
+    },
+    {
+      "epoch": 0.08243875584916047,
+      "grad_norm": 1.8575623540278854,
+      "learning_rate": 1.9856138743076324e-05,
+      "loss": 1.1912,
+      "step": 599
+    },
+    {
+      "epoch": 0.08257638315441784,
+      "grad_norm": 1.9755032706377214,
+      "learning_rate": 1.9855384402669918e-05,
+      "loss": 1.2628,
+      "step": 600
+    },
+    {
+      "epoch": 0.0827140104596752,
+      "grad_norm": 1.7478271222634836,
+      "learning_rate": 1.9854628104133537e-05,
+      "loss": 1.2031,
+      "step": 601
+    },
+    {
+      "epoch": 0.08285163776493257,
+      "grad_norm": 1.8133089383753362,
+      "learning_rate": 1.9853869847617446e-05,
+      "loss": 1.1753,
+      "step": 602
+    },
+    {
+      "epoch": 0.08298926507018993,
+      "grad_norm": 1.7121916845828211,
+      "learning_rate": 1.9853109633272304e-05,
+      "loss": 1.2198,
+      "step": 603
+    },
+    {
+      "epoch": 0.08312689237544729,
+      "grad_norm": 1.832810934571967,
+      "learning_rate": 1.9852347461249147e-05,
+      "loss": 1.2359,
+      "step": 604
+    },
+    {
+      "epoch": 0.08326451968070465,
+      "grad_norm": 1.9455698035097133,
+      "learning_rate": 1.985158333169942e-05,
+      "loss": 1.2741,
+      "step": 605
+    },
+    {
+      "epoch": 0.08340214698596202,
+      "grad_norm": 1.7785525766174886,
+      "learning_rate": 1.985081724477493e-05,
+      "loss": 1.2948,
+      "step": 606
+    },
+    {
+      "epoch": 0.08353977429121938,
+      "grad_norm": 1.8250397292848948,
+      "learning_rate": 1.98500492006279e-05,
+      "loss": 1.1642,
+      "step": 607
+    },
+    {
+      "epoch": 0.08367740159647674,
+      "grad_norm": 1.877612555176493,
+      "learning_rate": 1.984927919941093e-05,
+      "loss": 1.2648,
+      "step": 608
+    },
+    {
+      "epoch": 0.0838150289017341,
+      "grad_norm": 2.0300331667181064,
+      "learning_rate": 1.9848507241277e-05,
+      "loss": 1.2384,
+      "step": 609
+    },
+    {
+      "epoch": 0.08395265620699147,
+      "grad_norm": 1.8324365552440949,
+      "learning_rate": 1.9847733326379498e-05,
+      "loss": 1.2175,
+      "step": 610
+    },
+    {
+      "epoch": 0.08409028351224883,
+      "grad_norm": 1.7009951015478637,
+      "learning_rate": 1.9846957454872184e-05,
+      "loss": 1.1513,
+      "step": 611
+    },
+    {
+      "epoch": 0.08422791081750619,
+      "grad_norm": 1.8723007199515214,
+      "learning_rate": 1.984617962690921e-05,
+      "loss": 1.1792,
+      "step": 612
+    },
+    {
+      "epoch": 0.08436553812276355,
+      "grad_norm": 1.6333198481656388,
+      "learning_rate": 1.9845399842645122e-05,
+      "loss": 1.2576,
+      "step": 613
+    },
+    {
+      "epoch": 0.08450316542802092,
+      "grad_norm": 1.9587193634240436,
+      "learning_rate": 1.984461810223486e-05,
+      "loss": 1.197,
+      "step": 614
+    },
+    {
+      "epoch": 0.08464079273327828,
+      "grad_norm": 1.7852430846973624,
+      "learning_rate": 1.9843834405833737e-05,
+      "loss": 1.2481,
+      "step": 615
+    },
+    {
+      "epoch": 0.08477842003853564,
+      "grad_norm": 1.885921597105817,
+      "learning_rate": 1.9843048753597465e-05,
+      "loss": 1.3181,
+      "step": 616
+    },
+    {
+      "epoch": 0.084916047343793,
+      "grad_norm": 1.6990015177886433,
+      "learning_rate": 1.984226114568214e-05,
+      "loss": 1.22,
+      "step": 617
+    },
+    {
+      "epoch": 0.08505367464905036,
+      "grad_norm": 1.8980496281782069,
+      "learning_rate": 1.9841471582244255e-05,
+      "loss": 1.2611,
+      "step": 618
+    },
+    {
+      "epoch": 0.08519130195430774,
+      "grad_norm": 1.7915875141063173,
+      "learning_rate": 1.9840680063440677e-05,
+      "loss": 1.1858,
+      "step": 619
+    },
+    {
+      "epoch": 0.0853289292595651,
+      "grad_norm": 1.8936549672848455,
+      "learning_rate": 1.9839886589428677e-05,
+      "loss": 1.2093,
+      "step": 620
+    },
+    {
+      "epoch": 0.08546655656482247,
+      "grad_norm": 1.758208040838905,
+      "learning_rate": 1.9839091160365905e-05,
+      "loss": 1.2658,
+      "step": 621
+    },
+    {
+      "epoch": 0.08560418387007983,
+      "grad_norm": 1.7564253481193826,
+      "learning_rate": 1.9838293776410403e-05,
+      "loss": 1.1366,
+      "step": 622
+    },
+    {
+      "epoch": 0.08574181117533719,
+      "grad_norm": 2.243469023207016,
+      "learning_rate": 1.98374944377206e-05,
+      "loss": 1.2652,
+      "step": 623
+    },
+    {
+      "epoch": 0.08587943848059455,
+      "grad_norm": 1.7147136350412064,
+      "learning_rate": 1.983669314445531e-05,
+      "loss": 1.2963,
+      "step": 624
+    },
+    {
+      "epoch": 0.08601706578585191,
+      "grad_norm": 1.9159372904227485,
+      "learning_rate": 1.9835889896773745e-05,
+      "loss": 1.279,
+      "step": 625
+    },
+    {
+      "epoch": 0.08615469309110928,
+      "grad_norm": 1.9872855549008928,
+      "learning_rate": 1.9835084694835494e-05,
+      "loss": 1.3619,
+      "step": 626
+    },
+    {
+      "epoch": 0.08629232039636664,
+      "grad_norm": 1.731810190424292,
+      "learning_rate": 1.9834277538800542e-05,
+      "loss": 1.2183,
+      "step": 627
+    },
+    {
+      "epoch": 0.086429947701624,
+      "grad_norm": 1.9784673080145145,
+      "learning_rate": 1.983346842882926e-05,
+      "loss": 1.2106,
+      "step": 628
+    },
+    {
+      "epoch": 0.08656757500688136,
+      "grad_norm": 1.7866826871725954,
+      "learning_rate": 1.9832657365082405e-05,
+      "loss": 1.2775,
+      "step": 629
+    },
+    {
+      "epoch": 0.08670520231213873,
+      "grad_norm": 1.8406817147451717,
+      "learning_rate": 1.983184434772113e-05,
+      "loss": 1.2212,
+      "step": 630
+    },
+    {
+      "epoch": 0.08684282961739609,
+      "grad_norm": 1.9065171842861885,
+      "learning_rate": 1.9831029376906964e-05,
+      "loss": 1.2547,
+      "step": 631
+    },
+    {
+      "epoch": 0.08698045692265345,
+      "grad_norm": 1.7619078995339281,
+      "learning_rate": 1.9830212452801833e-05,
+      "loss": 1.2579,
+      "step": 632
+    },
+    {
+      "epoch": 0.08711808422791081,
+      "grad_norm": 1.8872222608480584,
+      "learning_rate": 1.9829393575568047e-05,
+      "loss": 1.2933,
+      "step": 633
+    },
+    {
+      "epoch": 0.08725571153316818,
+      "grad_norm": 1.8093014240197023,
+      "learning_rate": 1.9828572745368308e-05,
+      "loss": 1.1833,
+      "step": 634
+    },
+    {
+      "epoch": 0.08739333883842554,
+      "grad_norm": 1.7917366553721097,
+      "learning_rate": 1.9827749962365703e-05,
+      "loss": 1.1602,
+      "step": 635
+    },
+    {
+      "epoch": 0.0875309661436829,
+      "grad_norm": 1.7988128028268695,
+      "learning_rate": 1.982692522672371e-05,
+      "loss": 1.2033,
+      "step": 636
+    },
+    {
+      "epoch": 0.08766859344894026,
+      "grad_norm": 1.8909545385019588,
+      "learning_rate": 1.982609853860619e-05,
+      "loss": 1.2734,
+      "step": 637
+    },
+    {
+      "epoch": 0.08780622075419764,
+      "grad_norm": 1.918998366977536,
+      "learning_rate": 1.9825269898177393e-05,
+      "loss": 1.261,
+      "step": 638
+    },
+    {
+      "epoch": 0.087943848059455,
+      "grad_norm": 1.9258087631701541,
+      "learning_rate": 1.9824439305601962e-05,
+      "loss": 1.2354,
+      "step": 639
+    },
+    {
+      "epoch": 0.08808147536471236,
+      "grad_norm": 2.0900426506399303,
+      "learning_rate": 1.982360676104492e-05,
+      "loss": 1.2465,
+      "step": 640
+    },
+    {
+      "epoch": 0.08821910266996973,
+      "grad_norm": 1.948945569543138,
+      "learning_rate": 1.9822772264671686e-05,
+      "loss": 1.2968,
+      "step": 641
+    },
+    {
+      "epoch": 0.08835672997522709,
+      "grad_norm": 1.7982632978073727,
+      "learning_rate": 1.9821935816648066e-05,
+      "loss": 1.1876,
+      "step": 642
+    },
+    {
+      "epoch": 0.08849435728048445,
+      "grad_norm": 1.9484754134540905,
+      "learning_rate": 1.9821097417140244e-05,
+      "loss": 1.1904,
+      "step": 643
+    },
+    {
+      "epoch": 0.08863198458574181,
+      "grad_norm": 1.7276502992932183,
+      "learning_rate": 1.9820257066314802e-05,
+      "loss": 1.2259,
+      "step": 644
+    },
+    {
+      "epoch": 0.08876961189099918,
+      "grad_norm": 1.5885474147554055,
+      "learning_rate": 1.9819414764338703e-05,
+      "loss": 1.1687,
+      "step": 645
+    },
+    {
+      "epoch": 0.08890723919625654,
+      "grad_norm": 1.7560298017509959,
+      "learning_rate": 1.9818570511379308e-05,
+      "loss": 1.1898,
+      "step": 646
+    },
+    {
+      "epoch": 0.0890448665015139,
+      "grad_norm": 1.7793076010414537,
+      "learning_rate": 1.9817724307604352e-05,
+      "loss": 1.1515,
+      "step": 647
+    },
+    {
+      "epoch": 0.08918249380677126,
+      "grad_norm": 1.8593857721543998,
+      "learning_rate": 1.9816876153181966e-05,
+      "loss": 1.2463,
+      "step": 648
+    },
+    {
+      "epoch": 0.08932012111202862,
+      "grad_norm": 1.7594835472864392,
+      "learning_rate": 1.9816026048280667e-05,
+      "loss": 1.2119,
+      "step": 649
+    },
+    {
+      "epoch": 0.08945774841728599,
+      "grad_norm": 1.5350244591404703,
+      "learning_rate": 1.9815173993069358e-05,
+      "loss": 1.1324,
+      "step": 650
+    },
+    {
+      "epoch": 0.08959537572254335,
+      "grad_norm": 2.076378114591709,
+      "learning_rate": 1.9814319987717334e-05,
+      "loss": 1.2727,
+      "step": 651
+    },
+    {
+      "epoch": 0.08973300302780071,
+      "grad_norm": 1.8855989007056961,
+      "learning_rate": 1.9813464032394268e-05,
+      "loss": 1.267,
+      "step": 652
+    },
+    {
+      "epoch": 0.08987063033305807,
+      "grad_norm": 2.108034190234174,
+      "learning_rate": 1.9812606127270235e-05,
+      "loss": 1.1576,
+      "step": 653
+    },
+    {
+      "epoch": 0.09000825763831544,
+      "grad_norm": 1.9859441124596537,
+      "learning_rate": 1.981174627251568e-05,
+      "loss": 1.2892,
+      "step": 654
+    },
+    {
+      "epoch": 0.0901458849435728,
+      "grad_norm": 1.8345461010068997,
+      "learning_rate": 1.9810884468301447e-05,
+      "loss": 1.3095,
+      "step": 655
+    },
+    {
+      "epoch": 0.09028351224883016,
+      "grad_norm": 1.8743731884736723,
+      "learning_rate": 1.981002071479877e-05,
+      "loss": 1.1875,
+      "step": 656
+    },
+    {
+      "epoch": 0.09042113955408754,
+      "grad_norm": 1.794317154941314,
+      "learning_rate": 1.9809155012179263e-05,
+      "loss": 1.2998,
+      "step": 657
+    },
+    {
+      "epoch": 0.0905587668593449,
+      "grad_norm": 1.7443226042588642,
+      "learning_rate": 1.980828736061493e-05,
+      "loss": 1.2207,
+      "step": 658
+    },
+    {
+      "epoch": 0.09069639416460226,
+      "grad_norm": 1.5814225235293542,
+      "learning_rate": 1.9807417760278155e-05,
+      "loss": 1.1775,
+      "step": 659
+    },
+    {
+      "epoch": 0.09083402146985962,
+      "grad_norm": 1.8304642950928993,
+      "learning_rate": 1.980654621134172e-05,
+      "loss": 1.2466,
+      "step": 660
+    },
+    {
+      "epoch": 0.09097164877511699,
+      "grad_norm": 2.3211483028102315,
+      "learning_rate": 1.9805672713978792e-05,
+      "loss": 1.2705,
+      "step": 661
+    },
+    {
+      "epoch": 0.09110927608037435,
+      "grad_norm": 1.6929008555449716,
+      "learning_rate": 1.9804797268362917e-05,
+      "loss": 1.2016,
+      "step": 662
+    },
+    {
+      "epoch": 0.09124690338563171,
+      "grad_norm": 1.8285618639988,
+      "learning_rate": 1.980391987466804e-05,
+      "loss": 1.2918,
+      "step": 663
+    },
+    {
+      "epoch": 0.09138453069088907,
+      "grad_norm": 1.7383522645955118,
+      "learning_rate": 1.9803040533068486e-05,
+      "loss": 1.1953,
+      "step": 664
+    },
+    {
+      "epoch": 0.09152215799614644,
+      "grad_norm": 1.6763602778235207,
+      "learning_rate": 1.9802159243738965e-05,
+      "loss": 1.1899,
+      "step": 665
+    },
+    {
+      "epoch": 0.0916597853014038,
+      "grad_norm": 1.6983935385806757,
+      "learning_rate": 1.980127600685458e-05,
+      "loss": 1.1132,
+      "step": 666
+    },
+    {
+      "epoch": 0.09179741260666116,
+      "grad_norm": 1.8739719072599466,
+      "learning_rate": 1.980039082259082e-05,
+      "loss": 1.1821,
+      "step": 667
+    },
+    {
+      "epoch": 0.09193503991191852,
+      "grad_norm": 1.7070208391799184,
+      "learning_rate": 1.979950369112355e-05,
+      "loss": 1.2924,
+      "step": 668
+    },
+    {
+      "epoch": 0.09207266721717589,
+      "grad_norm": 1.7080292769872887,
+      "learning_rate": 1.9798614612629046e-05,
+      "loss": 1.2166,
+      "step": 669
+    },
+    {
+      "epoch": 0.09221029452243325,
+      "grad_norm": 1.7157127820778475,
+      "learning_rate": 1.979772358728394e-05,
+      "loss": 1.2743,
+      "step": 670
+    },
+    {
+      "epoch": 0.09234792182769061,
+      "grad_norm": 1.727731542149226,
+      "learning_rate": 1.9796830615265277e-05,
+      "loss": 1.2061,
+      "step": 671
+    },
+    {
+      "epoch": 0.09248554913294797,
+      "grad_norm": 1.7332766679140847,
+      "learning_rate": 1.979593569675047e-05,
+      "loss": 1.1293,
+      "step": 672
+    },
+    {
+      "epoch": 0.09262317643820533,
+      "grad_norm": 2.0207110871890643,
+      "learning_rate": 1.9795038831917336e-05,
+      "loss": 1.1878,
+      "step": 673
+    },
+    {
+      "epoch": 0.0927608037434627,
+      "grad_norm": 2.0379833263438303,
+      "learning_rate": 1.9794140020944065e-05,
+      "loss": 1.1672,
+      "step": 674
+    },
+    {
+      "epoch": 0.09289843104872006,
+      "grad_norm": 1.8842495706527753,
+      "learning_rate": 1.979323926400924e-05,
+      "loss": 1.2484,
+      "step": 675
+    },
+    {
+      "epoch": 0.09303605835397744,
+      "grad_norm": 1.8724062474041343,
+      "learning_rate": 1.9792336561291826e-05,
+      "loss": 1.1874,
+      "step": 676
+    },
+    {
+      "epoch": 0.0931736856592348,
+      "grad_norm": 1.7144139868458128,
+      "learning_rate": 1.979143191297118e-05,
+      "loss": 1.2042,
+      "step": 677
+    },
+    {
+      "epoch": 0.09331131296449216,
+      "grad_norm": 1.7374984338058037,
+      "learning_rate": 1.9790525319227044e-05,
+      "loss": 1.2094,
+      "step": 678
+    },
+    {
+      "epoch": 0.09344894026974952,
+      "grad_norm": 1.8866441966630032,
+      "learning_rate": 1.9789616780239546e-05,
+      "loss": 1.2202,
+      "step": 679
+    },
+    {
+      "epoch": 0.09358656757500688,
+      "grad_norm": 1.7595446646295216,
+      "learning_rate": 1.9788706296189196e-05,
+      "loss": 1.2561,
+      "step": 680
+    },
+    {
+      "epoch": 0.09372419488026425,
+      "grad_norm": 1.9510244119012567,
+      "learning_rate": 1.9787793867256897e-05,
+      "loss": 1.1178,
+      "step": 681
+    },
+    {
+      "epoch": 0.09386182218552161,
+      "grad_norm": 1.880639658252601,
+      "learning_rate": 1.978687949362394e-05,
+      "loss": 1.2798,
+      "step": 682
+    },
+    {
+      "epoch": 0.09399944949077897,
+      "grad_norm": 1.944500901743983,
+      "learning_rate": 1.9785963175471995e-05,
+      "loss": 1.2167,
+      "step": 683
+    },
+    {
+      "epoch": 0.09413707679603633,
+      "grad_norm": 1.9942131359447002,
+      "learning_rate": 1.978504491298312e-05,
+      "loss": 1.262,
+      "step": 684
+    },
+    {
+      "epoch": 0.0942747041012937,
+      "grad_norm": 1.697629518782316,
+      "learning_rate": 1.9784124706339763e-05,
+      "loss": 1.2152,
+      "step": 685
+    },
+    {
+      "epoch": 0.09441233140655106,
+      "grad_norm": 1.8902814321213357,
+      "learning_rate": 1.978320255572476e-05,
+      "loss": 1.2812,
+      "step": 686
+    },
+    {
+      "epoch": 0.09454995871180842,
+      "grad_norm": 1.848342889841255,
+      "learning_rate": 1.9782278461321324e-05,
+      "loss": 1.2622,
+      "step": 687
+    },
+    {
+      "epoch": 0.09468758601706578,
+      "grad_norm": 1.9161678038484973,
+      "learning_rate": 1.9781352423313066e-05,
+      "loss": 1.118,
+      "step": 688
+    },
+    {
+      "epoch": 0.09482521332232315,
+      "grad_norm": 1.9036310641252565,
+      "learning_rate": 1.9780424441883973e-05,
+      "loss": 1.1664,
+      "step": 689
+    },
+    {
+      "epoch": 0.09496284062758051,
+      "grad_norm": 1.765079882881183,
+      "learning_rate": 1.977949451721842e-05,
+      "loss": 1.2104,
+      "step": 690
+    },
+    {
+      "epoch": 0.09510046793283787,
+      "grad_norm": 1.8390557654507544,
+      "learning_rate": 1.9778562649501178e-05,
+      "loss": 1.1319,
+      "step": 691
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 1.845007639060345,
+      "learning_rate": 1.9777628838917384e-05,
+      "loss": 1.2095,
+      "step": 692
+    },
+    {
+      "epoch": 0.0953757225433526,
+      "grad_norm": 1.9765014798328004,
+      "learning_rate": 1.9776693085652586e-05,
+      "loss": 1.309,
+      "step": 693
+    },
+    {
+      "epoch": 0.09551334984860996,
+      "grad_norm": 1.6779809099621654,
+      "learning_rate": 1.97757553898927e-05,
+      "loss": 1.1817,
+      "step": 694
+    },
+    {
+      "epoch": 0.09565097715386733,
+      "grad_norm": 1.6871863828426394,
+      "learning_rate": 1.9774815751824032e-05,
+      "loss": 1.2152,
+      "step": 695
+    },
+    {
+      "epoch": 0.0957886044591247,
+      "grad_norm": 1.742631167516675,
+      "learning_rate": 1.977387417163328e-05,
+      "loss": 1.205,
+      "step": 696
+    },
+    {
+      "epoch": 0.09592623176438206,
+      "grad_norm": 1.723068237013323,
+      "learning_rate": 1.9772930649507517e-05,
+      "loss": 1.1354,
+      "step": 697
+    },
+    {
+      "epoch": 0.09606385906963942,
+      "grad_norm": 1.9852375974753866,
+      "learning_rate": 1.9771985185634208e-05,
+      "loss": 1.1616,
+      "step": 698
+    },
+    {
+      "epoch": 0.09620148637489678,
+      "grad_norm": 2.4106687258470183,
+      "learning_rate": 1.977103778020121e-05,
+      "loss": 1.3333,
+      "step": 699
+    },
+    {
+      "epoch": 0.09633911368015415,
+      "grad_norm": 1.7530300741770577,
+      "learning_rate": 1.9770088433396757e-05,
+      "loss": 1.2055,
+      "step": 700
+    },
+    {
+      "epoch": 0.09647674098541151,
+      "grad_norm": 1.8967411690939935,
+      "learning_rate": 1.9769137145409466e-05,
+      "loss": 1.3456,
+      "step": 701
+    },
+    {
+      "epoch": 0.09661436829066887,
+      "grad_norm": 1.8158234589562774,
+      "learning_rate": 1.9768183916428353e-05,
+      "loss": 1.2027,
+      "step": 702
+    },
+    {
+      "epoch": 0.09675199559592623,
+      "grad_norm": 1.8892584425949879,
+      "learning_rate": 1.976722874664281e-05,
+      "loss": 1.2815,
+      "step": 703
+    },
+    {
+      "epoch": 0.0968896229011836,
+      "grad_norm": 1.7515117308536774,
+      "learning_rate": 1.9766271636242606e-05,
+      "loss": 1.223,
+      "step": 704
+    },
+    {
+      "epoch": 0.09702725020644096,
+      "grad_norm": 1.7168746036644345,
+      "learning_rate": 1.976531258541792e-05,
+      "loss": 1.132,
+      "step": 705
+    },
+    {
+      "epoch": 0.09716487751169832,
+      "grad_norm": 1.8208770612527814,
+      "learning_rate": 1.9764351594359295e-05,
+      "loss": 1.2828,
+      "step": 706
+    },
+    {
+      "epoch": 0.09730250481695568,
+      "grad_norm": 1.7735239200978077,
+      "learning_rate": 1.9763388663257665e-05,
+      "loss": 1.1977,
+      "step": 707
+    },
+    {
+      "epoch": 0.09744013212221304,
+      "grad_norm": 1.8943610560320723,
+      "learning_rate": 1.9762423792304355e-05,
+      "loss": 1.2475,
+      "step": 708
+    },
+    {
+      "epoch": 0.0975777594274704,
+      "grad_norm": 1.7374472409077133,
+      "learning_rate": 1.976145698169107e-05,
+      "loss": 1.1813,
+      "step": 709
+    },
+    {
+      "epoch": 0.09771538673272777,
+      "grad_norm": 1.8365119293894638,
+      "learning_rate": 1.9760488231609903e-05,
+      "loss": 1.2114,
+      "step": 710
+    },
+    {
+      "epoch": 0.09785301403798513,
+      "grad_norm": 1.7555311528848458,
+      "learning_rate": 1.975951754225333e-05,
+      "loss": 1.1959,
+      "step": 711
+    },
+    {
+      "epoch": 0.0979906413432425,
+      "grad_norm": 1.9808748840913646,
+      "learning_rate": 1.975854491381422e-05,
+      "loss": 1.1611,
+      "step": 712
+    },
+    {
+      "epoch": 0.09812826864849986,
+      "grad_norm": 1.5992386760621216,
+      "learning_rate": 1.9757570346485807e-05,
+      "loss": 1.1579,
+      "step": 713
+    },
+    {
+      "epoch": 0.09826589595375723,
+      "grad_norm": 1.701042550628502,
+      "learning_rate": 1.975659384046174e-05,
+      "loss": 1.207,
+      "step": 714
+    },
+    {
+      "epoch": 0.0984035232590146,
+      "grad_norm": 1.9661190987595265,
+      "learning_rate": 1.9755615395936026e-05,
+      "loss": 1.2784,
+      "step": 715
+    },
+    {
+      "epoch": 0.09854115056427196,
+      "grad_norm": 1.9156973558469255,
+      "learning_rate": 1.9754635013103072e-05,
+      "loss": 1.2755,
+      "step": 716
+    },
+    {
+      "epoch": 0.09867877786952932,
+      "grad_norm": 1.9986525269566224,
+      "learning_rate": 1.975365269215767e-05,
+      "loss": 1.2136,
+      "step": 717
+    },
+    {
+      "epoch": 0.09881640517478668,
+      "grad_norm": 1.8910712486238896,
+      "learning_rate": 1.9752668433294992e-05,
+      "loss": 1.1378,
+      "step": 718
+    },
+    {
+      "epoch": 0.09895403248004404,
+      "grad_norm": 2.057682622037617,
+      "learning_rate": 1.975168223671059e-05,
+      "loss": 1.2798,
+      "step": 719
+    },
+    {
+      "epoch": 0.0990916597853014,
+      "grad_norm": 1.9251974933693452,
+      "learning_rate": 1.975069410260042e-05,
+      "loss": 1.183,
+      "step": 720
+    },
+    {
+      "epoch": 0.09922928709055877,
+      "grad_norm": 1.637918280432621,
+      "learning_rate": 1.97497040311608e-05,
+      "loss": 1.1388,
+      "step": 721
+    },
+    {
+      "epoch": 0.09936691439581613,
+      "grad_norm": 1.6647602186504662,
+      "learning_rate": 1.9748712022588453e-05,
+      "loss": 1.1798,
+      "step": 722
+    },
+    {
+      "epoch": 0.09950454170107349,
+      "grad_norm": 1.959191109201414,
+      "learning_rate": 1.974771807708047e-05,
+      "loss": 1.3067,
+      "step": 723
+    },
+    {
+      "epoch": 0.09964216900633086,
+      "grad_norm": 1.8601822721385304,
+      "learning_rate": 1.9746722194834334e-05,
+      "loss": 1.2668,
+      "step": 724
+    },
+    {
+      "epoch": 0.09977979631158822,
+      "grad_norm": 1.8000742464283541,
+      "learning_rate": 1.974572437604792e-05,
+      "loss": 1.1486,
+      "step": 725
+    },
+    {
+      "epoch": 0.09991742361684558,
+      "grad_norm": 1.9399314993453969,
+      "learning_rate": 1.9744724620919474e-05,
+      "loss": 1.2432,
+      "step": 726
+    },
+    {
+      "epoch": 0.10005505092210294,
+      "grad_norm": 1.7784740151465612,
+      "learning_rate": 1.9743722929647634e-05,
+      "loss": 1.2243,
+      "step": 727
+    },
+    {
+      "epoch": 0.1001926782273603,
+      "grad_norm": 2.042182417173032,
+      "learning_rate": 1.974271930243143e-05,
+      "loss": 1.1788,
+      "step": 728
+    },
+    {
+      "epoch": 0.10033030553261767,
+      "grad_norm": 1.917460258303292,
+      "learning_rate": 1.9741713739470263e-05,
+      "loss": 1.2277,
+      "step": 729
+    },
+    {
+      "epoch": 0.10046793283787503,
+      "grad_norm": 1.8038060517866223,
+      "learning_rate": 1.974070624096392e-05,
+      "loss": 1.1389,
+      "step": 730
+    },
+    {
+      "epoch": 0.10060556014313239,
+      "grad_norm": 1.8977446811372967,
+      "learning_rate": 1.9739696807112592e-05,
+      "loss": 1.2265,
+      "step": 731
+    },
+    {
+      "epoch": 0.10074318744838975,
+      "grad_norm": 1.7868537758492014,
+      "learning_rate": 1.9738685438116825e-05,
+      "loss": 1.2211,
+      "step": 732
+    },
+    {
+      "epoch": 0.10088081475364713,
+      "grad_norm": 2.0106842945571124,
+      "learning_rate": 1.973767213417757e-05,
+      "loss": 1.2148,
+      "step": 733
+    },
+    {
+      "epoch": 0.10101844205890449,
+      "grad_norm": 1.6712782823250862,
+      "learning_rate": 1.9736656895496157e-05,
+      "loss": 1.1554,
+      "step": 734
+    },
+    {
+      "epoch": 0.10115606936416185,
+      "grad_norm": 1.5643888114139286,
+      "learning_rate": 1.97356397222743e-05,
+      "loss": 1.1379,
+      "step": 735
+    },
+    {
+      "epoch": 0.10129369666941922,
+      "grad_norm": 2.114090341942714,
+      "learning_rate": 1.9734620614714094e-05,
+      "loss": 1.2676,
+      "step": 736
+    },
+    {
+      "epoch": 0.10143132397467658,
+      "grad_norm": 1.7656795822169238,
+      "learning_rate": 1.9733599573018027e-05,
+      "loss": 1.234,
+      "step": 737
+    },
+    {
+      "epoch": 0.10156895127993394,
+      "grad_norm": 2.0119765189067524,
+      "learning_rate": 1.973257659738896e-05,
+      "loss": 1.2152,
+      "step": 738
+    },
+    {
+      "epoch": 0.1017065785851913,
+      "grad_norm": 1.7799545473837703,
+      "learning_rate": 1.9731551688030147e-05,
+      "loss": 1.2504,
+      "step": 739
+    },
+    {
+      "epoch": 0.10184420589044867,
+      "grad_norm": 2.0072899277008935,
+      "learning_rate": 1.973052484514523e-05,
+      "loss": 1.2618,
+      "step": 740
+    },
+    {
+      "epoch": 0.10198183319570603,
+      "grad_norm": 1.7754229916643216,
+      "learning_rate": 1.9729496068938216e-05,
+      "loss": 1.2147,
+      "step": 741
+    },
+    {
+      "epoch": 0.10211946050096339,
+      "grad_norm": 1.7440777510285383,
+      "learning_rate": 1.972846535961352e-05,
+      "loss": 1.2779,
+      "step": 742
+    },
+    {
+      "epoch": 0.10225708780622075,
+      "grad_norm": 1.75526064957948,
+      "learning_rate": 1.9727432717375924e-05,
+      "loss": 1.1549,
+      "step": 743
+    },
+    {
+      "epoch": 0.10239471511147812,
+      "grad_norm": 1.6249605696946492,
+      "learning_rate": 1.97263981424306e-05,
+      "loss": 1.1791,
+      "step": 744
+    },
+    {
+      "epoch": 0.10253234241673548,
+      "grad_norm": 1.6986864781343542,
+      "learning_rate": 1.9725361634983103e-05,
+      "loss": 1.2508,
+      "step": 745
+    },
+    {
+      "epoch": 0.10266996972199284,
+      "grad_norm": 1.93764499091909,
+      "learning_rate": 1.972432319523938e-05,
+      "loss": 1.1811,
+      "step": 746
+    },
+    {
+      "epoch": 0.1028075970272502,
+      "grad_norm": 2.0230995540669796,
+      "learning_rate": 1.9723282823405748e-05,
+      "loss": 1.2663,
+      "step": 747
+    },
+    {
+      "epoch": 0.10294522433250756,
+      "grad_norm": 1.730298986036247,
+      "learning_rate": 1.9722240519688915e-05,
+      "loss": 1.1682,
+      "step": 748
+    },
+    {
+      "epoch": 0.10308285163776493,
+      "grad_norm": 1.7089365539586916,
+      "learning_rate": 1.9721196284295973e-05,
+      "loss": 1.1776,
+      "step": 749
+    },
+    {
+      "epoch": 0.10322047894302229,
+      "grad_norm": 1.7837131574908531,
+      "learning_rate": 1.9720150117434397e-05,
+      "loss": 1.0927,
+      "step": 750
+    },
+    {
+      "epoch": 0.10335810624827965,
+      "grad_norm": 1.8373385477618713,
+      "learning_rate": 1.971910201931205e-05,
+      "loss": 1.2086,
+      "step": 751
+    },
+    {
+      "epoch": 0.10349573355353703,
+      "grad_norm": 1.7111699441748365,
+      "learning_rate": 1.971805199013717e-05,
+      "loss": 1.2304,
+      "step": 752
+    },
+    {
+      "epoch": 0.10363336085879439,
+      "grad_norm": 1.9201801896746795,
+      "learning_rate": 1.971700003011839e-05,
+      "loss": 1.1513,
+      "step": 753
+    },
+    {
+      "epoch": 0.10377098816405175,
+      "grad_norm": 1.7856567369268712,
+      "learning_rate": 1.9715946139464708e-05,
+      "loss": 1.2368,
+      "step": 754
+    },
+    {
+      "epoch": 0.10390861546930912,
+      "grad_norm": 1.7045737803004442,
+      "learning_rate": 1.971489031838553e-05,
+      "loss": 1.2169,
+      "step": 755
+    },
+    {
+      "epoch": 0.10404624277456648,
+      "grad_norm": 1.8956621696738971,
+      "learning_rate": 1.9713832567090623e-05,
+      "loss": 1.2508,
+      "step": 756
+    },
+    {
+      "epoch": 0.10418387007982384,
+      "grad_norm": 1.8207264687685447,
+      "learning_rate": 1.9712772885790158e-05,
+      "loss": 1.1835,
+      "step": 757
+    },
+    {
+      "epoch": 0.1043214973850812,
+      "grad_norm": 1.711579234121396,
+      "learning_rate": 1.9711711274694672e-05,
+      "loss": 1.1762,
+      "step": 758
+    },
+    {
+      "epoch": 0.10445912469033856,
+      "grad_norm": 1.6961355602286854,
+      "learning_rate": 1.9710647734015095e-05,
+      "loss": 1.2785,
+      "step": 759
+    },
+    {
+      "epoch": 0.10459675199559593,
+      "grad_norm": 2.005026216754533,
+      "learning_rate": 1.9709582263962733e-05,
+      "loss": 1.223,
+      "step": 760
+    },
+    {
+      "epoch": 0.10473437930085329,
+      "grad_norm": 1.6812742381859451,
+      "learning_rate": 1.970851486474929e-05,
+      "loss": 1.2093,
+      "step": 761
+    },
+    {
+      "epoch": 0.10487200660611065,
+      "grad_norm": 1.6338131791444162,
+      "learning_rate": 1.9707445536586838e-05,
+      "loss": 1.1787,
+      "step": 762
+    },
+    {
+      "epoch": 0.10500963391136801,
+      "grad_norm": 1.9276818871701158,
+      "learning_rate": 1.970637427968784e-05,
+      "loss": 1.1688,
+      "step": 763
+    },
+    {
+      "epoch": 0.10514726121662538,
+      "grad_norm": 1.7096468041785715,
+      "learning_rate": 1.9705301094265133e-05,
+      "loss": 1.1396,
+      "step": 764
+    },
+    {
+      "epoch": 0.10528488852188274,
+      "grad_norm": 1.67731357024276,
+      "learning_rate": 1.9704225980531955e-05,
+      "loss": 1.1899,
+      "step": 765
+    },
+    {
+      "epoch": 0.1054225158271401,
+      "grad_norm": 1.7965973099396828,
+      "learning_rate": 1.9703148938701908e-05,
+      "loss": 1.2839,
+      "step": 766
+    },
+    {
+      "epoch": 0.10556014313239746,
+      "grad_norm": 1.8123593135723155,
+      "learning_rate": 1.970206996898899e-05,
+      "loss": 1.2097,
+      "step": 767
+    },
+    {
+      "epoch": 0.10569777043765483,
+      "grad_norm": 1.649070274547035,
+      "learning_rate": 1.9700989071607576e-05,
+      "loss": 1.1442,
+      "step": 768
+    },
+    {
+      "epoch": 0.10583539774291219,
+      "grad_norm": 1.6642024215214961,
+      "learning_rate": 1.969990624677242e-05,
+      "loss": 1.1348,
+      "step": 769
+    },
+    {
+      "epoch": 0.10597302504816955,
+      "grad_norm": 1.7995298999157636,
+      "learning_rate": 1.969882149469868e-05,
+      "loss": 1.2297,
+      "step": 770
+    },
+    {
+      "epoch": 0.10611065235342693,
+      "grad_norm": 1.814648402489889,
+      "learning_rate": 1.9697734815601864e-05,
+      "loss": 1.2754,
+      "step": 771
+    },
+    {
+      "epoch": 0.10624827965868429,
+      "grad_norm": 1.7078896126295318,
+      "learning_rate": 1.969664620969789e-05,
+      "loss": 1.1533,
+      "step": 772
+    },
+    {
+      "epoch": 0.10638590696394165,
+      "grad_norm": 1.963466038072907,
+      "learning_rate": 1.9695555677203046e-05,
+      "loss": 1.1985,
+      "step": 773
+    },
+    {
+      "epoch": 0.10652353426919901,
+      "grad_norm": 1.6990478184049804,
+      "learning_rate": 1.969446321833401e-05,
+      "loss": 1.2044,
+      "step": 774
+    },
+    {
+      "epoch": 0.10666116157445638,
+      "grad_norm": 1.6851565048110029,
+      "learning_rate": 1.969336883330783e-05,
+      "loss": 1.1992,
+      "step": 775
+    },
+    {
+      "epoch": 0.10679878887971374,
+      "grad_norm": 1.8198013206758212,
+      "learning_rate": 1.9692272522341957e-05,
+      "loss": 1.215,
+      "step": 776
+    },
+    {
+      "epoch": 0.1069364161849711,
+      "grad_norm": 1.8414563705435112,
+      "learning_rate": 1.96911742856542e-05,
+      "loss": 1.165,
+      "step": 777
+    },
+    {
+      "epoch": 0.10707404349022846,
+      "grad_norm": 1.562134767351314,
+      "learning_rate": 1.9690074123462772e-05,
+      "loss": 1.1559,
+      "step": 778
+    },
+    {
+      "epoch": 0.10721167079548583,
+      "grad_norm": 1.7005654833028276,
+      "learning_rate": 1.968897203598626e-05,
+      "loss": 1.1267,
+      "step": 779
+    },
+    {
+      "epoch": 0.10734929810074319,
+      "grad_norm": 2.0035774578012786,
+      "learning_rate": 1.9687868023443633e-05,
+      "loss": 1.3623,
+      "step": 780
+    },
+    {
+      "epoch": 0.10748692540600055,
+      "grad_norm": 1.9407029596539918,
+      "learning_rate": 1.968676208605424e-05,
+      "loss": 1.2352,
+      "step": 781
+    },
+    {
+      "epoch": 0.10762455271125791,
+      "grad_norm": 1.9187328905612822,
+      "learning_rate": 1.9685654224037817e-05,
+      "loss": 1.2461,
+      "step": 782
+    },
+    {
+      "epoch": 0.10776218001651527,
+      "grad_norm": 2.4778910930837164,
+      "learning_rate": 1.968454443761448e-05,
+      "loss": 1.2034,
+      "step": 783
+    },
+    {
+      "epoch": 0.10789980732177264,
+      "grad_norm": 1.8364560332503237,
+      "learning_rate": 1.9683432727004736e-05,
+      "loss": 1.2181,
+      "step": 784
+    },
+    {
+      "epoch": 0.10803743462703,
+      "grad_norm": 1.8310485288297873,
+      "learning_rate": 1.968231909242946e-05,
+      "loss": 1.2354,
+      "step": 785
+    },
+    {
+      "epoch": 0.10817506193228736,
+      "grad_norm": 1.7264948258887771,
+      "learning_rate": 1.9681203534109915e-05,
+      "loss": 1.171,
+      "step": 786
+    },
+    {
+      "epoch": 0.10831268923754472,
+      "grad_norm": 1.7016499010742772,
+      "learning_rate": 1.9680086052267747e-05,
+      "loss": 1.2477,
+      "step": 787
+    },
+    {
+      "epoch": 0.10845031654280209,
+      "grad_norm": 1.8246020603939477,
+      "learning_rate": 1.967896664712499e-05,
+      "loss": 1.2836,
+      "step": 788
+    },
+    {
+      "epoch": 0.10858794384805945,
+      "grad_norm": 1.884603169691201,
+      "learning_rate": 1.967784531890405e-05,
+      "loss": 1.1921,
+      "step": 789
+    },
+    {
+      "epoch": 0.10872557115331682,
+      "grad_norm": 1.7542162102528964,
+      "learning_rate": 1.967672206782772e-05,
+      "loss": 1.1712,
+      "step": 790
+    },
+    {
+      "epoch": 0.10886319845857419,
+      "grad_norm": 1.818258118080152,
+      "learning_rate": 1.967559689411917e-05,
+      "loss": 1.2835,
+      "step": 791
+    },
+    {
+      "epoch": 0.10900082576383155,
+      "grad_norm": 1.7748266843569085,
+      "learning_rate": 1.967446979800197e-05,
+      "loss": 1.2189,
+      "step": 792
+    },
+    {
+      "epoch": 0.10913845306908891,
+      "grad_norm": 2.0508140240013106,
+      "learning_rate": 1.9673340779700045e-05,
+      "loss": 1.176,
+      "step": 793
+    },
+    {
+      "epoch": 0.10927608037434627,
+      "grad_norm": 1.7440930909016372,
+      "learning_rate": 1.967220983943772e-05,
+      "loss": 1.0881,
+      "step": 794
+    },
+    {
+      "epoch": 0.10941370767960364,
+      "grad_norm": 1.8089669284261072,
+      "learning_rate": 1.9671076977439704e-05,
+      "loss": 1.2417,
+      "step": 795
+    },
+    {
+      "epoch": 0.109551334984861,
+      "grad_norm": 1.8817810959088215,
+      "learning_rate": 1.966994219393107e-05,
+      "loss": 1.271,
+      "step": 796
+    },
+    {
+      "epoch": 0.10968896229011836,
+      "grad_norm": 1.8696471343247156,
+      "learning_rate": 1.966880548913729e-05,
+      "loss": 1.2113,
+      "step": 797
+    },
+    {
+      "epoch": 0.10982658959537572,
+      "grad_norm": 1.8233629280841457,
+      "learning_rate": 1.966766686328421e-05,
+      "loss": 1.2153,
+      "step": 798
+    },
+    {
+      "epoch": 0.10996421690063309,
+      "grad_norm": 1.9364882446464375,
+      "learning_rate": 1.9666526316598063e-05,
+      "loss": 1.2581,
+      "step": 799
+    },
+    {
+      "epoch": 0.11010184420589045,
+      "grad_norm": 1.748060179594741,
+      "learning_rate": 1.966538384930546e-05,
+      "loss": 1.2059,
+      "step": 800
+    },
+    {
+      "epoch": 0.11023947151114781,
+      "grad_norm": 1.6224895433154127,
+      "learning_rate": 1.9664239461633384e-05,
+      "loss": 1.1624,
+      "step": 801
+    },
+    {
+      "epoch": 0.11037709881640517,
+      "grad_norm": 1.5471068768537146,
+      "learning_rate": 1.966309315380922e-05,
+      "loss": 1.1836,
+      "step": 802
+    },
+    {
+      "epoch": 0.11051472612166253,
+      "grad_norm": 1.996415501924688,
+      "learning_rate": 1.9661944926060714e-05,
+      "loss": 1.3271,
+      "step": 803
+    },
+    {
+      "epoch": 0.1106523534269199,
+      "grad_norm": 1.8500775643092564,
+      "learning_rate": 1.9660794778616016e-05,
+      "loss": 1.2281,
+      "step": 804
+    },
+    {
+      "epoch": 0.11078998073217726,
+      "grad_norm": 1.6000750212089965,
+      "learning_rate": 1.9659642711703634e-05,
+      "loss": 1.1316,
+      "step": 805
+    },
+    {
+      "epoch": 0.11092760803743462,
+      "grad_norm": 1.8759961343323412,
+      "learning_rate": 1.965848872555247e-05,
+      "loss": 1.1781,
+      "step": 806
+    },
+    {
+      "epoch": 0.11106523534269198,
+      "grad_norm": 1.670197242669125,
+      "learning_rate": 1.965733282039181e-05,
+      "loss": 1.1585,
+      "step": 807
+    },
+    {
+      "epoch": 0.11120286264794935,
+      "grad_norm": 1.9309876152607959,
+      "learning_rate": 1.9656174996451313e-05,
+      "loss": 1.1772,
+      "step": 808
+    },
+    {
+      "epoch": 0.11134048995320672,
+      "grad_norm": 1.6680067793506097,
+      "learning_rate": 1.9655015253961018e-05,
+      "loss": 1.199,
+      "step": 809
+    },
+    {
+      "epoch": 0.11147811725846409,
+      "grad_norm": 1.9119723618363742,
+      "learning_rate": 1.965385359315136e-05,
+      "loss": 1.1388,
+      "step": 810
+    },
+    {
+      "epoch": 0.11161574456372145,
+      "grad_norm": 1.8613221549664547,
+      "learning_rate": 1.9652690014253134e-05,
+      "loss": 1.1879,
+      "step": 811
+    },
+    {
+      "epoch": 0.11175337186897881,
+      "grad_norm": 1.6390392855243738,
+      "learning_rate": 1.965152451749754e-05,
+      "loss": 1.1766,
+      "step": 812
+    },
+    {
+      "epoch": 0.11189099917423617,
+      "grad_norm": 1.7462595359826287,
+      "learning_rate": 1.9650357103116134e-05,
+      "loss": 1.1712,
+      "step": 813
+    },
+    {
+      "epoch": 0.11202862647949353,
+      "grad_norm": 1.9498657499696128,
+      "learning_rate": 1.9649187771340873e-05,
+      "loss": 1.1652,
+      "step": 814
+    },
+    {
+      "epoch": 0.1121662537847509,
+      "grad_norm": 1.8671501162070752,
+      "learning_rate": 1.9648016522404087e-05,
+      "loss": 1.2681,
+      "step": 815
+    },
+    {
+      "epoch": 0.11230388109000826,
+      "grad_norm": 1.703099533861835,
+      "learning_rate": 1.9646843356538482e-05,
+      "loss": 1.2402,
+      "step": 816
+    },
+    {
+      "epoch": 0.11244150839526562,
+      "grad_norm": 1.707723108808178,
+      "learning_rate": 1.9645668273977154e-05,
+      "loss": 1.2,
+      "step": 817
+    },
+    {
+      "epoch": 0.11257913570052298,
+      "grad_norm": 1.8617622271902463,
+      "learning_rate": 1.9644491274953573e-05,
+      "loss": 1.1954,
+      "step": 818
+    },
+    {
+      "epoch": 0.11271676300578035,
+      "grad_norm": 1.9961433759016127,
+      "learning_rate": 1.9643312359701596e-05,
+      "loss": 1.2775,
+      "step": 819
+    },
+    {
+      "epoch": 0.11285439031103771,
+      "grad_norm": 1.8725230314491939,
+      "learning_rate": 1.9642131528455458e-05,
+      "loss": 1.2394,
+      "step": 820
+    },
+    {
+      "epoch": 0.11299201761629507,
+      "grad_norm": 1.8993790330750357,
+      "learning_rate": 1.9640948781449766e-05,
+      "loss": 1.2574,
+      "step": 821
+    },
+    {
+      "epoch": 0.11312964492155243,
+      "grad_norm": 1.61326478689415,
+      "learning_rate": 1.9639764118919525e-05,
+      "loss": 1.1587,
+      "step": 822
+    },
+    {
+      "epoch": 0.1132672722268098,
+      "grad_norm": 1.9097744955803853,
+      "learning_rate": 1.963857754110011e-05,
+      "loss": 1.2792,
+      "step": 823
+    },
+    {
+      "epoch": 0.11340489953206716,
+      "grad_norm": 1.8198049781596035,
+      "learning_rate": 1.963738904822727e-05,
+      "loss": 1.2117,
+      "step": 824
+    },
+    {
+      "epoch": 0.11354252683732452,
+      "grad_norm": 1.6698221562481035,
+      "learning_rate": 1.963619864053715e-05,
+      "loss": 1.2298,
+      "step": 825
+    },
+    {
+      "epoch": 0.11368015414258188,
+      "grad_norm": 2.22089543543673,
+      "learning_rate": 1.9635006318266264e-05,
+      "loss": 1.2443,
+      "step": 826
+    },
+    {
+      "epoch": 0.11381778144783924,
+      "grad_norm": 1.8790557310023606,
+      "learning_rate": 1.9633812081651514e-05,
+      "loss": 1.1758,
+      "step": 827
+    },
+    {
+      "epoch": 0.11395540875309662,
+      "grad_norm": 1.905489590669244,
+      "learning_rate": 1.9632615930930175e-05,
+      "loss": 1.1858,
+      "step": 828
+    },
+    {
+      "epoch": 0.11409303605835398,
+      "grad_norm": 1.8798053582950898,
+      "learning_rate": 1.9631417866339908e-05,
+      "loss": 1.2721,
+      "step": 829
+    },
+    {
+      "epoch": 0.11423066336361135,
+      "grad_norm": 1.7110812505620618,
+      "learning_rate": 1.9630217888118748e-05,
+      "loss": 1.257,
+      "step": 830
+    },
+    {
+      "epoch": 0.11436829066886871,
+      "grad_norm": 1.6684295565096017,
+      "learning_rate": 1.9629015996505117e-05,
+      "loss": 1.1245,
+      "step": 831
+    },
+    {
+      "epoch": 0.11450591797412607,
+      "grad_norm": 1.8124179377767673,
+      "learning_rate": 1.9627812191737815e-05,
+      "loss": 1.2686,
+      "step": 832
+    },
+    {
+      "epoch": 0.11464354527938343,
+      "grad_norm": 1.7700020488521024,
+      "learning_rate": 1.9626606474056022e-05,
+      "loss": 1.264,
+      "step": 833
+    },
+    {
+      "epoch": 0.1147811725846408,
+      "grad_norm": 1.6888144526293025,
+      "learning_rate": 1.9625398843699295e-05,
+      "loss": 1.2051,
+      "step": 834
+    },
+    {
+      "epoch": 0.11491879988989816,
+      "grad_norm": 1.785257425435577,
+      "learning_rate": 1.9624189300907574e-05,
+      "loss": 1.1827,
+      "step": 835
+    },
+    {
+      "epoch": 0.11505642719515552,
+      "grad_norm": 1.8808853577203786,
+      "learning_rate": 1.9622977845921185e-05,
+      "loss": 1.2914,
+      "step": 836
+    },
+    {
+      "epoch": 0.11519405450041288,
+      "grad_norm": 1.7027648780128342,
+      "learning_rate": 1.962176447898082e-05,
+      "loss": 1.1537,
+      "step": 837
+    },
+    {
+      "epoch": 0.11533168180567024,
+      "grad_norm": 1.8105159301384934,
+      "learning_rate": 1.9620549200327558e-05,
+      "loss": 1.1222,
+      "step": 838
+    },
+    {
+      "epoch": 0.1154693091109276,
+      "grad_norm": 1.8510744983305099,
+      "learning_rate": 1.9619332010202862e-05,
+      "loss": 1.1916,
+      "step": 839
+    },
+    {
+      "epoch": 0.11560693641618497,
+      "grad_norm": 1.9288719901878486,
+      "learning_rate": 1.961811290884857e-05,
+      "loss": 1.2782,
+      "step": 840
+    },
+    {
+      "epoch": 0.11574456372144233,
+      "grad_norm": 1.7036963394850868,
+      "learning_rate": 1.96168918965069e-05,
+      "loss": 1.142,
+      "step": 841
+    },
+    {
+      "epoch": 0.1158821910266997,
+      "grad_norm": 1.8908257807963516,
+      "learning_rate": 1.9615668973420453e-05,
+      "loss": 1.242,
+      "step": 842
+    },
+    {
+      "epoch": 0.11601981833195706,
+      "grad_norm": 1.6996979182194627,
+      "learning_rate": 1.9614444139832204e-05,
+      "loss": 1.1526,
+      "step": 843
+    },
+    {
+      "epoch": 0.11615744563721442,
+      "grad_norm": 1.81368215009564,
+      "learning_rate": 1.9613217395985508e-05,
+      "loss": 1.1692,
+      "step": 844
+    },
+    {
+      "epoch": 0.11629507294247178,
+      "grad_norm": 1.9160781015939727,
+      "learning_rate": 1.961198874212411e-05,
+      "loss": 1.1775,
+      "step": 845
+    },
+    {
+      "epoch": 0.11643270024772914,
+      "grad_norm": 1.9475045493186849,
+      "learning_rate": 1.9610758178492123e-05,
+      "loss": 1.2796,
+      "step": 846
+    },
+    {
+      "epoch": 0.11657032755298652,
+      "grad_norm": 1.668364768270674,
+      "learning_rate": 1.9609525705334047e-05,
+      "loss": 1.1427,
+      "step": 847
+    },
+    {
+      "epoch": 0.11670795485824388,
+      "grad_norm": 1.701056052059771,
+      "learning_rate": 1.9608291322894745e-05,
+      "loss": 1.1939,
+      "step": 848
+    },
+    {
+      "epoch": 0.11684558216350124,
+      "grad_norm": 1.6741979979033652,
+      "learning_rate": 1.9607055031419485e-05,
+      "loss": 1.1607,
+      "step": 849
+    },
+    {
+      "epoch": 0.1169832094687586,
+      "grad_norm": 1.8334524030645067,
+      "learning_rate": 1.9605816831153897e-05,
+      "loss": 1.2256,
+      "step": 850
+    },
+    {
+      "epoch": 0.11712083677401597,
+      "grad_norm": 1.764896912502412,
+      "learning_rate": 1.960457672234399e-05,
+      "loss": 1.2586,
+      "step": 851
+    },
+    {
+      "epoch": 0.11725846407927333,
+      "grad_norm": 1.7490870606132587,
+      "learning_rate": 1.9603334705236165e-05,
+      "loss": 1.3306,
+      "step": 852
+    },
+    {
+      "epoch": 0.11739609138453069,
+      "grad_norm": 1.7589331806171191,
+      "learning_rate": 1.9602090780077184e-05,
+      "loss": 1.2887,
+      "step": 853
+    },
+    {
+      "epoch": 0.11753371868978806,
+      "grad_norm": 1.7531140945576942,
+      "learning_rate": 1.960084494711421e-05,
+      "loss": 1.2094,
+      "step": 854
+    },
+    {
+      "epoch": 0.11767134599504542,
+      "grad_norm": 1.7381167836118683,
+      "learning_rate": 1.9599597206594764e-05,
+      "loss": 1.188,
+      "step": 855
+    },
+    {
+      "epoch": 0.11780897330030278,
+      "grad_norm": 1.662611953461017,
+      "learning_rate": 1.9598347558766755e-05,
+      "loss": 1.078,
+      "step": 856
+    },
+    {
+      "epoch": 0.11794660060556014,
+      "grad_norm": 1.7853197150648044,
+      "learning_rate": 1.9597096003878474e-05,
+      "loss": 1.1884,
+      "step": 857
+    },
+    {
+      "epoch": 0.1180842279108175,
+      "grad_norm": 1.812426785010667,
+      "learning_rate": 1.9595842542178588e-05,
+      "loss": 1.2952,
+      "step": 858
+    },
+    {
+      "epoch": 0.11822185521607487,
+      "grad_norm": 1.9771318816184522,
+      "learning_rate": 1.959458717391614e-05,
+      "loss": 1.1888,
+      "step": 859
+    },
+    {
+      "epoch": 0.11835948252133223,
+      "grad_norm": 1.9590285727958894,
+      "learning_rate": 1.9593329899340556e-05,
+      "loss": 1.1648,
+      "step": 860
+    },
+    {
+      "epoch": 0.11849710982658959,
+      "grad_norm": 1.7975882711701672,
+      "learning_rate": 1.9592070718701645e-05,
+      "loss": 1.1918,
+      "step": 861
+    },
+    {
+      "epoch": 0.11863473713184695,
+      "grad_norm": 1.7638369780117387,
+      "learning_rate": 1.9590809632249576e-05,
+      "loss": 1.199,
+      "step": 862
+    },
+    {
+      "epoch": 0.11877236443710432,
+      "grad_norm": 1.9692409783066018,
+      "learning_rate": 1.958954664023492e-05,
+      "loss": 1.2083,
+      "step": 863
+    },
+    {
+      "epoch": 0.11890999174236168,
+      "grad_norm": 1.9069685008249782,
+      "learning_rate": 1.9588281742908615e-05,
+      "loss": 1.2375,
+      "step": 864
+    },
+    {
+      "epoch": 0.11904761904761904,
+      "grad_norm": 1.7885446339451163,
+      "learning_rate": 1.9587014940521975e-05,
+      "loss": 1.1749,
+      "step": 865
+    },
+    {
+      "epoch": 0.11918524635287642,
+      "grad_norm": 2.0044570723063897,
+      "learning_rate": 1.9585746233326698e-05,
+      "loss": 1.1707,
+      "step": 866
+    },
+    {
+      "epoch": 0.11932287365813378,
+      "grad_norm": 2.055830830005792,
+      "learning_rate": 1.9584475621574857e-05,
+      "loss": 1.2536,
+      "step": 867
+    },
+    {
+      "epoch": 0.11946050096339114,
+      "grad_norm": 1.910676245488409,
+      "learning_rate": 1.958320310551891e-05,
+      "loss": 1.1936,
+      "step": 868
+    },
+    {
+      "epoch": 0.1195981282686485,
+      "grad_norm": 1.748345875820702,
+      "learning_rate": 1.9581928685411685e-05,
+      "loss": 1.2188,
+      "step": 869
+    },
+    {
+      "epoch": 0.11973575557390587,
+      "grad_norm": 1.701262009569421,
+      "learning_rate": 1.958065236150639e-05,
+      "loss": 1.2505,
+      "step": 870
+    },
+    {
+      "epoch": 0.11987338287916323,
+      "grad_norm": 1.720350519101937,
+      "learning_rate": 1.957937413405662e-05,
+      "loss": 1.1504,
+      "step": 871
+    },
+    {
+      "epoch": 0.12001101018442059,
+      "grad_norm": 1.8301356647355744,
+      "learning_rate": 1.9578094003316336e-05,
+      "loss": 1.3008,
+      "step": 872
+    },
+    {
+      "epoch": 0.12014863748967795,
+      "grad_norm": 1.74662350534866,
+      "learning_rate": 1.9576811969539884e-05,
+      "loss": 1.1875,
+      "step": 873
+    },
+    {
+      "epoch": 0.12028626479493532,
+      "grad_norm": 1.8750771111823648,
+      "learning_rate": 1.957552803298198e-05,
+      "loss": 1.1388,
+      "step": 874
+    },
+    {
+      "epoch": 0.12042389210019268,
+      "grad_norm": 1.7952243704730921,
+      "learning_rate": 1.9574242193897735e-05,
+      "loss": 1.1582,
+      "step": 875
+    },
+    {
+      "epoch": 0.12056151940545004,
+      "grad_norm": 1.5682299961618378,
+      "learning_rate": 1.9572954452542627e-05,
+      "loss": 1.1697,
+      "step": 876
+    },
+    {
+      "epoch": 0.1206991467107074,
+      "grad_norm": 1.7516962879332203,
+      "learning_rate": 1.9571664809172505e-05,
+      "loss": 1.1978,
+      "step": 877
+    },
+    {
+      "epoch": 0.12083677401596477,
+      "grad_norm": 1.852355924708167,
+      "learning_rate": 1.9570373264043606e-05,
+      "loss": 1.2418,
+      "step": 878
+    },
+    {
+      "epoch": 0.12097440132122213,
+      "grad_norm": 1.7535538721339785,
+      "learning_rate": 1.9569079817412545e-05,
+      "loss": 1.2343,
+      "step": 879
+    },
+    {
+      "epoch": 0.12111202862647949,
+      "grad_norm": 1.7169762275226343,
+      "learning_rate": 1.956778446953631e-05,
+      "loss": 1.2458,
+      "step": 880
+    },
+    {
+      "epoch": 0.12124965593173685,
+      "grad_norm": 1.786448507139269,
+      "learning_rate": 1.956648722067227e-05,
+      "loss": 1.2379,
+      "step": 881
+    },
+    {
+      "epoch": 0.12138728323699421,
+      "grad_norm": 1.8872721138716886,
+      "learning_rate": 1.956518807107817e-05,
+      "loss": 1.2703,
+      "step": 882
+    },
+    {
+      "epoch": 0.12152491054225158,
+      "grad_norm": 1.8933606860245056,
+      "learning_rate": 1.9563887021012135e-05,
+      "loss": 1.2002,
+      "step": 883
+    },
+    {
+      "epoch": 0.12166253784750894,
+      "grad_norm": 1.8799858929558066,
+      "learning_rate": 1.9562584070732663e-05,
+      "loss": 1.2304,
+      "step": 884
+    },
+    {
+      "epoch": 0.12180016515276632,
+      "grad_norm": 1.9863110537050936,
+      "learning_rate": 1.9561279220498632e-05,
+      "loss": 1.273,
+      "step": 885
+    },
+    {
+      "epoch": 0.12193779245802368,
+      "grad_norm": 1.4976881244914186,
+      "learning_rate": 1.9559972470569302e-05,
+      "loss": 1.1797,
+      "step": 886
+    },
+    {
+      "epoch": 0.12207541976328104,
+      "grad_norm": 1.6519086895173287,
+      "learning_rate": 1.9558663821204302e-05,
+      "loss": 1.1606,
+      "step": 887
+    },
+    {
+      "epoch": 0.1222130470685384,
+      "grad_norm": 1.6979986986458808,
+      "learning_rate": 1.9557353272663648e-05,
+      "loss": 1.2686,
+      "step": 888
+    },
+    {
+      "epoch": 0.12235067437379576,
+      "grad_norm": 1.8682733886495013,
+      "learning_rate": 1.955604082520772e-05,
+      "loss": 1.2198,
+      "step": 889
+    },
+    {
+      "epoch": 0.12248830167905313,
+      "grad_norm": 1.5445401032171855,
+      "learning_rate": 1.955472647909729e-05,
+      "loss": 1.133,
+      "step": 890
+    },
+    {
+      "epoch": 0.12262592898431049,
+      "grad_norm": 1.7580662185546172,
+      "learning_rate": 1.95534102345935e-05,
+      "loss": 1.1769,
+      "step": 891
+    },
+    {
+      "epoch": 0.12276355628956785,
+      "grad_norm": 1.7562581012535048,
+      "learning_rate": 1.9552092091957864e-05,
+      "loss": 1.1413,
+      "step": 892
+    },
+    {
+      "epoch": 0.12290118359482521,
+      "grad_norm": 1.7279215983630576,
+      "learning_rate": 1.9550772051452287e-05,
+      "loss": 1.3172,
+      "step": 893
+    },
+    {
+      "epoch": 0.12303881090008258,
+      "grad_norm": 1.7076100369758198,
+      "learning_rate": 1.954945011333904e-05,
+      "loss": 1.1432,
+      "step": 894
+    },
+    {
+      "epoch": 0.12317643820533994,
+      "grad_norm": 1.673233751341388,
+      "learning_rate": 1.954812627788077e-05,
+      "loss": 1.1622,
+      "step": 895
+    },
+    {
+      "epoch": 0.1233140655105973,
+      "grad_norm": 1.7419266562126345,
+      "learning_rate": 1.954680054534051e-05,
+      "loss": 1.1963,
+      "step": 896
+    },
+    {
+      "epoch": 0.12345169281585466,
+      "grad_norm": 1.8171513358468807,
+      "learning_rate": 1.9545472915981664e-05,
+      "loss": 1.2056,
+      "step": 897
+    },
+    {
+      "epoch": 0.12358932012111203,
+      "grad_norm": 1.6679337355766228,
+      "learning_rate": 1.9544143390068012e-05,
+      "loss": 1.1497,
+      "step": 898
+    },
+    {
+      "epoch": 0.12372694742636939,
+      "grad_norm": 1.94580127851496,
+      "learning_rate": 1.9542811967863714e-05,
+      "loss": 1.2018,
+      "step": 899
+    },
+    {
+      "epoch": 0.12386457473162675,
+      "grad_norm": 1.8241949534980415,
+      "learning_rate": 1.9541478649633304e-05,
+      "loss": 1.2464,
+      "step": 900
+    },
+    {
+      "epoch": 0.12400220203688411,
+      "grad_norm": 1.936540773230824,
+      "learning_rate": 1.9540143435641693e-05,
+      "loss": 1.2172,
+      "step": 901
+    },
+    {
+      "epoch": 0.12413982934214148,
+      "grad_norm": 1.876067854726912,
+      "learning_rate": 1.9538806326154173e-05,
+      "loss": 1.1921,
+      "step": 902
+    },
+    {
+      "epoch": 0.12427745664739884,
+      "grad_norm": 1.7665301315509787,
+      "learning_rate": 1.9537467321436413e-05,
+      "loss": 1.1737,
+      "step": 903
+    },
+    {
+      "epoch": 0.12441508395265621,
+      "grad_norm": 1.8162055564067796,
+      "learning_rate": 1.9536126421754444e-05,
+      "loss": 1.0952,
+      "step": 904
+    },
+    {
+      "epoch": 0.12455271125791358,
+      "grad_norm": 1.7296780190435752,
+      "learning_rate": 1.9534783627374693e-05,
+      "loss": 1.1698,
+      "step": 905
+    },
+    {
+      "epoch": 0.12469033856317094,
+      "grad_norm": 1.671988218076766,
+      "learning_rate": 1.953343893856395e-05,
+      "loss": 1.2054,
+      "step": 906
+    },
+    {
+      "epoch": 0.1248279658684283,
+      "grad_norm": 1.9083236270954649,
+      "learning_rate": 1.953209235558939e-05,
+      "loss": 1.1741,
+      "step": 907
+    },
+    {
+      "epoch": 0.12496559317368566,
+      "grad_norm": 1.7519943695966007,
+      "learning_rate": 1.9530743878718556e-05,
+      "loss": 1.2594,
+      "step": 908
+    },
+    {
+      "epoch": 0.125103220478943,
+      "grad_norm": 1.6984133546060982,
+      "learning_rate": 1.952939350821938e-05,
+      "loss": 1.2143,
+      "step": 909
+    },
+    {
+      "epoch": 0.1252408477842004,
+      "grad_norm": 1.634057158730061,
+      "learning_rate": 1.9528041244360154e-05,
+      "loss": 1.1355,
+      "step": 910
+    },
+    {
+      "epoch": 0.12537847508945774,
+      "grad_norm": 1.9614450274222952,
+      "learning_rate": 1.952668708740956e-05,
+      "loss": 1.2045,
+      "step": 911
+    },
+    {
+      "epoch": 0.1255161023947151,
+      "grad_norm": 1.6451820967591815,
+      "learning_rate": 1.9525331037636644e-05,
+      "loss": 1.226,
+      "step": 912
+    },
+    {
+      "epoch": 0.12565372969997246,
+      "grad_norm": 1.7079880857151484,
+      "learning_rate": 1.952397309531084e-05,
+      "loss": 1.2498,
+      "step": 913
+    },
+    {
+      "epoch": 0.12579135700522984,
+      "grad_norm": 1.898090558307993,
+      "learning_rate": 1.9522613260701953e-05,
+      "loss": 1.2303,
+      "step": 914
+    },
+    {
+      "epoch": 0.1259289843104872,
+      "grad_norm": 1.9464490987711276,
+      "learning_rate": 1.952125153408016e-05,
+      "loss": 1.2242,
+      "step": 915
+    },
+    {
+      "epoch": 0.12606661161574456,
+      "grad_norm": 1.7066171121219222,
+      "learning_rate": 1.9519887915716017e-05,
+      "loss": 1.1619,
+      "step": 916
+    },
+    {
+      "epoch": 0.12620423892100194,
+      "grad_norm": 1.867252243384439,
+      "learning_rate": 1.951852240588046e-05,
+      "loss": 1.2599,
+      "step": 917
+    },
+    {
+      "epoch": 0.1263418662262593,
+      "grad_norm": 1.6361455203299031,
+      "learning_rate": 1.9517155004844795e-05,
+      "loss": 1.155,
+      "step": 918
+    },
+    {
+      "epoch": 0.12647949353151666,
+      "grad_norm": 1.795034711430814,
+      "learning_rate": 1.951578571288071e-05,
+      "loss": 1.1376,
+      "step": 919
+    },
+    {
+      "epoch": 0.126617120836774,
+      "grad_norm": 1.7023381840888188,
+      "learning_rate": 1.9514414530260256e-05,
+      "loss": 1.2517,
+      "step": 920
+    },
+    {
+      "epoch": 0.1267547481420314,
+      "grad_norm": 2.0873406467605684,
+      "learning_rate": 1.9513041457255873e-05,
+      "loss": 1.2014,
+      "step": 921
+    },
+    {
+      "epoch": 0.12689237544728874,
+      "grad_norm": 1.7184294618811753,
+      "learning_rate": 1.9511666494140375e-05,
+      "loss": 1.1628,
+      "step": 922
+    },
+    {
+      "epoch": 0.1270300027525461,
+      "grad_norm": 1.7379198521780002,
+      "learning_rate": 1.9510289641186945e-05,
+      "loss": 1.1927,
+      "step": 923
+    },
+    {
+      "epoch": 0.12716763005780346,
+      "grad_norm": 1.6348611550642247,
+      "learning_rate": 1.950891089866914e-05,
+      "loss": 1.126,
+      "step": 924
+    },
+    {
+      "epoch": 0.12730525736306084,
+      "grad_norm": 2.001623064603782,
+      "learning_rate": 1.9507530266860906e-05,
+      "loss": 1.2412,
+      "step": 925
+    },
+    {
+      "epoch": 0.12744288466831818,
+      "grad_norm": 1.7738383139429934,
+      "learning_rate": 1.950614774603655e-05,
+      "loss": 1.2028,
+      "step": 926
+    },
+    {
+      "epoch": 0.12758051197357556,
+      "grad_norm": 1.933072542887908,
+      "learning_rate": 1.9504763336470766e-05,
+      "loss": 1.2251,
+      "step": 927
+    },
+    {
+      "epoch": 0.1277181392788329,
+      "grad_norm": 1.7999483387018915,
+      "learning_rate": 1.950337703843861e-05,
+      "loss": 1.2151,
+      "step": 928
+    },
+    {
+      "epoch": 0.12785576658409029,
+      "grad_norm": 1.6487682772734311,
+      "learning_rate": 1.950198885221552e-05,
+      "loss": 1.1147,
+      "step": 929
+    },
+    {
+      "epoch": 0.12799339388934763,
+      "grad_norm": 1.8730587695500056,
+      "learning_rate": 1.950059877807732e-05,
+      "loss": 1.1684,
+      "step": 930
+    },
+    {
+      "epoch": 0.128131021194605,
+      "grad_norm": 1.8528040632627034,
+      "learning_rate": 1.949920681630019e-05,
+      "loss": 1.2139,
+      "step": 931
+    },
+    {
+      "epoch": 0.12826864849986236,
+      "grad_norm": 1.6732805206115529,
+      "learning_rate": 1.9497812967160692e-05,
+      "loss": 1.2013,
+      "step": 932
+    },
+    {
+      "epoch": 0.12840627580511974,
+      "grad_norm": 1.8608366539603611,
+      "learning_rate": 1.949641723093577e-05,
+      "loss": 1.268,
+      "step": 933
+    },
+    {
+      "epoch": 0.1285439031103771,
+      "grad_norm": 1.8669876562808079,
+      "learning_rate": 1.949501960790274e-05,
+      "loss": 1.1791,
+      "step": 934
+    },
+    {
+      "epoch": 0.12868153041563446,
+      "grad_norm": 1.7427210829656057,
+      "learning_rate": 1.949362009833928e-05,
+      "loss": 1.2205,
+      "step": 935
+    },
+    {
+      "epoch": 0.12881915772089184,
+      "grad_norm": 1.7347219291429874,
+      "learning_rate": 1.9492218702523464e-05,
+      "loss": 1.1405,
+      "step": 936
+    },
+    {
+      "epoch": 0.12895678502614918,
+      "grad_norm": 1.8575892328881445,
+      "learning_rate": 1.9490815420733724e-05,
+      "loss": 1.2235,
+      "step": 937
+    },
+    {
+      "epoch": 0.12909441233140656,
+      "grad_norm": 1.8638912828402099,
+      "learning_rate": 1.9489410253248876e-05,
+      "loss": 1.0953,
+      "step": 938
+    },
+    {
+      "epoch": 0.1292320396366639,
+      "grad_norm": 1.775823391274239,
+      "learning_rate": 1.9488003200348105e-05,
+      "loss": 1.1953,
+      "step": 939
+    },
+    {
+      "epoch": 0.12936966694192129,
+      "grad_norm": 1.6915162545378302,
+      "learning_rate": 1.9486594262310977e-05,
+      "loss": 1.1773,
+      "step": 940
+    },
+    {
+      "epoch": 0.12950729424717863,
+      "grad_norm": 1.6747112639953017,
+      "learning_rate": 1.9485183439417428e-05,
+      "loss": 1.2225,
+      "step": 941
+    },
+    {
+      "epoch": 0.129644921552436,
+      "grad_norm": 2.0704571179317313,
+      "learning_rate": 1.9483770731947766e-05,
+      "loss": 1.2667,
+      "step": 942
+    },
+    {
+      "epoch": 0.12978254885769336,
+      "grad_norm": 1.7495707875070545,
+      "learning_rate": 1.948235614018268e-05,
+      "loss": 1.1347,
+      "step": 943
+    },
+    {
+      "epoch": 0.12992017616295073,
+      "grad_norm": 1.5889053769632995,
+      "learning_rate": 1.9480939664403223e-05,
+      "loss": 1.14,
+      "step": 944
+    },
+    {
+      "epoch": 0.13005780346820808,
+      "grad_norm": 1.999426728163562,
+      "learning_rate": 1.9479521304890837e-05,
+      "loss": 1.2003,
+      "step": 945
+    },
+    {
+      "epoch": 0.13019543077346546,
+      "grad_norm": 1.7824427710009412,
+      "learning_rate": 1.947810106192733e-05,
+      "loss": 1.1148,
+      "step": 946
+    },
+    {
+      "epoch": 0.1303330580787228,
+      "grad_norm": 2.0298392691296203,
+      "learning_rate": 1.947667893579488e-05,
+      "loss": 1.2918,
+      "step": 947
+    },
+    {
+      "epoch": 0.13047068538398018,
+      "grad_norm": 1.7003766456052392,
+      "learning_rate": 1.947525492677605e-05,
+      "loss": 1.1613,
+      "step": 948
+    },
+    {
+      "epoch": 0.13060831268923753,
+      "grad_norm": 1.7999058410871331,
+      "learning_rate": 1.9473829035153767e-05,
+      "loss": 1.1968,
+      "step": 949
+    },
+    {
+      "epoch": 0.1307459399944949,
+      "grad_norm": 2.076816770750908,
+      "learning_rate": 1.9472401261211337e-05,
+      "loss": 1.2655,
+      "step": 950
+    },
+    {
+      "epoch": 0.13088356729975226,
+      "grad_norm": 1.8597615323640324,
+      "learning_rate": 1.9470971605232442e-05,
+      "loss": 1.2028,
+      "step": 951
+    },
+    {
+      "epoch": 0.13102119460500963,
+      "grad_norm": 1.7375982906603942,
+      "learning_rate": 1.946954006750113e-05,
+      "loss": 1.1663,
+      "step": 952
+    },
+    {
+      "epoch": 0.131158821910267,
+      "grad_norm": 1.8234967342291553,
+      "learning_rate": 1.946810664830183e-05,
+      "loss": 1.2284,
+      "step": 953
+    },
+    {
+      "epoch": 0.13129644921552436,
+      "grad_norm": 1.9554474505837507,
+      "learning_rate": 1.9466671347919344e-05,
+      "loss": 1.2314,
+      "step": 954
+    },
+    {
+      "epoch": 0.13143407652078173,
+      "grad_norm": 1.9156508013180402,
+      "learning_rate": 1.9465234166638846e-05,
+      "loss": 1.177,
+      "step": 955
+    },
+    {
+      "epoch": 0.13157170382603908,
+      "grad_norm": 1.7085134496264502,
+      "learning_rate": 1.9463795104745884e-05,
+      "loss": 1.152,
+      "step": 956
+    },
+    {
+      "epoch": 0.13170933113129646,
+      "grad_norm": 1.6434138197626693,
+      "learning_rate": 1.946235416252638e-05,
+      "loss": 1.1768,
+      "step": 957
+    },
+    {
+      "epoch": 0.1318469584365538,
+      "grad_norm": 1.8287900367511738,
+      "learning_rate": 1.946091134026663e-05,
+      "loss": 1.0793,
+      "step": 958
+    },
+    {
+      "epoch": 0.13198458574181118,
+      "grad_norm": 1.7780833774465106,
+      "learning_rate": 1.94594666382533e-05,
+      "loss": 1.1682,
+      "step": 959
+    },
+    {
+      "epoch": 0.13212221304706853,
+      "grad_norm": 1.7610874079193026,
+      "learning_rate": 1.9458020056773437e-05,
+      "loss": 1.1199,
+      "step": 960
+    },
+    {
+      "epoch": 0.1322598403523259,
+      "grad_norm": 1.8662459632903046,
+      "learning_rate": 1.9456571596114453e-05,
+      "loss": 1.1771,
+      "step": 961
+    },
+    {
+      "epoch": 0.13239746765758326,
+      "grad_norm": 1.5986172070234737,
+      "learning_rate": 1.9455121256564142e-05,
+      "loss": 1.1181,
+      "step": 962
+    },
+    {
+      "epoch": 0.13253509496284063,
+      "grad_norm": 1.5864292859221902,
+      "learning_rate": 1.9453669038410662e-05,
+      "loss": 1.1469,
+      "step": 963
+    },
+    {
+      "epoch": 0.13267272226809798,
+      "grad_norm": 1.6643144323686463,
+      "learning_rate": 1.945221494194255e-05,
+      "loss": 1.1307,
+      "step": 964
+    },
+    {
+      "epoch": 0.13281034957335536,
+      "grad_norm": 1.710399941688845,
+      "learning_rate": 1.945075896744872e-05,
+      "loss": 1.2325,
+      "step": 965
+    },
+    {
+      "epoch": 0.1329479768786127,
+      "grad_norm": 1.8138100319122905,
+      "learning_rate": 1.9449301115218446e-05,
+      "loss": 1.17,
+      "step": 966
+    },
+    {
+      "epoch": 0.13308560418387008,
+      "grad_norm": 1.6987073161560877,
+      "learning_rate": 1.9447841385541387e-05,
+      "loss": 1.199,
+      "step": 967
+    },
+    {
+      "epoch": 0.13322323148912743,
+      "grad_norm": 1.8201087419553204,
+      "learning_rate": 1.9446379778707573e-05,
+      "loss": 1.1858,
+      "step": 968
+    },
+    {
+      "epoch": 0.1333608587943848,
+      "grad_norm": 1.9144000429448347,
+      "learning_rate": 1.9444916295007404e-05,
+      "loss": 1.2228,
+      "step": 969
+    },
+    {
+      "epoch": 0.13349848609964216,
+      "grad_norm": 1.7404283608545148,
+      "learning_rate": 1.9443450934731653e-05,
+      "loss": 1.1596,
+      "step": 970
+    },
+    {
+      "epoch": 0.13363611340489953,
+      "grad_norm": 1.7346457983269885,
+      "learning_rate": 1.944198369817147e-05,
+      "loss": 1.1319,
+      "step": 971
+    },
+    {
+      "epoch": 0.1337737407101569,
+      "grad_norm": 1.831135867258422,
+      "learning_rate": 1.944051458561837e-05,
+      "loss": 1.1445,
+      "step": 972
+    },
+    {
+      "epoch": 0.13391136801541426,
+      "grad_norm": 1.5902069722950272,
+      "learning_rate": 1.9439043597364252e-05,
+      "loss": 1.1545,
+      "step": 973
+    },
+    {
+      "epoch": 0.13404899532067163,
+      "grad_norm": 1.744437434324077,
+      "learning_rate": 1.9437570733701375e-05,
+      "loss": 1.2082,
+      "step": 974
+    },
+    {
+      "epoch": 0.13418662262592898,
+      "grad_norm": 1.9540738678049723,
+      "learning_rate": 1.943609599492238e-05,
+      "loss": 1.2003,
+      "step": 975
+    },
+    {
+      "epoch": 0.13432424993118636,
+      "grad_norm": 1.7992368627666153,
+      "learning_rate": 1.9434619381320278e-05,
+      "loss": 1.1986,
+      "step": 976
+    },
+    {
+      "epoch": 0.1344618772364437,
+      "grad_norm": 1.8787671340662577,
+      "learning_rate": 1.9433140893188448e-05,
+      "loss": 1.1974,
+      "step": 977
+    },
+    {
+      "epoch": 0.13459950454170108,
+      "grad_norm": 1.942763566283457,
+      "learning_rate": 1.943166053082065e-05,
+      "loss": 1.1802,
+      "step": 978
+    },
+    {
+      "epoch": 0.13473713184695843,
+      "grad_norm": 1.77450404026138,
+      "learning_rate": 1.943017829451101e-05,
+      "loss": 1.2281,
+      "step": 979
+    },
+    {
+      "epoch": 0.1348747591522158,
+      "grad_norm": 1.984280537582193,
+      "learning_rate": 1.9428694184554028e-05,
+      "loss": 1.1355,
+      "step": 980
+    },
+    {
+      "epoch": 0.13501238645747315,
+      "grad_norm": 1.7601278338710733,
+      "learning_rate": 1.9427208201244574e-05,
+      "loss": 1.1676,
+      "step": 981
+    },
+    {
+      "epoch": 0.13515001376273053,
+      "grad_norm": 1.797922221760212,
+      "learning_rate": 1.9425720344877898e-05,
+      "loss": 1.1582,
+      "step": 982
+    },
+    {
+      "epoch": 0.13528764106798788,
+      "grad_norm": 1.7285323922114277,
+      "learning_rate": 1.942423061574961e-05,
+      "loss": 1.1811,
+      "step": 983
+    },
+    {
+      "epoch": 0.13542526837324526,
+      "grad_norm": 2.08796500588785,
+      "learning_rate": 1.9422739014155706e-05,
+      "loss": 1.1933,
+      "step": 984
+    },
+    {
+      "epoch": 0.1355628956785026,
+      "grad_norm": 1.9537189972531754,
+      "learning_rate": 1.9421245540392542e-05,
+      "loss": 1.2037,
+      "step": 985
+    },
+    {
+      "epoch": 0.13570052298375998,
+      "grad_norm": 1.596405297410078,
+      "learning_rate": 1.9419750194756852e-05,
+      "loss": 1.21,
+      "step": 986
+    },
+    {
+      "epoch": 0.13583815028901733,
+      "grad_norm": 1.8410900344787045,
+      "learning_rate": 1.941825297754574e-05,
+      "loss": 1.256,
+      "step": 987
+    },
+    {
+      "epoch": 0.1359757775942747,
+      "grad_norm": 1.7058783634179229,
+      "learning_rate": 1.9416753889056683e-05,
+      "loss": 1.1779,
+      "step": 988
+    },
+    {
+      "epoch": 0.13611340489953205,
+      "grad_norm": 1.5744881145235696,
+      "learning_rate": 1.941525292958753e-05,
+      "loss": 1.08,
+      "step": 989
+    },
+    {
+      "epoch": 0.13625103220478943,
+      "grad_norm": 1.8437728403828717,
+      "learning_rate": 1.94137500994365e-05,
+      "loss": 1.2129,
+      "step": 990
+    },
+    {
+      "epoch": 0.1363886595100468,
+      "grad_norm": 1.7826378706870099,
+      "learning_rate": 1.9412245398902187e-05,
+      "loss": 1.1938,
+      "step": 991
+    },
+    {
+      "epoch": 0.13652628681530415,
+      "grad_norm": 1.6310487316330167,
+      "learning_rate": 1.9410738828283552e-05,
+      "loss": 1.137,
+      "step": 992
+    },
+    {
+      "epoch": 0.13666391412056153,
+      "grad_norm": 1.8128725023210575,
+      "learning_rate": 1.940923038787993e-05,
+      "loss": 1.084,
+      "step": 993
+    },
+    {
+      "epoch": 0.13680154142581888,
+      "grad_norm": 1.7791035332214813,
+      "learning_rate": 1.940772007799103e-05,
+      "loss": 1.1933,
+      "step": 994
+    },
+    {
+      "epoch": 0.13693916873107626,
+      "grad_norm": 1.6534117925347271,
+      "learning_rate": 1.9406207898916927e-05,
+      "loss": 1.1888,
+      "step": 995
+    },
+    {
+      "epoch": 0.1370767960363336,
+      "grad_norm": 1.6779203161696332,
+      "learning_rate": 1.9404693850958072e-05,
+      "loss": 1.155,
+      "step": 996
+    },
+    {
+      "epoch": 0.13721442334159098,
+      "grad_norm": 1.8769608977013257,
+      "learning_rate": 1.9403177934415286e-05,
+      "loss": 1.287,
+      "step": 997
+    },
+    {
+      "epoch": 0.13735205064684833,
+      "grad_norm": 2.083083583318729,
+      "learning_rate": 1.9401660149589757e-05,
+      "loss": 1.1051,
+      "step": 998
+    },
+    {
+      "epoch": 0.1374896779521057,
+      "grad_norm": 1.7496090381720109,
+      "learning_rate": 1.9400140496783052e-05,
+      "loss": 1.2172,
+      "step": 999
+    },
+    {
+      "epoch": 0.13762730525736305,
+      "grad_norm": 1.8305695694032829,
+      "learning_rate": 1.9398618976297104e-05,
+      "loss": 1.2164,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13776493256262043,
+      "grad_norm": 1.9222872600593532,
+      "learning_rate": 1.939709558843422e-05,
+      "loss": 1.1451,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13790255986787778,
+      "grad_norm": 1.7281968622028878,
+      "learning_rate": 1.9395570333497074e-05,
+      "loss": 1.2143,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13804018717313515,
+      "grad_norm": 1.6635202909868703,
+      "learning_rate": 1.9394043211788715e-05,
+      "loss": 1.1511,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1381778144783925,
+      "grad_norm": 1.7759548107742285,
+      "learning_rate": 1.939251422361256e-05,
+      "loss": 1.214,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13831544178364988,
+      "grad_norm": 1.7051114919664374,
+      "learning_rate": 1.9390983369272397e-05,
+      "loss": 1.2002,
+      "step": 1005
+    },
+    {
+      "epoch": 0.13845306908890723,
+      "grad_norm": 1.9916929530575822,
+      "learning_rate": 1.9389450649072387e-05,
+      "loss": 1.1544,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1385906963941646,
+      "grad_norm": 1.7018306325073047,
+      "learning_rate": 1.938791606331706e-05,
+      "loss": 1.128,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13872832369942195,
+      "grad_norm": 1.7620122935853855,
+      "learning_rate": 1.9386379612311322e-05,
+      "loss": 1.2104,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13886595100467933,
+      "grad_norm": 1.758463166097322,
+      "learning_rate": 1.9384841296360436e-05,
+      "loss": 1.198,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1390035783099367,
+      "grad_norm": 1.7035755842904703,
+      "learning_rate": 1.938330111577005e-05,
+      "loss": 1.137,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13914120561519405,
+      "grad_norm": 1.6180985550871507,
+      "learning_rate": 1.9381759070846177e-05,
+      "loss": 1.1109,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13927883292045143,
+      "grad_norm": 1.8239650085082386,
+      "learning_rate": 1.93802151618952e-05,
+      "loss": 1.1805,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13941646022570878,
+      "grad_norm": 1.717971985154755,
+      "learning_rate": 1.9378669389223867e-05,
+      "loss": 1.1397,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13955408753096615,
+      "grad_norm": 1.8742652820598187,
+      "learning_rate": 1.937712175313931e-05,
+      "loss": 1.0707,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1396917148362235,
+      "grad_norm": 1.716158338179538,
+      "learning_rate": 1.937557225394902e-05,
+      "loss": 1.1402,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13982934214148088,
+      "grad_norm": 1.8125418435452227,
+      "learning_rate": 1.9374020891960858e-05,
+      "loss": 1.1629,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13996696944673823,
+      "grad_norm": 1.830710050842697,
+      "learning_rate": 1.9372467667483065e-05,
+      "loss": 1.2176,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1401045967519956,
+      "grad_norm": 1.782669261878356,
+      "learning_rate": 1.9370912580824238e-05,
+      "loss": 1.2282,
+      "step": 1018
+    },
+    {
+      "epoch": 0.14024222405725295,
+      "grad_norm": 1.7579993267498053,
+      "learning_rate": 1.9369355632293352e-05,
+      "loss": 1.1382,
+      "step": 1019
+    },
+    {
+      "epoch": 0.14037985136251033,
+      "grad_norm": 1.9767420983376895,
+      "learning_rate": 1.936779682219976e-05,
+      "loss": 1.217,
+      "step": 1020
+    },
+    {
+      "epoch": 0.14051747866776768,
+      "grad_norm": 1.7926710583251622,
+      "learning_rate": 1.936623615085317e-05,
+      "loss": 1.1541,
+      "step": 1021
+    },
+    {
+      "epoch": 0.14065510597302505,
+      "grad_norm": 2.0166884806822205,
+      "learning_rate": 1.9364673618563666e-05,
+      "loss": 1.3269,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1407927332782824,
+      "grad_norm": 2.251048067645687,
+      "learning_rate": 1.9363109225641702e-05,
+      "loss": 1.2945,
+      "step": 1023
+    },
+    {
+      "epoch": 0.14093036058353978,
+      "grad_norm": 1.7266424840449794,
+      "learning_rate": 1.93615429723981e-05,
+      "loss": 1.2156,
+      "step": 1024
+    },
+    {
+      "epoch": 0.14106798788879713,
+      "grad_norm": 1.7653596668641565,
+      "learning_rate": 1.9359974859144058e-05,
+      "loss": 1.2482,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1412056151940545,
+      "grad_norm": 1.6493715042919326,
+      "learning_rate": 1.9358404886191133e-05,
+      "loss": 1.1414,
+      "step": 1026
+    },
+    {
+      "epoch": 0.14134324249931185,
+      "grad_norm": 1.63504392731604,
+      "learning_rate": 1.9356833053851262e-05,
+      "loss": 1.225,
+      "step": 1027
+    },
+    {
+      "epoch": 0.14148086980456923,
+      "grad_norm": 1.7665428026053909,
+      "learning_rate": 1.9355259362436743e-05,
+      "loss": 1.1353,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1416184971098266,
+      "grad_norm": 1.6125876395540197,
+      "learning_rate": 1.935368381226025e-05,
+      "loss": 1.2844,
+      "step": 1029
+    },
+    {
+      "epoch": 0.14175612441508395,
+      "grad_norm": 1.788981634479404,
+      "learning_rate": 1.935210640363482e-05,
+      "loss": 1.2099,
+      "step": 1030
+    },
+    {
+      "epoch": 0.14189375172034133,
+      "grad_norm": 1.726604615163737,
+      "learning_rate": 1.935052713687386e-05,
+      "loss": 1.1622,
+      "step": 1031
+    },
+    {
+      "epoch": 0.14203137902559868,
+      "grad_norm": 1.743455890007309,
+      "learning_rate": 1.934894601229116e-05,
+      "loss": 1.1364,
+      "step": 1032
+    },
+    {
+      "epoch": 0.14216900633085605,
+      "grad_norm": 1.9621735073421336,
+      "learning_rate": 1.9347363030200857e-05,
+      "loss": 1.1423,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1423066336361134,
+      "grad_norm": 1.758988591245746,
+      "learning_rate": 1.934577819091747e-05,
+      "loss": 1.1739,
+      "step": 1034
+    },
+    {
+      "epoch": 0.14244426094137078,
+      "grad_norm": 1.751718249895641,
+      "learning_rate": 1.9344191494755887e-05,
+      "loss": 1.2405,
+      "step": 1035
+    },
+    {
+      "epoch": 0.14258188824662812,
+      "grad_norm": 1.8790472774464573,
+      "learning_rate": 1.9342602942031364e-05,
+      "loss": 1.1736,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1427195155518855,
+      "grad_norm": 1.8532339534859714,
+      "learning_rate": 1.934101253305952e-05,
+      "loss": 1.2269,
+      "step": 1037
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.732238789058539,
+      "learning_rate": 1.933942026815635e-05,
+      "loss": 1.1781,
+      "step": 1038
+    },
+    {
+      "epoch": 0.14299477016240023,
+      "grad_norm": 1.9106705025407693,
+      "learning_rate": 1.933782614763822e-05,
+      "loss": 1.3124,
+      "step": 1039
+    },
+    {
+      "epoch": 0.14313239746765757,
+      "grad_norm": 1.7809880613524174,
+      "learning_rate": 1.933623017182185e-05,
+      "loss": 1.2129,
+      "step": 1040
+    },
+    {
+      "epoch": 0.14327002477291495,
+      "grad_norm": 2.293840668653045,
+      "learning_rate": 1.9334632341024352e-05,
+      "loss": 1.2869,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1434076520781723,
+      "grad_norm": 1.8962266658647235,
+      "learning_rate": 1.9333032655563176e-05,
+      "loss": 1.1879,
+      "step": 1042
+    },
+    {
+      "epoch": 0.14354527938342967,
+      "grad_norm": 1.7333168283422837,
+      "learning_rate": 1.933143111575617e-05,
+      "loss": 1.2069,
+      "step": 1043
+    },
+    {
+      "epoch": 0.14368290668868702,
+      "grad_norm": 1.7512778750804503,
+      "learning_rate": 1.9329827721921535e-05,
+      "loss": 1.0966,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1438205339939444,
+      "grad_norm": 1.7731853258870127,
+      "learning_rate": 1.9328222474377843e-05,
+      "loss": 1.1012,
+      "step": 1045
+    },
+    {
+      "epoch": 0.14395816129920175,
+      "grad_norm": 1.6706797510131859,
+      "learning_rate": 1.9326615373444036e-05,
+      "loss": 1.178,
+      "step": 1046
+    },
+    {
+      "epoch": 0.14409578860445912,
+      "grad_norm": 1.7886250745714387,
+      "learning_rate": 1.9325006419439418e-05,
+      "loss": 1.1862,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1442334159097165,
+      "grad_norm": 1.7842012031483538,
+      "learning_rate": 1.9323395612683673e-05,
+      "loss": 1.2039,
+      "step": 1048
+    },
+    {
+      "epoch": 0.14437104321497385,
+      "grad_norm": 1.766246892959719,
+      "learning_rate": 1.9321782953496843e-05,
+      "loss": 1.166,
+      "step": 1049
+    },
+    {
+      "epoch": 0.14450867052023122,
+      "grad_norm": 1.5490139118158635,
+      "learning_rate": 1.932016844219934e-05,
+      "loss": 1.1899,
+      "step": 1050
+    },
+    {
+      "epoch": 0.14464629782548857,
+      "grad_norm": 1.8267749256918715,
+      "learning_rate": 1.9318552079111946e-05,
+      "loss": 1.2621,
+      "step": 1051
+    },
+    {
+      "epoch": 0.14478392513074595,
+      "grad_norm": 1.8740076347849273,
+      "learning_rate": 1.9316933864555814e-05,
+      "loss": 1.149,
+      "step": 1052
+    },
+    {
+      "epoch": 0.1449215524360033,
+      "grad_norm": 1.6240384842519076,
+      "learning_rate": 1.9315313798852456e-05,
+      "loss": 1.1747,
+      "step": 1053
+    },
+    {
+      "epoch": 0.14505917974126067,
+      "grad_norm": 1.9158754137865142,
+      "learning_rate": 1.931369188232376e-05,
+      "loss": 1.181,
+      "step": 1054
+    },
+    {
+      "epoch": 0.14519680704651802,
+      "grad_norm": 1.6661373349137558,
+      "learning_rate": 1.931206811529198e-05,
+      "loss": 1.214,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1453344343517754,
+      "grad_norm": 1.6873691015904142,
+      "learning_rate": 1.9310442498079732e-05,
+      "loss": 1.0654,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14547206165703275,
+      "grad_norm": 1.878988417018608,
+      "learning_rate": 1.9308815031010003e-05,
+      "loss": 1.1782,
+      "step": 1057
+    },
+    {
+      "epoch": 0.14560968896229012,
+      "grad_norm": 1.7727817859829433,
+      "learning_rate": 1.930718571440615e-05,
+      "loss": 1.1697,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14574731626754747,
+      "grad_norm": 1.8044338444917591,
+      "learning_rate": 1.93055545485919e-05,
+      "loss": 1.2463,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14588494357280485,
+      "grad_norm": 1.8348774430262578,
+      "learning_rate": 1.9303921533891342e-05,
+      "loss": 1.1561,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1460225708780622,
+      "grad_norm": 1.8579746835310416,
+      "learning_rate": 1.9302286670628932e-05,
+      "loss": 1.1746,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14616019818331957,
+      "grad_norm": 1.7584891340215778,
+      "learning_rate": 1.930064995912949e-05,
+      "loss": 1.2167,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14629782548857692,
+      "grad_norm": 1.9271754680079283,
+      "learning_rate": 1.929901139971822e-05,
+      "loss": 1.2323,
+      "step": 1063
+    },
+    {
+      "epoch": 0.1464354527938343,
+      "grad_norm": 1.6735516321392738,
+      "learning_rate": 1.9297370992720667e-05,
+      "loss": 1.1562,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14657308009909165,
+      "grad_norm": 1.660688131026209,
+      "learning_rate": 1.929572873846277e-05,
+      "loss": 1.1151,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14671070740434902,
+      "grad_norm": 1.6417273457225148,
+      "learning_rate": 1.9294084637270816e-05,
+      "loss": 1.1849,
+      "step": 1066
+    },
+    {
+      "epoch": 0.1468483347096064,
+      "grad_norm": 1.7055508407109583,
+      "learning_rate": 1.9292438689471466e-05,
+      "loss": 1.1513,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14698596201486375,
+      "grad_norm": 1.680377650619904,
+      "learning_rate": 1.929079089539175e-05,
+      "loss": 1.1348,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14712358932012112,
+      "grad_norm": 2.0644459502440644,
+      "learning_rate": 1.9289141255359055e-05,
+      "loss": 1.1607,
+      "step": 1069
+    },
+    {
+      "epoch": 0.14726121662537847,
+      "grad_norm": 1.9181268286509217,
+      "learning_rate": 1.9287489769701154e-05,
+      "loss": 1.2297,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14739884393063585,
+      "grad_norm": 1.961396147867366,
+      "learning_rate": 1.9285836438746163e-05,
+      "loss": 1.168,
+      "step": 1071
+    },
+    {
+      "epoch": 0.1475364712358932,
+      "grad_norm": 1.6400062968449955,
+      "learning_rate": 1.9284181262822582e-05,
+      "loss": 1.1445,
+      "step": 1072
+    },
+    {
+      "epoch": 0.14767409854115057,
+      "grad_norm": 1.802438324944088,
+      "learning_rate": 1.928252424225927e-05,
+      "loss": 1.0839,
+      "step": 1073
+    },
+    {
+      "epoch": 0.14781172584640792,
+      "grad_norm": 1.884699344357473,
+      "learning_rate": 1.9280865377385458e-05,
+      "loss": 1.1297,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1479493531516653,
+      "grad_norm": 1.5926213909405698,
+      "learning_rate": 1.9279204668530736e-05,
+      "loss": 1.1156,
+      "step": 1075
+    },
+    {
+      "epoch": 0.14808698045692265,
+      "grad_norm": 1.7060801070475513,
+      "learning_rate": 1.9277542116025063e-05,
+      "loss": 1.1051,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14822460776218002,
+      "grad_norm": 1.693883986145678,
+      "learning_rate": 1.927587772019877e-05,
+      "loss": 1.1408,
+      "step": 1077
+    },
+    {
+      "epoch": 0.14836223506743737,
+      "grad_norm": 1.821795209016418,
+      "learning_rate": 1.9274211481382542e-05,
+      "loss": 1.1327,
+      "step": 1078
+    },
+    {
+      "epoch": 0.14849986237269475,
+      "grad_norm": 1.710208067331486,
+      "learning_rate": 1.9272543399907446e-05,
+      "loss": 1.1074,
+      "step": 1079
+    },
+    {
+      "epoch": 0.1486374896779521,
+      "grad_norm": 1.781492477275213,
+      "learning_rate": 1.9270873476104902e-05,
+      "loss": 1.1232,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14877511698320947,
+      "grad_norm": 1.814575671002813,
+      "learning_rate": 1.9269201710306706e-05,
+      "loss": 1.2064,
+      "step": 1081
+    },
+    {
+      "epoch": 0.14891274428846682,
+      "grad_norm": 1.8733754758388939,
+      "learning_rate": 1.926752810284501e-05,
+      "loss": 1.1649,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1490503715937242,
+      "grad_norm": 1.7540829513207346,
+      "learning_rate": 1.9265852654052337e-05,
+      "loss": 1.1768,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14918799889898154,
+      "grad_norm": 1.6666321966271487,
+      "learning_rate": 1.926417536426158e-05,
+      "loss": 1.1761,
+      "step": 1084
+    },
+    {
+      "epoch": 0.14932562620423892,
+      "grad_norm": 2.1235814795968357,
+      "learning_rate": 1.9262496233805986e-05,
+      "loss": 1.192,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1494632535094963,
+      "grad_norm": 1.627920618863355,
+      "learning_rate": 1.9260815263019185e-05,
+      "loss": 1.1553,
+      "step": 1086
+    },
+    {
+      "epoch": 0.14960088081475365,
+      "grad_norm": 1.6342504526960437,
+      "learning_rate": 1.9259132452235157e-05,
+      "loss": 1.2189,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14973850812001102,
+      "grad_norm": 1.8156515782588236,
+      "learning_rate": 1.9257447801788254e-05,
+      "loss": 1.1686,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14987613542526837,
+      "grad_norm": 1.8845872414845555,
+      "learning_rate": 1.925576131201319e-05,
+      "loss": 1.1894,
+      "step": 1089
+    },
+    {
+      "epoch": 0.15001376273052575,
+      "grad_norm": 1.7534399727616279,
+      "learning_rate": 1.9254072983245057e-05,
+      "loss": 1.1432,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1501513900357831,
+      "grad_norm": 1.8862758263807577,
+      "learning_rate": 1.9252382815819295e-05,
+      "loss": 1.1742,
+      "step": 1091
+    },
+    {
+      "epoch": 0.15028901734104047,
+      "grad_norm": 1.6962041993965917,
+      "learning_rate": 1.925069081007172e-05,
+      "loss": 1.1694,
+      "step": 1092
+    },
+    {
+      "epoch": 0.15042664464629782,
+      "grad_norm": 1.7126127380678224,
+      "learning_rate": 1.9248996966338504e-05,
+      "loss": 1.1106,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1505642719515552,
+      "grad_norm": 1.7778475151056303,
+      "learning_rate": 1.92473012849562e-05,
+      "loss": 1.1778,
+      "step": 1094
+    },
+    {
+      "epoch": 0.15070189925681254,
+      "grad_norm": 1.6788842778221431,
+      "learning_rate": 1.924560376626171e-05,
+      "loss": 1.1781,
+      "step": 1095
+    },
+    {
+      "epoch": 0.15083952656206992,
+      "grad_norm": 1.8919662578352594,
+      "learning_rate": 1.9243904410592314e-05,
+      "loss": 1.1802,
+      "step": 1096
+    },
+    {
+      "epoch": 0.15097715386732727,
+      "grad_norm": 1.7223160681474352,
+      "learning_rate": 1.9242203218285644e-05,
+      "loss": 1.1919,
+      "step": 1097
+    },
+    {
+      "epoch": 0.15111478117258464,
+      "grad_norm": 1.8451220900734726,
+      "learning_rate": 1.9240500189679706e-05,
+      "loss": 1.1771,
+      "step": 1098
+    },
+    {
+      "epoch": 0.151252408477842,
+      "grad_norm": 1.7973510421571919,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 1.2209,
+      "step": 1099
+    },
+    {
+      "epoch": 0.15139003578309937,
+      "grad_norm": 1.673102273571798,
+      "learning_rate": 1.9237088624923866e-05,
+      "loss": 1.1634,
+      "step": 1100
+    },
+    {
+      "epoch": 0.15152766308835672,
+      "grad_norm": 1.662903547827202,
+      "learning_rate": 1.9235380089451796e-05,
+      "loss": 1.1413,
+      "step": 1101
+    },
+    {
+      "epoch": 0.1516652903936141,
+      "grad_norm": 1.8351188693541205,
+      "learning_rate": 1.9233669719036122e-05,
+      "loss": 1.181,
+      "step": 1102
+    },
+    {
+      "epoch": 0.15180291769887144,
+      "grad_norm": 1.7578699356040393,
+      "learning_rate": 1.923195751401667e-05,
+      "loss": 1.183,
+      "step": 1103
+    },
+    {
+      "epoch": 0.15194054500412882,
+      "grad_norm": 1.877239388456024,
+      "learning_rate": 1.923024347473363e-05,
+      "loss": 1.2347,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1520781723093862,
+      "grad_norm": 1.6931098624796037,
+      "learning_rate": 1.922852760152756e-05,
+      "loss": 1.2622,
+      "step": 1105
+    },
+    {
+      "epoch": 0.15221579961464354,
+      "grad_norm": 1.6173988489502167,
+      "learning_rate": 1.9226809894739383e-05,
+      "loss": 1.1361,
+      "step": 1106
+    },
+    {
+      "epoch": 0.15235342691990092,
+      "grad_norm": 1.597422562520179,
+      "learning_rate": 1.9225090354710377e-05,
+      "loss": 1.1046,
+      "step": 1107
+    },
+    {
+      "epoch": 0.15249105422515827,
+      "grad_norm": 1.7597677458259218,
+      "learning_rate": 1.9223368981782196e-05,
+      "loss": 1.0444,
+      "step": 1108
+    },
+    {
+      "epoch": 0.15262868153041564,
+      "grad_norm": 1.6349361096354922,
+      "learning_rate": 1.9221645776296855e-05,
+      "loss": 1.1541,
+      "step": 1109
+    },
+    {
+      "epoch": 0.152766308835673,
+      "grad_norm": 2.0712384877647545,
+      "learning_rate": 1.9219920738596728e-05,
+      "loss": 1.2074,
+      "step": 1110
+    },
+    {
+      "epoch": 0.15290393614093037,
+      "grad_norm": 1.7190146324682027,
+      "learning_rate": 1.9218193869024558e-05,
+      "loss": 1.1441,
+      "step": 1111
+    },
+    {
+      "epoch": 0.15304156344618772,
+      "grad_norm": 1.8584502305855388,
+      "learning_rate": 1.9216465167923447e-05,
+      "loss": 1.0596,
+      "step": 1112
+    },
+    {
+      "epoch": 0.1531791907514451,
+      "grad_norm": 1.70788262946192,
+      "learning_rate": 1.921473463563687e-05,
+      "loss": 1.1082,
+      "step": 1113
+    },
+    {
+      "epoch": 0.15331681805670244,
+      "grad_norm": 1.7450804091735537,
+      "learning_rate": 1.9213002272508655e-05,
+      "loss": 1.1796,
+      "step": 1114
+    },
+    {
+      "epoch": 0.15345444536195982,
+      "grad_norm": 1.8436541996446891,
+      "learning_rate": 1.9211268078883e-05,
+      "loss": 1.1605,
+      "step": 1115
+    },
+    {
+      "epoch": 0.15359207266721717,
+      "grad_norm": 1.7762221323965188,
+      "learning_rate": 1.9209532055104464e-05,
+      "loss": 1.1714,
+      "step": 1116
+    },
+    {
+      "epoch": 0.15372969997247454,
+      "grad_norm": 1.7697579876112075,
+      "learning_rate": 1.920779420151798e-05,
+      "loss": 1.1769,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1538673272777319,
+      "grad_norm": 1.9362753815275267,
+      "learning_rate": 1.9206054518468823e-05,
+      "loss": 1.145,
+      "step": 1118
+    },
+    {
+      "epoch": 0.15400495458298927,
+      "grad_norm": 1.8211496788316504,
+      "learning_rate": 1.9204313006302654e-05,
+      "loss": 1.2133,
+      "step": 1119
+    },
+    {
+      "epoch": 0.15414258188824662,
+      "grad_norm": 1.7131108977526066,
+      "learning_rate": 1.920256966536548e-05,
+      "loss": 1.2251,
+      "step": 1120
+    },
+    {
+      "epoch": 0.154280209193504,
+      "grad_norm": 1.6369555605884978,
+      "learning_rate": 1.920082449600369e-05,
+      "loss": 1.1679,
+      "step": 1121
+    },
+    {
+      "epoch": 0.15441783649876134,
+      "grad_norm": 1.568505566958367,
+      "learning_rate": 1.919907749856401e-05,
+      "loss": 1.1307,
+      "step": 1122
+    },
+    {
+      "epoch": 0.15455546380401872,
+      "grad_norm": 1.7271908182126796,
+      "learning_rate": 1.9197328673393558e-05,
+      "loss": 1.1814,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1546930911092761,
+      "grad_norm": 1.8720518552940209,
+      "learning_rate": 1.9195578020839795e-05,
+      "loss": 1.2299,
+      "step": 1124
+    },
+    {
+      "epoch": 0.15483071841453344,
+      "grad_norm": 1.800848239215077,
+      "learning_rate": 1.9193825541250554e-05,
+      "loss": 1.1586,
+      "step": 1125
+    },
+    {
+      "epoch": 0.15496834571979082,
+      "grad_norm": 1.7856939922639687,
+      "learning_rate": 1.9192071234974027e-05,
+      "loss": 1.2087,
+      "step": 1126
+    },
+    {
+      "epoch": 0.15510597302504817,
+      "grad_norm": 1.8142270394472655,
+      "learning_rate": 1.919031510235877e-05,
+      "loss": 1.1881,
+      "step": 1127
+    },
+    {
+      "epoch": 0.15524360033030554,
+      "grad_norm": 2.0015206557273095,
+      "learning_rate": 1.9188557143753706e-05,
+      "loss": 1.1957,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1553812276355629,
+      "grad_norm": 1.6866156058746307,
+      "learning_rate": 1.918679735950811e-05,
+      "loss": 1.1477,
+      "step": 1129
+    },
+    {
+      "epoch": 0.15551885494082027,
+      "grad_norm": 1.8906385222782192,
+      "learning_rate": 1.9185035749971638e-05,
+      "loss": 1.1886,
+      "step": 1130
+    },
+    {
+      "epoch": 0.15565648224607762,
+      "grad_norm": 1.5737492122406012,
+      "learning_rate": 1.9183272315494294e-05,
+      "loss": 1.0834,
+      "step": 1131
+    },
+    {
+      "epoch": 0.155794109551335,
+      "grad_norm": 1.72114040270019,
+      "learning_rate": 1.918150705642644e-05,
+      "loss": 1.1165,
+      "step": 1132
+    },
+    {
+      "epoch": 0.15593173685659234,
+      "grad_norm": 1.9928451339861477,
+      "learning_rate": 1.9179739973118815e-05,
+      "loss": 1.2278,
+      "step": 1133
+    },
+    {
+      "epoch": 0.15606936416184972,
+      "grad_norm": 1.9067862677065102,
+      "learning_rate": 1.9177971065922517e-05,
+      "loss": 1.1819,
+      "step": 1134
+    },
+    {
+      "epoch": 0.15620699146710706,
+      "grad_norm": 1.6415217453617363,
+      "learning_rate": 1.9176200335189e-05,
+      "loss": 1.1651,
+      "step": 1135
+    },
+    {
+      "epoch": 0.15634461877236444,
+      "grad_norm": 1.7038795309168628,
+      "learning_rate": 1.917442778127009e-05,
+      "loss": 1.1599,
+      "step": 1136
+    },
+    {
+      "epoch": 0.1564822460776218,
+      "grad_norm": 1.6945769298594398,
+      "learning_rate": 1.9172653404517957e-05,
+      "loss": 1.181,
+      "step": 1137
+    },
+    {
+      "epoch": 0.15661987338287917,
+      "grad_norm": 1.7415109485560851,
+      "learning_rate": 1.9170877205285157e-05,
+      "loss": 1.2343,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15675750068813651,
+      "grad_norm": 1.6511722458961884,
+      "learning_rate": 1.9169099183924588e-05,
+      "loss": 1.1697,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1568951279933939,
+      "grad_norm": 1.8991494689279091,
+      "learning_rate": 1.9167319340789524e-05,
+      "loss": 1.2807,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15703275529865124,
+      "grad_norm": 1.8301084928792084,
+      "learning_rate": 1.9165537676233598e-05,
+      "loss": 1.2145,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15717038260390861,
+      "grad_norm": 1.5765563019301203,
+      "learning_rate": 1.9163754190610794e-05,
+      "loss": 1.1565,
+      "step": 1142
+    },
+    {
+      "epoch": 0.157308009909166,
+      "grad_norm": 1.780969213194246,
+      "learning_rate": 1.916196888427547e-05,
+      "loss": 1.2189,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15744563721442334,
+      "grad_norm": 1.8029724059973156,
+      "learning_rate": 1.9160181757582343e-05,
+      "loss": 1.1583,
+      "step": 1144
+    },
+    {
+      "epoch": 0.15758326451968072,
+      "grad_norm": 1.6385068951646138,
+      "learning_rate": 1.915839281088649e-05,
+      "loss": 1.1617,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15772089182493806,
+      "grad_norm": 1.8547153541130317,
+      "learning_rate": 1.915660204454335e-05,
+      "loss": 1.13,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15785851913019544,
+      "grad_norm": 1.8340859392542355,
+      "learning_rate": 1.9154809458908725e-05,
+      "loss": 1.1341,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1579961464354528,
+      "grad_norm": 1.9289024952068476,
+      "learning_rate": 1.9153015054338775e-05,
+      "loss": 1.0964,
+      "step": 1148
+    },
+    {
+      "epoch": 0.15813377374071017,
+      "grad_norm": 2.3353063408216173,
+      "learning_rate": 1.9151218831190026e-05,
+      "loss": 1.2917,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1582714010459675,
+      "grad_norm": 1.9065187723715638,
+      "learning_rate": 1.9149420789819363e-05,
+      "loss": 1.1229,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1584090283512249,
+      "grad_norm": 1.8838328948951106,
+      "learning_rate": 1.914762093058403e-05,
+      "loss": 1.21,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15854665565648224,
+      "grad_norm": 1.9340852966292466,
+      "learning_rate": 1.9145819253841633e-05,
+      "loss": 1.1442,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15868428296173961,
+      "grad_norm": 1.7331393305817628,
+      "learning_rate": 1.9144015759950147e-05,
+      "loss": 1.2471,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15882191026699696,
+      "grad_norm": 1.5973441891568578,
+      "learning_rate": 1.9142210449267897e-05,
+      "loss": 1.1188,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15895953757225434,
+      "grad_norm": 1.549708320140468,
+      "learning_rate": 1.9140403322153568e-05,
+      "loss": 1.1523,
+      "step": 1155
+    },
+    {
+      "epoch": 0.1590971648775117,
+      "grad_norm": 1.6848260781347864,
+      "learning_rate": 1.913859437896622e-05,
+      "loss": 1.1453,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15923479218276906,
+      "grad_norm": 1.839655842737148,
+      "learning_rate": 1.9136783620065268e-05,
+      "loss": 1.1423,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1593724194880264,
+      "grad_norm": 1.760337683218225,
+      "learning_rate": 1.9134971045810475e-05,
+      "loss": 1.2196,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1595100467932838,
+      "grad_norm": 1.7876502708734725,
+      "learning_rate": 1.9133156656561978e-05,
+      "loss": 1.1391,
+      "step": 1159
+    },
+    {
+      "epoch": 0.15964767409854114,
+      "grad_norm": 1.6716825317924568,
+      "learning_rate": 1.9131340452680276e-05,
+      "loss": 1.1313,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1597853014037985,
+      "grad_norm": 1.8481667076136052,
+      "learning_rate": 1.912952243452622e-05,
+      "loss": 1.1971,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1599229287090559,
+      "grad_norm": 1.7821914493313937,
+      "learning_rate": 1.9127702602461025e-05,
+      "loss": 1.1545,
+      "step": 1162
+    },
+    {
+      "epoch": 0.16006055601431324,
+      "grad_norm": 1.6759202493310554,
+      "learning_rate": 1.9125880956846264e-05,
+      "loss": 1.238,
+      "step": 1163
+    },
+    {
+      "epoch": 0.16019818331957061,
+      "grad_norm": 1.8706385103342449,
+      "learning_rate": 1.912405749804388e-05,
+      "loss": 1.2643,
+      "step": 1164
+    },
+    {
+      "epoch": 0.16033581062482796,
+      "grad_norm": 1.764663560639773,
+      "learning_rate": 1.9122232226416168e-05,
+      "loss": 1.1404,
+      "step": 1165
+    },
+    {
+      "epoch": 0.16047343793008534,
+      "grad_norm": 1.7051372452431504,
+      "learning_rate": 1.9120405142325782e-05,
+      "loss": 1.1924,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1606110652353427,
+      "grad_norm": 1.8277927801193208,
+      "learning_rate": 1.9118576246135738e-05,
+      "loss": 1.2084,
+      "step": 1167
+    },
+    {
+      "epoch": 0.16074869254060006,
+      "grad_norm": 1.731419163435174,
+      "learning_rate": 1.9116745538209413e-05,
+      "loss": 1.2567,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1608863198458574,
+      "grad_norm": 1.7366494842037918,
+      "learning_rate": 1.9114913018910544e-05,
+      "loss": 1.1153,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1610239471511148,
+      "grad_norm": 1.7096956288944254,
+      "learning_rate": 1.9113078688603225e-05,
+      "loss": 1.1912,
+      "step": 1170
+    },
+    {
+      "epoch": 0.16116157445637214,
+      "grad_norm": 1.8688828483884679,
+      "learning_rate": 1.911124254765192e-05,
+      "loss": 1.1745,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1612992017616295,
+      "grad_norm": 1.7694918672894087,
+      "learning_rate": 1.9109404596421436e-05,
+      "loss": 1.1868,
+      "step": 1172
+    },
+    {
+      "epoch": 0.16143682906688686,
+      "grad_norm": 1.7162555917332398,
+      "learning_rate": 1.9107564835276955e-05,
+      "loss": 1.1354,
+      "step": 1173
+    },
+    {
+      "epoch": 0.16157445637214424,
+      "grad_norm": 1.796469216251665,
+      "learning_rate": 1.910572326458401e-05,
+      "loss": 1.1491,
+      "step": 1174
+    },
+    {
+      "epoch": 0.16171208367740159,
+      "grad_norm": 1.8153470943521934,
+      "learning_rate": 1.9103879884708496e-05,
+      "loss": 1.2759,
+      "step": 1175
+    },
+    {
+      "epoch": 0.16184971098265896,
+      "grad_norm": 1.6584917976749105,
+      "learning_rate": 1.9102034696016667e-05,
+      "loss": 1.1787,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1619873382879163,
+      "grad_norm": 1.720254603321273,
+      "learning_rate": 1.910018769887514e-05,
+      "loss": 1.1175,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1621249655931737,
+      "grad_norm": 1.8927521261492168,
+      "learning_rate": 1.9098338893650883e-05,
+      "loss": 1.1794,
+      "step": 1178
+    },
+    {
+      "epoch": 0.16226259289843104,
+      "grad_norm": 1.815195981529941,
+      "learning_rate": 1.9096488280711228e-05,
+      "loss": 1.2181,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1624002202036884,
+      "grad_norm": 1.9401431716349582,
+      "learning_rate": 1.9094635860423875e-05,
+      "loss": 1.2432,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1625378475089458,
+      "grad_norm": 1.6216445585990626,
+      "learning_rate": 1.9092781633156863e-05,
+      "loss": 1.144,
+      "step": 1181
+    },
+    {
+      "epoch": 0.16267547481420314,
+      "grad_norm": 1.8310199994715557,
+      "learning_rate": 1.9090925599278608e-05,
+      "loss": 1.1435,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1628131021194605,
+      "grad_norm": 1.8154804524673684,
+      "learning_rate": 1.9089067759157876e-05,
+      "loss": 1.201,
+      "step": 1183
+    },
+    {
+      "epoch": 0.16295072942471786,
+      "grad_norm": 1.5943740126400703,
+      "learning_rate": 1.9087208113163803e-05,
+      "loss": 1.2086,
+      "step": 1184
+    },
+    {
+      "epoch": 0.16308835672997524,
+      "grad_norm": 1.7887876493343156,
+      "learning_rate": 1.9085346661665866e-05,
+      "loss": 1.1996,
+      "step": 1185
+    },
+    {
+      "epoch": 0.16322598403523259,
+      "grad_norm": 1.674001168610742,
+      "learning_rate": 1.9083483405033908e-05,
+      "loss": 1.1983,
+      "step": 1186
+    },
+    {
+      "epoch": 0.16336361134048996,
+      "grad_norm": 1.8094104887070412,
+      "learning_rate": 1.908161834363814e-05,
+      "loss": 1.1624,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1635012386457473,
+      "grad_norm": 1.6646968399994766,
+      "learning_rate": 1.9079751477849117e-05,
+      "loss": 1.1684,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1636388659510047,
+      "grad_norm": 1.585954761073073,
+      "learning_rate": 1.9077882808037766e-05,
+      "loss": 1.1459,
+      "step": 1189
+    },
+    {
+      "epoch": 0.16377649325626203,
+      "grad_norm": 1.5928228775041737,
+      "learning_rate": 1.9076012334575366e-05,
+      "loss": 1.1387,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1639141205615194,
+      "grad_norm": 1.6036192388573898,
+      "learning_rate": 1.907414005783355e-05,
+      "loss": 1.1259,
+      "step": 1191
+    },
+    {
+      "epoch": 0.16405174786677676,
+      "grad_norm": 1.8016017870511996,
+      "learning_rate": 1.9072265978184315e-05,
+      "loss": 1.1368,
+      "step": 1192
+    },
+    {
+      "epoch": 0.16418937517203414,
+      "grad_norm": 1.8314631312719303,
+      "learning_rate": 1.9070390096000016e-05,
+      "loss": 1.1657,
+      "step": 1193
+    },
+    {
+      "epoch": 0.16432700247729148,
+      "grad_norm": 1.665241550837254,
+      "learning_rate": 1.906851241165337e-05,
+      "loss": 1.158,
+      "step": 1194
+    },
+    {
+      "epoch": 0.16446462978254886,
+      "grad_norm": 1.6646890439441453,
+      "learning_rate": 1.9066632925517437e-05,
+      "loss": 1.071,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1646022570878062,
+      "grad_norm": 1.9083295487952538,
+      "learning_rate": 1.9064751637965653e-05,
+      "loss": 1.1955,
+      "step": 1196
+    },
+    {
+      "epoch": 0.16473988439306358,
+      "grad_norm": 1.7149647264657328,
+      "learning_rate": 1.90628685493718e-05,
+      "loss": 1.0902,
+      "step": 1197
+    },
+    {
+      "epoch": 0.16487751169832093,
+      "grad_norm": 1.9228771282746275,
+      "learning_rate": 1.906098366011002e-05,
+      "loss": 1.2401,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1650151390035783,
+      "grad_norm": 1.769875710395189,
+      "learning_rate": 1.9059096970554824e-05,
+      "loss": 1.2453,
+      "step": 1199
+    },
+    {
+      "epoch": 0.16515276630883569,
+      "grad_norm": 1.663309308617753,
+      "learning_rate": 1.9057208481081063e-05,
+      "loss": 1.1216,
+      "step": 1200
+    },
+    {
+      "epoch": 0.16529039361409303,
+      "grad_norm": 1.7652471117275161,
+      "learning_rate": 1.9055318192063955e-05,
+      "loss": 1.1684,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1654280209193504,
+      "grad_norm": 1.9152250010725427,
+      "learning_rate": 1.905342610387908e-05,
+      "loss": 1.2469,
+      "step": 1202
+    },
+    {
+      "epoch": 0.16556564822460776,
+      "grad_norm": 1.9016327773445862,
+      "learning_rate": 1.9051532216902366e-05,
+      "loss": 1.1416,
+      "step": 1203
+    },
+    {
+      "epoch": 0.16570327552986513,
+      "grad_norm": 1.7598960794938714,
+      "learning_rate": 1.9049636531510098e-05,
+      "loss": 1.185,
+      "step": 1204
+    },
+    {
+      "epoch": 0.16584090283512248,
+      "grad_norm": 1.6619027849908046,
+      "learning_rate": 1.904773904807893e-05,
+      "loss": 1.1055,
+      "step": 1205
+    },
+    {
+      "epoch": 0.16597853014037986,
+      "grad_norm": 1.6524499458290003,
+      "learning_rate": 1.9045839766985864e-05,
+      "loss": 1.2159,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1661161574456372,
+      "grad_norm": 1.7770658011682856,
+      "learning_rate": 1.9043938688608258e-05,
+      "loss": 1.1731,
+      "step": 1207
+    },
+    {
+      "epoch": 0.16625378475089458,
+      "grad_norm": 1.8758344754586038,
+      "learning_rate": 1.9042035813323835e-05,
+      "loss": 1.227,
+      "step": 1208
+    },
+    {
+      "epoch": 0.16639141205615193,
+      "grad_norm": 1.6260721375736185,
+      "learning_rate": 1.9040131141510666e-05,
+      "loss": 1.1785,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1665290393614093,
+      "grad_norm": 1.6559796751256615,
+      "learning_rate": 1.9038224673547185e-05,
+      "loss": 1.1827,
+      "step": 1210
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 1.90419068138968,
+      "learning_rate": 1.903631640981218e-05,
+      "loss": 1.1749,
+      "step": 1211
+    },
+    {
+      "epoch": 0.16680429397192403,
+      "grad_norm": 1.6017989655001967,
+      "learning_rate": 1.90344063506848e-05,
+      "loss": 1.0195,
+      "step": 1212
+    },
+    {
+      "epoch": 0.16694192127718138,
+      "grad_norm": 1.69109864820019,
+      "learning_rate": 1.9032494496544545e-05,
+      "loss": 1.1208,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16707954858243876,
+      "grad_norm": 1.8735570256230418,
+      "learning_rate": 1.9030580847771278e-05,
+      "loss": 1.1963,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1672171758876961,
+      "grad_norm": 1.8926080826017415,
+      "learning_rate": 1.9028665404745208e-05,
+      "loss": 1.1991,
+      "step": 1215
+    },
+    {
+      "epoch": 0.16735480319295348,
+      "grad_norm": 1.8016349622727303,
+      "learning_rate": 1.9026748167846914e-05,
+      "loss": 1.2143,
+      "step": 1216
+    },
+    {
+      "epoch": 0.16749243049821083,
+      "grad_norm": 1.793061437836532,
+      "learning_rate": 1.9024829137457317e-05,
+      "loss": 1.087,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1676300578034682,
+      "grad_norm": 1.7054045851447461,
+      "learning_rate": 1.902290831395771e-05,
+      "loss": 1.1403,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16776768510872558,
+      "grad_norm": 1.5680307244180376,
+      "learning_rate": 1.902098569772973e-05,
+      "loss": 1.1575,
+      "step": 1219
+    },
+    {
+      "epoch": 0.16790531241398293,
+      "grad_norm": 1.8117692514637878,
+      "learning_rate": 1.901906128915538e-05,
+      "loss": 1.1163,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1680429397192403,
+      "grad_norm": 1.9217914895114794,
+      "learning_rate": 1.9017135088617004e-05,
+      "loss": 1.158,
+      "step": 1221
+    },
+    {
+      "epoch": 0.16818056702449766,
+      "grad_norm": 1.6999795797673225,
+      "learning_rate": 1.901520709649732e-05,
+      "loss": 1.1332,
+      "step": 1222
+    },
+    {
+      "epoch": 0.16831819432975503,
+      "grad_norm": 1.7156560391626765,
+      "learning_rate": 1.901327731317939e-05,
+      "loss": 1.2244,
+      "step": 1223
+    },
+    {
+      "epoch": 0.16845582163501238,
+      "grad_norm": 1.6647902896309577,
+      "learning_rate": 1.9011345739046636e-05,
+      "loss": 1.1432,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16859344894026976,
+      "grad_norm": 1.7821199514561332,
+      "learning_rate": 1.9009412374482835e-05,
+      "loss": 1.2725,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1687310762455271,
+      "grad_norm": 1.8497540614531893,
+      "learning_rate": 1.9007477219872122e-05,
+      "loss": 1.2248,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16886870355078448,
+      "grad_norm": 1.7965090156589574,
+      "learning_rate": 1.900554027559898e-05,
+      "loss": 1.1234,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16900633085604183,
+      "grad_norm": 1.6874676730461604,
+      "learning_rate": 1.9003601542048263e-05,
+      "loss": 1.1601,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1691439581612992,
+      "grad_norm": 1.7693791881956678,
+      "learning_rate": 1.900166101960516e-05,
+      "loss": 1.1662,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16928158546655656,
+      "grad_norm": 1.6998545420256448,
+      "learning_rate": 1.8999718708655238e-05,
+      "loss": 1.2309,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16941921277181393,
+      "grad_norm": 1.62366962173597,
+      "learning_rate": 1.89977746095844e-05,
+      "loss": 1.0779,
+      "step": 1231
+    },
+    {
+      "epoch": 0.16955684007707128,
+      "grad_norm": 1.8222242228271763,
+      "learning_rate": 1.899582872277891e-05,
+      "loss": 1.1856,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16969446738232866,
+      "grad_norm": 1.9442680573982873,
+      "learning_rate": 1.8993881048625396e-05,
+      "loss": 1.013,
+      "step": 1233
+    },
+    {
+      "epoch": 0.169832094687586,
+      "grad_norm": 1.555422513678515,
+      "learning_rate": 1.899193158751083e-05,
+      "loss": 1.1317,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16996972199284338,
+      "grad_norm": 1.8836160756597673,
+      "learning_rate": 1.898998033982254e-05,
+      "loss": 1.1097,
+      "step": 1235
+    },
+    {
+      "epoch": 0.17010734929810073,
+      "grad_norm": 1.8048326379285833,
+      "learning_rate": 1.8988027305948218e-05,
+      "loss": 1.159,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1702449766033581,
+      "grad_norm": 1.7627950397515593,
+      "learning_rate": 1.8986072486275904e-05,
+      "loss": 1.1384,
+      "step": 1237
+    },
+    {
+      "epoch": 0.17038260390861548,
+      "grad_norm": 1.8043614168999946,
+      "learning_rate": 1.898411588119399e-05,
+      "loss": 1.2142,
+      "step": 1238
+    },
+    {
+      "epoch": 0.17052023121387283,
+      "grad_norm": 1.7052618399596813,
+      "learning_rate": 1.8982157491091232e-05,
+      "loss": 1.1984,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1706578585191302,
+      "grad_norm": 1.6310255014342303,
+      "learning_rate": 1.8980197316356733e-05,
+      "loss": 1.1473,
+      "step": 1240
+    },
+    {
+      "epoch": 0.17079548582438756,
+      "grad_norm": 1.563935091402885,
+      "learning_rate": 1.897823535737995e-05,
+      "loss": 1.0674,
+      "step": 1241
+    },
+    {
+      "epoch": 0.17093311312964493,
+      "grad_norm": 1.5520983951857958,
+      "learning_rate": 1.89762716145507e-05,
+      "loss": 1.0586,
+      "step": 1242
+    },
+    {
+      "epoch": 0.17107074043490228,
+      "grad_norm": 1.7490032067166692,
+      "learning_rate": 1.8974306088259158e-05,
+      "loss": 1.0755,
+      "step": 1243
+    },
+    {
+      "epoch": 0.17120836774015966,
+      "grad_norm": 1.8694016472313402,
+      "learning_rate": 1.8972338778895837e-05,
+      "loss": 1.1823,
+      "step": 1244
+    },
+    {
+      "epoch": 0.171345995045417,
+      "grad_norm": 1.8057959975355817,
+      "learning_rate": 1.897036968685162e-05,
+      "loss": 1.1623,
+      "step": 1245
+    },
+    {
+      "epoch": 0.17148362235067438,
+      "grad_norm": 1.8170912877968763,
+      "learning_rate": 1.8968398812517734e-05,
+      "loss": 1.0887,
+      "step": 1246
+    },
+    {
+      "epoch": 0.17162124965593173,
+      "grad_norm": 1.7243067133256853,
+      "learning_rate": 1.896642615628577e-05,
+      "loss": 1.1716,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1717588769611891,
+      "grad_norm": 1.8691466518813193,
+      "learning_rate": 1.896445171854767e-05,
+      "loss": 1.2415,
+      "step": 1248
+    },
+    {
+      "epoch": 0.17189650426644645,
+      "grad_norm": 1.9905927401860763,
+      "learning_rate": 1.8962475499695715e-05,
+      "loss": 1.2241,
+      "step": 1249
+    },
+    {
+      "epoch": 0.17203413157170383,
+      "grad_norm": 1.6072388676311713,
+      "learning_rate": 1.8960497500122565e-05,
+      "loss": 1.1763,
+      "step": 1250
+    },
+    {
+      "epoch": 0.17217175887696118,
+      "grad_norm": 1.7307767900030802,
+      "learning_rate": 1.895851772022122e-05,
+      "loss": 1.251,
+      "step": 1251
+    },
+    {
+      "epoch": 0.17230938618221855,
+      "grad_norm": 1.8075546907865503,
+      "learning_rate": 1.895653616038503e-05,
+      "loss": 1.244,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1724470134874759,
+      "grad_norm": 1.7073247573453674,
+      "learning_rate": 1.895455282100771e-05,
+      "loss": 1.212,
+      "step": 1253
+    },
+    {
+      "epoch": 0.17258464079273328,
+      "grad_norm": 1.6861808448896867,
+      "learning_rate": 1.8952567702483314e-05,
+      "loss": 1.1134,
+      "step": 1254
+    },
+    {
+      "epoch": 0.17272226809799063,
+      "grad_norm": 1.559343840245276,
+      "learning_rate": 1.8950580805206264e-05,
+      "loss": 1.1116,
+      "step": 1255
+    },
+    {
+      "epoch": 0.172859895403248,
+      "grad_norm": 1.5853992306502935,
+      "learning_rate": 1.894859212957133e-05,
+      "loss": 1.2251,
+      "step": 1256
+    },
+    {
+      "epoch": 0.17299752270850538,
+      "grad_norm": 1.7629676423451743,
+      "learning_rate": 1.8946601675973632e-05,
+      "loss": 1.0989,
+      "step": 1257
+    },
+    {
+      "epoch": 0.17313515001376273,
+      "grad_norm": 1.7091134488396544,
+      "learning_rate": 1.894460944480865e-05,
+      "loss": 1.138,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1732727773190201,
+      "grad_norm": 1.7438958072158701,
+      "learning_rate": 1.8942615436472203e-05,
+      "loss": 1.2311,
+      "step": 1259
+    },
+    {
+      "epoch": 0.17341040462427745,
+      "grad_norm": 1.6785772050025798,
+      "learning_rate": 1.8940619651360484e-05,
+      "loss": 1.089,
+      "step": 1260
+    },
+    {
+      "epoch": 0.17354803192953483,
+      "grad_norm": 1.7805768293703217,
+      "learning_rate": 1.893862208987002e-05,
+      "loss": 1.2063,
+      "step": 1261
+    },
+    {
+      "epoch": 0.17368565923479218,
+      "grad_norm": 1.767612901092861,
+      "learning_rate": 1.8936622752397708e-05,
+      "loss": 1.1608,
+      "step": 1262
+    },
+    {
+      "epoch": 0.17382328654004955,
+      "grad_norm": 1.7606099723303654,
+      "learning_rate": 1.893462163934078e-05,
+      "loss": 1.1485,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1739609138453069,
+      "grad_norm": 1.9642517108874387,
+      "learning_rate": 1.893261875109684e-05,
+      "loss": 1.2192,
+      "step": 1264
+    },
+    {
+      "epoch": 0.17409854115056428,
+      "grad_norm": 1.7487943259888767,
+      "learning_rate": 1.8930614088063823e-05,
+      "loss": 1.12,
+      "step": 1265
+    },
+    {
+      "epoch": 0.17423616845582163,
+      "grad_norm": 1.6893141759300843,
+      "learning_rate": 1.8928607650640037e-05,
+      "loss": 1.1507,
+      "step": 1266
+    },
+    {
+      "epoch": 0.174373795761079,
+      "grad_norm": 1.6147309819890632,
+      "learning_rate": 1.8926599439224124e-05,
+      "loss": 1.2136,
+      "step": 1267
+    },
+    {
+      "epoch": 0.17451142306633635,
+      "grad_norm": 1.9625141971687832,
+      "learning_rate": 1.89245894542151e-05,
+      "loss": 1.1445,
+      "step": 1268
+    },
+    {
+      "epoch": 0.17464905037159373,
+      "grad_norm": 1.8082180482141077,
+      "learning_rate": 1.8922577696012317e-05,
+      "loss": 1.2045,
+      "step": 1269
+    },
+    {
+      "epoch": 0.17478667767685108,
+      "grad_norm": 1.6485283003357651,
+      "learning_rate": 1.8920564165015477e-05,
+      "loss": 1.1654,
+      "step": 1270
+    },
+    {
+      "epoch": 0.17492430498210845,
+      "grad_norm": 1.7884947317470825,
+      "learning_rate": 1.891854886162465e-05,
+      "loss": 1.1848,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1750619322873658,
+      "grad_norm": 1.7535664229004793,
+      "learning_rate": 1.8916531786240243e-05,
+      "loss": 1.1656,
+      "step": 1272
+    },
+    {
+      "epoch": 0.17519955959262318,
+      "grad_norm": 1.5008181517724697,
+      "learning_rate": 1.891451293926303e-05,
+      "loss": 1.1512,
+      "step": 1273
+    },
+    {
+      "epoch": 0.17533718689788053,
+      "grad_norm": 1.7017101035828872,
+      "learning_rate": 1.8912492321094117e-05,
+      "loss": 1.1512,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1754748142031379,
+      "grad_norm": 1.7661096321611374,
+      "learning_rate": 1.891046993213498e-05,
+      "loss": 1.2335,
+      "step": 1275
+    },
+    {
+      "epoch": 0.17561244150839528,
+      "grad_norm": 1.9033019707546557,
+      "learning_rate": 1.890844577278744e-05,
+      "loss": 1.156,
+      "step": 1276
+    },
+    {
+      "epoch": 0.17575006881365263,
+      "grad_norm": 1.6646980906776103,
+      "learning_rate": 1.890641984345367e-05,
+      "loss": 1.149,
+      "step": 1277
+    },
+    {
+      "epoch": 0.17588769611891,
+      "grad_norm": 1.9127021313646553,
+      "learning_rate": 1.8904392144536187e-05,
+      "loss": 1.2671,
+      "step": 1278
+    },
+    {
+      "epoch": 0.17602532342416735,
+      "grad_norm": 1.786222995126419,
+      "learning_rate": 1.8902362676437876e-05,
+      "loss": 1.1817,
+      "step": 1279
+    },
+    {
+      "epoch": 0.17616295072942473,
+      "grad_norm": 1.8407894657074446,
+      "learning_rate": 1.890033143956196e-05,
+      "loss": 1.1658,
+      "step": 1280
+    },
+    {
+      "epoch": 0.17630057803468208,
+      "grad_norm": 1.7872220499616576,
+      "learning_rate": 1.889829843431202e-05,
+      "loss": 1.1884,
+      "step": 1281
+    },
+    {
+      "epoch": 0.17643820533993945,
+      "grad_norm": 1.8052835353599588,
+      "learning_rate": 1.8896263661091986e-05,
+      "loss": 1.1714,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1765758326451968,
+      "grad_norm": 1.7854605477143324,
+      "learning_rate": 1.889422712030614e-05,
+      "loss": 1.1467,
+      "step": 1283
+    },
+    {
+      "epoch": 0.17671345995045418,
+      "grad_norm": 1.8278469103832629,
+      "learning_rate": 1.889218881235911e-05,
+      "loss": 1.2026,
+      "step": 1284
+    },
+    {
+      "epoch": 0.17685108725571153,
+      "grad_norm": 1.6175291682475705,
+      "learning_rate": 1.8890148737655883e-05,
+      "loss": 1.0917,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1769887145609689,
+      "grad_norm": 1.6946899292545088,
+      "learning_rate": 1.8888106896601803e-05,
+      "loss": 1.1542,
+      "step": 1286
+    },
+    {
+      "epoch": 0.17712634186622625,
+      "grad_norm": 1.8590474940446629,
+      "learning_rate": 1.888606328960254e-05,
+      "loss": 1.1743,
+      "step": 1287
+    },
+    {
+      "epoch": 0.17726396917148363,
+      "grad_norm": 1.8468986505771823,
+      "learning_rate": 1.888401791706414e-05,
+      "loss": 1.1482,
+      "step": 1288
+    },
+    {
+      "epoch": 0.17740159647674097,
+      "grad_norm": 1.695164655444383,
+      "learning_rate": 1.888197077939299e-05,
+      "loss": 1.1621,
+      "step": 1289
+    },
+    {
+      "epoch": 0.17753922378199835,
+      "grad_norm": 1.6479109288906586,
+      "learning_rate": 1.887992187699583e-05,
+      "loss": 1.1036,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1776768510872557,
+      "grad_norm": 1.828556680449922,
+      "learning_rate": 1.887787121027974e-05,
+      "loss": 1.1818,
+      "step": 1291
+    },
+    {
+      "epoch": 0.17781447839251308,
+      "grad_norm": 1.8302543225443302,
+      "learning_rate": 1.887581877965217e-05,
+      "loss": 1.1916,
+      "step": 1292
+    },
+    {
+      "epoch": 0.17795210569777042,
+      "grad_norm": 1.8513660691283713,
+      "learning_rate": 1.88737645855209e-05,
+      "loss": 1.1142,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1780897330030278,
+      "grad_norm": 1.6905249442918073,
+      "learning_rate": 1.887170862829408e-05,
+      "loss": 1.1957,
+      "step": 1294
+    },
+    {
+      "epoch": 0.17822736030828518,
+      "grad_norm": 1.7357174245832458,
+      "learning_rate": 1.8869650908380196e-05,
+      "loss": 1.1342,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17836498761354252,
+      "grad_norm": 1.718182362722972,
+      "learning_rate": 1.8867591426188084e-05,
+      "loss": 1.14,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1785026149187999,
+      "grad_norm": 1.6606057961257188,
+      "learning_rate": 1.8865530182126947e-05,
+      "loss": 1.2008,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17864024222405725,
+      "grad_norm": 1.8648644334662043,
+      "learning_rate": 1.8863467176606315e-05,
+      "loss": 1.3018,
+      "step": 1298
+    },
+    {
+      "epoch": 0.17877786952931463,
+      "grad_norm": 1.7374865475563974,
+      "learning_rate": 1.886140241003608e-05,
+      "loss": 1.1765,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17891549683457197,
+      "grad_norm": 1.763254211971766,
+      "learning_rate": 1.885933588282649e-05,
+      "loss": 1.2579,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17905312413982935,
+      "grad_norm": 1.883305768809243,
+      "learning_rate": 1.885726759538813e-05,
+      "loss": 1.1837,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1791907514450867,
+      "grad_norm": 1.533172977654244,
+      "learning_rate": 1.8855197548131938e-05,
+      "loss": 1.0219,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17932837875034408,
+      "grad_norm": 1.7348673710435476,
+      "learning_rate": 1.8853125741469212e-05,
+      "loss": 1.209,
+      "step": 1303
+    },
+    {
+      "epoch": 0.17946600605560142,
+      "grad_norm": 1.7165100110994085,
+      "learning_rate": 1.8851052175811587e-05,
+      "loss": 1.1811,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1796036333608588,
+      "grad_norm": 1.7687152793989553,
+      "learning_rate": 1.884897685157105e-05,
+      "loss": 1.1187,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17974126066611615,
+      "grad_norm": 1.5178890000689444,
+      "learning_rate": 1.8846899769159945e-05,
+      "loss": 1.0949,
+      "step": 1306
+    },
+    {
+      "epoch": 0.17987888797137352,
+      "grad_norm": 1.8314478797136182,
+      "learning_rate": 1.884482092899095e-05,
+      "loss": 1.135,
+      "step": 1307
+    },
+    {
+      "epoch": 0.18001651527663087,
+      "grad_norm": 1.8669879250925665,
+      "learning_rate": 1.8842740331477116e-05,
+      "loss": 1.1846,
+      "step": 1308
+    },
+    {
+      "epoch": 0.18015414258188825,
+      "grad_norm": 1.6245078516656457,
+      "learning_rate": 1.884065797703182e-05,
+      "loss": 1.133,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1802917698871456,
+      "grad_norm": 2.1141938380423846,
+      "learning_rate": 1.88385738660688e-05,
+      "loss": 1.3486,
+      "step": 1310
+    },
+    {
+      "epoch": 0.18042939719240297,
+      "grad_norm": 1.7572692345075003,
+      "learning_rate": 1.8836487999002136e-05,
+      "loss": 1.1952,
+      "step": 1311
+    },
+    {
+      "epoch": 0.18056702449766032,
+      "grad_norm": 1.7089898825007488,
+      "learning_rate": 1.883440037624627e-05,
+      "loss": 1.1283,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1807046518029177,
+      "grad_norm": 1.7969319275632885,
+      "learning_rate": 1.8832310998215975e-05,
+      "loss": 1.2225,
+      "step": 1313
+    },
+    {
+      "epoch": 0.18084227910817507,
+      "grad_norm": 1.837975467830707,
+      "learning_rate": 1.8830219865326386e-05,
+      "loss": 1.2251,
+      "step": 1314
+    },
+    {
+      "epoch": 0.18097990641343242,
+      "grad_norm": 1.7986163843209684,
+      "learning_rate": 1.8828126977992984e-05,
+      "loss": 1.1381,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1811175337186898,
+      "grad_norm": 1.6292621695023655,
+      "learning_rate": 1.8826032336631598e-05,
+      "loss": 1.2114,
+      "step": 1316
+    },
+    {
+      "epoch": 0.18125516102394715,
+      "grad_norm": 1.7507397304605983,
+      "learning_rate": 1.8823935941658396e-05,
+      "loss": 1.1796,
+      "step": 1317
+    },
+    {
+      "epoch": 0.18139278832920452,
+      "grad_norm": 1.7041657609668732,
+      "learning_rate": 1.8821837793489912e-05,
+      "loss": 1.1318,
+      "step": 1318
+    },
+    {
+      "epoch": 0.18153041563446187,
+      "grad_norm": 1.7979476039567885,
+      "learning_rate": 1.8819737892543018e-05,
+      "loss": 1.1972,
+      "step": 1319
+    },
+    {
+      "epoch": 0.18166804293971925,
+      "grad_norm": 1.804581707685713,
+      "learning_rate": 1.881763623923493e-05,
+      "loss": 1.165,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1818056702449766,
+      "grad_norm": 1.5937108838679828,
+      "learning_rate": 1.8815532833983226e-05,
+      "loss": 1.2164,
+      "step": 1321
+    },
+    {
+      "epoch": 0.18194329755023397,
+      "grad_norm": 1.9142066558573876,
+      "learning_rate": 1.8813427677205815e-05,
+      "loss": 1.1715,
+      "step": 1322
+    },
+    {
+      "epoch": 0.18208092485549132,
+      "grad_norm": 1.8834601707699743,
+      "learning_rate": 1.881132076932097e-05,
+      "loss": 1.2342,
+      "step": 1323
+    },
+    {
+      "epoch": 0.1822185521607487,
+      "grad_norm": 1.8048255607689512,
+      "learning_rate": 1.88092121107473e-05,
+      "loss": 1.1372,
+      "step": 1324
+    },
+    {
+      "epoch": 0.18235617946600605,
+      "grad_norm": 1.613665127291824,
+      "learning_rate": 1.8807101701903773e-05,
+      "loss": 1.1692,
+      "step": 1325
+    },
+    {
+      "epoch": 0.18249380677126342,
+      "grad_norm": 1.9695684202270132,
+      "learning_rate": 1.880498954320969e-05,
+      "loss": 1.184,
+      "step": 1326
+    },
+    {
+      "epoch": 0.18263143407652077,
+      "grad_norm": 1.767869229084148,
+      "learning_rate": 1.8802875635084708e-05,
+      "loss": 1.1809,
+      "step": 1327
+    },
+    {
+      "epoch": 0.18276906138177815,
+      "grad_norm": 1.8020965822905677,
+      "learning_rate": 1.8800759977948844e-05,
+      "loss": 1.0834,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1829066886870355,
+      "grad_norm": 1.6091152412115173,
+      "learning_rate": 1.8798642572222437e-05,
+      "loss": 1.0841,
+      "step": 1329
+    },
+    {
+      "epoch": 0.18304431599229287,
+      "grad_norm": 1.6864269655680584,
+      "learning_rate": 1.8796523418326193e-05,
+      "loss": 1.0901,
+      "step": 1330
+    },
+    {
+      "epoch": 0.18318194329755022,
+      "grad_norm": 1.9344097860333702,
+      "learning_rate": 1.8794402516681153e-05,
+      "loss": 1.1277,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1833195706028076,
+      "grad_norm": 1.6634023661180872,
+      "learning_rate": 1.879227986770872e-05,
+      "loss": 1.1867,
+      "step": 1332
+    },
+    {
+      "epoch": 0.18345719790806497,
+      "grad_norm": 1.8378425885234546,
+      "learning_rate": 1.879015547183063e-05,
+      "loss": 1.2014,
+      "step": 1333
+    },
+    {
+      "epoch": 0.18359482521332232,
+      "grad_norm": 1.7514022464196024,
+      "learning_rate": 1.878802932946897e-05,
+      "loss": 1.1555,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1837324525185797,
+      "grad_norm": 1.8422191997613435,
+      "learning_rate": 1.8785901441046178e-05,
+      "loss": 1.213,
+      "step": 1335
+    },
+    {
+      "epoch": 0.18387007982383705,
+      "grad_norm": 1.5228323238414818,
+      "learning_rate": 1.8783771806985035e-05,
+      "loss": 1.1858,
+      "step": 1336
+    },
+    {
+      "epoch": 0.18400770712909442,
+      "grad_norm": 1.5988258305602032,
+      "learning_rate": 1.878164042770867e-05,
+      "loss": 1.045,
+      "step": 1337
+    },
+    {
+      "epoch": 0.18414533443435177,
+      "grad_norm": 1.919091080769259,
+      "learning_rate": 1.8779507303640557e-05,
+      "loss": 1.132,
+      "step": 1338
+    },
+    {
+      "epoch": 0.18428296173960915,
+      "grad_norm": 2.432965915591615,
+      "learning_rate": 1.8777372435204526e-05,
+      "loss": 1.0891,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1844205890448665,
+      "grad_norm": 1.7950952441408612,
+      "learning_rate": 1.877523582282474e-05,
+      "loss": 1.1232,
+      "step": 1340
+    },
+    {
+      "epoch": 0.18455821635012387,
+      "grad_norm": 1.8445019801893467,
+      "learning_rate": 1.877309746692572e-05,
+      "loss": 1.1971,
+      "step": 1341
+    },
+    {
+      "epoch": 0.18469584365538122,
+      "grad_norm": 1.6882410744905951,
+      "learning_rate": 1.8770957367932318e-05,
+      "loss": 1.1237,
+      "step": 1342
+    },
+    {
+      "epoch": 0.1848334709606386,
+      "grad_norm": 1.9528681955209033,
+      "learning_rate": 1.876881552626975e-05,
+      "loss": 1.2132,
+      "step": 1343
+    },
+    {
+      "epoch": 0.18497109826589594,
+      "grad_norm": 1.7398127168625208,
+      "learning_rate": 1.8766671942363566e-05,
+      "loss": 1.1674,
+      "step": 1344
+    },
+    {
+      "epoch": 0.18510872557115332,
+      "grad_norm": 1.7402533363358288,
+      "learning_rate": 1.8764526616639675e-05,
+      "loss": 1.1232,
+      "step": 1345
+    },
+    {
+      "epoch": 0.18524635287641067,
+      "grad_norm": 1.71242725300125,
+      "learning_rate": 1.8762379549524318e-05,
+      "loss": 1.1123,
+      "step": 1346
+    },
+    {
+      "epoch": 0.18538398018166805,
+      "grad_norm": 1.6092102332589888,
+      "learning_rate": 1.8760230741444086e-05,
+      "loss": 1.0973,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1855216074869254,
+      "grad_norm": 1.667671721285342,
+      "learning_rate": 1.8758080192825917e-05,
+      "loss": 1.16,
+      "step": 1348
+    },
+    {
+      "epoch": 0.18565923479218277,
+      "grad_norm": 1.7004956906942499,
+      "learning_rate": 1.8755927904097105e-05,
+      "loss": 1.1257,
+      "step": 1349
+    },
+    {
+      "epoch": 0.18579686209744012,
+      "grad_norm": 1.6074914674499095,
+      "learning_rate": 1.875377387568527e-05,
+      "loss": 1.0837,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1859344894026975,
+      "grad_norm": 1.9916854166724056,
+      "learning_rate": 1.8751618108018396e-05,
+      "loss": 1.1069,
+      "step": 1351
+    },
+    {
+      "epoch": 0.18607211670795487,
+      "grad_norm": 1.6729894212690504,
+      "learning_rate": 1.8749460601524796e-05,
+      "loss": 1.1603,
+      "step": 1352
+    },
+    {
+      "epoch": 0.18620974401321222,
+      "grad_norm": 1.8747355316824585,
+      "learning_rate": 1.8747301356633145e-05,
+      "loss": 1.1904,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1863473713184696,
+      "grad_norm": 1.9292232760193697,
+      "learning_rate": 1.8745140373772452e-05,
+      "loss": 1.1354,
+      "step": 1354
+    },
+    {
+      "epoch": 0.18648499862372694,
+      "grad_norm": 1.9161753248927433,
+      "learning_rate": 1.874297765337207e-05,
+      "loss": 1.0891,
+      "step": 1355
+    },
+    {
+      "epoch": 0.18662262592898432,
+      "grad_norm": 1.7380786286486736,
+      "learning_rate": 1.8740813195861713e-05,
+      "loss": 1.0881,
+      "step": 1356
+    },
+    {
+      "epoch": 0.18676025323424167,
+      "grad_norm": 1.8106954572929312,
+      "learning_rate": 1.873864700167142e-05,
+      "loss": 1.1598,
+      "step": 1357
+    },
+    {
+      "epoch": 0.18689788053949905,
+      "grad_norm": 1.7807308826423203,
+      "learning_rate": 1.8736479071231587e-05,
+      "loss": 1.2506,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1870355078447564,
+      "grad_norm": 1.6481367892060674,
+      "learning_rate": 1.8734309404972946e-05,
+      "loss": 1.1045,
+      "step": 1359
+    },
+    {
+      "epoch": 0.18717313515001377,
+      "grad_norm": 1.699384823592173,
+      "learning_rate": 1.8732138003326597e-05,
+      "loss": 1.0641,
+      "step": 1360
+    },
+    {
+      "epoch": 0.18731076245527112,
+      "grad_norm": 1.7372377477945322,
+      "learning_rate": 1.872996486672395e-05,
+      "loss": 1.1346,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1874483897605285,
+      "grad_norm": 1.7395479090438313,
+      "learning_rate": 1.8727789995596784e-05,
+      "loss": 1.1675,
+      "step": 1362
+    },
+    {
+      "epoch": 0.18758601706578584,
+      "grad_norm": 1.8600412255048144,
+      "learning_rate": 1.872561339037722e-05,
+      "loss": 1.1278,
+      "step": 1363
+    },
+    {
+      "epoch": 0.18772364437104322,
+      "grad_norm": 1.6258840477338212,
+      "learning_rate": 1.872343505149771e-05,
+      "loss": 1.0879,
+      "step": 1364
+    },
+    {
+      "epoch": 0.18786127167630057,
+      "grad_norm": 1.721394025625272,
+      "learning_rate": 1.8721254979391072e-05,
+      "loss": 1.2177,
+      "step": 1365
+    },
+    {
+      "epoch": 0.18799889898155794,
+      "grad_norm": 1.7347207094336736,
+      "learning_rate": 1.871907317449045e-05,
+      "loss": 1.2197,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1881365262868153,
+      "grad_norm": 1.836533971747441,
+      "learning_rate": 1.8716889637229334e-05,
+      "loss": 1.1957,
+      "step": 1367
+    },
+    {
+      "epoch": 0.18827415359207267,
+      "grad_norm": 1.7688364946797064,
+      "learning_rate": 1.871470436804157e-05,
+      "loss": 1.1222,
+      "step": 1368
+    },
+    {
+      "epoch": 0.18841178089733002,
+      "grad_norm": 1.5318304831719753,
+      "learning_rate": 1.871251736736134e-05,
+      "loss": 1.1265,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1885494082025874,
+      "grad_norm": 1.63723809086378,
+      "learning_rate": 1.8710328635623172e-05,
+      "loss": 1.1687,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18868703550784477,
+      "grad_norm": 1.7732303396194524,
+      "learning_rate": 1.8708138173261936e-05,
+      "loss": 1.2927,
+      "step": 1371
+    },
+    {
+      "epoch": 0.18882466281310212,
+      "grad_norm": 1.5929481110051298,
+      "learning_rate": 1.8705945980712845e-05,
+      "loss": 1.1643,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1889622901183595,
+      "grad_norm": 1.6634775350897766,
+      "learning_rate": 1.870375205841146e-05,
+      "loss": 1.1021,
+      "step": 1373
+    },
+    {
+      "epoch": 0.18909991742361684,
+      "grad_norm": 1.7255356183089035,
+      "learning_rate": 1.8701556406793682e-05,
+      "loss": 1.1641,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18923754472887422,
+      "grad_norm": 1.6461185944999166,
+      "learning_rate": 1.8699359026295755e-05,
+      "loss": 1.1327,
+      "step": 1375
+    },
+    {
+      "epoch": 0.18937517203413157,
+      "grad_norm": 1.5037595609581154,
+      "learning_rate": 1.8697159917354275e-05,
+      "loss": 1.1332,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18951279933938894,
+      "grad_norm": 1.7487701950460637,
+      "learning_rate": 1.869495908040617e-05,
+      "loss": 1.1696,
+      "step": 1377
+    },
+    {
+      "epoch": 0.1896504266446463,
+      "grad_norm": 1.8554712375099518,
+      "learning_rate": 1.8692756515888714e-05,
+      "loss": 1.1761,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18978805394990367,
+      "grad_norm": 1.73223124422111,
+      "learning_rate": 1.869055222423953e-05,
+      "loss": 1.1074,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18992568125516102,
+      "grad_norm": 1.728830299557569,
+      "learning_rate": 1.868834620589658e-05,
+      "loss": 1.1494,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1900633085604184,
+      "grad_norm": 1.5698559674756527,
+      "learning_rate": 1.868613846129817e-05,
+      "loss": 1.1132,
+      "step": 1381
+    },
+    {
+      "epoch": 0.19020093586567574,
+      "grad_norm": 1.866817330061516,
+      "learning_rate": 1.8683928990882946e-05,
+      "loss": 1.2559,
+      "step": 1382
+    },
+    {
+      "epoch": 0.19033856317093312,
+      "grad_norm": 1.7623575350668947,
+      "learning_rate": 1.8681717795089903e-05,
+      "loss": 1.1889,
+      "step": 1383
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 1.474895816521599,
+      "learning_rate": 1.8679504874358376e-05,
+      "loss": 1.2057,
+      "step": 1384
+    },
+    {
+      "epoch": 0.19061381778144784,
+      "grad_norm": 1.576435309019052,
+      "learning_rate": 1.8677290229128034e-05,
+      "loss": 1.1238,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1907514450867052,
+      "grad_norm": 1.9662905088005784,
+      "learning_rate": 1.867507385983891e-05,
+      "loss": 1.2325,
+      "step": 1386
+    },
+    {
+      "epoch": 0.19088907239196257,
+      "grad_norm": 1.6019653318570228,
+      "learning_rate": 1.8672855766931354e-05,
+      "loss": 1.1174,
+      "step": 1387
+    },
+    {
+      "epoch": 0.19102669969721991,
+      "grad_norm": 1.5541417147207683,
+      "learning_rate": 1.8670635950846077e-05,
+      "loss": 1.0783,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1911643270024773,
+      "grad_norm": 1.9478059186171508,
+      "learning_rate": 1.8668414412024124e-05,
+      "loss": 1.1729,
+      "step": 1389
+    },
+    {
+      "epoch": 0.19130195430773467,
+      "grad_norm": 1.9248375436355005,
+      "learning_rate": 1.8666191150906888e-05,
+      "loss": 1.2866,
+      "step": 1390
+    },
+    {
+      "epoch": 0.19143958161299202,
+      "grad_norm": 1.6685985209711145,
+      "learning_rate": 1.8663966167936093e-05,
+      "loss": 1.2177,
+      "step": 1391
+    },
+    {
+      "epoch": 0.1915772089182494,
+      "grad_norm": 1.7826729917476112,
+      "learning_rate": 1.866173946355382e-05,
+      "loss": 1.1118,
+      "step": 1392
+    },
+    {
+      "epoch": 0.19171483622350674,
+      "grad_norm": 1.8903543004752952,
+      "learning_rate": 1.865951103820248e-05,
+      "loss": 1.2123,
+      "step": 1393
+    },
+    {
+      "epoch": 0.19185246352876412,
+      "grad_norm": 1.7949355284410375,
+      "learning_rate": 1.8657280892324833e-05,
+      "loss": 1.165,
+      "step": 1394
+    },
+    {
+      "epoch": 0.19199009083402147,
+      "grad_norm": 1.790311976743421,
+      "learning_rate": 1.865504902636398e-05,
+      "loss": 1.2449,
+      "step": 1395
+    },
+    {
+      "epoch": 0.19212771813927884,
+      "grad_norm": 1.8683394734432548,
+      "learning_rate": 1.865281544076336e-05,
+      "loss": 1.1502,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1922653454445362,
+      "grad_norm": 1.6800404273773077,
+      "learning_rate": 1.8650580135966755e-05,
+      "loss": 1.1591,
+      "step": 1397
+    },
+    {
+      "epoch": 0.19240297274979357,
+      "grad_norm": 1.87458100075439,
+      "learning_rate": 1.8648343112418287e-05,
+      "loss": 1.1836,
+      "step": 1398
+    },
+    {
+      "epoch": 0.19254060005505091,
+      "grad_norm": 1.7034136822126607,
+      "learning_rate": 1.864610437056243e-05,
+      "loss": 1.1262,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1926782273603083,
+      "grad_norm": 1.6799033596699697,
+      "learning_rate": 1.8643863910843986e-05,
+      "loss": 1.1333,
+      "step": 1400
+    },
+    {
+      "epoch": 0.19281585466556564,
+      "grad_norm": 1.8758567176964704,
+      "learning_rate": 1.8641621733708103e-05,
+      "loss": 1.1017,
+      "step": 1401
+    },
+    {
+      "epoch": 0.19295348197082302,
+      "grad_norm": 1.750418138836148,
+      "learning_rate": 1.8639377839600274e-05,
+      "loss": 1.1213,
+      "step": 1402
+    },
+    {
+      "epoch": 0.19309110927608036,
+      "grad_norm": 1.9846878188428023,
+      "learning_rate": 1.863713222896633e-05,
+      "loss": 1.2122,
+      "step": 1403
+    },
+    {
+      "epoch": 0.19322873658133774,
+      "grad_norm": 2.0712533831888607,
+      "learning_rate": 1.8634884902252438e-05,
+      "loss": 1.1835,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1933663638865951,
+      "grad_norm": 1.7025072088794955,
+      "learning_rate": 1.8632635859905118e-05,
+      "loss": 1.1099,
+      "step": 1405
+    },
+    {
+      "epoch": 0.19350399119185246,
+      "grad_norm": 1.8184803734768709,
+      "learning_rate": 1.863038510237122e-05,
+      "loss": 1.0742,
+      "step": 1406
+    },
+    {
+      "epoch": 0.1936416184971098,
+      "grad_norm": 1.7729785930947264,
+      "learning_rate": 1.8628132630097936e-05,
+      "loss": 1.1315,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1937792458023672,
+      "grad_norm": 1.9028281421901685,
+      "learning_rate": 1.8625878443532808e-05,
+      "loss": 1.261,
+      "step": 1408
+    },
+    {
+      "epoch": 0.19391687310762457,
+      "grad_norm": 1.729115802343711,
+      "learning_rate": 1.8623622543123705e-05,
+      "loss": 1.2271,
+      "step": 1409
+    },
+    {
+      "epoch": 0.19405450041288191,
+      "grad_norm": 1.8999807364485768,
+      "learning_rate": 1.862136492931885e-05,
+      "loss": 1.2311,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1941921277181393,
+      "grad_norm": 1.7067465313617611,
+      "learning_rate": 1.8619105602566797e-05,
+      "loss": 1.1575,
+      "step": 1411
+    },
+    {
+      "epoch": 0.19432975502339664,
+      "grad_norm": 1.727526623999419,
+      "learning_rate": 1.8616844563316444e-05,
+      "loss": 1.0584,
+      "step": 1412
+    },
+    {
+      "epoch": 0.19446738232865401,
+      "grad_norm": 1.9034677298256828,
+      "learning_rate": 1.8614581812017026e-05,
+      "loss": 1.1266,
+      "step": 1413
+    },
+    {
+      "epoch": 0.19460500963391136,
+      "grad_norm": 1.6851083411792012,
+      "learning_rate": 1.8612317349118125e-05,
+      "loss": 1.1844,
+      "step": 1414
+    },
+    {
+      "epoch": 0.19474263693916874,
+      "grad_norm": 1.8460375479801978,
+      "learning_rate": 1.8610051175069658e-05,
+      "loss": 1.2395,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1948802642444261,
+      "grad_norm": 1.8675280164641854,
+      "learning_rate": 1.8607783290321876e-05,
+      "loss": 1.1573,
+      "step": 1416
+    },
+    {
+      "epoch": 0.19501789154968346,
+      "grad_norm": 1.887737228426763,
+      "learning_rate": 1.8605513695325385e-05,
+      "loss": 1.1635,
+      "step": 1417
+    },
+    {
+      "epoch": 0.1951555188549408,
+      "grad_norm": 1.5812824623528716,
+      "learning_rate": 1.860324239053112e-05,
+      "loss": 1.0703,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1952931461601982,
+      "grad_norm": 1.9122706301212489,
+      "learning_rate": 1.8600969376390354e-05,
+      "loss": 1.2402,
+      "step": 1419
+    },
+    {
+      "epoch": 0.19543077346545554,
+      "grad_norm": 1.8098456965620218,
+      "learning_rate": 1.8598694653354712e-05,
+      "loss": 1.1187,
+      "step": 1420
+    },
+    {
+      "epoch": 0.1955684007707129,
+      "grad_norm": 1.6370595709752755,
+      "learning_rate": 1.8596418221876143e-05,
+      "loss": 1.1025,
+      "step": 1421
+    },
+    {
+      "epoch": 0.19570602807597026,
+      "grad_norm": 1.6850778271488844,
+      "learning_rate": 1.8594140082406946e-05,
+      "loss": 1.1944,
+      "step": 1422
+    },
+    {
+      "epoch": 0.19584365538122764,
+      "grad_norm": 1.630424868183961,
+      "learning_rate": 1.8591860235399758e-05,
+      "loss": 1.1156,
+      "step": 1423
+    },
+    {
+      "epoch": 0.195981282686485,
+      "grad_norm": 1.6978019669830673,
+      "learning_rate": 1.8589578681307548e-05,
+      "loss": 1.168,
+      "step": 1424
+    },
+    {
+      "epoch": 0.19611890999174236,
+      "grad_norm": 2.0712098179296796,
+      "learning_rate": 1.858729542058363e-05,
+      "loss": 1.1526,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1962565372969997,
+      "grad_norm": 1.702740143532348,
+      "learning_rate": 1.8585010453681665e-05,
+      "loss": 1.1472,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1963941646022571,
+      "grad_norm": 1.7197165767658424,
+      "learning_rate": 1.8582723781055637e-05,
+      "loss": 1.1622,
+      "step": 1427
+    },
+    {
+      "epoch": 0.19653179190751446,
+      "grad_norm": 1.760041143670627,
+      "learning_rate": 1.8580435403159875e-05,
+      "loss": 1.0441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1966694192127718,
+      "grad_norm": 1.553361656885028,
+      "learning_rate": 1.8578145320449055e-05,
+      "loss": 1.1338,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1968070465180292,
+      "grad_norm": 1.7111594284557685,
+      "learning_rate": 1.857585353337818e-05,
+      "loss": 1.1329,
+      "step": 1430
+    },
+    {
+      "epoch": 0.19694467382328654,
+      "grad_norm": 1.679618901409045,
+      "learning_rate": 1.85735600424026e-05,
+      "loss": 1.123,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1970823011285439,
+      "grad_norm": 1.4946698972002228,
+      "learning_rate": 1.8571264847978e-05,
+      "loss": 1.057,
+      "step": 1432
+    },
+    {
+      "epoch": 0.19721992843380126,
+      "grad_norm": 1.5939227998254473,
+      "learning_rate": 1.85689679505604e-05,
+      "loss": 1.1829,
+      "step": 1433
+    },
+    {
+      "epoch": 0.19735755573905864,
+      "grad_norm": 2.048369781363598,
+      "learning_rate": 1.856666935060617e-05,
+      "loss": 1.2109,
+      "step": 1434
+    },
+    {
+      "epoch": 0.197495183044316,
+      "grad_norm": 1.5599801101418271,
+      "learning_rate": 1.8564369048572e-05,
+      "loss": 1.0624,
+      "step": 1435
+    },
+    {
+      "epoch": 0.19763281034957336,
+      "grad_norm": 1.7181288027230486,
+      "learning_rate": 1.8562067044914935e-05,
+      "loss": 1.1065,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1977704376548307,
+      "grad_norm": 1.6160246748302163,
+      "learning_rate": 1.8559763340092348e-05,
+      "loss": 1.1735,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1979080649600881,
+      "grad_norm": 1.6668393092741962,
+      "learning_rate": 1.8557457934561956e-05,
+      "loss": 1.1315,
+      "step": 1438
+    },
+    {
+      "epoch": 0.19804569226534544,
+      "grad_norm": 1.7296865550108709,
+      "learning_rate": 1.8555150828781815e-05,
+      "loss": 1.1274,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1981833195706028,
+      "grad_norm": 1.653796195649246,
+      "learning_rate": 1.8552842023210303e-05,
+      "loss": 1.1277,
+      "step": 1440
+    },
+    {
+      "epoch": 0.19832094687586016,
+      "grad_norm": 1.776718986461215,
+      "learning_rate": 1.8550531518306165e-05,
+      "loss": 1.1705,
+      "step": 1441
+    },
+    {
+      "epoch": 0.19845857418111754,
+      "grad_norm": 1.5619998802309825,
+      "learning_rate": 1.8548219314528455e-05,
+      "loss": 1.0974,
+      "step": 1442
+    },
+    {
+      "epoch": 0.19859620148637488,
+      "grad_norm": 1.74124357188006,
+      "learning_rate": 1.8545905412336582e-05,
+      "loss": 1.1785,
+      "step": 1443
+    },
+    {
+      "epoch": 0.19873382879163226,
+      "grad_norm": 1.6661727338002352,
+      "learning_rate": 1.8543589812190284e-05,
+      "loss": 1.1966,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1988714560968896,
+      "grad_norm": 1.5680752755518732,
+      "learning_rate": 1.8541272514549637e-05,
+      "loss": 1.096,
+      "step": 1445
+    },
+    {
+      "epoch": 0.19900908340214699,
+      "grad_norm": 1.6334510961202302,
+      "learning_rate": 1.853895351987506e-05,
+      "loss": 1.1594,
+      "step": 1446
+    },
+    {
+      "epoch": 0.19914671070740436,
+      "grad_norm": 1.7849088492686243,
+      "learning_rate": 1.8536632828627303e-05,
+      "loss": 1.1868,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1992843380126617,
+      "grad_norm": 1.6494135324323504,
+      "learning_rate": 1.853431044126746e-05,
+      "loss": 1.0938,
+      "step": 1448
+    },
+    {
+      "epoch": 0.1994219653179191,
+      "grad_norm": 1.7806091447157413,
+      "learning_rate": 1.853198635825695e-05,
+      "loss": 1.1449,
+      "step": 1449
+    },
+    {
+      "epoch": 0.19955959262317643,
+      "grad_norm": 1.635779114478088,
+      "learning_rate": 1.852966058005754e-05,
+      "loss": 1.1249,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1996972199284338,
+      "grad_norm": 1.6197146437437697,
+      "learning_rate": 1.8527333107131336e-05,
+      "loss": 1.2314,
+      "step": 1451
+    },
+    {
+      "epoch": 0.19983484723369116,
+      "grad_norm": 1.7534321757255151,
+      "learning_rate": 1.8525003939940767e-05,
+      "loss": 1.1674,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19997247453894854,
+      "grad_norm": 1.7818009599403286,
+      "learning_rate": 1.852267307894861e-05,
+      "loss": 1.0936,
+      "step": 1453
+    },
+    {
+      "epoch": 0.20011010184420588,
+      "grad_norm": 1.814397905873886,
+      "learning_rate": 1.8520340524617976e-05,
+      "loss": 1.1068,
+      "step": 1454
+    },
+    {
+      "epoch": 0.20024772914946326,
+      "grad_norm": 1.818530314058793,
+      "learning_rate": 1.851800627741231e-05,
+      "loss": 1.2086,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2003853564547206,
+      "grad_norm": 1.7067755781520677,
+      "learning_rate": 1.8515670337795395e-05,
+      "loss": 1.1804,
+      "step": 1456
+    },
+    {
+      "epoch": 0.20052298375997799,
+      "grad_norm": 1.7644386314039746,
+      "learning_rate": 1.851333270623135e-05,
+      "loss": 1.1891,
+      "step": 1457
+    },
+    {
+      "epoch": 0.20066061106523533,
+      "grad_norm": 1.659899609931917,
+      "learning_rate": 1.851099338318463e-05,
+      "loss": 1.1942,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2007982383704927,
+      "grad_norm": 1.8032983252222388,
+      "learning_rate": 1.850865236912003e-05,
+      "loss": 1.1726,
+      "step": 1459
+    },
+    {
+      "epoch": 0.20093586567575006,
+      "grad_norm": 1.8819256036773617,
+      "learning_rate": 1.8506309664502677e-05,
+      "loss": 1.1104,
+      "step": 1460
+    },
+    {
+      "epoch": 0.20107349298100743,
+      "grad_norm": 1.5615530173714074,
+      "learning_rate": 1.850396526979803e-05,
+      "loss": 1.0727,
+      "step": 1461
+    },
+    {
+      "epoch": 0.20121112028626478,
+      "grad_norm": 1.6889936496309117,
+      "learning_rate": 1.8501619185471892e-05,
+      "loss": 1.2714,
+      "step": 1462
+    },
+    {
+      "epoch": 0.20134874759152216,
+      "grad_norm": 1.6447418620446248,
+      "learning_rate": 1.8499271411990393e-05,
+      "loss": 1.1642,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2014863748967795,
+      "grad_norm": 1.8079970925014657,
+      "learning_rate": 1.849692194982001e-05,
+      "loss": 1.1666,
+      "step": 1464
+    },
+    {
+      "epoch": 0.20162400220203688,
+      "grad_norm": 1.9881441777965214,
+      "learning_rate": 1.8494570799427546e-05,
+      "loss": 1.2401,
+      "step": 1465
+    },
+    {
+      "epoch": 0.20176162950729426,
+      "grad_norm": 1.7507568876818547,
+      "learning_rate": 1.849221796128014e-05,
+      "loss": 1.0601,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2018992568125516,
+      "grad_norm": 1.666751208017914,
+      "learning_rate": 1.848986343584527e-05,
+      "loss": 1.1561,
+      "step": 1467
+    },
+    {
+      "epoch": 0.20203688411780898,
+      "grad_norm": 1.7431047779107722,
+      "learning_rate": 1.848750722359075e-05,
+      "loss": 1.1831,
+      "step": 1468
+    },
+    {
+      "epoch": 0.20217451142306633,
+      "grad_norm": 1.7704530634393503,
+      "learning_rate": 1.8485149324984727e-05,
+      "loss": 1.1704,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2023121387283237,
+      "grad_norm": 1.6244326440711412,
+      "learning_rate": 1.8482789740495683e-05,
+      "loss": 1.0498,
+      "step": 1470
+    },
+    {
+      "epoch": 0.20244976603358106,
+      "grad_norm": 1.6526028101202077,
+      "learning_rate": 1.8480428470592434e-05,
+      "loss": 1.0975,
+      "step": 1471
+    },
+    {
+      "epoch": 0.20258739333883843,
+      "grad_norm": 1.5858887814994425,
+      "learning_rate": 1.8478065515744127e-05,
+      "loss": 1.092,
+      "step": 1472
+    },
+    {
+      "epoch": 0.20272502064409578,
+      "grad_norm": 1.6222728951109036,
+      "learning_rate": 1.847570087642026e-05,
+      "loss": 1.1763,
+      "step": 1473
+    },
+    {
+      "epoch": 0.20286264794935316,
+      "grad_norm": 1.7067187705398859,
+      "learning_rate": 1.847333455309065e-05,
+      "loss": 1.0824,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2030002752546105,
+      "grad_norm": 1.8514270786862388,
+      "learning_rate": 1.8470966546225442e-05,
+      "loss": 1.1826,
+      "step": 1475
+    },
+    {
+      "epoch": 0.20313790255986788,
+      "grad_norm": 1.6495514950651438,
+      "learning_rate": 1.8468596856295145e-05,
+      "loss": 1.1695,
+      "step": 1476
+    },
+    {
+      "epoch": 0.20327552986512523,
+      "grad_norm": 1.7043857334280428,
+      "learning_rate": 1.8466225483770574e-05,
+      "loss": 1.1265,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2034131571703826,
+      "grad_norm": 1.7505422894365834,
+      "learning_rate": 1.846385242912289e-05,
+      "loss": 1.2209,
+      "step": 1478
+    },
+    {
+      "epoch": 0.20355078447563996,
+      "grad_norm": 1.8372738039420844,
+      "learning_rate": 1.8461477692823582e-05,
+      "loss": 1.1598,
+      "step": 1479
+    },
+    {
+      "epoch": 0.20368841178089733,
+      "grad_norm": 1.747337738666948,
+      "learning_rate": 1.8459101275344482e-05,
+      "loss": 1.1163,
+      "step": 1480
+    },
+    {
+      "epoch": 0.20382603908615468,
+      "grad_norm": 1.68488409175198,
+      "learning_rate": 1.8456723177157753e-05,
+      "loss": 1.1136,
+      "step": 1481
+    },
+    {
+      "epoch": 0.20396366639141206,
+      "grad_norm": 1.7814469235023478,
+      "learning_rate": 1.845434339873589e-05,
+      "loss": 1.1094,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2041012936966694,
+      "grad_norm": 1.531909168755628,
+      "learning_rate": 1.8451961940551716e-05,
+      "loss": 1.1319,
+      "step": 1483
+    },
+    {
+      "epoch": 0.20423892100192678,
+      "grad_norm": 1.7078744515602595,
+      "learning_rate": 1.8449578803078402e-05,
+      "loss": 1.1861,
+      "step": 1484
+    },
+    {
+      "epoch": 0.20437654830718416,
+      "grad_norm": 1.7890446991180813,
+      "learning_rate": 1.8447193986789442e-05,
+      "loss": 1.1799,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2045141756124415,
+      "grad_norm": 1.6853028501671425,
+      "learning_rate": 1.8444807492158668e-05,
+      "loss": 1.1074,
+      "step": 1486
+    },
+    {
+      "epoch": 0.20465180291769888,
+      "grad_norm": 1.7548640045548987,
+      "learning_rate": 1.844241931966024e-05,
+      "loss": 1.1582,
+      "step": 1487
+    },
+    {
+      "epoch": 0.20478943022295623,
+      "grad_norm": 1.5582333318690997,
+      "learning_rate": 1.8440029469768656e-05,
+      "loss": 1.1635,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2049270575282136,
+      "grad_norm": 1.7127795758252853,
+      "learning_rate": 1.843763794295875e-05,
+      "loss": 1.1038,
+      "step": 1489
+    },
+    {
+      "epoch": 0.20506468483347096,
+      "grad_norm": 1.7348784455099946,
+      "learning_rate": 1.8435244739705683e-05,
+      "loss": 1.0809,
+      "step": 1490
+    },
+    {
+      "epoch": 0.20520231213872833,
+      "grad_norm": 1.7950534366288966,
+      "learning_rate": 1.8432849860484953e-05,
+      "loss": 1.1106,
+      "step": 1491
+    },
+    {
+      "epoch": 0.20533993944398568,
+      "grad_norm": 1.828597656454292,
+      "learning_rate": 1.843045330577239e-05,
+      "loss": 1.1708,
+      "step": 1492
+    },
+    {
+      "epoch": 0.20547756674924306,
+      "grad_norm": 1.6432822710817476,
+      "learning_rate": 1.8428055076044152e-05,
+      "loss": 1.2025,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2056151940545004,
+      "grad_norm": 1.7223027931756794,
+      "learning_rate": 1.842565517177674e-05,
+      "loss": 1.1548,
+      "step": 1494
+    },
+    {
+      "epoch": 0.20575282135975778,
+      "grad_norm": 1.7781453526983706,
+      "learning_rate": 1.842325359344698e-05,
+      "loss": 1.2065,
+      "step": 1495
+    },
+    {
+      "epoch": 0.20589044866501513,
+      "grad_norm": 1.7876647041451774,
+      "learning_rate": 1.842085034153203e-05,
+      "loss": 1.0787,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2060280759702725,
+      "grad_norm": 1.6396241831269793,
+      "learning_rate": 1.841844541650939e-05,
+      "loss": 1.1253,
+      "step": 1497
+    },
+    {
+      "epoch": 0.20616570327552985,
+      "grad_norm": 1.8184726392000505,
+      "learning_rate": 1.841603881885688e-05,
+      "loss": 1.1897,
+      "step": 1498
+    },
+    {
+      "epoch": 0.20630333058078723,
+      "grad_norm": 1.7051772473613152,
+      "learning_rate": 1.841363054905266e-05,
+      "loss": 1.0871,
+      "step": 1499
+    },
+    {
+      "epoch": 0.20644095788604458,
+      "grad_norm": 1.5922567486806138,
+      "learning_rate": 1.8411220607575217e-05,
+      "loss": 1.1079,
+      "step": 1500
+    },
+    {
+      "epoch": 0.20657858519130196,
+      "grad_norm": 1.7135429844538697,
+      "learning_rate": 1.8408808994903383e-05,
+      "loss": 1.0997,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2067162124965593,
+      "grad_norm": 1.6265002061853238,
+      "learning_rate": 1.8406395711516302e-05,
+      "loss": 1.1642,
+      "step": 1502
+    },
+    {
+      "epoch": 0.20685383980181668,
+      "grad_norm": 1.723867090123497,
+      "learning_rate": 1.840398075789346e-05,
+      "loss": 1.1241,
+      "step": 1503
+    },
+    {
+      "epoch": 0.20699146710707406,
+      "grad_norm": 1.8314962145741862,
+      "learning_rate": 1.8401564134514685e-05,
+      "loss": 1.2169,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2071290944123314,
+      "grad_norm": 1.6989073528314984,
+      "learning_rate": 1.8399145841860118e-05,
+      "loss": 1.1078,
+      "step": 1505
+    },
+    {
+      "epoch": 0.20726672171758878,
+      "grad_norm": 1.4408169308052605,
+      "learning_rate": 1.8396725880410248e-05,
+      "loss": 1.0885,
+      "step": 1506
+    },
+    {
+      "epoch": 0.20740434902284613,
+      "grad_norm": 1.672325123770154,
+      "learning_rate": 1.839430425064588e-05,
+      "loss": 1.1501,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2075419763281035,
+      "grad_norm": 1.706428530051366,
+      "learning_rate": 1.8391880953048165e-05,
+      "loss": 1.0881,
+      "step": 1508
+    },
+    {
+      "epoch": 0.20767960363336085,
+      "grad_norm": 1.9072069516622783,
+      "learning_rate": 1.8389455988098578e-05,
+      "loss": 1.1422,
+      "step": 1509
+    },
+    {
+      "epoch": 0.20781723093861823,
+      "grad_norm": 1.5962664859905822,
+      "learning_rate": 1.838702935627892e-05,
+      "loss": 1.0953,
+      "step": 1510
+    },
+    {
+      "epoch": 0.20795485824387558,
+      "grad_norm": 1.9522962180375956,
+      "learning_rate": 1.8384601058071342e-05,
+      "loss": 1.1023,
+      "step": 1511
+    },
+    {
+      "epoch": 0.20809248554913296,
+      "grad_norm": 1.8589532820394925,
+      "learning_rate": 1.83821710939583e-05,
+      "loss": 1.2127,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2082301128543903,
+      "grad_norm": 1.6217516367365552,
+      "learning_rate": 1.8379739464422605e-05,
+      "loss": 1.136,
+      "step": 1513
+    },
+    {
+      "epoch": 0.20836774015964768,
+      "grad_norm": 1.8127800191192904,
+      "learning_rate": 1.8377306169947387e-05,
+      "loss": 1.1016,
+      "step": 1514
+    },
+    {
+      "epoch": 0.20850536746490503,
+      "grad_norm": 1.8585550749728161,
+      "learning_rate": 1.83748712110161e-05,
+      "loss": 1.1603,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2086429947701624,
+      "grad_norm": 1.565547222205009,
+      "learning_rate": 1.8372434588112546e-05,
+      "loss": 1.099,
+      "step": 1516
+    },
+    {
+      "epoch": 0.20878062207541975,
+      "grad_norm": 1.7742271598531,
+      "learning_rate": 1.8369996301720845e-05,
+      "loss": 1.1852,
+      "step": 1517
+    },
+    {
+      "epoch": 0.20891824938067713,
+      "grad_norm": 1.6813180716757246,
+      "learning_rate": 1.836755635232545e-05,
+      "loss": 1.1303,
+      "step": 1518
+    },
+    {
+      "epoch": 0.20905587668593448,
+      "grad_norm": 1.6220405153113173,
+      "learning_rate": 1.8365114740411153e-05,
+      "loss": 1.1468,
+      "step": 1519
+    },
+    {
+      "epoch": 0.20919350399119185,
+      "grad_norm": 1.947273421725539,
+      "learning_rate": 1.8362671466463057e-05,
+      "loss": 1.1162,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2093311312964492,
+      "grad_norm": 1.927623344348797,
+      "learning_rate": 1.8360226530966617e-05,
+      "loss": 1.213,
+      "step": 1521
+    },
+    {
+      "epoch": 0.20946875860170658,
+      "grad_norm": 1.6617161482620493,
+      "learning_rate": 1.83577799344076e-05,
+      "loss": 1.1008,
+      "step": 1522
+    },
+    {
+      "epoch": 0.20960638590696395,
+      "grad_norm": 1.5286888332597184,
+      "learning_rate": 1.8355331677272117e-05,
+      "loss": 1.1545,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2097440132122213,
+      "grad_norm": 1.7621532694185635,
+      "learning_rate": 1.8352881760046604e-05,
+      "loss": 1.1773,
+      "step": 1524
+    },
+    {
+      "epoch": 0.20988164051747868,
+      "grad_norm": 1.5593393942279592,
+      "learning_rate": 1.8350430183217823e-05,
+      "loss": 1.1041,
+      "step": 1525
+    },
+    {
+      "epoch": 0.21001926782273603,
+      "grad_norm": 1.8885665356344805,
+      "learning_rate": 1.8347976947272868e-05,
+      "loss": 1.1647,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2101568951279934,
+      "grad_norm": 1.7268926701560279,
+      "learning_rate": 1.8345522052699164e-05,
+      "loss": 1.1733,
+      "step": 1527
+    },
+    {
+      "epoch": 0.21029452243325075,
+      "grad_norm": 1.888184466498517,
+      "learning_rate": 1.834306549998447e-05,
+      "loss": 1.2119,
+      "step": 1528
+    },
+    {
+      "epoch": 0.21043214973850813,
+      "grad_norm": 1.5873141171628622,
+      "learning_rate": 1.8340607289616863e-05,
+      "loss": 1.0735,
+      "step": 1529
+    },
+    {
+      "epoch": 0.21056977704376548,
+      "grad_norm": 1.6578126571316665,
+      "learning_rate": 1.8338147422084756e-05,
+      "loss": 1.1526,
+      "step": 1530
+    },
+    {
+      "epoch": 0.21070740434902285,
+      "grad_norm": 2.0205953466974895,
+      "learning_rate": 1.8335685897876896e-05,
+      "loss": 1.2828,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2108450316542802,
+      "grad_norm": 1.7549753092008,
+      "learning_rate": 1.833322271748235e-05,
+      "loss": 1.0157,
+      "step": 1532
+    },
+    {
+      "epoch": 0.21098265895953758,
+      "grad_norm": 1.6156389796454735,
+      "learning_rate": 1.833075788139052e-05,
+      "loss": 1.0719,
+      "step": 1533
+    },
+    {
+      "epoch": 0.21112028626479493,
+      "grad_norm": 1.690397089079198,
+      "learning_rate": 1.8328291390091135e-05,
+      "loss": 1.1228,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2112579135700523,
+      "grad_norm": 1.7700883055225065,
+      "learning_rate": 1.832582324407425e-05,
+      "loss": 1.1462,
+      "step": 1535
+    },
+    {
+      "epoch": 0.21139554087530965,
+      "grad_norm": 1.9287872263338504,
+      "learning_rate": 1.8323353443830258e-05,
+      "loss": 1.1415,
+      "step": 1536
+    },
+    {
+      "epoch": 0.21153316818056703,
+      "grad_norm": 1.8012065321892141,
+      "learning_rate": 1.832088198984987e-05,
+      "loss": 1.151,
+      "step": 1537
+    },
+    {
+      "epoch": 0.21167079548582438,
+      "grad_norm": 1.7326192395221385,
+      "learning_rate": 1.831840888262413e-05,
+      "loss": 1.1847,
+      "step": 1538
+    },
+    {
+      "epoch": 0.21180842279108175,
+      "grad_norm": 1.757066715589881,
+      "learning_rate": 1.8315934122644413e-05,
+      "loss": 1.1218,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2119460500963391,
+      "grad_norm": 1.6913213061764916,
+      "learning_rate": 1.8313457710402415e-05,
+      "loss": 1.1215,
+      "step": 1540
+    },
+    {
+      "epoch": 0.21208367740159648,
+      "grad_norm": 1.7245041209060892,
+      "learning_rate": 1.8310979646390172e-05,
+      "loss": 1.2049,
+      "step": 1541
+    },
+    {
+      "epoch": 0.21222130470685385,
+      "grad_norm": 1.959165607462594,
+      "learning_rate": 1.830849993110004e-05,
+      "loss": 1.243,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2123589320121112,
+      "grad_norm": 1.8056087058705172,
+      "learning_rate": 1.8306018565024698e-05,
+      "loss": 1.2385,
+      "step": 1543
+    },
+    {
+      "epoch": 0.21249655931736858,
+      "grad_norm": 1.9096284034025444,
+      "learning_rate": 1.8303535548657166e-05,
+      "loss": 1.212,
+      "step": 1544
+    },
+    {
+      "epoch": 0.21263418662262593,
+      "grad_norm": 1.8045486437431584,
+      "learning_rate": 1.8301050882490786e-05,
+      "loss": 1.1449,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2127718139278833,
+      "grad_norm": 1.7002331488190596,
+      "learning_rate": 1.8298564567019223e-05,
+      "loss": 1.143,
+      "step": 1546
+    },
+    {
+      "epoch": 0.21290944123314065,
+      "grad_norm": 1.7084133374422172,
+      "learning_rate": 1.8296076602736472e-05,
+      "loss": 1.149,
+      "step": 1547
+    },
+    {
+      "epoch": 0.21304706853839803,
+      "grad_norm": 1.771061620507088,
+      "learning_rate": 1.8293586990136865e-05,
+      "loss": 1.1654,
+      "step": 1548
+    },
+    {
+      "epoch": 0.21318469584365538,
+      "grad_norm": 1.7036337900487009,
+      "learning_rate": 1.829109572971505e-05,
+      "loss": 1.0914,
+      "step": 1549
+    },
+    {
+      "epoch": 0.21332232314891275,
+      "grad_norm": 1.6639024702304208,
+      "learning_rate": 1.8288602821966005e-05,
+      "loss": 1.1988,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2134599504541701,
+      "grad_norm": 1.7420115550828694,
+      "learning_rate": 1.8286108267385038e-05,
+      "loss": 1.0555,
+      "step": 1551
+    },
+    {
+      "epoch": 0.21359757775942748,
+      "grad_norm": 1.8470119120613782,
+      "learning_rate": 1.828361206646778e-05,
+      "loss": 1.231,
+      "step": 1552
+    },
+    {
+      "epoch": 0.21373520506468482,
+      "grad_norm": 1.6290196984186365,
+      "learning_rate": 1.8281114219710198e-05,
+      "loss": 1.1714,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2138728323699422,
+      "grad_norm": 1.8226191688186457,
+      "learning_rate": 1.8278614727608576e-05,
+      "loss": 1.0612,
+      "step": 1554
+    },
+    {
+      "epoch": 0.21401045967519955,
+      "grad_norm": 1.7457363196785947,
+      "learning_rate": 1.8276113590659532e-05,
+      "loss": 1.1341,
+      "step": 1555
+    },
+    {
+      "epoch": 0.21414808698045693,
+      "grad_norm": 1.7397466820341942,
+      "learning_rate": 1.8273610809360003e-05,
+      "loss": 1.1432,
+      "step": 1556
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.4539936444210937,
+      "learning_rate": 1.827110638420726e-05,
+      "loss": 1.1304,
+      "step": 1557
+    },
+    {
+      "epoch": 0.21442334159097165,
+      "grad_norm": 1.860034165797952,
+      "learning_rate": 1.8268600315698893e-05,
+      "loss": 1.2328,
+      "step": 1558
+    },
+    {
+      "epoch": 0.214560968896229,
+      "grad_norm": 1.8535194553246104,
+      "learning_rate": 1.8266092604332833e-05,
+      "loss": 1.1377,
+      "step": 1559
+    },
+    {
+      "epoch": 0.21469859620148637,
+      "grad_norm": 1.9104603102347049,
+      "learning_rate": 1.8263583250607324e-05,
+      "loss": 1.1389,
+      "step": 1560
+    },
+    {
+      "epoch": 0.21483622350674375,
+      "grad_norm": 1.5901665682693031,
+      "learning_rate": 1.8261072255020936e-05,
+      "loss": 1.075,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2149738508120011,
+      "grad_norm": 1.763983087281282,
+      "learning_rate": 1.8258559618072575e-05,
+      "loss": 1.1511,
+      "step": 1562
+    },
+    {
+      "epoch": 0.21511147811725848,
+      "grad_norm": 1.9360440778496673,
+      "learning_rate": 1.8256045340261462e-05,
+      "loss": 1.1467,
+      "step": 1563
+    },
+    {
+      "epoch": 0.21524910542251582,
+      "grad_norm": 1.5805731408101602,
+      "learning_rate": 1.8253529422087156e-05,
+      "loss": 1.1088,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2153867327277732,
+      "grad_norm": 1.6558575141928982,
+      "learning_rate": 1.8251011864049533e-05,
+      "loss": 1.1368,
+      "step": 1565
+    },
+    {
+      "epoch": 0.21552436003303055,
+      "grad_norm": 1.6922532002590815,
+      "learning_rate": 1.8248492666648795e-05,
+      "loss": 1.2594,
+      "step": 1566
+    },
+    {
+      "epoch": 0.21566198733828792,
+      "grad_norm": 1.7302516661250806,
+      "learning_rate": 1.8245971830385475e-05,
+      "loss": 1.1611,
+      "step": 1567
+    },
+    {
+      "epoch": 0.21579961464354527,
+      "grad_norm": 1.631442286253137,
+      "learning_rate": 1.8243449355760424e-05,
+      "loss": 1.1594,
+      "step": 1568
+    },
+    {
+      "epoch": 0.21593724194880265,
+      "grad_norm": 1.7023564513198883,
+      "learning_rate": 1.8240925243274827e-05,
+      "loss": 1.0893,
+      "step": 1569
+    },
+    {
+      "epoch": 0.21607486925406,
+      "grad_norm": 1.917024172190429,
+      "learning_rate": 1.8238399493430196e-05,
+      "loss": 1.2098,
+      "step": 1570
+    },
+    {
+      "epoch": 0.21621249655931737,
+      "grad_norm": 1.7051953655629273,
+      "learning_rate": 1.8235872106728353e-05,
+      "loss": 1.1684,
+      "step": 1571
+    },
+    {
+      "epoch": 0.21635012386457472,
+      "grad_norm": 1.7879858543743854,
+      "learning_rate": 1.8233343083671458e-05,
+      "loss": 1.115,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2164877511698321,
+      "grad_norm": 1.5461768752435598,
+      "learning_rate": 1.8230812424761996e-05,
+      "loss": 1.1345,
+      "step": 1573
+    },
+    {
+      "epoch": 0.21662537847508945,
+      "grad_norm": 1.9390451026137865,
+      "learning_rate": 1.8228280130502768e-05,
+      "loss": 1.0945,
+      "step": 1574
+    },
+    {
+      "epoch": 0.21676300578034682,
+      "grad_norm": 1.7756756796425552,
+      "learning_rate": 1.8225746201396916e-05,
+      "loss": 1.0549,
+      "step": 1575
+    },
+    {
+      "epoch": 0.21690063308560417,
+      "grad_norm": 1.92496170616975,
+      "learning_rate": 1.822321063794789e-05,
+      "loss": 1.1697,
+      "step": 1576
+    },
+    {
+      "epoch": 0.21703826039086155,
+      "grad_norm": 2.044080448664165,
+      "learning_rate": 1.8220673440659478e-05,
+      "loss": 1.1937,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2171758876961189,
+      "grad_norm": 1.8467012622420291,
+      "learning_rate": 1.821813461003578e-05,
+      "loss": 1.0525,
+      "step": 1578
+    },
+    {
+      "epoch": 0.21731351500137627,
+      "grad_norm": 1.8675649167212505,
+      "learning_rate": 1.821559414658123e-05,
+      "loss": 1.168,
+      "step": 1579
+    },
+    {
+      "epoch": 0.21745114230663365,
+      "grad_norm": 1.5867764837479763,
+      "learning_rate": 1.8213052050800576e-05,
+      "loss": 1.1193,
+      "step": 1580
+    },
+    {
+      "epoch": 0.217588769611891,
+      "grad_norm": 1.713416189112341,
+      "learning_rate": 1.821050832319891e-05,
+      "loss": 1.1247,
+      "step": 1581
+    },
+    {
+      "epoch": 0.21772639691714837,
+      "grad_norm": 1.8171916110356148,
+      "learning_rate": 1.8207962964281628e-05,
+      "loss": 1.1131,
+      "step": 1582
+    },
+    {
+      "epoch": 0.21786402422240572,
+      "grad_norm": 1.8508707904053736,
+      "learning_rate": 1.8205415974554457e-05,
+      "loss": 1.2271,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2180016515276631,
+      "grad_norm": 1.7052287655903968,
+      "learning_rate": 1.8202867354523456e-05,
+      "loss": 1.0835,
+      "step": 1584
+    },
+    {
+      "epoch": 0.21813927883292045,
+      "grad_norm": 1.7970039997551248,
+      "learning_rate": 1.8200317104694996e-05,
+      "loss": 1.1495,
+      "step": 1585
+    },
+    {
+      "epoch": 0.21827690613817782,
+      "grad_norm": 1.6290904310607954,
+      "learning_rate": 1.8197765225575775e-05,
+      "loss": 1.1988,
+      "step": 1586
+    },
+    {
+      "epoch": 0.21841453344343517,
+      "grad_norm": 1.7329535725632388,
+      "learning_rate": 1.8195211717672824e-05,
+      "loss": 1.084,
+      "step": 1587
+    },
+    {
+      "epoch": 0.21855216074869255,
+      "grad_norm": 1.7904691428629729,
+      "learning_rate": 1.8192656581493477e-05,
+      "loss": 1.1504,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2186897880539499,
+      "grad_norm": 1.6602481514724015,
+      "learning_rate": 1.8190099817545417e-05,
+      "loss": 1.0842,
+      "step": 1589
+    },
+    {
+      "epoch": 0.21882741535920727,
+      "grad_norm": 1.812665397402033,
+      "learning_rate": 1.8187541426336633e-05,
+      "loss": 1.1046,
+      "step": 1590
+    },
+    {
+      "epoch": 0.21896504266446462,
+      "grad_norm": 1.6644078046092825,
+      "learning_rate": 1.8184981408375446e-05,
+      "loss": 1.1364,
+      "step": 1591
+    },
+    {
+      "epoch": 0.219102669969722,
+      "grad_norm": 1.5819879253636904,
+      "learning_rate": 1.8182419764170488e-05,
+      "loss": 1.101,
+      "step": 1592
+    },
+    {
+      "epoch": 0.21924029727497935,
+      "grad_norm": 1.7027993648543982,
+      "learning_rate": 1.8179856494230733e-05,
+      "loss": 1.1888,
+      "step": 1593
+    },
+    {
+      "epoch": 0.21937792458023672,
+      "grad_norm": 1.5536096620982303,
+      "learning_rate": 1.817729159906546e-05,
+      "loss": 1.0872,
+      "step": 1594
+    },
+    {
+      "epoch": 0.21951555188549407,
+      "grad_norm": 1.6273583506645348,
+      "learning_rate": 1.817472507918428e-05,
+      "loss": 1.1041,
+      "step": 1595
+    },
+    {
+      "epoch": 0.21965317919075145,
+      "grad_norm": 2.0190241769108725,
+      "learning_rate": 1.8172156935097128e-05,
+      "loss": 1.1486,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2197908064960088,
+      "grad_norm": 1.8070844815637965,
+      "learning_rate": 1.8169587167314256e-05,
+      "loss": 1.1795,
+      "step": 1597
+    },
+    {
+      "epoch": 0.21992843380126617,
+      "grad_norm": 1.6757726525636913,
+      "learning_rate": 1.8167015776346247e-05,
+      "loss": 1.0699,
+      "step": 1598
+    },
+    {
+      "epoch": 0.22006606110652355,
+      "grad_norm": 1.7205978885405127,
+      "learning_rate": 1.8164442762703997e-05,
+      "loss": 1.0641,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2202036884117809,
+      "grad_norm": 1.7197692892889012,
+      "learning_rate": 1.8161868126898724e-05,
+      "loss": 1.0994,
+      "step": 1600
+    },
+    {
+      "epoch": 0.22034131571703827,
+      "grad_norm": 1.7451825827952954,
+      "learning_rate": 1.8159291869441983e-05,
+      "loss": 1.1776,
+      "step": 1601
+    },
+    {
+      "epoch": 0.22047894302229562,
+      "grad_norm": 1.5855995861965546,
+      "learning_rate": 1.8156713990845634e-05,
+      "loss": 1.1708,
+      "step": 1602
+    },
+    {
+      "epoch": 0.220616570327553,
+      "grad_norm": 1.7286592742332065,
+      "learning_rate": 1.8154134491621872e-05,
+      "loss": 1.0867,
+      "step": 1603
+    },
+    {
+      "epoch": 0.22075419763281035,
+      "grad_norm": 1.5632218989109004,
+      "learning_rate": 1.81515533722832e-05,
+      "loss": 1.0661,
+      "step": 1604
+    },
+    {
+      "epoch": 0.22089182493806772,
+      "grad_norm": 1.7540732699619548,
+      "learning_rate": 1.8148970633342455e-05,
+      "loss": 1.2064,
+      "step": 1605
+    },
+    {
+      "epoch": 0.22102945224332507,
+      "grad_norm": 1.735277303926088,
+      "learning_rate": 1.8146386275312794e-05,
+      "loss": 1.239,
+      "step": 1606
+    },
+    {
+      "epoch": 0.22116707954858245,
+      "grad_norm": 1.8149977517130742,
+      "learning_rate": 1.814380029870769e-05,
+      "loss": 1.1415,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2213047068538398,
+      "grad_norm": 1.7857315295419747,
+      "learning_rate": 1.814121270404095e-05,
+      "loss": 1.0724,
+      "step": 1608
+    },
+    {
+      "epoch": 0.22144233415909717,
+      "grad_norm": 1.8341942323440683,
+      "learning_rate": 1.813862349182668e-05,
+      "loss": 1.1581,
+      "step": 1609
+    },
+    {
+      "epoch": 0.22157996146435452,
+      "grad_norm": 1.6069606129956517,
+      "learning_rate": 1.8136032662579327e-05,
+      "loss": 1.0713,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2217175887696119,
+      "grad_norm": 1.6935991289726544,
+      "learning_rate": 1.813344021681366e-05,
+      "loss": 1.1341,
+      "step": 1611
+    },
+    {
+      "epoch": 0.22185521607486924,
+      "grad_norm": 1.6958093513761305,
+      "learning_rate": 1.813084615504475e-05,
+      "loss": 1.1456,
+      "step": 1612
+    },
+    {
+      "epoch": 0.22199284338012662,
+      "grad_norm": 1.6761222278281145,
+      "learning_rate": 1.8128250477788013e-05,
+      "loss": 1.138,
+      "step": 1613
+    },
+    {
+      "epoch": 0.22213047068538397,
+      "grad_norm": 1.8352346147759513,
+      "learning_rate": 1.8125653185559172e-05,
+      "loss": 1.1889,
+      "step": 1614
+    },
+    {
+      "epoch": 0.22226809799064134,
+      "grad_norm": 1.7498444094730516,
+      "learning_rate": 1.8123054278874265e-05,
+      "loss": 1.1317,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2224057252958987,
+      "grad_norm": 1.8331573379026291,
+      "learning_rate": 1.8120453758249666e-05,
+      "loss": 1.1407,
+      "step": 1616
+    },
+    {
+      "epoch": 0.22254335260115607,
+      "grad_norm": 1.5965125766434676,
+      "learning_rate": 1.8117851624202066e-05,
+      "loss": 1.1146,
+      "step": 1617
+    },
+    {
+      "epoch": 0.22268097990641345,
+      "grad_norm": 1.7178446365597235,
+      "learning_rate": 1.8115247877248467e-05,
+      "loss": 1.1153,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2228186072116708,
+      "grad_norm": 1.6264172791093963,
+      "learning_rate": 1.81126425179062e-05,
+      "loss": 1.0713,
+      "step": 1619
+    },
+    {
+      "epoch": 0.22295623451692817,
+      "grad_norm": 1.7403875441102825,
+      "learning_rate": 1.8110035546692916e-05,
+      "loss": 1.1302,
+      "step": 1620
+    },
+    {
+      "epoch": 0.22309386182218552,
+      "grad_norm": 1.7085348143703882,
+      "learning_rate": 1.8107426964126582e-05,
+      "loss": 1.1751,
+      "step": 1621
+    },
+    {
+      "epoch": 0.2232314891274429,
+      "grad_norm": 1.7313479055296241,
+      "learning_rate": 1.810481677072549e-05,
+      "loss": 1.1809,
+      "step": 1622
+    },
+    {
+      "epoch": 0.22336911643270024,
+      "grad_norm": 1.6083332549467435,
+      "learning_rate": 1.8102204967008248e-05,
+      "loss": 1.1015,
+      "step": 1623
+    },
+    {
+      "epoch": 0.22350674373795762,
+      "grad_norm": 1.6946653305301262,
+      "learning_rate": 1.809959155349379e-05,
+      "loss": 1.174,
+      "step": 1624
+    },
+    {
+      "epoch": 0.22364437104321497,
+      "grad_norm": 1.7378314664176662,
+      "learning_rate": 1.8096976530701355e-05,
+      "loss": 1.0629,
+      "step": 1625
+    },
+    {
+      "epoch": 0.22378199834847234,
+      "grad_norm": 1.9925117096343832,
+      "learning_rate": 1.809435989915052e-05,
+      "loss": 1.1711,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2239196256537297,
+      "grad_norm": 1.7181537764630712,
+      "learning_rate": 1.8091741659361174e-05,
+      "loss": 1.1892,
+      "step": 1627
+    },
+    {
+      "epoch": 0.22405725295898707,
+      "grad_norm": 1.8915709308265245,
+      "learning_rate": 1.8089121811853524e-05,
+      "loss": 1.1496,
+      "step": 1628
+    },
+    {
+      "epoch": 0.22419488026424442,
+      "grad_norm": 1.6467886002259564,
+      "learning_rate": 1.80865003571481e-05,
+      "loss": 1.1147,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2243325075695018,
+      "grad_norm": 1.9053031378795866,
+      "learning_rate": 1.8083877295765747e-05,
+      "loss": 1.1964,
+      "step": 1630
+    },
+    {
+      "epoch": 0.22447013487475914,
+      "grad_norm": 1.6584585277036383,
+      "learning_rate": 1.808125262822763e-05,
+      "loss": 1.0619,
+      "step": 1631
+    },
+    {
+      "epoch": 0.22460776218001652,
+      "grad_norm": 1.6652663449547,
+      "learning_rate": 1.8078626355055237e-05,
+      "loss": 1.1513,
+      "step": 1632
+    },
+    {
+      "epoch": 0.22474538948527387,
+      "grad_norm": 1.6850856893787831,
+      "learning_rate": 1.807599847677037e-05,
+      "loss": 1.1037,
+      "step": 1633
+    },
+    {
+      "epoch": 0.22488301679053124,
+      "grad_norm": 1.553141328096441,
+      "learning_rate": 1.8073368993895156e-05,
+      "loss": 1.1213,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2250206440957886,
+      "grad_norm": 1.5988938473188312,
+      "learning_rate": 1.8070737906952032e-05,
+      "loss": 1.1238,
+      "step": 1635
+    },
+    {
+      "epoch": 0.22515827140104597,
+      "grad_norm": 1.675872074662097,
+      "learning_rate": 1.8068105216463763e-05,
+      "loss": 1.1291,
+      "step": 1636
+    },
+    {
+      "epoch": 0.22529589870630334,
+      "grad_norm": 1.690113613026726,
+      "learning_rate": 1.806547092295343e-05,
+      "loss": 1.1218,
+      "step": 1637
+    },
+    {
+      "epoch": 0.2254335260115607,
+      "grad_norm": 1.6988441918926727,
+      "learning_rate": 1.806283502694443e-05,
+      "loss": 1.1164,
+      "step": 1638
+    },
+    {
+      "epoch": 0.22557115331681807,
+      "grad_norm": 1.6505951186490517,
+      "learning_rate": 1.8060197528960477e-05,
+      "loss": 1.1791,
+      "step": 1639
+    },
+    {
+      "epoch": 0.22570878062207542,
+      "grad_norm": 1.530847327358255,
+      "learning_rate": 1.8057558429525606e-05,
+      "loss": 1.083,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2258464079273328,
+      "grad_norm": 1.6136463362922857,
+      "learning_rate": 1.805491772916417e-05,
+      "loss": 1.0571,
+      "step": 1641
+    },
+    {
+      "epoch": 0.22598403523259014,
+      "grad_norm": 1.7586532372893589,
+      "learning_rate": 1.8052275428400847e-05,
+      "loss": 1.1361,
+      "step": 1642
+    },
+    {
+      "epoch": 0.22612166253784752,
+      "grad_norm": 1.798567796317372,
+      "learning_rate": 1.804963152776061e-05,
+      "loss": 1.1165,
+      "step": 1643
+    },
+    {
+      "epoch": 0.22625928984310487,
+      "grad_norm": 1.6552751794382337,
+      "learning_rate": 1.8046986027768786e-05,
+      "loss": 1.2078,
+      "step": 1644
+    },
+    {
+      "epoch": 0.22639691714836224,
+      "grad_norm": 1.684940520040137,
+      "learning_rate": 1.8044338928950984e-05,
+      "loss": 1.1351,
+      "step": 1645
+    },
+    {
+      "epoch": 0.2265345444536196,
+      "grad_norm": 1.7231810168868344,
+      "learning_rate": 1.8041690231833155e-05,
+      "loss": 1.1588,
+      "step": 1646
+    },
+    {
+      "epoch": 0.22667217175887697,
+      "grad_norm": 1.7913198400351107,
+      "learning_rate": 1.8039039936941552e-05,
+      "loss": 1.1893,
+      "step": 1647
+    },
+    {
+      "epoch": 0.22680979906413432,
+      "grad_norm": 1.710869214797614,
+      "learning_rate": 1.8036388044802757e-05,
+      "loss": 1.0214,
+      "step": 1648
+    },
+    {
+      "epoch": 0.2269474263693917,
+      "grad_norm": 1.7303803714431538,
+      "learning_rate": 1.8033734555943664e-05,
+      "loss": 1.1621,
+      "step": 1649
+    },
+    {
+      "epoch": 0.22708505367464904,
+      "grad_norm": 1.576122750144607,
+      "learning_rate": 1.8031079470891484e-05,
+      "loss": 1.0444,
+      "step": 1650
+    },
+    {
+      "epoch": 0.22722268097990642,
+      "grad_norm": 1.6440388581442575,
+      "learning_rate": 1.802842279017375e-05,
+      "loss": 1.1104,
+      "step": 1651
+    },
+    {
+      "epoch": 0.22736030828516376,
+      "grad_norm": 1.9321884922241224,
+      "learning_rate": 1.80257645143183e-05,
+      "loss": 1.1893,
+      "step": 1652
+    },
+    {
+      "epoch": 0.22749793559042114,
+      "grad_norm": 1.9124544975213824,
+      "learning_rate": 1.8023104643853306e-05,
+      "loss": 1.21,
+      "step": 1653
+    },
+    {
+      "epoch": 0.2276355628956785,
+      "grad_norm": 1.7601561234349525,
+      "learning_rate": 1.802044317930724e-05,
+      "loss": 1.2091,
+      "step": 1654
+    },
+    {
+      "epoch": 0.22777319020093587,
+      "grad_norm": 1.8840219619968523,
+      "learning_rate": 1.8017780121208904e-05,
+      "loss": 1.1552,
+      "step": 1655
+    },
+    {
+      "epoch": 0.22791081750619324,
+      "grad_norm": 1.9125127967624584,
+      "learning_rate": 1.8015115470087406e-05,
+      "loss": 1.1246,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2280484448114506,
+      "grad_norm": 1.744455341669251,
+      "learning_rate": 1.8012449226472183e-05,
+      "loss": 1.0935,
+      "step": 1657
+    },
+    {
+      "epoch": 0.22818607211670797,
+      "grad_norm": 1.8047687061914486,
+      "learning_rate": 1.800978139089298e-05,
+      "loss": 1.1582,
+      "step": 1658
+    },
+    {
+      "epoch": 0.22832369942196531,
+      "grad_norm": 1.5547045382261586,
+      "learning_rate": 1.800711196387985e-05,
+      "loss": 1.0452,
+      "step": 1659
+    },
+    {
+      "epoch": 0.2284613267272227,
+      "grad_norm": 1.7317420896338862,
+      "learning_rate": 1.8004440945963183e-05,
+      "loss": 1.1062,
+      "step": 1660
+    },
+    {
+      "epoch": 0.22859895403248004,
+      "grad_norm": 1.882230688934405,
+      "learning_rate": 1.8001768337673666e-05,
+      "loss": 1.1374,
+      "step": 1661
+    },
+    {
+      "epoch": 0.22873658133773742,
+      "grad_norm": 1.7430189425194837,
+      "learning_rate": 1.7999094139542314e-05,
+      "loss": 1.0869,
+      "step": 1662
+    },
+    {
+      "epoch": 0.22887420864299476,
+      "grad_norm": 1.9399043495698964,
+      "learning_rate": 1.799641835210045e-05,
+      "loss": 1.2778,
+      "step": 1663
+    },
+    {
+      "epoch": 0.22901183594825214,
+      "grad_norm": 1.739356226106059,
+      "learning_rate": 1.7993740975879723e-05,
+      "loss": 1.1216,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2291494632535095,
+      "grad_norm": 1.7647917368559702,
+      "learning_rate": 1.7991062011412082e-05,
+      "loss": 1.1627,
+      "step": 1665
+    },
+    {
+      "epoch": 0.22928709055876687,
+      "grad_norm": 1.788992704743257,
+      "learning_rate": 1.7988381459229805e-05,
+      "loss": 1.1253,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2294247178640242,
+      "grad_norm": 1.696551967846157,
+      "learning_rate": 1.7985699319865482e-05,
+      "loss": 1.204,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2295623451692816,
+      "grad_norm": 1.675371556258455,
+      "learning_rate": 1.798301559385202e-05,
+      "loss": 1.1638,
+      "step": 1668
+    },
+    {
+      "epoch": 0.22969997247453894,
+      "grad_norm": 1.7986542760406026,
+      "learning_rate": 1.798033028172263e-05,
+      "loss": 1.2111,
+      "step": 1669
+    },
+    {
+      "epoch": 0.22983759977979631,
+      "grad_norm": 1.6549489513564422,
+      "learning_rate": 1.7977643384010857e-05,
+      "loss": 1.1311,
+      "step": 1670
+    },
+    {
+      "epoch": 0.22997522708505366,
+      "grad_norm": 1.7051886617491292,
+      "learning_rate": 1.797495490125054e-05,
+      "loss": 1.1077,
+      "step": 1671
+    },
+    {
+      "epoch": 0.23011285439031104,
+      "grad_norm": 1.775160498838214,
+      "learning_rate": 1.7972264833975855e-05,
+      "loss": 1.11,
+      "step": 1672
+    },
+    {
+      "epoch": 0.2302504816955684,
+      "grad_norm": 1.631387923703715,
+      "learning_rate": 1.7969573182721274e-05,
+      "loss": 1.0707,
+      "step": 1673
+    },
+    {
+      "epoch": 0.23038810900082576,
+      "grad_norm": 1.8948070801858465,
+      "learning_rate": 1.796687994802159e-05,
+      "loss": 1.0727,
+      "step": 1674
+    },
+    {
+      "epoch": 0.23052573630608314,
+      "grad_norm": 1.7578574325444796,
+      "learning_rate": 1.796418513041192e-05,
+      "loss": 1.127,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2306633636113405,
+      "grad_norm": 1.8512279075885227,
+      "learning_rate": 1.7961488730427676e-05,
+      "loss": 1.1991,
+      "step": 1676
+    },
+    {
+      "epoch": 0.23080099091659786,
+      "grad_norm": 1.6378387664600542,
+      "learning_rate": 1.7958790748604606e-05,
+      "loss": 1.1975,
+      "step": 1677
+    },
+    {
+      "epoch": 0.2309386182218552,
+      "grad_norm": 1.9852581709771042,
+      "learning_rate": 1.795609118547876e-05,
+      "loss": 1.1893,
+      "step": 1678
+    },
+    {
+      "epoch": 0.2310762455271126,
+      "grad_norm": 1.9269688084361234,
+      "learning_rate": 1.7953390041586498e-05,
+      "loss": 1.1835,
+      "step": 1679
+    },
+    {
+      "epoch": 0.23121387283236994,
+      "grad_norm": 1.5918493841210304,
+      "learning_rate": 1.7950687317464508e-05,
+      "loss": 1.0727,
+      "step": 1680
+    },
+    {
+      "epoch": 0.23135150013762731,
+      "grad_norm": 1.5215432325396774,
+      "learning_rate": 1.7947983013649782e-05,
+      "loss": 1.0635,
+      "step": 1681
+    },
+    {
+      "epoch": 0.23148912744288466,
+      "grad_norm": 1.6579038189657276,
+      "learning_rate": 1.7945277130679626e-05,
+      "loss": 1.1081,
+      "step": 1682
+    },
+    {
+      "epoch": 0.23162675474814204,
+      "grad_norm": 1.718435770413427,
+      "learning_rate": 1.794256966909166e-05,
+      "loss": 1.1586,
+      "step": 1683
+    },
+    {
+      "epoch": 0.2317643820533994,
+      "grad_norm": 1.7016249037951952,
+      "learning_rate": 1.7939860629423825e-05,
+      "loss": 1.1402,
+      "step": 1684
+    },
+    {
+      "epoch": 0.23190200935865676,
+      "grad_norm": 1.8262633292243078,
+      "learning_rate": 1.793715001221437e-05,
+      "loss": 1.1017,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2320396366639141,
+      "grad_norm": 1.618967024926263,
+      "learning_rate": 1.7934437818001852e-05,
+      "loss": 1.0977,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2321772639691715,
+      "grad_norm": 1.7713821763891748,
+      "learning_rate": 1.7931724047325153e-05,
+      "loss": 1.1092,
+      "step": 1687
+    },
+    {
+      "epoch": 0.23231489127442884,
+      "grad_norm": 2.0180152414963892,
+      "learning_rate": 1.7929008700723458e-05,
+      "loss": 1.1163,
+      "step": 1688
+    },
+    {
+      "epoch": 0.2324525185796862,
+      "grad_norm": 1.6999766299919146,
+      "learning_rate": 1.7926291778736268e-05,
+      "loss": 1.1064,
+      "step": 1689
+    },
+    {
+      "epoch": 0.23259014588494356,
+      "grad_norm": 1.581692299559519,
+      "learning_rate": 1.7923573281903405e-05,
+      "loss": 1.0942,
+      "step": 1690
+    },
+    {
+      "epoch": 0.23272777319020094,
+      "grad_norm": 1.8377777004402367,
+      "learning_rate": 1.792085321076499e-05,
+      "loss": 1.1935,
+      "step": 1691
+    },
+    {
+      "epoch": 0.23286540049545829,
+      "grad_norm": 1.8731560145907848,
+      "learning_rate": 1.7918131565861467e-05,
+      "loss": 1.1627,
+      "step": 1692
+    },
+    {
+      "epoch": 0.23300302780071566,
+      "grad_norm": 2.1207514325989307,
+      "learning_rate": 1.791540834773359e-05,
+      "loss": 1.1886,
+      "step": 1693
+    },
+    {
+      "epoch": 0.23314065510597304,
+      "grad_norm": 1.793611054440342,
+      "learning_rate": 1.7912683556922427e-05,
+      "loss": 1.167,
+      "step": 1694
+    },
+    {
+      "epoch": 0.2332782824112304,
+      "grad_norm": 1.5967269888859532,
+      "learning_rate": 1.790995719396935e-05,
+      "loss": 1.0345,
+      "step": 1695
+    },
+    {
+      "epoch": 0.23341590971648776,
+      "grad_norm": 1.7055682388892068,
+      "learning_rate": 1.7907229259416054e-05,
+      "loss": 1.1263,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2335535370217451,
+      "grad_norm": 1.7982275495397122,
+      "learning_rate": 1.7904499753804545e-05,
+      "loss": 1.1289,
+      "step": 1697
+    },
+    {
+      "epoch": 0.2336911643270025,
+      "grad_norm": 1.6748987152843873,
+      "learning_rate": 1.7901768677677133e-05,
+      "loss": 1.1822,
+      "step": 1698
+    },
+    {
+      "epoch": 0.23382879163225984,
+      "grad_norm": 1.9468172953078995,
+      "learning_rate": 1.789903603157645e-05,
+      "loss": 1.0993,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2339664189375172,
+      "grad_norm": 1.5614646546908204,
+      "learning_rate": 1.7896301816045432e-05,
+      "loss": 1.1602,
+      "step": 1700
+    },
+    {
+      "epoch": 0.23410404624277456,
+      "grad_norm": 1.8238678230838912,
+      "learning_rate": 1.7893566031627334e-05,
+      "loss": 1.0884,
+      "step": 1701
+    },
+    {
+      "epoch": 0.23424167354803194,
+      "grad_norm": 1.6917074885540846,
+      "learning_rate": 1.7890828678865718e-05,
+      "loss": 1.1268,
+      "step": 1702
+    },
+    {
+      "epoch": 0.23437930085328929,
+      "grad_norm": 1.7522222890678387,
+      "learning_rate": 1.7888089758304456e-05,
+      "loss": 1.0702,
+      "step": 1703
+    },
+    {
+      "epoch": 0.23451692815854666,
+      "grad_norm": 1.730583005433053,
+      "learning_rate": 1.7885349270487734e-05,
+      "loss": 1.1509,
+      "step": 1704
+    },
+    {
+      "epoch": 0.234654555463804,
+      "grad_norm": 1.539723969935983,
+      "learning_rate": 1.7882607215960045e-05,
+      "loss": 1.05,
+      "step": 1705
+    },
+    {
+      "epoch": 0.23479218276906139,
+      "grad_norm": 1.9019974211243587,
+      "learning_rate": 1.787986359526621e-05,
+      "loss": 1.2123,
+      "step": 1706
+    },
+    {
+      "epoch": 0.23492981007431873,
+      "grad_norm": 1.8340668640848363,
+      "learning_rate": 1.7877118408951344e-05,
+      "loss": 1.1598,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2350674373795761,
+      "grad_norm": 1.9155636592600638,
+      "learning_rate": 1.7874371657560873e-05,
+      "loss": 1.2227,
+      "step": 1708
+    },
+    {
+      "epoch": 0.23520506468483346,
+      "grad_norm": 2.1013823915162315,
+      "learning_rate": 1.7871623341640544e-05,
+      "loss": 1.0773,
+      "step": 1709
+    },
+    {
+      "epoch": 0.23534269199009084,
+      "grad_norm": 1.6051504099784308,
+      "learning_rate": 1.7868873461736403e-05,
+      "loss": 1.121,
+      "step": 1710
+    },
+    {
+      "epoch": 0.23548031929534818,
+      "grad_norm": 1.7025121600518087,
+      "learning_rate": 1.786612201839482e-05,
+      "loss": 1.0359,
+      "step": 1711
+    },
+    {
+      "epoch": 0.23561794660060556,
+      "grad_norm": 1.621082942340268,
+      "learning_rate": 1.786336901216247e-05,
+      "loss": 1.0986,
+      "step": 1712
+    },
+    {
+      "epoch": 0.23575557390586294,
+      "grad_norm": 1.6948684009112633,
+      "learning_rate": 1.7860614443586335e-05,
+      "loss": 1.1539,
+      "step": 1713
+    },
+    {
+      "epoch": 0.23589320121112028,
+      "grad_norm": 1.7480132482375463,
+      "learning_rate": 1.7857858313213708e-05,
+      "loss": 1.2192,
+      "step": 1714
+    },
+    {
+      "epoch": 0.23603082851637766,
+      "grad_norm": 1.7424339918228735,
+      "learning_rate": 1.78551006215922e-05,
+      "loss": 1.037,
+      "step": 1715
+    },
+    {
+      "epoch": 0.236168455821635,
+      "grad_norm": 1.8265710045217458,
+      "learning_rate": 1.785234136926972e-05,
+      "loss": 1.1737,
+      "step": 1716
+    },
+    {
+      "epoch": 0.23630608312689239,
+      "grad_norm": 2.047749495646647,
+      "learning_rate": 1.7849580556794497e-05,
+      "loss": 1.1365,
+      "step": 1717
+    },
+    {
+      "epoch": 0.23644371043214973,
+      "grad_norm": 1.9504541302155423,
+      "learning_rate": 1.7846818184715066e-05,
+      "loss": 1.2045,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2365813377374071,
+      "grad_norm": 1.5822998818423637,
+      "learning_rate": 1.7844054253580273e-05,
+      "loss": 1.2185,
+      "step": 1719
+    },
+    {
+      "epoch": 0.23671896504266446,
+      "grad_norm": 1.6801087724145576,
+      "learning_rate": 1.7841288763939275e-05,
+      "loss": 1.0221,
+      "step": 1720
+    },
+    {
+      "epoch": 0.23685659234792183,
+      "grad_norm": 2.01985975743288,
+      "learning_rate": 1.7838521716341533e-05,
+      "loss": 1.1546,
+      "step": 1721
+    },
+    {
+      "epoch": 0.23699421965317918,
+      "grad_norm": 1.710092564866613,
+      "learning_rate": 1.7835753111336824e-05,
+      "loss": 1.0679,
+      "step": 1722
+    },
+    {
+      "epoch": 0.23713184695843656,
+      "grad_norm": 1.661332263650292,
+      "learning_rate": 1.7832982949475234e-05,
+      "loss": 1.156,
+      "step": 1723
+    },
+    {
+      "epoch": 0.2372694742636939,
+      "grad_norm": 1.6201194721499923,
+      "learning_rate": 1.783021123130715e-05,
+      "loss": 1.0921,
+      "step": 1724
+    },
+    {
+      "epoch": 0.23740710156895128,
+      "grad_norm": 1.7033396008297428,
+      "learning_rate": 1.782743795738328e-05,
+      "loss": 1.1713,
+      "step": 1725
+    },
+    {
+      "epoch": 0.23754472887420863,
+      "grad_norm": 1.8041058199837072,
+      "learning_rate": 1.7824663128254637e-05,
+      "loss": 1.108,
+      "step": 1726
+    },
+    {
+      "epoch": 0.237682356179466,
+      "grad_norm": 1.808408225356363,
+      "learning_rate": 1.7821886744472536e-05,
+      "loss": 1.2065,
+      "step": 1727
+    },
+    {
+      "epoch": 0.23781998348472336,
+      "grad_norm": 1.9414012651841512,
+      "learning_rate": 1.7819108806588607e-05,
+      "loss": 1.0435,
+      "step": 1728
+    },
+    {
+      "epoch": 0.23795761078998073,
+      "grad_norm": 1.751495399278274,
+      "learning_rate": 1.7816329315154793e-05,
+      "loss": 1.1618,
+      "step": 1729
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 1.617623808380777,
+      "learning_rate": 1.7813548270723334e-05,
+      "loss": 1.1935,
+      "step": 1730
+    },
+    {
+      "epoch": 0.23823286540049546,
+      "grad_norm": 1.8355865927617843,
+      "learning_rate": 1.7810765673846795e-05,
+      "loss": 1.145,
+      "step": 1731
+    },
+    {
+      "epoch": 0.23837049270575283,
+      "grad_norm": 1.6919422102902557,
+      "learning_rate": 1.780798152507803e-05,
+      "loss": 1.0804,
+      "step": 1732
+    },
+    {
+      "epoch": 0.23850812001101018,
+      "grad_norm": 1.812005149233321,
+      "learning_rate": 1.7805195824970216e-05,
+      "loss": 1.0597,
+      "step": 1733
+    },
+    {
+      "epoch": 0.23864574731626756,
+      "grad_norm": 1.6626148917342687,
+      "learning_rate": 1.7802408574076833e-05,
+      "loss": 0.9991,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2387833746215249,
+      "grad_norm": 1.5978212011692006,
+      "learning_rate": 1.7799619772951673e-05,
+      "loss": 1.1143,
+      "step": 1735
+    },
+    {
+      "epoch": 0.23892100192678228,
+      "grad_norm": 2.077605111959503,
+      "learning_rate": 1.7796829422148825e-05,
+      "loss": 1.1434,
+      "step": 1736
+    },
+    {
+      "epoch": 0.23905862923203963,
+      "grad_norm": 1.7663264100475409,
+      "learning_rate": 1.77940375222227e-05,
+      "loss": 1.1655,
+      "step": 1737
+    },
+    {
+      "epoch": 0.239196256537297,
+      "grad_norm": 1.7357272956589496,
+      "learning_rate": 1.7791244073728008e-05,
+      "loss": 1.1665,
+      "step": 1738
+    },
+    {
+      "epoch": 0.23933388384255436,
+      "grad_norm": 1.5790612588969983,
+      "learning_rate": 1.7788449077219767e-05,
+      "loss": 1.088,
+      "step": 1739
+    },
+    {
+      "epoch": 0.23947151114781173,
+      "grad_norm": 1.7792981498748501,
+      "learning_rate": 1.778565253325331e-05,
+      "loss": 1.1267,
+      "step": 1740
+    },
+    {
+      "epoch": 0.23960913845306908,
+      "grad_norm": 1.579683568444218,
+      "learning_rate": 1.7782854442384264e-05,
+      "loss": 1.118,
+      "step": 1741
+    },
+    {
+      "epoch": 0.23974676575832646,
+      "grad_norm": 1.8751291707523239,
+      "learning_rate": 1.778005480516858e-05,
+      "loss": 1.1884,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2398843930635838,
+      "grad_norm": 1.9604220287781875,
+      "learning_rate": 1.7777253622162503e-05,
+      "loss": 1.1912,
+      "step": 1743
+    },
+    {
+      "epoch": 0.24002202036884118,
+      "grad_norm": 1.6530576950514995,
+      "learning_rate": 1.777445089392259e-05,
+      "loss": 1.1238,
+      "step": 1744
+    },
+    {
+      "epoch": 0.24015964767409853,
+      "grad_norm": 1.567419009084569,
+      "learning_rate": 1.7771646621005705e-05,
+      "loss": 1.1204,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2402972749793559,
+      "grad_norm": 1.6986231329416253,
+      "learning_rate": 1.7768840803969017e-05,
+      "loss": 1.056,
+      "step": 1746
+    },
+    {
+      "epoch": 0.24043490228461326,
+      "grad_norm": 1.786000280692303,
+      "learning_rate": 1.776603344337001e-05,
+      "loss": 1.1674,
+      "step": 1747
+    },
+    {
+      "epoch": 0.24057252958987063,
+      "grad_norm": 1.641755517898788,
+      "learning_rate": 1.776322453976646e-05,
+      "loss": 1.1018,
+      "step": 1748
+    },
+    {
+      "epoch": 0.24071015689512798,
+      "grad_norm": 1.6380778310943016,
+      "learning_rate": 1.776041409371646e-05,
+      "loss": 1.1236,
+      "step": 1749
+    },
+    {
+      "epoch": 0.24084778420038536,
+      "grad_norm": 1.739723979288783,
+      "learning_rate": 1.7757602105778406e-05,
+      "loss": 1.1955,
+      "step": 1750
+    },
+    {
+      "epoch": 0.24098541150564273,
+      "grad_norm": 1.76387877774985,
+      "learning_rate": 1.7754788576511006e-05,
+      "loss": 1.106,
+      "step": 1751
+    },
+    {
+      "epoch": 0.24112303881090008,
+      "grad_norm": 1.6283618773579944,
+      "learning_rate": 1.775197350647327e-05,
+      "loss": 1.1125,
+      "step": 1752
+    },
+    {
+      "epoch": 0.24126066611615746,
+      "grad_norm": 1.679213175860945,
+      "learning_rate": 1.774915689622451e-05,
+      "loss": 1.0208,
+      "step": 1753
+    },
+    {
+      "epoch": 0.2413982934214148,
+      "grad_norm": 1.947309066825703,
+      "learning_rate": 1.774633874632435e-05,
+      "loss": 1.1273,
+      "step": 1754
+    },
+    {
+      "epoch": 0.24153592072667218,
+      "grad_norm": 1.7200878643105664,
+      "learning_rate": 1.7743519057332708e-05,
+      "loss": 1.0674,
+      "step": 1755
+    },
+    {
+      "epoch": 0.24167354803192953,
+      "grad_norm": 1.8521793056897395,
+      "learning_rate": 1.7740697829809836e-05,
+      "loss": 1.0499,
+      "step": 1756
+    },
+    {
+      "epoch": 0.2418111753371869,
+      "grad_norm": 1.5229558547256976,
+      "learning_rate": 1.7737875064316262e-05,
+      "loss": 1.1348,
+      "step": 1757
+    },
+    {
+      "epoch": 0.24194880264244426,
+      "grad_norm": 1.7463338829142465,
+      "learning_rate": 1.773505076141283e-05,
+      "loss": 1.1131,
+      "step": 1758
+    },
+    {
+      "epoch": 0.24208642994770163,
+      "grad_norm": 1.7529322996695358,
+      "learning_rate": 1.7732224921660695e-05,
+      "loss": 1.1788,
+      "step": 1759
+    },
+    {
+      "epoch": 0.24222405725295898,
+      "grad_norm": 1.9078328549853052,
+      "learning_rate": 1.7729397545621313e-05,
+      "loss": 1.0732,
+      "step": 1760
+    },
+    {
+      "epoch": 0.24236168455821636,
+      "grad_norm": 1.583998590871363,
+      "learning_rate": 1.772656863385644e-05,
+      "loss": 1.0956,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2424993118634737,
+      "grad_norm": 1.6085521340961173,
+      "learning_rate": 1.7723738186928145e-05,
+      "loss": 1.0702,
+      "step": 1762
+    },
+    {
+      "epoch": 0.24263693916873108,
+      "grad_norm": 1.660338393967526,
+      "learning_rate": 1.7720906205398796e-05,
+      "loss": 1.0874,
+      "step": 1763
+    },
+    {
+      "epoch": 0.24277456647398843,
+      "grad_norm": 1.7522154900552314,
+      "learning_rate": 1.7718072689831075e-05,
+      "loss": 1.1249,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2429121937792458,
+      "grad_norm": 1.7146455173963568,
+      "learning_rate": 1.771523764078796e-05,
+      "loss": 1.2018,
+      "step": 1765
+    },
+    {
+      "epoch": 0.24304982108450315,
+      "grad_norm": 1.6725748171596602,
+      "learning_rate": 1.7712401058832733e-05,
+      "loss": 1.1392,
+      "step": 1766
+    },
+    {
+      "epoch": 0.24318744838976053,
+      "grad_norm": 2.016352780039768,
+      "learning_rate": 1.770956294452899e-05,
+      "loss": 1.1621,
+      "step": 1767
+    },
+    {
+      "epoch": 0.24332507569501788,
+      "grad_norm": 1.9127677565112473,
+      "learning_rate": 1.770672329844062e-05,
+      "loss": 1.1615,
+      "step": 1768
+    },
+    {
+      "epoch": 0.24346270300027525,
+      "grad_norm": 1.5673178992976797,
+      "learning_rate": 1.7703882121131823e-05,
+      "loss": 1.0958,
+      "step": 1769
+    },
+    {
+      "epoch": 0.24360033030553263,
+      "grad_norm": 1.7136909594958947,
+      "learning_rate": 1.770103941316711e-05,
+      "loss": 1.1469,
+      "step": 1770
+    },
+    {
+      "epoch": 0.24373795761078998,
+      "grad_norm": 1.8443646656748942,
+      "learning_rate": 1.7698195175111274e-05,
+      "loss": 1.1596,
+      "step": 1771
+    },
+    {
+      "epoch": 0.24387558491604736,
+      "grad_norm": 1.796621081323455,
+      "learning_rate": 1.7695349407529435e-05,
+      "loss": 1.0918,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2440132122213047,
+      "grad_norm": 1.842177623984622,
+      "learning_rate": 1.7692502110987008e-05,
+      "loss": 1.1804,
+      "step": 1773
+    },
+    {
+      "epoch": 0.24415083952656208,
+      "grad_norm": 1.6055570479918668,
+      "learning_rate": 1.768965328604971e-05,
+      "loss": 1.0056,
+      "step": 1774
+    },
+    {
+      "epoch": 0.24428846683181943,
+      "grad_norm": 1.8405870371494335,
+      "learning_rate": 1.7686802933283565e-05,
+      "loss": 1.1986,
+      "step": 1775
+    },
+    {
+      "epoch": 0.2444260941370768,
+      "grad_norm": 1.8188710479664265,
+      "learning_rate": 1.7683951053254897e-05,
+      "loss": 1.1542,
+      "step": 1776
+    },
+    {
+      "epoch": 0.24456372144233415,
+      "grad_norm": 1.7447280017502376,
+      "learning_rate": 1.7681097646530333e-05,
+      "loss": 1.1149,
+      "step": 1777
+    },
+    {
+      "epoch": 0.24470134874759153,
+      "grad_norm": 1.7455038033445627,
+      "learning_rate": 1.7678242713676813e-05,
+      "loss": 1.099,
+      "step": 1778
+    },
+    {
+      "epoch": 0.24483897605284888,
+      "grad_norm": 1.8439559013551687,
+      "learning_rate": 1.7675386255261573e-05,
+      "loss": 1.1831,
+      "step": 1779
+    },
+    {
+      "epoch": 0.24497660335810625,
+      "grad_norm": 1.9287421775691638,
+      "learning_rate": 1.7672528271852142e-05,
+      "loss": 1.1994,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2451142306633636,
+      "grad_norm": 1.8005419955025663,
+      "learning_rate": 1.7669668764016373e-05,
+      "loss": 1.1393,
+      "step": 1781
+    },
+    {
+      "epoch": 0.24525185796862098,
+      "grad_norm": 1.7344359222205779,
+      "learning_rate": 1.7666807732322407e-05,
+      "loss": 1.1027,
+      "step": 1782
+    },
+    {
+      "epoch": 0.24538948527387833,
+      "grad_norm": 1.7840168421975116,
+      "learning_rate": 1.766394517733869e-05,
+      "loss": 1.1139,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2455271125791357,
+      "grad_norm": 1.6058689359369729,
+      "learning_rate": 1.7661081099633976e-05,
+      "loss": 1.0813,
+      "step": 1784
+    },
+    {
+      "epoch": 0.24566473988439305,
+      "grad_norm": 1.8362292539707685,
+      "learning_rate": 1.7658215499777314e-05,
+      "loss": 1.1848,
+      "step": 1785
+    },
+    {
+      "epoch": 0.24580236718965043,
+      "grad_norm": 1.7537342790161152,
+      "learning_rate": 1.7655348378338062e-05,
+      "loss": 1.136,
+      "step": 1786
+    },
+    {
+      "epoch": 0.24593999449490778,
+      "grad_norm": 2.1129762107038954,
+      "learning_rate": 1.765247973588588e-05,
+      "loss": 1.1678,
+      "step": 1787
+    },
+    {
+      "epoch": 0.24607762180016515,
+      "grad_norm": 1.5753264715939579,
+      "learning_rate": 1.7649609572990724e-05,
+      "loss": 1.0935,
+      "step": 1788
+    },
+    {
+      "epoch": 0.24621524910542253,
+      "grad_norm": 2.049864212593271,
+      "learning_rate": 1.7646737890222855e-05,
+      "loss": 1.174,
+      "step": 1789
+    },
+    {
+      "epoch": 0.24635287641067988,
+      "grad_norm": 1.788181790436539,
+      "learning_rate": 1.7643864688152843e-05,
+      "loss": 1.2012,
+      "step": 1790
+    },
+    {
+      "epoch": 0.24649050371593725,
+      "grad_norm": 1.77933594059868,
+      "learning_rate": 1.764098996735155e-05,
+      "loss": 1.191,
+      "step": 1791
+    },
+    {
+      "epoch": 0.2466281310211946,
+      "grad_norm": 1.7041495631736205,
+      "learning_rate": 1.7638113728390145e-05,
+      "loss": 1.169,
+      "step": 1792
+    },
+    {
+      "epoch": 0.24676575832645198,
+      "grad_norm": 1.816556475933002,
+      "learning_rate": 1.7635235971840094e-05,
+      "loss": 1.2465,
+      "step": 1793
+    },
+    {
+      "epoch": 0.24690338563170933,
+      "grad_norm": 1.7270719146788693,
+      "learning_rate": 1.7632356698273173e-05,
+      "loss": 1.0703,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2470410129369667,
+      "grad_norm": 1.603438942487467,
+      "learning_rate": 1.7629475908261452e-05,
+      "loss": 1.137,
+      "step": 1795
+    },
+    {
+      "epoch": 0.24717864024222405,
+      "grad_norm": 1.8229029454550567,
+      "learning_rate": 1.76265936023773e-05,
+      "loss": 1.1492,
+      "step": 1796
+    },
+    {
+      "epoch": 0.24731626754748143,
+      "grad_norm": 1.6678589107159985,
+      "learning_rate": 1.7623709781193398e-05,
+      "loss": 1.0467,
+      "step": 1797
+    },
+    {
+      "epoch": 0.24745389485273878,
+      "grad_norm": 1.778150510217301,
+      "learning_rate": 1.762082444528272e-05,
+      "loss": 1.1571,
+      "step": 1798
+    },
+    {
+      "epoch": 0.24759152215799615,
+      "grad_norm": 1.7514895290860197,
+      "learning_rate": 1.7617937595218543e-05,
+      "loss": 1.1012,
+      "step": 1799
+    },
+    {
+      "epoch": 0.2477291494632535,
+      "grad_norm": 1.7639957804035473,
+      "learning_rate": 1.7615049231574444e-05,
+      "loss": 1.1498,
+      "step": 1800
+    },
+    {
+      "epoch": 0.24786677676851088,
+      "grad_norm": 1.7612904039185773,
+      "learning_rate": 1.7612159354924303e-05,
+      "loss": 1.1481,
+      "step": 1801
+    },
+    {
+      "epoch": 0.24800440407376823,
+      "grad_norm": 1.6596626519137907,
+      "learning_rate": 1.7609267965842295e-05,
+      "loss": 1.1745,
+      "step": 1802
+    },
+    {
+      "epoch": 0.2481420313790256,
+      "grad_norm": 1.7333992480054492,
+      "learning_rate": 1.76063750649029e-05,
+      "loss": 1.1346,
+      "step": 1803
+    },
+    {
+      "epoch": 0.24827965868428295,
+      "grad_norm": 1.6346919657368744,
+      "learning_rate": 1.7603480652680902e-05,
+      "loss": 1.0676,
+      "step": 1804
+    },
+    {
+      "epoch": 0.24841728598954033,
+      "grad_norm": 1.7488661076830252,
+      "learning_rate": 1.7600584729751378e-05,
+      "loss": 1.1304,
+      "step": 1805
+    },
+    {
+      "epoch": 0.24855491329479767,
+      "grad_norm": 1.786724520974546,
+      "learning_rate": 1.759768729668971e-05,
+      "loss": 1.1333,
+      "step": 1806
+    },
+    {
+      "epoch": 0.24869254060005505,
+      "grad_norm": 1.6234807324161031,
+      "learning_rate": 1.7594788354071576e-05,
+      "loss": 1.1911,
+      "step": 1807
+    },
+    {
+      "epoch": 0.24883016790531243,
+      "grad_norm": 1.6948443820262118,
+      "learning_rate": 1.7591887902472956e-05,
+      "loss": 1.0733,
+      "step": 1808
+    },
+    {
+      "epoch": 0.24896779521056978,
+      "grad_norm": 1.8254423990604678,
+      "learning_rate": 1.7588985942470134e-05,
+      "loss": 1.129,
+      "step": 1809
+    },
+    {
+      "epoch": 0.24910542251582715,
+      "grad_norm": 1.7044848393490044,
+      "learning_rate": 1.7586082474639688e-05,
+      "loss": 1.0491,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2492430498210845,
+      "grad_norm": 1.759765225806777,
+      "learning_rate": 1.7583177499558492e-05,
+      "loss": 1.0945,
+      "step": 1811
+    },
+    {
+      "epoch": 0.24938067712634188,
+      "grad_norm": 1.795122659422111,
+      "learning_rate": 1.7580271017803734e-05,
+      "loss": 1.0906,
+      "step": 1812
+    },
+    {
+      "epoch": 0.24951830443159922,
+      "grad_norm": 1.6385124363587005,
+      "learning_rate": 1.7577363029952885e-05,
+      "loss": 1.1158,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2496559317368566,
+      "grad_norm": 1.7672365190278323,
+      "learning_rate": 1.7574453536583724e-05,
+      "loss": 1.137,
+      "step": 1814
+    },
+    {
+      "epoch": 0.24979355904211395,
+      "grad_norm": 1.7832313084753144,
+      "learning_rate": 1.7571542538274328e-05,
+      "loss": 1.0508,
+      "step": 1815
+    },
+    {
+      "epoch": 0.24993118634737133,
+      "grad_norm": 1.8922905398070986,
+      "learning_rate": 1.7568630035603077e-05,
+      "loss": 1.1407,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2500688136526287,
+      "grad_norm": 1.632291884046593,
+      "learning_rate": 1.7565716029148637e-05,
+      "loss": 1.0914,
+      "step": 1817
+    },
+    {
+      "epoch": 0.250206440957886,
+      "grad_norm": 1.6580055410971883,
+      "learning_rate": 1.7562800519489986e-05,
+      "loss": 1.0831,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2503440682631434,
+      "grad_norm": 1.6553374689797562,
+      "learning_rate": 1.7559883507206396e-05,
+      "loss": 1.0976,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2504816955684008,
+      "grad_norm": 1.558621702138325,
+      "learning_rate": 1.7556964992877438e-05,
+      "loss": 1.0899,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2506193228736581,
+      "grad_norm": 1.6722964692257043,
+      "learning_rate": 1.7554044977082976e-05,
+      "loss": 1.0875,
+      "step": 1821
+    },
+    {
+      "epoch": 0.25075695017891547,
+      "grad_norm": 1.9679945908210024,
+      "learning_rate": 1.7551123460403185e-05,
+      "loss": 1.189,
+      "step": 1822
+    },
+    {
+      "epoch": 0.2508945774841729,
+      "grad_norm": 1.6997082578064715,
+      "learning_rate": 1.754820044341852e-05,
+      "loss": 1.0609,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2510322047894302,
+      "grad_norm": 1.6420027629874525,
+      "learning_rate": 1.7545275926709757e-05,
+      "loss": 1.0684,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2511698320946876,
+      "grad_norm": 1.8609431318755765,
+      "learning_rate": 1.7542349910857948e-05,
+      "loss": 1.1587,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2513074593999449,
+      "grad_norm": 1.8096304252143798,
+      "learning_rate": 1.753942239644445e-05,
+      "loss": 1.1493,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2514450867052023,
+      "grad_norm": 2.0046784926813617,
+      "learning_rate": 1.7536493384050933e-05,
+      "loss": 1.102,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2515827140104597,
+      "grad_norm": 1.851203881678385,
+      "learning_rate": 1.753356287425934e-05,
+      "loss": 1.1114,
+      "step": 1828
+    },
+    {
+      "epoch": 0.251720341315717,
+      "grad_norm": 1.7255736883436192,
+      "learning_rate": 1.753063086765193e-05,
+      "loss": 1.0925,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2518579686209744,
+      "grad_norm": 1.656822468561523,
+      "learning_rate": 1.752769736481125e-05,
+      "loss": 1.1612,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2519955959262318,
+      "grad_norm": 1.8980298591118834,
+      "learning_rate": 1.752476236632014e-05,
+      "loss": 1.0598,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2521332232314891,
+      "grad_norm": 1.5760309613380692,
+      "learning_rate": 1.752182587276176e-05,
+      "loss": 1.1381,
+      "step": 1832
+    },
+    {
+      "epoch": 0.25227085053674647,
+      "grad_norm": 1.9498514646245573,
+      "learning_rate": 1.7518887884719537e-05,
+      "loss": 1.233,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2524084778420039,
+      "grad_norm": 1.8737963372956097,
+      "learning_rate": 1.751594840277721e-05,
+      "loss": 1.1083,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2525461051472612,
+      "grad_norm": 1.7944138909615754,
+      "learning_rate": 1.7513007427518825e-05,
+      "loss": 1.0518,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2526837324525186,
+      "grad_norm": 1.6536725094690443,
+      "learning_rate": 1.75100649595287e-05,
+      "loss": 1.0873,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2528213597577759,
+      "grad_norm": 1.5604150150403506,
+      "learning_rate": 1.7507120999391472e-05,
+      "loss": 1.0738,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2529589870630333,
+      "grad_norm": 1.8311572579801987,
+      "learning_rate": 1.7504175547692062e-05,
+      "loss": 1.0598,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2530966143682907,
+      "grad_norm": 1.766369818127065,
+      "learning_rate": 1.7501228605015693e-05,
+      "loss": 1.1233,
+      "step": 1839
+    },
+    {
+      "epoch": 0.253234241673548,
+      "grad_norm": 1.9157399350091002,
+      "learning_rate": 1.7498280171947878e-05,
+      "loss": 1.0431,
+      "step": 1840
+    },
+    {
+      "epoch": 0.25337186897880537,
+      "grad_norm": 1.7488639694119437,
+      "learning_rate": 1.7495330249074436e-05,
+      "loss": 1.154,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2535094962840628,
+      "grad_norm": 1.6645540395612066,
+      "learning_rate": 1.749237883698147e-05,
+      "loss": 1.079,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2536471235893201,
+      "grad_norm": 1.9519412373818978,
+      "learning_rate": 1.748942593625539e-05,
+      "loss": 1.0737,
+      "step": 1843
+    },
+    {
+      "epoch": 0.25378475089457747,
+      "grad_norm": 1.5931762747784026,
+      "learning_rate": 1.74864715474829e-05,
+      "loss": 1.0757,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2539223781998348,
+      "grad_norm": 1.7338443337525924,
+      "learning_rate": 1.7483515671250985e-05,
+      "loss": 1.0541,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2540600055050922,
+      "grad_norm": 1.7749954920533604,
+      "learning_rate": 1.748055830814695e-05,
+      "loss": 1.0932,
+      "step": 1846
+    },
+    {
+      "epoch": 0.25419763281034957,
+      "grad_norm": 1.8638148363545926,
+      "learning_rate": 1.747759945875837e-05,
+      "loss": 1.1064,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2543352601156069,
+      "grad_norm": 1.701750955316755,
+      "learning_rate": 1.7474639123673143e-05,
+      "loss": 1.0945,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2544728874208643,
+      "grad_norm": 1.668473108038409,
+      "learning_rate": 1.7471677303479434e-05,
+      "loss": 1.1402,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2546105147261217,
+      "grad_norm": 2.0111304303497874,
+      "learning_rate": 1.746871399876572e-05,
+      "loss": 1.1714,
+      "step": 1850
+    },
+    {
+      "epoch": 0.254748142031379,
+      "grad_norm": 1.7446170195849178,
+      "learning_rate": 1.746574921012077e-05,
+      "loss": 1.1849,
+      "step": 1851
+    },
+    {
+      "epoch": 0.25488576933663637,
+      "grad_norm": 1.8354686515332905,
+      "learning_rate": 1.7462782938133654e-05,
+      "loss": 1.1637,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2550233966418938,
+      "grad_norm": 1.7283065251896477,
+      "learning_rate": 1.7459815183393715e-05,
+      "loss": 1.0342,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2551610239471511,
+      "grad_norm": 1.7929327603234388,
+      "learning_rate": 1.7456845946490616e-05,
+      "loss": 1.1544,
+      "step": 1854
+    },
+    {
+      "epoch": 0.25529865125240847,
+      "grad_norm": 1.6563598725343802,
+      "learning_rate": 1.7453875228014297e-05,
+      "loss": 1.0887,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2554362785576658,
+      "grad_norm": 1.5431662853314518,
+      "learning_rate": 1.745090302855501e-05,
+      "loss": 1.0594,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2555739058629232,
+      "grad_norm": 1.7604415629468957,
+      "learning_rate": 1.7447929348703277e-05,
+      "loss": 1.1918,
+      "step": 1857
+    },
+    {
+      "epoch": 0.25571153316818057,
+      "grad_norm": 1.8504348177930372,
+      "learning_rate": 1.7444954189049935e-05,
+      "loss": 1.2017,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2558491604734379,
+      "grad_norm": 1.7495303682377772,
+      "learning_rate": 1.7441977550186108e-05,
+      "loss": 1.0816,
+      "step": 1859
+    },
+    {
+      "epoch": 0.25598678777869527,
+      "grad_norm": 1.7496712194558393,
+      "learning_rate": 1.7438999432703204e-05,
+      "loss": 1.1264,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2561244150839527,
+      "grad_norm": 1.9967389602641927,
+      "learning_rate": 1.743601983719295e-05,
+      "loss": 1.1408,
+      "step": 1861
+    },
+    {
+      "epoch": 0.25626204238921,
+      "grad_norm": 1.946589071615279,
+      "learning_rate": 1.743303876424734e-05,
+      "loss": 1.0132,
+      "step": 1862
+    },
+    {
+      "epoch": 0.25639966969446737,
+      "grad_norm": 1.7169386728851386,
+      "learning_rate": 1.743005621445867e-05,
+      "loss": 1.0923,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2565372969997247,
+      "grad_norm": 1.9418484609179523,
+      "learning_rate": 1.7427072188419543e-05,
+      "loss": 1.2262,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2566749243049821,
+      "grad_norm": 1.8391477174173823,
+      "learning_rate": 1.742408668672283e-05,
+      "loss": 1.148,
+      "step": 1865
+    },
+    {
+      "epoch": 0.25681255161023947,
+      "grad_norm": 1.6740232730302724,
+      "learning_rate": 1.7421099709961728e-05,
+      "loss": 1.1178,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2569501789154968,
+      "grad_norm": 1.7754516552959174,
+      "learning_rate": 1.741811125872969e-05,
+      "loss": 1.1375,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2570878062207542,
+      "grad_norm": 1.689119959599771,
+      "learning_rate": 1.7415121333620487e-05,
+      "loss": 1.1193,
+      "step": 1868
+    },
+    {
+      "epoch": 0.25722543352601157,
+      "grad_norm": 1.7456331122525384,
+      "learning_rate": 1.7412129935228178e-05,
+      "loss": 1.1404,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2573630608312689,
+      "grad_norm": 1.650225735021971,
+      "learning_rate": 1.7409137064147113e-05,
+      "loss": 1.0337,
+      "step": 1870
+    },
+    {
+      "epoch": 0.25750068813652627,
+      "grad_norm": 1.7479327820885169,
+      "learning_rate": 1.7406142720971935e-05,
+      "loss": 1.0464,
+      "step": 1871
+    },
+    {
+      "epoch": 0.25763831544178367,
+      "grad_norm": 1.6226918681074909,
+      "learning_rate": 1.7403146906297578e-05,
+      "loss": 1.084,
+      "step": 1872
+    },
+    {
+      "epoch": 0.257775942747041,
+      "grad_norm": 1.6634160068776511,
+      "learning_rate": 1.7400149620719263e-05,
+      "loss": 1.1075,
+      "step": 1873
+    },
+    {
+      "epoch": 0.25791357005229837,
+      "grad_norm": 1.8103226752623403,
+      "learning_rate": 1.739715086483252e-05,
+      "loss": 1.1166,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2580511973575557,
+      "grad_norm": 1.5218059986267214,
+      "learning_rate": 1.7394150639233156e-05,
+      "loss": 1.0973,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2581888246628131,
+      "grad_norm": 1.8027747164134855,
+      "learning_rate": 1.7391148944517273e-05,
+      "loss": 1.0843,
+      "step": 1876
+    },
+    {
+      "epoch": 0.25832645196807047,
+      "grad_norm": 1.7543031268527447,
+      "learning_rate": 1.738814578128127e-05,
+      "loss": 1.1548,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2584640792733278,
+      "grad_norm": 1.853976847466797,
+      "learning_rate": 1.738514115012183e-05,
+      "loss": 1.0953,
+      "step": 1878
+    },
+    {
+      "epoch": 0.25860170657858517,
+      "grad_norm": 1.8301304204783642,
+      "learning_rate": 1.7382135051635938e-05,
+      "loss": 1.1785,
+      "step": 1879
+    },
+    {
+      "epoch": 0.25873933388384257,
+      "grad_norm": 1.6382201352979313,
+      "learning_rate": 1.7379127486420862e-05,
+      "loss": 1.1409,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2588769611890999,
+      "grad_norm": 1.6000603028954439,
+      "learning_rate": 1.7376118455074164e-05,
+      "loss": 1.02,
+      "step": 1881
+    },
+    {
+      "epoch": 0.25901458849435727,
+      "grad_norm": 1.6173426039588876,
+      "learning_rate": 1.7373107958193696e-05,
+      "loss": 1.072,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2591522157996146,
+      "grad_norm": 1.6268575255357909,
+      "learning_rate": 1.7370095996377603e-05,
+      "loss": 1.1221,
+      "step": 1883
+    },
+    {
+      "epoch": 0.259289843104872,
+      "grad_norm": 1.9562584502291203,
+      "learning_rate": 1.736708257022432e-05,
+      "loss": 1.2108,
+      "step": 1884
+    },
+    {
+      "epoch": 0.25942747041012937,
+      "grad_norm": 1.7357960430112962,
+      "learning_rate": 1.7364067680332576e-05,
+      "loss": 1.1334,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2595650977153867,
+      "grad_norm": 1.5534080212017654,
+      "learning_rate": 1.736105132730139e-05,
+      "loss": 1.0894,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2597027250206441,
+      "grad_norm": 1.7266812984526405,
+      "learning_rate": 1.7358033511730062e-05,
+      "loss": 1.1247,
+      "step": 1887
+    },
+    {
+      "epoch": 0.25984035232590147,
+      "grad_norm": 2.0424559838437606,
+      "learning_rate": 1.7355014234218197e-05,
+      "loss": 1.0143,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2599779796311588,
+      "grad_norm": 1.6648603786118623,
+      "learning_rate": 1.7351993495365684e-05,
+      "loss": 1.153,
+      "step": 1889
+    },
+    {
+      "epoch": 0.26011560693641617,
+      "grad_norm": 1.7203423342490476,
+      "learning_rate": 1.7348971295772704e-05,
+      "loss": 1.0809,
+      "step": 1890
+    },
+    {
+      "epoch": 0.26025323424167357,
+      "grad_norm": 1.7348369944966082,
+      "learning_rate": 1.7345947636039718e-05,
+      "loss": 1.0997,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2603908615469309,
+      "grad_norm": 1.6912364737994603,
+      "learning_rate": 1.7342922516767496e-05,
+      "loss": 1.0849,
+      "step": 1892
+    },
+    {
+      "epoch": 0.26052848885218827,
+      "grad_norm": 1.5411227531889038,
+      "learning_rate": 1.7339895938557083e-05,
+      "loss": 0.9714,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2606661161574456,
+      "grad_norm": 1.6607115962083139,
+      "learning_rate": 1.7336867902009817e-05,
+      "loss": 1.0174,
+      "step": 1894
+    },
+    {
+      "epoch": 0.260803743462703,
+      "grad_norm": 1.7198247937211633,
+      "learning_rate": 1.7333838407727333e-05,
+      "loss": 1.063,
+      "step": 1895
+    },
+    {
+      "epoch": 0.26094137076796037,
+      "grad_norm": 1.7832959376742208,
+      "learning_rate": 1.7330807456311546e-05,
+      "loss": 1.0814,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2610789980732177,
+      "grad_norm": 1.7676483045730174,
+      "learning_rate": 1.7327775048364662e-05,
+      "loss": 1.102,
+      "step": 1897
+    },
+    {
+      "epoch": 0.26121662537847506,
+      "grad_norm": 1.892594996999069,
+      "learning_rate": 1.7324741184489186e-05,
+      "loss": 1.068,
+      "step": 1898
+    },
+    {
+      "epoch": 0.26135425268373247,
+      "grad_norm": 1.80627369462942,
+      "learning_rate": 1.73217058652879e-05,
+      "loss": 1.0426,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2614918799889898,
+      "grad_norm": 1.7115252664649632,
+      "learning_rate": 1.7318669091363883e-05,
+      "loss": 1.0984,
+      "step": 1900
+    },
+    {
+      "epoch": 0.26162950729424717,
+      "grad_norm": 2.0929938760296367,
+      "learning_rate": 1.7315630863320496e-05,
+      "loss": 1.1519,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2617671345995045,
+      "grad_norm": 1.819354116769841,
+      "learning_rate": 1.7312591181761398e-05,
+      "loss": 1.1023,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2619047619047619,
+      "grad_norm": 1.7070935400563674,
+      "learning_rate": 1.730955004729053e-05,
+      "loss": 1.0986,
+      "step": 1903
+    },
+    {
+      "epoch": 0.26204238921001927,
+      "grad_norm": 1.741643697964033,
+      "learning_rate": 1.730650746051213e-05,
+      "loss": 1.0495,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2621800165152766,
+      "grad_norm": 1.6526578815370194,
+      "learning_rate": 1.7303463422030707e-05,
+      "loss": 1.0742,
+      "step": 1905
+    },
+    {
+      "epoch": 0.262317643820534,
+      "grad_norm": 1.7750252735521697,
+      "learning_rate": 1.7300417932451075e-05,
+      "loss": 1.1221,
+      "step": 1906
+    },
+    {
+      "epoch": 0.26245527112579137,
+      "grad_norm": 1.8543814840421027,
+      "learning_rate": 1.7297370992378332e-05,
+      "loss": 1.0815,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2625928984310487,
+      "grad_norm": 1.6647377985052554,
+      "learning_rate": 1.7294322602417867e-05,
+      "loss": 1.1205,
+      "step": 1908
+    },
+    {
+      "epoch": 0.26273052573630606,
+      "grad_norm": 1.6049803651717396,
+      "learning_rate": 1.7291272763175345e-05,
+      "loss": 1.0985,
+      "step": 1909
+    },
+    {
+      "epoch": 0.26286815304156347,
+      "grad_norm": 1.5902819523848442,
+      "learning_rate": 1.7288221475256735e-05,
+      "loss": 1.0641,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2630057803468208,
+      "grad_norm": 1.6597202670430593,
+      "learning_rate": 1.728516873926828e-05,
+      "loss": 1.0808,
+      "step": 1911
+    },
+    {
+      "epoch": 0.26314340765207817,
+      "grad_norm": 1.8568308356965246,
+      "learning_rate": 1.728211455581652e-05,
+      "loss": 1.0345,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2632810349573355,
+      "grad_norm": 1.707350611412565,
+      "learning_rate": 1.7279058925508284e-05,
+      "loss": 1.0661,
+      "step": 1913
+    },
+    {
+      "epoch": 0.2634186622625929,
+      "grad_norm": 1.8123911454567971,
+      "learning_rate": 1.7276001848950672e-05,
+      "loss": 1.0625,
+      "step": 1914
+    },
+    {
+      "epoch": 0.26355628956785027,
+      "grad_norm": 1.642921826885694,
+      "learning_rate": 1.7272943326751094e-05,
+      "loss": 1.1234,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2636939168731076,
+      "grad_norm": 1.7727727077282347,
+      "learning_rate": 1.726988335951723e-05,
+      "loss": 1.1071,
+      "step": 1916
+    },
+    {
+      "epoch": 0.26383154417836496,
+      "grad_norm": 1.5965378555642655,
+      "learning_rate": 1.7266821947857057e-05,
+      "loss": 1.0341,
+      "step": 1917
+    },
+    {
+      "epoch": 0.26396917148362237,
+      "grad_norm": 1.600991731819956,
+      "learning_rate": 1.7263759092378836e-05,
+      "loss": 1.1008,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2641067987888797,
+      "grad_norm": 1.7579453216645406,
+      "learning_rate": 1.7260694793691114e-05,
+      "loss": 1.1069,
+      "step": 1919
+    },
+    {
+      "epoch": 0.26424442609413706,
+      "grad_norm": 1.8507168254143103,
+      "learning_rate": 1.7257629052402725e-05,
+      "loss": 1.1686,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2643820533993944,
+      "grad_norm": 1.7954328413029912,
+      "learning_rate": 1.7254561869122786e-05,
+      "loss": 1.1135,
+      "step": 1921
+    },
+    {
+      "epoch": 0.2645196807046518,
+      "grad_norm": 1.7276186632125112,
+      "learning_rate": 1.7251493244460712e-05,
+      "loss": 1.099,
+      "step": 1922
+    },
+    {
+      "epoch": 0.26465730800990916,
+      "grad_norm": 1.8639922658622365,
+      "learning_rate": 1.724842317902619e-05,
+      "loss": 1.0965,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2647949353151665,
+      "grad_norm": 1.7063193990359053,
+      "learning_rate": 1.7245351673429203e-05,
+      "loss": 1.1179,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2649325626204239,
+      "grad_norm": 1.7454398811103677,
+      "learning_rate": 1.7242278728280015e-05,
+      "loss": 1.13,
+      "step": 1925
+    },
+    {
+      "epoch": 0.26507018992568127,
+      "grad_norm": 1.7712126868575098,
+      "learning_rate": 1.7239204344189182e-05,
+      "loss": 1.1748,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2652078172309386,
+      "grad_norm": 1.6891472900224795,
+      "learning_rate": 1.7236128521767538e-05,
+      "loss": 1.0694,
+      "step": 1927
+    },
+    {
+      "epoch": 0.26534544453619596,
+      "grad_norm": 1.8551599125563996,
+      "learning_rate": 1.7233051261626213e-05,
+      "loss": 1.1639,
+      "step": 1928
+    },
+    {
+      "epoch": 0.26548307184145337,
+      "grad_norm": 1.8096725132910423,
+      "learning_rate": 1.7229972564376605e-05,
+      "loss": 1.2175,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2656206991467107,
+      "grad_norm": 1.7783909314566977,
+      "learning_rate": 1.7226892430630423e-05,
+      "loss": 1.103,
+      "step": 1930
+    },
+    {
+      "epoch": 0.26575832645196806,
+      "grad_norm": 1.6251083925783627,
+      "learning_rate": 1.7223810860999636e-05,
+      "loss": 1.0898,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2658959537572254,
+      "grad_norm": 1.6207558005310294,
+      "learning_rate": 1.722072785609651e-05,
+      "loss": 1.13,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2660335810624828,
+      "grad_norm": 1.7116284866065754,
+      "learning_rate": 1.7217643416533605e-05,
+      "loss": 1.0881,
+      "step": 1933
+    },
+    {
+      "epoch": 0.26617120836774016,
+      "grad_norm": 1.9827725626372417,
+      "learning_rate": 1.7214557542923757e-05,
+      "loss": 1.1696,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2663088356729975,
+      "grad_norm": 2.108617487605124,
+      "learning_rate": 1.7211470235880076e-05,
+      "loss": 1.2439,
+      "step": 1935
+    },
+    {
+      "epoch": 0.26644646297825486,
+      "grad_norm": 1.5124398602229299,
+      "learning_rate": 1.7208381496015973e-05,
+      "loss": 1.0986,
+      "step": 1936
+    },
+    {
+      "epoch": 0.26658409028351226,
+      "grad_norm": 1.795482766766421,
+      "learning_rate": 1.720529132394514e-05,
+      "loss": 1.1961,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2667217175887696,
+      "grad_norm": 1.970789241045633,
+      "learning_rate": 1.7202199720281548e-05,
+      "loss": 1.2327,
+      "step": 1938
+    },
+    {
+      "epoch": 0.26685934489402696,
+      "grad_norm": 1.659935049798508,
+      "learning_rate": 1.719910668563946e-05,
+      "loss": 1.0677,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2669969721992843,
+      "grad_norm": 1.7713271888489008,
+      "learning_rate": 1.7196012220633423e-05,
+      "loss": 1.157,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2671345995045417,
+      "grad_norm": 1.5892225390764534,
+      "learning_rate": 1.7192916325878254e-05,
+      "loss": 1.0779,
+      "step": 1941
+    },
+    {
+      "epoch": 0.26727222680979906,
+      "grad_norm": 1.7702760603479895,
+      "learning_rate": 1.7189819001989075e-05,
+      "loss": 1.1171,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2674098541150564,
+      "grad_norm": 1.6059021618970897,
+      "learning_rate": 1.7186720249581277e-05,
+      "loss": 1.113,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2675474814203138,
+      "grad_norm": 2.005383153966915,
+      "learning_rate": 1.7183620069270545e-05,
+      "loss": 1.059,
+      "step": 1944
+    },
+    {
+      "epoch": 0.26768510872557116,
+      "grad_norm": 1.746512417994671,
+      "learning_rate": 1.7180518461672836e-05,
+      "loss": 1.0577,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2678227360308285,
+      "grad_norm": 1.6044823191779194,
+      "learning_rate": 1.7177415427404396e-05,
+      "loss": 1.1623,
+      "step": 1946
+    },
+    {
+      "epoch": 0.26796036333608586,
+      "grad_norm": 1.7603424617598378,
+      "learning_rate": 1.717431096708176e-05,
+      "loss": 0.9971,
+      "step": 1947
+    },
+    {
+      "epoch": 0.26809799064134326,
+      "grad_norm": 1.5225608378102733,
+      "learning_rate": 1.7171205081321742e-05,
+      "loss": 1.039,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2682356179466006,
+      "grad_norm": 1.8551322689402063,
+      "learning_rate": 1.716809777074144e-05,
+      "loss": 1.1384,
+      "step": 1949
+    },
+    {
+      "epoch": 0.26837324525185796,
+      "grad_norm": 1.7717513227515933,
+      "learning_rate": 1.7164989035958225e-05,
+      "loss": 1.0865,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2685108725571153,
+      "grad_norm": 1.6133262163783593,
+      "learning_rate": 1.716187887758977e-05,
+      "loss": 1.1506,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2686484998623727,
+      "grad_norm": 1.7626524384731777,
+      "learning_rate": 1.7158767296254017e-05,
+      "loss": 1.0407,
+      "step": 1952
+    },
+    {
+      "epoch": 0.26878612716763006,
+      "grad_norm": 1.8101465847649512,
+      "learning_rate": 1.7155654292569198e-05,
+      "loss": 1.0953,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2689237544728874,
+      "grad_norm": 1.7122761437446572,
+      "learning_rate": 1.7152539867153817e-05,
+      "loss": 0.9581,
+      "step": 1954
+    },
+    {
+      "epoch": 0.26906138177814476,
+      "grad_norm": 1.7473059609186494,
+      "learning_rate": 1.7149424020626676e-05,
+      "loss": 1.1008,
+      "step": 1955
+    },
+    {
+      "epoch": 0.26919900908340216,
+      "grad_norm": 2.047662105510371,
+      "learning_rate": 1.714630675360684e-05,
+      "loss": 1.139,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2693366363886595,
+      "grad_norm": 1.8116354563293793,
+      "learning_rate": 1.714318806671368e-05,
+      "loss": 1.1498,
+      "step": 1957
+    },
+    {
+      "epoch": 0.26947426369391686,
+      "grad_norm": 1.8657353341936453,
+      "learning_rate": 1.7140067960566833e-05,
+      "loss": 1.2056,
+      "step": 1958
+    },
+    {
+      "epoch": 0.2696118909991742,
+      "grad_norm": 1.993858766280044,
+      "learning_rate": 1.7136946435786214e-05,
+      "loss": 1.1834,
+      "step": 1959
+    },
+    {
+      "epoch": 0.2697495183044316,
+      "grad_norm": 1.6516881142660074,
+      "learning_rate": 1.7133823492992035e-05,
+      "loss": 0.9887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.26988714560968896,
+      "grad_norm": 1.6051890108291405,
+      "learning_rate": 1.713069913280478e-05,
+      "loss": 1.1352,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2700247729149463,
+      "grad_norm": 1.811981189197741,
+      "learning_rate": 1.7127573355845214e-05,
+      "loss": 1.0849,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2701624002202037,
+      "grad_norm": 1.9537132824913355,
+      "learning_rate": 1.7124446162734386e-05,
+      "loss": 1.1399,
+      "step": 1963
+    },
+    {
+      "epoch": 0.27030002752546106,
+      "grad_norm": 1.5747066062988588,
+      "learning_rate": 1.7121317554093634e-05,
+      "loss": 1.0836,
+      "step": 1964
+    },
+    {
+      "epoch": 0.2704376548307184,
+      "grad_norm": 1.7193288564484561,
+      "learning_rate": 1.711818753054456e-05,
+      "loss": 1.1238,
+      "step": 1965
+    },
+    {
+      "epoch": 0.27057528213597576,
+      "grad_norm": 1.681442305379545,
+      "learning_rate": 1.711505609270906e-05,
+      "loss": 1.0868,
+      "step": 1966
+    },
+    {
+      "epoch": 0.27071290944123316,
+      "grad_norm": 1.71362229032511,
+      "learning_rate": 1.7111923241209312e-05,
+      "loss": 1.1301,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2708505367464905,
+      "grad_norm": 1.772805656584986,
+      "learning_rate": 1.710878897666777e-05,
+      "loss": 1.1952,
+      "step": 1968
+    },
+    {
+      "epoch": 0.27098816405174786,
+      "grad_norm": 1.780145633181192,
+      "learning_rate": 1.7105653299707165e-05,
+      "loss": 1.0662,
+      "step": 1969
+    },
+    {
+      "epoch": 0.2711257913570052,
+      "grad_norm": 1.8325127966552426,
+      "learning_rate": 1.7102516210950514e-05,
+      "loss": 1.1697,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2712634186622626,
+      "grad_norm": 1.9325591083614,
+      "learning_rate": 1.709937771102111e-05,
+      "loss": 1.1291,
+      "step": 1971
+    },
+    {
+      "epoch": 0.27140104596751996,
+      "grad_norm": 1.798915549228028,
+      "learning_rate": 1.7096237800542543e-05,
+      "loss": 1.1337,
+      "step": 1972
+    },
+    {
+      "epoch": 0.2715386732727773,
+      "grad_norm": 1.7728295432391228,
+      "learning_rate": 1.7093096480138657e-05,
+      "loss": 1.1486,
+      "step": 1973
+    },
+    {
+      "epoch": 0.27167630057803466,
+      "grad_norm": 1.5303237245410317,
+      "learning_rate": 1.7089953750433595e-05,
+      "loss": 1.0136,
+      "step": 1974
+    },
+    {
+      "epoch": 0.27181392788329206,
+      "grad_norm": 2.044857472263054,
+      "learning_rate": 1.7086809612051776e-05,
+      "loss": 1.2353,
+      "step": 1975
+    },
+    {
+      "epoch": 0.2719515551885494,
+      "grad_norm": 1.8576738785882372,
+      "learning_rate": 1.7083664065617896e-05,
+      "loss": 1.1416,
+      "step": 1976
+    },
+    {
+      "epoch": 0.27208918249380676,
+      "grad_norm": 1.5763762699423334,
+      "learning_rate": 1.7080517111756927e-05,
+      "loss": 1.0329,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2722268097990641,
+      "grad_norm": 1.8272214769110426,
+      "learning_rate": 1.707736875109413e-05,
+      "loss": 1.1151,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2723644371043215,
+      "grad_norm": 1.6683754218281484,
+      "learning_rate": 1.7074218984255042e-05,
+      "loss": 1.1596,
+      "step": 1979
+    },
+    {
+      "epoch": 0.27250206440957886,
+      "grad_norm": 1.8800368708411355,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.1762,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2726396917148362,
+      "grad_norm": 1.6353189669205435,
+      "learning_rate": 1.706791523455153e-05,
+      "loss": 1.0275,
+      "step": 1981
+    },
+    {
+      "epoch": 0.2727773190200936,
+      "grad_norm": 1.5882897033146293,
+      "learning_rate": 1.7064761252939576e-05,
+      "loss": 1.029,
+      "step": 1982
+    },
+    {
+      "epoch": 0.27291494632535096,
+      "grad_norm": 1.6746140654096426,
+      "learning_rate": 1.7061605867656265e-05,
+      "loss": 1.0675,
+      "step": 1983
+    },
+    {
+      "epoch": 0.2730525736306083,
+      "grad_norm": 2.770223903014143,
+      "learning_rate": 1.7058449079328535e-05,
+      "loss": 1.108,
+      "step": 1984
+    },
+    {
+      "epoch": 0.27319020093586566,
+      "grad_norm": 1.6837115242135507,
+      "learning_rate": 1.7055290888583588e-05,
+      "loss": 1.0289,
+      "step": 1985
+    },
+    {
+      "epoch": 0.27332782824112306,
+      "grad_norm": 1.8272372111576105,
+      "learning_rate": 1.705213129604892e-05,
+      "loss": 1.1496,
+      "step": 1986
+    },
+    {
+      "epoch": 0.2734654555463804,
+      "grad_norm": 1.990414918090637,
+      "learning_rate": 1.7048970302352293e-05,
+      "loss": 1.1556,
+      "step": 1987
+    },
+    {
+      "epoch": 0.27360308285163776,
+      "grad_norm": 1.9616818140560504,
+      "learning_rate": 1.7045807908121764e-05,
+      "loss": 1.1601,
+      "step": 1988
+    },
+    {
+      "epoch": 0.2737407101568951,
+      "grad_norm": 1.8061629320836823,
+      "learning_rate": 1.7042644113985646e-05,
+      "loss": 1.0673,
+      "step": 1989
+    },
+    {
+      "epoch": 0.2738783374621525,
+      "grad_norm": 1.9462847703720496,
+      "learning_rate": 1.7039478920572546e-05,
+      "loss": 1.1882,
+      "step": 1990
+    },
+    {
+      "epoch": 0.27401596476740986,
+      "grad_norm": 1.561489046841805,
+      "learning_rate": 1.7036312328511345e-05,
+      "loss": 1.088,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2741535920726672,
+      "grad_norm": 1.7037076428665592,
+      "learning_rate": 1.70331443384312e-05,
+      "loss": 1.099,
+      "step": 1992
+    },
+    {
+      "epoch": 0.27429121937792456,
+      "grad_norm": 1.700474901653103,
+      "learning_rate": 1.7029974950961552e-05,
+      "loss": 1.0643,
+      "step": 1993
+    },
+    {
+      "epoch": 0.27442884668318196,
+      "grad_norm": 1.7313578169278945,
+      "learning_rate": 1.702680416673211e-05,
+      "loss": 1.0204,
+      "step": 1994
+    },
+    {
+      "epoch": 0.2745664739884393,
+      "grad_norm": 1.8549789928794156,
+      "learning_rate": 1.702363198637286e-05,
+      "loss": 1.1249,
+      "step": 1995
+    },
+    {
+      "epoch": 0.27470410129369666,
+      "grad_norm": 1.7640766559780332,
+      "learning_rate": 1.7020458410514083e-05,
+      "loss": 1.0,
+      "step": 1996
+    },
+    {
+      "epoch": 0.274841728598954,
+      "grad_norm": 1.603608528593956,
+      "learning_rate": 1.7017283439786317e-05,
+      "loss": 1.0562,
+      "step": 1997
+    },
+    {
+      "epoch": 0.2749793559042114,
+      "grad_norm": 2.0089059619933503,
+      "learning_rate": 1.7014107074820388e-05,
+      "loss": 1.0738,
+      "step": 1998
+    },
+    {
+      "epoch": 0.27511698320946876,
+      "grad_norm": 1.7113491987640808,
+      "learning_rate": 1.7010929316247397e-05,
+      "loss": 1.1221,
+      "step": 1999
+    },
+    {
+      "epoch": 0.2752546105147261,
+      "grad_norm": 1.6520336935761395,
+      "learning_rate": 1.7007750164698723e-05,
+      "loss": 1.0767,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2753922378199835,
+      "grad_norm": 1.7867215322466121,
+      "learning_rate": 1.7004569620806014e-05,
+      "loss": 1.1555,
+      "step": 2001
+    },
+    {
+      "epoch": 0.27552986512524086,
+      "grad_norm": 1.7537277899181478,
+      "learning_rate": 1.70013876852012e-05,
+      "loss": 1.1466,
+      "step": 2002
+    },
+    {
+      "epoch": 0.2756674924304982,
+      "grad_norm": 1.5679141104315972,
+      "learning_rate": 1.6998204358516497e-05,
+      "loss": 1.0114,
+      "step": 2003
+    },
+    {
+      "epoch": 0.27580511973575556,
+      "grad_norm": 1.9277787439763607,
+      "learning_rate": 1.699501964138438e-05,
+      "loss": 1.2311,
+      "step": 2004
+    },
+    {
+      "epoch": 0.27594274704101296,
+      "grad_norm": 1.6790631794101432,
+      "learning_rate": 1.6991833534437612e-05,
+      "loss": 1.1673,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2760803743462703,
+      "grad_norm": 1.76143559420636,
+      "learning_rate": 1.698864603830923e-05,
+      "loss": 1.0883,
+      "step": 2006
+    },
+    {
+      "epoch": 0.27621800165152766,
+      "grad_norm": 2.0215568945373263,
+      "learning_rate": 1.6985457153632543e-05,
+      "loss": 1.0502,
+      "step": 2007
+    },
+    {
+      "epoch": 0.276355628956785,
+      "grad_norm": 1.6028884654841207,
+      "learning_rate": 1.6982266881041143e-05,
+      "loss": 1.1212,
+      "step": 2008
+    },
+    {
+      "epoch": 0.2764932562620424,
+      "grad_norm": 1.7434382995298872,
+      "learning_rate": 1.6979075221168884e-05,
+      "loss": 1.1368,
+      "step": 2009
+    },
+    {
+      "epoch": 0.27663088356729976,
+      "grad_norm": 1.5856556318634336,
+      "learning_rate": 1.6975882174649915e-05,
+      "loss": 0.9787,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2767685108725571,
+      "grad_norm": 1.6314355809519026,
+      "learning_rate": 1.697268774211865e-05,
+      "loss": 1.0885,
+      "step": 2011
+    },
+    {
+      "epoch": 0.27690613817781445,
+      "grad_norm": 1.8896032766926456,
+      "learning_rate": 1.6969491924209774e-05,
+      "loss": 1.1325,
+      "step": 2012
+    },
+    {
+      "epoch": 0.27704376548307186,
+      "grad_norm": 1.7404251647478077,
+      "learning_rate": 1.6966294721558254e-05,
+      "loss": 1.1548,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2771813927883292,
+      "grad_norm": 2.398255128669592,
+      "learning_rate": 1.6963096134799333e-05,
+      "loss": 1.2102,
+      "step": 2014
+    },
+    {
+      "epoch": 0.27731902009358655,
+      "grad_norm": 1.6973110640252802,
+      "learning_rate": 1.6959896164568523e-05,
+      "loss": 1.1966,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2774566473988439,
+      "grad_norm": 1.9945316440579002,
+      "learning_rate": 1.6956694811501616e-05,
+      "loss": 1.2151,
+      "step": 2016
+    },
+    {
+      "epoch": 0.2775942747041013,
+      "grad_norm": 1.523829225233868,
+      "learning_rate": 1.6953492076234677e-05,
+      "loss": 0.9693,
+      "step": 2017
+    },
+    {
+      "epoch": 0.27773190200935866,
+      "grad_norm": 1.6497026518211393,
+      "learning_rate": 1.6950287959404045e-05,
+      "loss": 1.1018,
+      "step": 2018
+    },
+    {
+      "epoch": 0.277869529314616,
+      "grad_norm": 2.1331492739121183,
+      "learning_rate": 1.6947082461646332e-05,
+      "loss": 1.1367,
+      "step": 2019
+    },
+    {
+      "epoch": 0.2780071566198734,
+      "grad_norm": 1.7313574663210336,
+      "learning_rate": 1.6943875583598434e-05,
+      "loss": 1.0869,
+      "step": 2020
+    },
+    {
+      "epoch": 0.27814478392513076,
+      "grad_norm": 1.6446969823582942,
+      "learning_rate": 1.6940667325897507e-05,
+      "loss": 1.0628,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2782824112303881,
+      "grad_norm": 1.771172781755217,
+      "learning_rate": 1.693745768918099e-05,
+      "loss": 1.1357,
+      "step": 2022
+    },
+    {
+      "epoch": 0.27842003853564545,
+      "grad_norm": 1.717084998984212,
+      "learning_rate": 1.693424667408659e-05,
+      "loss": 1.1028,
+      "step": 2023
+    },
+    {
+      "epoch": 0.27855766584090286,
+      "grad_norm": 1.5905010474897845,
+      "learning_rate": 1.69310342812523e-05,
+      "loss": 1.0663,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2786952931461602,
+      "grad_norm": 1.6039250181365219,
+      "learning_rate": 1.692782051131637e-05,
+      "loss": 1.0653,
+      "step": 2025
+    },
+    {
+      "epoch": 0.27883292045141755,
+      "grad_norm": 1.7738622249834513,
+      "learning_rate": 1.692460536491734e-05,
+      "loss": 1.1121,
+      "step": 2026
+    },
+    {
+      "epoch": 0.2789705477566749,
+      "grad_norm": 1.667844574797634,
+      "learning_rate": 1.6921388842694008e-05,
+      "loss": 1.0626,
+      "step": 2027
+    },
+    {
+      "epoch": 0.2791081750619323,
+      "grad_norm": 1.565639110720571,
+      "learning_rate": 1.691817094528546e-05,
+      "loss": 1.0454,
+      "step": 2028
+    },
+    {
+      "epoch": 0.27924580236718965,
+      "grad_norm": 1.6397196513696386,
+      "learning_rate": 1.6914951673331043e-05,
+      "loss": 1.0442,
+      "step": 2029
+    },
+    {
+      "epoch": 0.279383429672447,
+      "grad_norm": 1.8331260842538097,
+      "learning_rate": 1.691173102747038e-05,
+      "loss": 1.0762,
+      "step": 2030
+    },
+    {
+      "epoch": 0.27952105697770435,
+      "grad_norm": 1.6156043947666379,
+      "learning_rate": 1.6908509008343377e-05,
+      "loss": 1.0451,
+      "step": 2031
+    },
+    {
+      "epoch": 0.27965868428296176,
+      "grad_norm": 1.8113405389011512,
+      "learning_rate": 1.69052856165902e-05,
+      "loss": 1.1375,
+      "step": 2032
+    },
+    {
+      "epoch": 0.2797963115882191,
+      "grad_norm": 1.7015895075657994,
+      "learning_rate": 1.6902060852851297e-05,
+      "loss": 1.121,
+      "step": 2033
+    },
+    {
+      "epoch": 0.27993393889347645,
+      "grad_norm": 1.6314739761852879,
+      "learning_rate": 1.689883471776738e-05,
+      "loss": 1.0972,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2800715661987338,
+      "grad_norm": 1.6141357725194707,
+      "learning_rate": 1.6895607211979438e-05,
+      "loss": 1.1153,
+      "step": 2035
+    },
+    {
+      "epoch": 0.2802091935039912,
+      "grad_norm": 1.8784245596494695,
+      "learning_rate": 1.6892378336128733e-05,
+      "loss": 1.0922,
+      "step": 2036
+    },
+    {
+      "epoch": 0.28034682080924855,
+      "grad_norm": 1.8360065183053507,
+      "learning_rate": 1.6889148090856798e-05,
+      "loss": 1.022,
+      "step": 2037
+    },
+    {
+      "epoch": 0.2804844481145059,
+      "grad_norm": 1.8498739749930087,
+      "learning_rate": 1.6885916476805445e-05,
+      "loss": 1.1159,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2806220754197633,
+      "grad_norm": 1.7397370249812885,
+      "learning_rate": 1.688268349461674e-05,
+      "loss": 1.1116,
+      "step": 2039
+    },
+    {
+      "epoch": 0.28075970272502065,
+      "grad_norm": 1.7478691208907988,
+      "learning_rate": 1.6879449144933043e-05,
+      "loss": 0.9518,
+      "step": 2040
+    },
+    {
+      "epoch": 0.280897330030278,
+      "grad_norm": 2.026744780936143,
+      "learning_rate": 1.6876213428396967e-05,
+      "loss": 1.1071,
+      "step": 2041
+    },
+    {
+      "epoch": 0.28103495733553535,
+      "grad_norm": 1.8082190117072885,
+      "learning_rate": 1.687297634565141e-05,
+      "loss": 1.0825,
+      "step": 2042
+    },
+    {
+      "epoch": 0.28117258464079276,
+      "grad_norm": 1.8343920831483869,
+      "learning_rate": 1.686973789733953e-05,
+      "loss": 1.1894,
+      "step": 2043
+    },
+    {
+      "epoch": 0.2813102119460501,
+      "grad_norm": 1.6593857308532538,
+      "learning_rate": 1.6866498084104774e-05,
+      "loss": 1.0536,
+      "step": 2044
+    },
+    {
+      "epoch": 0.28144783925130745,
+      "grad_norm": 2.3510775768472927,
+      "learning_rate": 1.6863256906590834e-05,
+      "loss": 1.13,
+      "step": 2045
+    },
+    {
+      "epoch": 0.2815854665565648,
+      "grad_norm": 1.908625469192015,
+      "learning_rate": 1.6860014365441695e-05,
+      "loss": 1.1122,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2817230938618222,
+      "grad_norm": 1.7133134367241183,
+      "learning_rate": 1.685677046130161e-05,
+      "loss": 1.0172,
+      "step": 2047
+    },
+    {
+      "epoch": 0.28186072116707955,
+      "grad_norm": 1.768728430140071,
+      "learning_rate": 1.6853525194815086e-05,
+      "loss": 1.1265,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2819983484723369,
+      "grad_norm": 1.6859237156226738,
+      "learning_rate": 1.6850278566626925e-05,
+      "loss": 1.1225,
+      "step": 2049
+    },
+    {
+      "epoch": 0.28213597577759425,
+      "grad_norm": 1.889295770954119,
+      "learning_rate": 1.6847030577382182e-05,
+      "loss": 1.2004,
+      "step": 2050
+    },
+    {
+      "epoch": 0.28227360308285165,
+      "grad_norm": 1.8893801408118909,
+      "learning_rate": 1.684378122772619e-05,
+      "loss": 1.0503,
+      "step": 2051
+    },
+    {
+      "epoch": 0.282411230388109,
+      "grad_norm": 1.660420819266096,
+      "learning_rate": 1.6840530518304547e-05,
+      "loss": 0.9928,
+      "step": 2052
+    },
+    {
+      "epoch": 0.28254885769336635,
+      "grad_norm": 1.927375167710676,
+      "learning_rate": 1.6837278449763126e-05,
+      "loss": 1.0799,
+      "step": 2053
+    },
+    {
+      "epoch": 0.2826864849986237,
+      "grad_norm": 1.6052413256281568,
+      "learning_rate": 1.6834025022748073e-05,
+      "loss": 1.1049,
+      "step": 2054
+    },
+    {
+      "epoch": 0.2828241123038811,
+      "grad_norm": 1.7558864199973336,
+      "learning_rate": 1.6830770237905794e-05,
+      "loss": 1.0905,
+      "step": 2055
+    },
+    {
+      "epoch": 0.28296173960913845,
+      "grad_norm": 1.9935096667523264,
+      "learning_rate": 1.6827514095882974e-05,
+      "loss": 1.1119,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2830993669143958,
+      "grad_norm": 1.6813325183346206,
+      "learning_rate": 1.6824256597326557e-05,
+      "loss": 1.0358,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2832369942196532,
+      "grad_norm": 1.7103888177110769,
+      "learning_rate": 1.682099774288377e-05,
+      "loss": 1.0622,
+      "step": 2058
+    },
+    {
+      "epoch": 0.28337462152491055,
+      "grad_norm": 1.676036343947673,
+      "learning_rate": 1.68177375332021e-05,
+      "loss": 1.0495,
+      "step": 2059
+    },
+    {
+      "epoch": 0.2835122488301679,
+      "grad_norm": 1.9400509676925761,
+      "learning_rate": 1.6814475968929308e-05,
+      "loss": 1.1378,
+      "step": 2060
+    },
+    {
+      "epoch": 0.28364987613542525,
+      "grad_norm": 1.9817733584747985,
+      "learning_rate": 1.681121305071342e-05,
+      "loss": 1.1447,
+      "step": 2061
+    },
+    {
+      "epoch": 0.28378750344068265,
+      "grad_norm": 1.5481459745731871,
+      "learning_rate": 1.6807948779202735e-05,
+      "loss": 1.0325,
+      "step": 2062
+    },
+    {
+      "epoch": 0.28392513074594,
+      "grad_norm": 1.7678235241275253,
+      "learning_rate": 1.680468315504582e-05,
+      "loss": 1.067,
+      "step": 2063
+    },
+    {
+      "epoch": 0.28406275805119735,
+      "grad_norm": 1.486088699376935,
+      "learning_rate": 1.6801416178891504e-05,
+      "loss": 1.165,
+      "step": 2064
+    },
+    {
+      "epoch": 0.2842003853564547,
+      "grad_norm": 1.9417658293501943,
+      "learning_rate": 1.6798147851388895e-05,
+      "loss": 1.1373,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2843380126617121,
+      "grad_norm": 2.018353271229725,
+      "learning_rate": 1.679487817318737e-05,
+      "loss": 1.1376,
+      "step": 2066
+    },
+    {
+      "epoch": 0.28447563996696945,
+      "grad_norm": 1.5940569546743422,
+      "learning_rate": 1.6791607144936556e-05,
+      "loss": 1.1078,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2846132672722268,
+      "grad_norm": 1.726469643801644,
+      "learning_rate": 1.678833476728637e-05,
+      "loss": 1.1575,
+      "step": 2068
+    },
+    {
+      "epoch": 0.28475089457748415,
+      "grad_norm": 1.7467337120359443,
+      "learning_rate": 1.6785061040886995e-05,
+      "loss": 1.0492,
+      "step": 2069
+    },
+    {
+      "epoch": 0.28488852188274155,
+      "grad_norm": 1.6427540995852143,
+      "learning_rate": 1.6781785966388863e-05,
+      "loss": 1.1052,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2850261491879989,
+      "grad_norm": 1.712182790593229,
+      "learning_rate": 1.6778509544442695e-05,
+      "loss": 1.0611,
+      "step": 2071
+    },
+    {
+      "epoch": 0.28516377649325625,
+      "grad_norm": 1.7508265125928628,
+      "learning_rate": 1.6775231775699464e-05,
+      "loss": 1.1772,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2853014037985136,
+      "grad_norm": 1.5271174131666998,
+      "learning_rate": 1.6771952660810426e-05,
+      "loss": 1.0645,
+      "step": 2073
+    },
+    {
+      "epoch": 0.285439031103771,
+      "grad_norm": 1.6501540475049745,
+      "learning_rate": 1.6768672200427092e-05,
+      "loss": 1.0302,
+      "step": 2074
+    },
+    {
+      "epoch": 0.28557665840902835,
+      "grad_norm": 1.7511606011008105,
+      "learning_rate": 1.6765390395201243e-05,
+      "loss": 1.1335,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.76017050570742,
+      "learning_rate": 1.6762107245784932e-05,
+      "loss": 1.148,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2858519130195431,
+      "grad_norm": 1.7510642322553143,
+      "learning_rate": 1.6758822752830473e-05,
+      "loss": 1.1483,
+      "step": 2077
+    },
+    {
+      "epoch": 0.28598954032480045,
+      "grad_norm": 1.783043410788411,
+      "learning_rate": 1.6755536916990453e-05,
+      "loss": 1.0983,
+      "step": 2078
+    },
+    {
+      "epoch": 0.2861271676300578,
+      "grad_norm": 1.7960061030180332,
+      "learning_rate": 1.675224973891772e-05,
+      "loss": 1.1555,
+      "step": 2079
+    },
+    {
+      "epoch": 0.28626479493531515,
+      "grad_norm": 1.6003936852187608,
+      "learning_rate": 1.6748961219265387e-05,
+      "loss": 1.1036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.28640242224057255,
+      "grad_norm": 1.605752343915571,
+      "learning_rate": 1.674567135868685e-05,
+      "loss": 1.1056,
+      "step": 2081
+    },
+    {
+      "epoch": 0.2865400495458299,
+      "grad_norm": 1.7049018525831086,
+      "learning_rate": 1.674238015783575e-05,
+      "loss": 1.0842,
+      "step": 2082
+    },
+    {
+      "epoch": 0.28667767685108725,
+      "grad_norm": 1.7292928903693676,
+      "learning_rate": 1.6739087617366006e-05,
+      "loss": 1.0618,
+      "step": 2083
+    },
+    {
+      "epoch": 0.2868153041563446,
+      "grad_norm": 1.9123822792078355,
+      "learning_rate": 1.6735793737931798e-05,
+      "loss": 1.1823,
+      "step": 2084
+    },
+    {
+      "epoch": 0.286952931461602,
+      "grad_norm": 1.8897127886437164,
+      "learning_rate": 1.6732498520187582e-05,
+      "loss": 1.0423,
+      "step": 2085
+    },
+    {
+      "epoch": 0.28709055876685935,
+      "grad_norm": 1.6758081769982054,
+      "learning_rate": 1.6729201964788066e-05,
+      "loss": 1.0977,
+      "step": 2086
+    },
+    {
+      "epoch": 0.2872281860721167,
+      "grad_norm": 1.6206252119432412,
+      "learning_rate": 1.6725904072388228e-05,
+      "loss": 1.0741,
+      "step": 2087
+    },
+    {
+      "epoch": 0.28736581337737405,
+      "grad_norm": 1.8853614573090982,
+      "learning_rate": 1.6722604843643324e-05,
+      "loss": 1.1017,
+      "step": 2088
+    },
+    {
+      "epoch": 0.28750344068263145,
+      "grad_norm": 1.5972405785196395,
+      "learning_rate": 1.6719304279208858e-05,
+      "loss": 1.0596,
+      "step": 2089
+    },
+    {
+      "epoch": 0.2876410679878888,
+      "grad_norm": 1.657682105825771,
+      "learning_rate": 1.6716002379740607e-05,
+      "loss": 1.1142,
+      "step": 2090
+    },
+    {
+      "epoch": 0.28777869529314615,
+      "grad_norm": 1.7818383910131304,
+      "learning_rate": 1.671269914589462e-05,
+      "loss": 1.1027,
+      "step": 2091
+    },
+    {
+      "epoch": 0.2879163225984035,
+      "grad_norm": 1.7278744896149576,
+      "learning_rate": 1.6709394578327195e-05,
+      "loss": 1.0935,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2880539499036609,
+      "grad_norm": 2.1172185681020412,
+      "learning_rate": 1.6706088677694913e-05,
+      "loss": 1.2641,
+      "step": 2093
+    },
+    {
+      "epoch": 0.28819157720891825,
+      "grad_norm": 1.6578770955831497,
+      "learning_rate": 1.6702781444654604e-05,
+      "loss": 1.1019,
+      "step": 2094
+    },
+    {
+      "epoch": 0.2883292045141756,
+      "grad_norm": 1.678905003613266,
+      "learning_rate": 1.6699472879863373e-05,
+      "loss": 1.1776,
+      "step": 2095
+    },
+    {
+      "epoch": 0.288466831819433,
+      "grad_norm": 1.7249513838734574,
+      "learning_rate": 1.6696162983978585e-05,
+      "loss": 1.099,
+      "step": 2096
+    },
+    {
+      "epoch": 0.28860445912469035,
+      "grad_norm": 1.7157457790469366,
+      "learning_rate": 1.6692851757657876e-05,
+      "loss": 1.0304,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2887420864299477,
+      "grad_norm": 1.751356486322032,
+      "learning_rate": 1.6689539201559137e-05,
+      "loss": 1.0388,
+      "step": 2098
+    },
+    {
+      "epoch": 0.28887971373520505,
+      "grad_norm": 1.6372315101728687,
+      "learning_rate": 1.668622531634053e-05,
+      "loss": 1.1452,
+      "step": 2099
+    },
+    {
+      "epoch": 0.28901734104046245,
+      "grad_norm": 2.1969869335082857,
+      "learning_rate": 1.668291010266047e-05,
+      "loss": 1.0927,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2891549683457198,
+      "grad_norm": 1.5655375643090481,
+      "learning_rate": 1.6679593561177664e-05,
+      "loss": 1.0737,
+      "step": 2101
+    },
+    {
+      "epoch": 0.28929259565097715,
+      "grad_norm": 1.6806373416439504,
+      "learning_rate": 1.667627569255104e-05,
+      "loss": 1.131,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2894302229562345,
+      "grad_norm": 1.693805495353178,
+      "learning_rate": 1.6672956497439828e-05,
+      "loss": 1.1175,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2895678502614919,
+      "grad_norm": 1.7734786165575591,
+      "learning_rate": 1.6669635976503505e-05,
+      "loss": 1.0966,
+      "step": 2104
+    },
+    {
+      "epoch": 0.28970547756674925,
+      "grad_norm": 1.7306698069346045,
+      "learning_rate": 1.6666314130401808e-05,
+      "loss": 1.0929,
+      "step": 2105
+    },
+    {
+      "epoch": 0.2898431048720066,
+      "grad_norm": 1.7871381038841916,
+      "learning_rate": 1.6662990959794746e-05,
+      "loss": 1.0415,
+      "step": 2106
+    },
+    {
+      "epoch": 0.28998073217726394,
+      "grad_norm": 1.6974647137783825,
+      "learning_rate": 1.6659666465342587e-05,
+      "loss": 1.0801,
+      "step": 2107
+    },
+    {
+      "epoch": 0.29011835948252135,
+      "grad_norm": 1.6806131179176573,
+      "learning_rate": 1.6656340647705862e-05,
+      "loss": 1.0692,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2902559867877787,
+      "grad_norm": 1.7812812653559127,
+      "learning_rate": 1.6653013507545364e-05,
+      "loss": 1.1454,
+      "step": 2109
+    },
+    {
+      "epoch": 0.29039361409303605,
+      "grad_norm": 1.7511479413227669,
+      "learning_rate": 1.6649685045522156e-05,
+      "loss": 1.04,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2905312413982934,
+      "grad_norm": 1.7277497700745206,
+      "learning_rate": 1.664635526229755e-05,
+      "loss": 1.1445,
+      "step": 2111
+    },
+    {
+      "epoch": 0.2906688687035508,
+      "grad_norm": 1.7180912328927473,
+      "learning_rate": 1.6643024158533135e-05,
+      "loss": 1.0931,
+      "step": 2112
+    },
+    {
+      "epoch": 0.29080649600880815,
+      "grad_norm": 1.5963932616255387,
+      "learning_rate": 1.663969173489075e-05,
+      "loss": 1.0567,
+      "step": 2113
+    },
+    {
+      "epoch": 0.2909441233140655,
+      "grad_norm": 1.56385242089709,
+      "learning_rate": 1.6636357992032507e-05,
+      "loss": 1.1286,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2910817506193229,
+      "grad_norm": 1.793848555515424,
+      "learning_rate": 1.663302293062077e-05,
+      "loss": 1.0721,
+      "step": 2115
+    },
+    {
+      "epoch": 0.29121937792458025,
+      "grad_norm": 1.79159613010735,
+      "learning_rate": 1.6629686551318173e-05,
+      "loss": 1.0812,
+      "step": 2116
+    },
+    {
+      "epoch": 0.2913570052298376,
+      "grad_norm": 1.8977632754286275,
+      "learning_rate": 1.6626348854787614e-05,
+      "loss": 1.1531,
+      "step": 2117
+    },
+    {
+      "epoch": 0.29149463253509494,
+      "grad_norm": 1.6853100707599935,
+      "learning_rate": 1.6623009841692236e-05,
+      "loss": 1.1202,
+      "step": 2118
+    },
+    {
+      "epoch": 0.29163225984035235,
+      "grad_norm": 1.8166463334233767,
+      "learning_rate": 1.6619669512695465e-05,
+      "loss": 1.1091,
+      "step": 2119
+    },
+    {
+      "epoch": 0.2917698871456097,
+      "grad_norm": 1.7284986176296433,
+      "learning_rate": 1.6616327868460973e-05,
+      "loss": 1.1024,
+      "step": 2120
+    },
+    {
+      "epoch": 0.29190751445086704,
+      "grad_norm": 1.9886611448999614,
+      "learning_rate": 1.6612984909652702e-05,
+      "loss": 1.1393,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2920451417561244,
+      "grad_norm": 1.6460826258669519,
+      "learning_rate": 1.6609640636934855e-05,
+      "loss": 1.0939,
+      "step": 2122
+    },
+    {
+      "epoch": 0.2921827690613818,
+      "grad_norm": 1.8468487323550347,
+      "learning_rate": 1.6606295050971885e-05,
+      "loss": 1.1459,
+      "step": 2123
+    },
+    {
+      "epoch": 0.29232039636663915,
+      "grad_norm": 1.8380348752082594,
+      "learning_rate": 1.660294815242852e-05,
+      "loss": 1.175,
+      "step": 2124
+    },
+    {
+      "epoch": 0.2924580236718965,
+      "grad_norm": 1.733619427261147,
+      "learning_rate": 1.6599599941969745e-05,
+      "loss": 1.1573,
+      "step": 2125
+    },
+    {
+      "epoch": 0.29259565097715384,
+      "grad_norm": 1.7694586840593325,
+      "learning_rate": 1.6596250420260796e-05,
+      "loss": 1.0829,
+      "step": 2126
+    },
+    {
+      "epoch": 0.29273327828241125,
+      "grad_norm": 1.6709911061553078,
+      "learning_rate": 1.659289958796718e-05,
+      "loss": 1.0377,
+      "step": 2127
+    },
+    {
+      "epoch": 0.2928709055876686,
+      "grad_norm": 1.6445240434433102,
+      "learning_rate": 1.6589547445754666e-05,
+      "loss": 1.0932,
+      "step": 2128
+    },
+    {
+      "epoch": 0.29300853289292594,
+      "grad_norm": 1.531996212617955,
+      "learning_rate": 1.6586193994289278e-05,
+      "loss": 1.0377,
+      "step": 2129
+    },
+    {
+      "epoch": 0.2931461601981833,
+      "grad_norm": 1.6920612129847745,
+      "learning_rate": 1.6582839234237292e-05,
+      "loss": 1.1508,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2932837875034407,
+      "grad_norm": 1.7545497398416776,
+      "learning_rate": 1.6579483166265262e-05,
+      "loss": 1.1011,
+      "step": 2131
+    },
+    {
+      "epoch": 0.29342141480869804,
+      "grad_norm": 1.6543622278050902,
+      "learning_rate": 1.657612579103999e-05,
+      "loss": 1.0769,
+      "step": 2132
+    },
+    {
+      "epoch": 0.2935590421139554,
+      "grad_norm": 1.8200202780595547,
+      "learning_rate": 1.6572767109228546e-05,
+      "loss": 1.1694,
+      "step": 2133
+    },
+    {
+      "epoch": 0.2936966694192128,
+      "grad_norm": 2.006285325817001,
+      "learning_rate": 1.6569407121498243e-05,
+      "loss": 1.2456,
+      "step": 2134
+    },
+    {
+      "epoch": 0.29383429672447015,
+      "grad_norm": 1.7388708192443536,
+      "learning_rate": 1.656604582851667e-05,
+      "loss": 1.1126,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2939719240297275,
+      "grad_norm": 1.9920297585358244,
+      "learning_rate": 1.6562683230951675e-05,
+      "loss": 1.1244,
+      "step": 2136
+    },
+    {
+      "epoch": 0.29410955133498484,
+      "grad_norm": 1.6311193082702358,
+      "learning_rate": 1.6559319329471352e-05,
+      "loss": 1.1316,
+      "step": 2137
+    },
+    {
+      "epoch": 0.29424717864024225,
+      "grad_norm": 1.8577968575926889,
+      "learning_rate": 1.6555954124744066e-05,
+      "loss": 1.1373,
+      "step": 2138
+    },
+    {
+      "epoch": 0.2943848059454996,
+      "grad_norm": 1.6870503392496519,
+      "learning_rate": 1.6552587617438435e-05,
+      "loss": 1.0251,
+      "step": 2139
+    },
+    {
+      "epoch": 0.29452243325075694,
+      "grad_norm": 1.715884648282807,
+      "learning_rate": 1.654921980822334e-05,
+      "loss": 1.0618,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2946600605560143,
+      "grad_norm": 1.8122712789499047,
+      "learning_rate": 1.654585069776792e-05,
+      "loss": 1.1109,
+      "step": 2141
+    },
+    {
+      "epoch": 0.2947976878612717,
+      "grad_norm": 1.7953016507353854,
+      "learning_rate": 1.654248028674157e-05,
+      "loss": 1.1484,
+      "step": 2142
+    },
+    {
+      "epoch": 0.29493531516652904,
+      "grad_norm": 1.5006542931792108,
+      "learning_rate": 1.6539108575813944e-05,
+      "loss": 1.0793,
+      "step": 2143
+    },
+    {
+      "epoch": 0.2950729424717864,
+      "grad_norm": 1.7215379876342598,
+      "learning_rate": 1.653573556565495e-05,
+      "loss": 1.049,
+      "step": 2144
+    },
+    {
+      "epoch": 0.29521056977704374,
+      "grad_norm": 1.7149317964778832,
+      "learning_rate": 1.6532361256934767e-05,
+      "loss": 1.0026,
+      "step": 2145
+    },
+    {
+      "epoch": 0.29534819708230114,
+      "grad_norm": 1.9238708956149693,
+      "learning_rate": 1.652898565032382e-05,
+      "loss": 1.1993,
+      "step": 2146
+    },
+    {
+      "epoch": 0.2954858243875585,
+      "grad_norm": 1.7780401240763801,
+      "learning_rate": 1.6525608746492794e-05,
+      "loss": 1.1176,
+      "step": 2147
+    },
+    {
+      "epoch": 0.29562345169281584,
+      "grad_norm": 1.8108252207537217,
+      "learning_rate": 1.6522230546112642e-05,
+      "loss": 1.0875,
+      "step": 2148
+    },
+    {
+      "epoch": 0.2957610789980732,
+      "grad_norm": 1.6437813213185566,
+      "learning_rate": 1.6518851049854555e-05,
+      "loss": 1.1163,
+      "step": 2149
+    },
+    {
+      "epoch": 0.2958987063033306,
+      "grad_norm": 1.8870554238769743,
+      "learning_rate": 1.6515470258390002e-05,
+      "loss": 1.0098,
+      "step": 2150
+    },
+    {
+      "epoch": 0.29603633360858794,
+      "grad_norm": 1.6117151758740642,
+      "learning_rate": 1.651208817239069e-05,
+      "loss": 1.1571,
+      "step": 2151
+    },
+    {
+      "epoch": 0.2961739609138453,
+      "grad_norm": 2.095468072153884,
+      "learning_rate": 1.6508704792528604e-05,
+      "loss": 1.0987,
+      "step": 2152
+    },
+    {
+      "epoch": 0.2963115882191027,
+      "grad_norm": 1.9865657748429184,
+      "learning_rate": 1.6505320119475966e-05,
+      "loss": 1.172,
+      "step": 2153
+    },
+    {
+      "epoch": 0.29644921552436004,
+      "grad_norm": 1.764098261595099,
+      "learning_rate": 1.650193415390527e-05,
+      "loss": 1.0615,
+      "step": 2154
+    },
+    {
+      "epoch": 0.2965868428296174,
+      "grad_norm": 1.6973956180479857,
+      "learning_rate": 1.649854689648926e-05,
+      "loss": 1.2213,
+      "step": 2155
+    },
+    {
+      "epoch": 0.29672447013487474,
+      "grad_norm": 1.8127625711383928,
+      "learning_rate": 1.6495158347900937e-05,
+      "loss": 1.1535,
+      "step": 2156
+    },
+    {
+      "epoch": 0.29686209744013214,
+      "grad_norm": 1.7039888646601242,
+      "learning_rate": 1.6491768508813558e-05,
+      "loss": 1.0719,
+      "step": 2157
+    },
+    {
+      "epoch": 0.2969997247453895,
+      "grad_norm": 1.6593005465021127,
+      "learning_rate": 1.6488377379900643e-05,
+      "loss": 1.1093,
+      "step": 2158
+    },
+    {
+      "epoch": 0.29713735205064684,
+      "grad_norm": 1.7216106082146327,
+      "learning_rate": 1.6484984961835956e-05,
+      "loss": 1.0818,
+      "step": 2159
+    },
+    {
+      "epoch": 0.2972749793559042,
+      "grad_norm": 1.632503223181387,
+      "learning_rate": 1.6481591255293525e-05,
+      "loss": 1.12,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2974126066611616,
+      "grad_norm": 1.884893814508176,
+      "learning_rate": 1.647819626094764e-05,
+      "loss": 1.1713,
+      "step": 2161
+    },
+    {
+      "epoch": 0.29755023396641894,
+      "grad_norm": 1.8644278591125731,
+      "learning_rate": 1.6474799979472828e-05,
+      "loss": 1.1693,
+      "step": 2162
+    },
+    {
+      "epoch": 0.2976878612716763,
+      "grad_norm": 1.6593990825486917,
+      "learning_rate": 1.647140241154389e-05,
+      "loss": 1.0791,
+      "step": 2163
+    },
+    {
+      "epoch": 0.29782548857693364,
+      "grad_norm": 1.8568199320582006,
+      "learning_rate": 1.6468003557835886e-05,
+      "loss": 1.114,
+      "step": 2164
+    },
+    {
+      "epoch": 0.29796311588219104,
+      "grad_norm": 1.6454675420236906,
+      "learning_rate": 1.646460341902411e-05,
+      "loss": 1.1515,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2981007431874484,
+      "grad_norm": 1.5830823243476428,
+      "learning_rate": 1.646120199578412e-05,
+      "loss": 1.1014,
+      "step": 2166
+    },
+    {
+      "epoch": 0.29823837049270574,
+      "grad_norm": 1.9572069521366735,
+      "learning_rate": 1.645779928879174e-05,
+      "loss": 1.1042,
+      "step": 2167
+    },
+    {
+      "epoch": 0.2983759977979631,
+      "grad_norm": 1.7596426280324715,
+      "learning_rate": 1.645439529872304e-05,
+      "loss": 1.1576,
+      "step": 2168
+    },
+    {
+      "epoch": 0.2985136251032205,
+      "grad_norm": 1.7289258948218793,
+      "learning_rate": 1.6450990026254343e-05,
+      "loss": 1.0765,
+      "step": 2169
+    },
+    {
+      "epoch": 0.29865125240847784,
+      "grad_norm": 1.7355700279786292,
+      "learning_rate": 1.6447583472062234e-05,
+      "loss": 1.1058,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2987888797137352,
+      "grad_norm": 1.716961753595389,
+      "learning_rate": 1.6444175636823547e-05,
+      "loss": 1.0671,
+      "step": 2171
+    },
+    {
+      "epoch": 0.2989265070189926,
+      "grad_norm": 1.7436806458474103,
+      "learning_rate": 1.644076652121537e-05,
+      "loss": 1.0839,
+      "step": 2172
+    },
+    {
+      "epoch": 0.29906413432424994,
+      "grad_norm": 1.5676958407958548,
+      "learning_rate": 1.6437356125915052e-05,
+      "loss": 1.0675,
+      "step": 2173
+    },
+    {
+      "epoch": 0.2992017616295073,
+      "grad_norm": 1.6711782723401658,
+      "learning_rate": 1.643394445160019e-05,
+      "loss": 1.1298,
+      "step": 2174
+    },
+    {
+      "epoch": 0.29933938893476464,
+      "grad_norm": 1.6439680415014082,
+      "learning_rate": 1.6430531498948633e-05,
+      "loss": 1.0751,
+      "step": 2175
+    },
+    {
+      "epoch": 0.29947701624002204,
+      "grad_norm": 1.6635855031244835,
+      "learning_rate": 1.6427117268638495e-05,
+      "loss": 1.095,
+      "step": 2176
+    },
+    {
+      "epoch": 0.2996146435452794,
+      "grad_norm": 2.042829635238149,
+      "learning_rate": 1.642370176134813e-05,
+      "loss": 1.081,
+      "step": 2177
+    },
+    {
+      "epoch": 0.29975227085053674,
+      "grad_norm": 1.745096513996818,
+      "learning_rate": 1.6420284977756158e-05,
+      "loss": 0.9773,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2998898981557941,
+      "grad_norm": 1.8572861864472607,
+      "learning_rate": 1.6416866918541446e-05,
+      "loss": 1.0224,
+      "step": 2179
+    },
+    {
+      "epoch": 0.3000275254610515,
+      "grad_norm": 1.711667156535208,
+      "learning_rate": 1.6413447584383114e-05,
+      "loss": 1.0908,
+      "step": 2180
+    },
+    {
+      "epoch": 0.30016515276630884,
+      "grad_norm": 1.6025759139010358,
+      "learning_rate": 1.641002697596054e-05,
+      "loss": 1.0985,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3003027800715662,
+      "grad_norm": 1.7769781874005735,
+      "learning_rate": 1.6406605093953345e-05,
+      "loss": 1.1335,
+      "step": 2182
+    },
+    {
+      "epoch": 0.30044040737682354,
+      "grad_norm": 1.6739004727427294,
+      "learning_rate": 1.6403181939041417e-05,
+      "loss": 1.0949,
+      "step": 2183
+    },
+    {
+      "epoch": 0.30057803468208094,
+      "grad_norm": 1.8531095080188655,
+      "learning_rate": 1.6399757511904888e-05,
+      "loss": 1.1202,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3007156619873383,
+      "grad_norm": 1.746308885201986,
+      "learning_rate": 1.639633181322414e-05,
+      "loss": 1.0776,
+      "step": 2185
+    },
+    {
+      "epoch": 0.30085328929259564,
+      "grad_norm": 1.698615720350092,
+      "learning_rate": 1.6392904843679823e-05,
+      "loss": 1.1175,
+      "step": 2186
+    },
+    {
+      "epoch": 0.300990916597853,
+      "grad_norm": 1.8932228350098081,
+      "learning_rate": 1.6389476603952822e-05,
+      "loss": 1.0837,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3011285439031104,
+      "grad_norm": 1.8700853274993954,
+      "learning_rate": 1.638604709472428e-05,
+      "loss": 1.141,
+      "step": 2188
+    },
+    {
+      "epoch": 0.30126617120836774,
+      "grad_norm": 1.6316886887307018,
+      "learning_rate": 1.63826163166756e-05,
+      "loss": 1.0856,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3014037985136251,
+      "grad_norm": 1.7691010349200074,
+      "learning_rate": 1.637918427048842e-05,
+      "loss": 1.1018,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3015414258188825,
+      "grad_norm": 1.5749834042975361,
+      "learning_rate": 1.637575095684465e-05,
+      "loss": 1.165,
+      "step": 2191
+    },
+    {
+      "epoch": 0.30167905312413984,
+      "grad_norm": 1.7723958149247103,
+      "learning_rate": 1.637231637642644e-05,
+      "loss": 1.0807,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3018166804293972,
+      "grad_norm": 1.670026288605514,
+      "learning_rate": 1.6368880529916192e-05,
+      "loss": 1.1037,
+      "step": 2193
+    },
+    {
+      "epoch": 0.30195430773465454,
+      "grad_norm": 1.6994838348732557,
+      "learning_rate": 1.636544341799656e-05,
+      "loss": 1.0256,
+      "step": 2194
+    },
+    {
+      "epoch": 0.30209193503991194,
+      "grad_norm": 1.9471310342303108,
+      "learning_rate": 1.6362005041350462e-05,
+      "loss": 1.1392,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3022295623451693,
+      "grad_norm": 1.7249503502425678,
+      "learning_rate": 1.6358565400661045e-05,
+      "loss": 1.0925,
+      "step": 2196
+    },
+    {
+      "epoch": 0.30236718965042664,
+      "grad_norm": 1.6631911375870054,
+      "learning_rate": 1.635512449661172e-05,
+      "loss": 1.0531,
+      "step": 2197
+    },
+    {
+      "epoch": 0.302504816955684,
+      "grad_norm": 1.74940569141512,
+      "learning_rate": 1.6351682329886155e-05,
+      "loss": 1.0969,
+      "step": 2198
+    },
+    {
+      "epoch": 0.3026424442609414,
+      "grad_norm": 1.61721086385767,
+      "learning_rate": 1.634823890116825e-05,
+      "loss": 1.07,
+      "step": 2199
+    },
+    {
+      "epoch": 0.30278007156619874,
+      "grad_norm": 1.694969940640828,
+      "learning_rate": 1.6344794211142176e-05,
+      "loss": 1.0358,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3029176988714561,
+      "grad_norm": 1.800868565113983,
+      "learning_rate": 1.6341348260492345e-05,
+      "loss": 1.1247,
+      "step": 2201
+    },
+    {
+      "epoch": 0.30305532617671344,
+      "grad_norm": 1.6831847817808754,
+      "learning_rate": 1.6337901049903418e-05,
+      "loss": 1.0509,
+      "step": 2202
+    },
+    {
+      "epoch": 0.30319295348197084,
+      "grad_norm": 1.7824390073896086,
+      "learning_rate": 1.6334452580060305e-05,
+      "loss": 1.1158,
+      "step": 2203
+    },
+    {
+      "epoch": 0.3033305807872282,
+      "grad_norm": 1.7308634254944903,
+      "learning_rate": 1.6331002851648178e-05,
+      "loss": 1.1346,
+      "step": 2204
+    },
+    {
+      "epoch": 0.30346820809248554,
+      "grad_norm": 1.6582720993973892,
+      "learning_rate": 1.6327551865352442e-05,
+      "loss": 1.0143,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3036058353977429,
+      "grad_norm": 1.8849399992819307,
+      "learning_rate": 1.6324099621858768e-05,
+      "loss": 1.1521,
+      "step": 2206
+    },
+    {
+      "epoch": 0.3037434627030003,
+      "grad_norm": 1.6056172104066964,
+      "learning_rate": 1.6320646121853065e-05,
+      "loss": 1.1414,
+      "step": 2207
+    },
+    {
+      "epoch": 0.30388109000825764,
+      "grad_norm": 1.7702731689466813,
+      "learning_rate": 1.6317191366021502e-05,
+      "loss": 1.1071,
+      "step": 2208
+    },
+    {
+      "epoch": 0.304018717313515,
+      "grad_norm": 1.9420385981894535,
+      "learning_rate": 1.631373535505048e-05,
+      "loss": 1.1815,
+      "step": 2209
+    },
+    {
+      "epoch": 0.3041563446187724,
+      "grad_norm": 1.785972796997425,
+      "learning_rate": 1.6310278089626672e-05,
+      "loss": 1.0727,
+      "step": 2210
+    },
+    {
+      "epoch": 0.30429397192402974,
+      "grad_norm": 1.8586894968951926,
+      "learning_rate": 1.6306819570436988e-05,
+      "loss": 1.1304,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3044315992292871,
+      "grad_norm": 1.661374049824449,
+      "learning_rate": 1.6303359798168584e-05,
+      "loss": 1.1286,
+      "step": 2212
+    },
+    {
+      "epoch": 0.30456922653454443,
+      "grad_norm": 1.5390374764803356,
+      "learning_rate": 1.6299898773508875e-05,
+      "loss": 1.0984,
+      "step": 2213
+    },
+    {
+      "epoch": 0.30470685383980184,
+      "grad_norm": 1.8998446491509091,
+      "learning_rate": 1.6296436497145514e-05,
+      "loss": 1.1048,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3048444811450592,
+      "grad_norm": 1.717422029097654,
+      "learning_rate": 1.629297296976641e-05,
+      "loss": 1.0921,
+      "step": 2215
+    },
+    {
+      "epoch": 0.30498210845031654,
+      "grad_norm": 1.9147297403392152,
+      "learning_rate": 1.628950819205972e-05,
+      "loss": 1.0979,
+      "step": 2216
+    },
+    {
+      "epoch": 0.3051197357555739,
+      "grad_norm": 2.1365262779059946,
+      "learning_rate": 1.628604216471384e-05,
+      "loss": 1.0939,
+      "step": 2217
+    },
+    {
+      "epoch": 0.3052573630608313,
+      "grad_norm": 1.6266334310651613,
+      "learning_rate": 1.6282574888417433e-05,
+      "loss": 1.0722,
+      "step": 2218
+    },
+    {
+      "epoch": 0.30539499036608864,
+      "grad_norm": 1.8071929402245814,
+      "learning_rate": 1.6279106363859395e-05,
+      "loss": 1.1031,
+      "step": 2219
+    },
+    {
+      "epoch": 0.305532617671346,
+      "grad_norm": 1.6130987316196415,
+      "learning_rate": 1.627563659172887e-05,
+      "loss": 1.0878,
+      "step": 2220
+    },
+    {
+      "epoch": 0.30567024497660333,
+      "grad_norm": 1.84737530188381,
+      "learning_rate": 1.6272165572715263e-05,
+      "loss": 1.0674,
+      "step": 2221
+    },
+    {
+      "epoch": 0.30580787228186074,
+      "grad_norm": 1.7736845027614323,
+      "learning_rate": 1.6268693307508213e-05,
+      "loss": 1.1688,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3059454995871181,
+      "grad_norm": 1.803235048461757,
+      "learning_rate": 1.626521979679761e-05,
+      "loss": 1.1824,
+      "step": 2223
+    },
+    {
+      "epoch": 0.30608312689237543,
+      "grad_norm": 1.5839709871366263,
+      "learning_rate": 1.6261745041273595e-05,
+      "loss": 1.1278,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3062207541976328,
+      "grad_norm": 1.7732457074217556,
+      "learning_rate": 1.6258269041626555e-05,
+      "loss": 1.1044,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3063583815028902,
+      "grad_norm": 1.7835034933126916,
+      "learning_rate": 1.6254791798547122e-05,
+      "loss": 1.0824,
+      "step": 2226
+    },
+    {
+      "epoch": 0.30649600880814754,
+      "grad_norm": 1.5554097238362798,
+      "learning_rate": 1.6251313312726177e-05,
+      "loss": 1.1209,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3066336361134049,
+      "grad_norm": 1.6990032978058978,
+      "learning_rate": 1.624783358485485e-05,
+      "loss": 1.1452,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3067712634186623,
+      "grad_norm": 1.738735025735243,
+      "learning_rate": 1.6244352615624514e-05,
+      "loss": 1.1642,
+      "step": 2229
+    },
+    {
+      "epoch": 0.30690889072391964,
+      "grad_norm": 1.7024162924238175,
+      "learning_rate": 1.6240870405726786e-05,
+      "loss": 1.059,
+      "step": 2230
+    },
+    {
+      "epoch": 0.307046518029177,
+      "grad_norm": 1.8289147813991757,
+      "learning_rate": 1.6237386955853536e-05,
+      "loss": 1.0577,
+      "step": 2231
+    },
+    {
+      "epoch": 0.30718414533443433,
+      "grad_norm": 1.9145491609028986,
+      "learning_rate": 1.623390226669688e-05,
+      "loss": 1.0868,
+      "step": 2232
+    },
+    {
+      "epoch": 0.30732177263969174,
+      "grad_norm": 1.6728680762237125,
+      "learning_rate": 1.623041633894918e-05,
+      "loss": 1.0151,
+      "step": 2233
+    },
+    {
+      "epoch": 0.3074593999449491,
+      "grad_norm": 1.8793449209981985,
+      "learning_rate": 1.6226929173303034e-05,
+      "loss": 1.064,
+      "step": 2234
+    },
+    {
+      "epoch": 0.30759702725020643,
+      "grad_norm": 1.735271770532169,
+      "learning_rate": 1.62234407704513e-05,
+      "loss": 1.0983,
+      "step": 2235
+    },
+    {
+      "epoch": 0.3077346545554638,
+      "grad_norm": 1.8815878041058784,
+      "learning_rate": 1.6219951131087076e-05,
+      "loss": 1.0307,
+      "step": 2236
+    },
+    {
+      "epoch": 0.3078722818607212,
+      "grad_norm": 1.8211030615560844,
+      "learning_rate": 1.62164602559037e-05,
+      "loss": 1.1275,
+      "step": 2237
+    },
+    {
+      "epoch": 0.30800990916597853,
+      "grad_norm": 1.5460532431063154,
+      "learning_rate": 1.621296814559477e-05,
+      "loss": 1.0331,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3081475364712359,
+      "grad_norm": 1.5546189096400589,
+      "learning_rate": 1.6209474800854114e-05,
+      "loss": 1.0785,
+      "step": 2239
+    },
+    {
+      "epoch": 0.30828516377649323,
+      "grad_norm": 1.7309889418156763,
+      "learning_rate": 1.620598022237581e-05,
+      "loss": 1.1228,
+      "step": 2240
+    },
+    {
+      "epoch": 0.30842279108175064,
+      "grad_norm": 1.8284616743870008,
+      "learning_rate": 1.620248441085419e-05,
+      "loss": 1.2583,
+      "step": 2241
+    },
+    {
+      "epoch": 0.308560418387008,
+      "grad_norm": 1.8788026617379256,
+      "learning_rate": 1.619898736698382e-05,
+      "loss": 1.1232,
+      "step": 2242
+    },
+    {
+      "epoch": 0.30869804569226533,
+      "grad_norm": 1.9290747067652312,
+      "learning_rate": 1.6195489091459515e-05,
+      "loss": 1.2494,
+      "step": 2243
+    },
+    {
+      "epoch": 0.3088356729975227,
+      "grad_norm": 1.6694916878253072,
+      "learning_rate": 1.6191989584976332e-05,
+      "loss": 1.0606,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3089733003027801,
+      "grad_norm": 1.915135087540746,
+      "learning_rate": 1.6188488848229584e-05,
+      "loss": 1.0291,
+      "step": 2245
+    },
+    {
+      "epoch": 0.30911092760803743,
+      "grad_norm": 1.7434048860870541,
+      "learning_rate": 1.6184986881914808e-05,
+      "loss": 1.0229,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3092485549132948,
+      "grad_norm": 2.0144486003700126,
+      "learning_rate": 1.6181483686727802e-05,
+      "loss": 1.1263,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3093861822185522,
+      "grad_norm": 1.9089431424538068,
+      "learning_rate": 1.6177979263364606e-05,
+      "loss": 1.0712,
+      "step": 2248
+    },
+    {
+      "epoch": 0.30952380952380953,
+      "grad_norm": 2.0708949144574684,
+      "learning_rate": 1.6174473612521496e-05,
+      "loss": 1.1152,
+      "step": 2249
+    },
+    {
+      "epoch": 0.3096614368290669,
+      "grad_norm": 1.68696118189307,
+      "learning_rate": 1.6170966734894996e-05,
+      "loss": 1.1885,
+      "step": 2250
+    },
+    {
+      "epoch": 0.30979906413432423,
+      "grad_norm": 1.863931555526344,
+      "learning_rate": 1.6167458631181877e-05,
+      "loss": 1.1366,
+      "step": 2251
+    },
+    {
+      "epoch": 0.30993669143958164,
+      "grad_norm": 1.8456116696089355,
+      "learning_rate": 1.6163949302079153e-05,
+      "loss": 1.1017,
+      "step": 2252
+    },
+    {
+      "epoch": 0.310074318744839,
+      "grad_norm": 1.5578508265517148,
+      "learning_rate": 1.616043874828408e-05,
+      "loss": 1.0415,
+      "step": 2253
+    },
+    {
+      "epoch": 0.31021194605009633,
+      "grad_norm": 1.882952307713398,
+      "learning_rate": 1.6156926970494154e-05,
+      "loss": 1.1674,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3103495733553537,
+      "grad_norm": 1.9021672831497831,
+      "learning_rate": 1.6153413969407113e-05,
+      "loss": 1.0952,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3104872006606111,
+      "grad_norm": 1.8568328774281189,
+      "learning_rate": 1.6149899745720952e-05,
+      "loss": 1.1495,
+      "step": 2256
+    },
+    {
+      "epoch": 0.31062482796586843,
+      "grad_norm": 1.7894773883954638,
+      "learning_rate": 1.6146384300133895e-05,
+      "loss": 1.0494,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3107624552711258,
+      "grad_norm": 1.690909862036309,
+      "learning_rate": 1.614286763334441e-05,
+      "loss": 1.1476,
+      "step": 2258
+    },
+    {
+      "epoch": 0.31090008257638313,
+      "grad_norm": 1.6403130211689063,
+      "learning_rate": 1.6139349746051215e-05,
+      "loss": 1.1188,
+      "step": 2259
+    },
+    {
+      "epoch": 0.31103770988164053,
+      "grad_norm": 1.6755032293719765,
+      "learning_rate": 1.6135830638953264e-05,
+      "loss": 1.0986,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3111753371868979,
+      "grad_norm": 1.7511703818391924,
+      "learning_rate": 1.6132310312749754e-05,
+      "loss": 1.0851,
+      "step": 2261
+    },
+    {
+      "epoch": 0.31131296449215523,
+      "grad_norm": 1.6767572025242836,
+      "learning_rate": 1.612878876814013e-05,
+      "loss": 1.0885,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3114505917974126,
+      "grad_norm": 1.6584814649289747,
+      "learning_rate": 1.612526600582407e-05,
+      "loss": 1.1916,
+      "step": 2263
+    },
+    {
+      "epoch": 0.31158821910267,
+      "grad_norm": 1.8397010163800933,
+      "learning_rate": 1.6121742026501496e-05,
+      "loss": 1.0342,
+      "step": 2264
+    },
+    {
+      "epoch": 0.31172584640792733,
+      "grad_norm": 1.8445151982548667,
+      "learning_rate": 1.6118216830872583e-05,
+      "loss": 1.0583,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3118634737131847,
+      "grad_norm": 1.7132185187576148,
+      "learning_rate": 1.6114690419637736e-05,
+      "loss": 1.0047,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3120011010184421,
+      "grad_norm": 1.6089838698094958,
+      "learning_rate": 1.6111162793497604e-05,
+      "loss": 1.0095,
+      "step": 2267
+    },
+    {
+      "epoch": 0.31213872832369943,
+      "grad_norm": 1.5652485444464184,
+      "learning_rate": 1.6107633953153075e-05,
+      "loss": 1.0992,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3122763556289568,
+      "grad_norm": 1.7125184318109963,
+      "learning_rate": 1.6104103899305284e-05,
+      "loss": 1.1326,
+      "step": 2269
+    },
+    {
+      "epoch": 0.31241398293421413,
+      "grad_norm": 1.8968762591900499,
+      "learning_rate": 1.6100572632655603e-05,
+      "loss": 1.0976,
+      "step": 2270
+    },
+    {
+      "epoch": 0.31255161023947153,
+      "grad_norm": 1.8648520467517655,
+      "learning_rate": 1.609704015390565e-05,
+      "loss": 1.0975,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3126892375447289,
+      "grad_norm": 1.6195633596302104,
+      "learning_rate": 1.609350646375728e-05,
+      "loss": 1.0805,
+      "step": 2272
+    },
+    {
+      "epoch": 0.31282686484998623,
+      "grad_norm": 1.8936009498165187,
+      "learning_rate": 1.608997156291258e-05,
+      "loss": 1.1756,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3129644921552436,
+      "grad_norm": 1.9339836397657018,
+      "learning_rate": 1.60864354520739e-05,
+      "loss": 1.0214,
+      "step": 2274
+    },
+    {
+      "epoch": 0.313102119460501,
+      "grad_norm": 1.764479598542925,
+      "learning_rate": 1.608289813194381e-05,
+      "loss": 1.0112,
+      "step": 2275
+    },
+    {
+      "epoch": 0.31323974676575833,
+      "grad_norm": 1.6743033308276218,
+      "learning_rate": 1.6079359603225123e-05,
+      "loss": 1.1987,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3133773740710157,
+      "grad_norm": 1.7292733751922802,
+      "learning_rate": 1.60758198666209e-05,
+      "loss": 1.0511,
+      "step": 2277
+    },
+    {
+      "epoch": 0.31351500137627303,
+      "grad_norm": 1.763973526400724,
+      "learning_rate": 1.6072278922834442e-05,
+      "loss": 1.0588,
+      "step": 2278
+    },
+    {
+      "epoch": 0.31365262868153043,
+      "grad_norm": 1.7162165761275257,
+      "learning_rate": 1.606873677256928e-05,
+      "loss": 1.0871,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3137902559867878,
+      "grad_norm": 1.6769339809803583,
+      "learning_rate": 1.6065193416529195e-05,
+      "loss": 0.9978,
+      "step": 2280
+    },
+    {
+      "epoch": 0.31392788329204513,
+      "grad_norm": 1.9054773302687735,
+      "learning_rate": 1.6061648855418197e-05,
+      "loss": 1.1643,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3140655105973025,
+      "grad_norm": 1.6416762172112254,
+      "learning_rate": 1.605810308994055e-05,
+      "loss": 1.0335,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3142031379025599,
+      "grad_norm": 1.803399031523385,
+      "learning_rate": 1.6054556120800742e-05,
+      "loss": 1.148,
+      "step": 2283
+    },
+    {
+      "epoch": 0.31434076520781723,
+      "grad_norm": 1.6220559659085125,
+      "learning_rate": 1.6051007948703517e-05,
+      "loss": 1.0901,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3144783925130746,
+      "grad_norm": 2.0672646693819425,
+      "learning_rate": 1.6047458574353838e-05,
+      "loss": 1.0439,
+      "step": 2285
+    },
+    {
+      "epoch": 0.314616019818332,
+      "grad_norm": 1.6291604438891474,
+      "learning_rate": 1.604390799845692e-05,
+      "loss": 0.9976,
+      "step": 2286
+    },
+    {
+      "epoch": 0.31475364712358933,
+      "grad_norm": 1.8960126114828377,
+      "learning_rate": 1.604035622171822e-05,
+      "loss": 1.1045,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3148912744288467,
+      "grad_norm": 1.731972933116561,
+      "learning_rate": 1.6036803244843415e-05,
+      "loss": 1.0936,
+      "step": 2288
+    },
+    {
+      "epoch": 0.315028901734104,
+      "grad_norm": 1.6255213666812534,
+      "learning_rate": 1.6033249068538445e-05,
+      "loss": 1.0364,
+      "step": 2289
+    },
+    {
+      "epoch": 0.31516652903936143,
+      "grad_norm": 1.6970566255347188,
+      "learning_rate": 1.602969369350947e-05,
+      "loss": 1.0918,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3153041563446188,
+      "grad_norm": 1.8604537566046033,
+      "learning_rate": 1.6026137120462895e-05,
+      "loss": 1.1146,
+      "step": 2291
+    },
+    {
+      "epoch": 0.31544178364987613,
+      "grad_norm": 1.6135813274240594,
+      "learning_rate": 1.6022579350105362e-05,
+      "loss": 1.038,
+      "step": 2292
+    },
+    {
+      "epoch": 0.3155794109551335,
+      "grad_norm": 1.6749503165053843,
+      "learning_rate": 1.6019020383143754e-05,
+      "loss": 1.1293,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3157170382603909,
+      "grad_norm": 1.6840539808506523,
+      "learning_rate": 1.6015460220285185e-05,
+      "loss": 1.1406,
+      "step": 2294
+    },
+    {
+      "epoch": 0.31585466556564823,
+      "grad_norm": 1.7136770477155474,
+      "learning_rate": 1.601189886223702e-05,
+      "loss": 1.1485,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3159922928709056,
+      "grad_norm": 1.5846586128283668,
+      "learning_rate": 1.6008336309706835e-05,
+      "loss": 1.0802,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3161299201761629,
+      "grad_norm": 1.7663962216175761,
+      "learning_rate": 1.6004772563402477e-05,
+      "loss": 1.0915,
+      "step": 2297
+    },
+    {
+      "epoch": 0.31626754748142033,
+      "grad_norm": 1.6904754200515562,
+      "learning_rate": 1.6001207624032006e-05,
+      "loss": 1.0453,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3164051747866777,
+      "grad_norm": 1.9891265184616433,
+      "learning_rate": 1.5997641492303727e-05,
+      "loss": 1.1462,
+      "step": 2299
+    },
+    {
+      "epoch": 0.316542802091935,
+      "grad_norm": 1.772704527602838,
+      "learning_rate": 1.599407416892618e-05,
+      "loss": 1.1113,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3166804293971924,
+      "grad_norm": 1.7766902012776271,
+      "learning_rate": 1.599050565460815e-05,
+      "loss": 1.0587,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3168180567024498,
+      "grad_norm": 1.933262242545844,
+      "learning_rate": 1.5986935950058642e-05,
+      "loss": 1.1241,
+      "step": 2302
+    },
+    {
+      "epoch": 0.31695568400770713,
+      "grad_norm": 1.800019927433424,
+      "learning_rate": 1.5983365055986917e-05,
+      "loss": 1.0814,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3170933113129645,
+      "grad_norm": 1.5983183734035724,
+      "learning_rate": 1.597979297310246e-05,
+      "loss": 1.0935,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3172309386182219,
+      "grad_norm": 1.9125475956060354,
+      "learning_rate": 1.597621970211499e-05,
+      "loss": 1.0988,
+      "step": 2305
+    },
+    {
+      "epoch": 0.31736856592347923,
+      "grad_norm": 1.9020757124520562,
+      "learning_rate": 1.5972645243734473e-05,
+      "loss": 1.1055,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3175061932287366,
+      "grad_norm": 1.7643985733187741,
+      "learning_rate": 1.59690695986711e-05,
+      "loss": 1.1126,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3176438205339939,
+      "grad_norm": 1.5091678716845236,
+      "learning_rate": 1.5965492767635307e-05,
+      "loss": 1.0118,
+      "step": 2308
+    },
+    {
+      "epoch": 0.31778144783925133,
+      "grad_norm": 1.860448777853755,
+      "learning_rate": 1.5961914751337762e-05,
+      "loss": 0.9792,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3179190751445087,
+      "grad_norm": 1.7426759550871298,
+      "learning_rate": 1.5958335550489367e-05,
+      "loss": 1.0956,
+      "step": 2310
+    },
+    {
+      "epoch": 0.318056702449766,
+      "grad_norm": 1.6914039313364324,
+      "learning_rate": 1.5954755165801253e-05,
+      "loss": 1.0355,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3181943297550234,
+      "grad_norm": 1.7801941735940796,
+      "learning_rate": 1.59511735979848e-05,
+      "loss": 1.1175,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3183319570602808,
+      "grad_norm": 1.6082071287109962,
+      "learning_rate": 1.594759084775162e-05,
+      "loss": 1.1161,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3184695843655381,
+      "grad_norm": 1.9456554729076283,
+      "learning_rate": 1.594400691581355e-05,
+      "loss": 1.1416,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3186072116707955,
+      "grad_norm": 1.7272363550957333,
+      "learning_rate": 1.594042180288267e-05,
+      "loss": 1.1477,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3187448389760528,
+      "grad_norm": 3.5664810739271253,
+      "learning_rate": 1.5936835509671294e-05,
+      "loss": 1.0083,
+      "step": 2316
+    },
+    {
+      "epoch": 0.31888246628131023,
+      "grad_norm": 1.8179687995631697,
+      "learning_rate": 1.5933248036891968e-05,
+      "loss": 1.1404,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3190200935865676,
+      "grad_norm": 1.7027646526491853,
+      "learning_rate": 1.5929659385257475e-05,
+      "loss": 1.08,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3191577208918249,
+      "grad_norm": 1.6086468336865665,
+      "learning_rate": 1.5926069555480827e-05,
+      "loss": 1.1195,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3192953481970823,
+      "grad_norm": 1.5863347472287999,
+      "learning_rate": 1.592247854827528e-05,
+      "loss": 1.0912,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3194329755023397,
+      "grad_norm": 1.7912733824826224,
+      "learning_rate": 1.5918886364354315e-05,
+      "loss": 1.1714,
+      "step": 2321
+    },
+    {
+      "epoch": 0.319570602807597,
+      "grad_norm": 1.8109376049888988,
+      "learning_rate": 1.591529300443165e-05,
+      "loss": 1.1253,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3197082301128544,
+      "grad_norm": 2.10375077606398,
+      "learning_rate": 1.591169846922124e-05,
+      "loss": 1.1378,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3198458574181118,
+      "grad_norm": 1.6289270735227497,
+      "learning_rate": 1.5908102759437265e-05,
+      "loss": 1.0542,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3199834847233691,
+      "grad_norm": 2.13222515842592,
+      "learning_rate": 1.5904505875794144e-05,
+      "loss": 1.1582,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3201211120286265,
+      "grad_norm": 1.6649241454636934,
+      "learning_rate": 1.590090781900653e-05,
+      "loss": 0.9699,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3202587393338838,
+      "grad_norm": 1.9519304229349628,
+      "learning_rate": 1.5897308589789305e-05,
+      "loss": 1.1089,
+      "step": 2327
+    },
+    {
+      "epoch": 0.32039636663914123,
+      "grad_norm": 1.6650331588854406,
+      "learning_rate": 1.589370818885759e-05,
+      "loss": 1.0378,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3205339939443986,
+      "grad_norm": 1.9348901998083594,
+      "learning_rate": 1.5890106616926732e-05,
+      "loss": 1.0754,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3206716212496559,
+      "grad_norm": 1.5012382844947767,
+      "learning_rate": 1.5886503874712316e-05,
+      "loss": 0.995,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3208092485549133,
+      "grad_norm": 1.661694356030018,
+      "learning_rate": 1.5882899962930162e-05,
+      "loss": 0.9706,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3209468758601707,
+      "grad_norm": 1.6725339026758324,
+      "learning_rate": 1.587929488229631e-05,
+      "loss": 1.0437,
+      "step": 2332
+    },
+    {
+      "epoch": 0.321084503165428,
+      "grad_norm": 1.754075685346653,
+      "learning_rate": 1.5875688633527043e-05,
+      "loss": 1.143,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3212221304706854,
+      "grad_norm": 1.9898848614835465,
+      "learning_rate": 1.5872081217338874e-05,
+      "loss": 1.1093,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3213597577759427,
+      "grad_norm": 1.5104178094667415,
+      "learning_rate": 1.586847263444855e-05,
+      "loss": 1.0481,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3214973850812001,
+      "grad_norm": 2.018293111903957,
+      "learning_rate": 1.5864862885573042e-05,
+      "loss": 1.1189,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3216350123864575,
+      "grad_norm": 1.7661490130223527,
+      "learning_rate": 1.5861251971429557e-05,
+      "loss": 1.1409,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3217726396917148,
+      "grad_norm": 1.7062105848401936,
+      "learning_rate": 1.5857639892735544e-05,
+      "loss": 1.0966,
+      "step": 2338
+    },
+    {
+      "epoch": 0.32191026699697217,
+      "grad_norm": 1.76628155603058,
+      "learning_rate": 1.5854026650208662e-05,
+      "loss": 1.0969,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3220478943022296,
+      "grad_norm": 1.6356783012090528,
+      "learning_rate": 1.5850412244566817e-05,
+      "loss": 1.077,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3221855216074869,
+      "grad_norm": 1.6955593658789085,
+      "learning_rate": 1.5846796676528147e-05,
+      "loss": 1.0417,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3223231489127443,
+      "grad_norm": 1.8493747638151574,
+      "learning_rate": 1.5843179946811006e-05,
+      "loss": 1.1204,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3224607762180017,
+      "grad_norm": 1.8759195177691446,
+      "learning_rate": 1.5839562056133998e-05,
+      "loss": 1.0453,
+      "step": 2343
+    },
+    {
+      "epoch": 0.322598403523259,
+      "grad_norm": 1.7085931154027267,
+      "learning_rate": 1.5835943005215944e-05,
+      "loss": 0.993,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3227360308285164,
+      "grad_norm": 1.6818874493546516,
+      "learning_rate": 1.5832322794775898e-05,
+      "loss": 1.0608,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3228736581337737,
+      "grad_norm": 1.9166687559531714,
+      "learning_rate": 1.582870142553315e-05,
+      "loss": 1.1488,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3230112854390311,
+      "grad_norm": 1.566279373070242,
+      "learning_rate": 1.5825078898207214e-05,
+      "loss": 1.0584,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3231489127442885,
+      "grad_norm": 1.7198993460098657,
+      "learning_rate": 1.5821455213517837e-05,
+      "loss": 1.0155,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3232865400495458,
+      "grad_norm": 1.8606493326206026,
+      "learning_rate": 1.5817830372184998e-05,
+      "loss": 1.1371,
+      "step": 2349
+    },
+    {
+      "epoch": 0.32342416735480317,
+      "grad_norm": 1.7388856212878667,
+      "learning_rate": 1.5814204374928898e-05,
+      "loss": 1.0515,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3235617946600606,
+      "grad_norm": 1.6649017289299055,
+      "learning_rate": 1.581057722246998e-05,
+      "loss": 1.0911,
+      "step": 2351
+    },
+    {
+      "epoch": 0.3236994219653179,
+      "grad_norm": 1.7682101828216725,
+      "learning_rate": 1.5806948915528904e-05,
+      "loss": 1.1168,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3238370492705753,
+      "grad_norm": 1.7690607414438981,
+      "learning_rate": 1.580331945482657e-05,
+      "loss": 1.0604,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3239746765758326,
+      "grad_norm": 1.7270337931770436,
+      "learning_rate": 1.5799688841084097e-05,
+      "loss": 1.096,
+      "step": 2354
+    },
+    {
+      "epoch": 0.32411230388109,
+      "grad_norm": 1.7735729383154624,
+      "learning_rate": 1.5796057075022838e-05,
+      "loss": 1.1673,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3242499311863474,
+      "grad_norm": 2.04313966196195,
+      "learning_rate": 1.579242415736438e-05,
+      "loss": 1.0997,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3243875584916047,
+      "grad_norm": 1.7400971187375815,
+      "learning_rate": 1.5788790088830534e-05,
+      "loss": 1.0251,
+      "step": 2357
+    },
+    {
+      "epoch": 0.32452518579686207,
+      "grad_norm": 1.6832618591094404,
+      "learning_rate": 1.5785154870143333e-05,
+      "loss": 1.0825,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3246628131021195,
+      "grad_norm": 1.8769334318375996,
+      "learning_rate": 1.5781518502025054e-05,
+      "loss": 1.1911,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3248004404073768,
+      "grad_norm": 1.7201128259880378,
+      "learning_rate": 1.5777880985198185e-05,
+      "loss": 1.0818,
+      "step": 2360
+    },
+    {
+      "epoch": 0.32493806771263417,
+      "grad_norm": 1.927904506595857,
+      "learning_rate": 1.5774242320385456e-05,
+      "loss": 1.0753,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3250756950178916,
+      "grad_norm": 1.6518787098038445,
+      "learning_rate": 1.5770602508309818e-05,
+      "loss": 1.0347,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3252133223231489,
+      "grad_norm": 1.7254303827610937,
+      "learning_rate": 1.5766961549694452e-05,
+      "loss": 1.09,
+      "step": 2363
+    },
+    {
+      "epoch": 0.32535094962840627,
+      "grad_norm": 1.8468023746410842,
+      "learning_rate": 1.576331944526277e-05,
+      "loss": 1.1002,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3254885769336636,
+      "grad_norm": 1.7760414595909948,
+      "learning_rate": 1.5759676195738403e-05,
+      "loss": 1.1114,
+      "step": 2365
+    },
+    {
+      "epoch": 0.325626204238921,
+      "grad_norm": 1.70808927899885,
+      "learning_rate": 1.5756031801845218e-05,
+      "loss": 1.1066,
+      "step": 2366
+    },
+    {
+      "epoch": 0.3257638315441784,
+      "grad_norm": 1.7937495633770215,
+      "learning_rate": 1.5752386264307307e-05,
+      "loss": 1.05,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3259014588494357,
+      "grad_norm": 1.7795041548322201,
+      "learning_rate": 1.5748739583848983e-05,
+      "loss": 1.0677,
+      "step": 2368
+    },
+    {
+      "epoch": 0.32603908615469307,
+      "grad_norm": 1.7302835811021178,
+      "learning_rate": 1.57450917611948e-05,
+      "loss": 1.0605,
+      "step": 2369
+    },
+    {
+      "epoch": 0.3261767134599505,
+      "grad_norm": 1.6602865270467608,
+      "learning_rate": 1.5741442797069523e-05,
+      "loss": 1.0686,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3263143407652078,
+      "grad_norm": 1.5257812599981984,
+      "learning_rate": 1.5737792692198155e-05,
+      "loss": 1.0961,
+      "step": 2371
+    },
+    {
+      "epoch": 0.32645196807046517,
+      "grad_norm": 1.6016563903527217,
+      "learning_rate": 1.573414144730592e-05,
+      "loss": 1.1306,
+      "step": 2372
+    },
+    {
+      "epoch": 0.3265895953757225,
+      "grad_norm": 1.7141437081719049,
+      "learning_rate": 1.5730489063118274e-05,
+      "loss": 1.0809,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3267272226809799,
+      "grad_norm": 1.7166592463678925,
+      "learning_rate": 1.572683554036089e-05,
+      "loss": 1.0676,
+      "step": 2374
+    },
+    {
+      "epoch": 0.32686484998623727,
+      "grad_norm": 2.004562180474064,
+      "learning_rate": 1.5723180879759675e-05,
+      "loss": 1.0014,
+      "step": 2375
+    },
+    {
+      "epoch": 0.3270024772914946,
+      "grad_norm": 1.6586025150990276,
+      "learning_rate": 1.5719525082040764e-05,
+      "loss": 1.1033,
+      "step": 2376
+    },
+    {
+      "epoch": 0.32714010459675197,
+      "grad_norm": 1.7811987714300934,
+      "learning_rate": 1.571586814793051e-05,
+      "loss": 1.1586,
+      "step": 2377
+    },
+    {
+      "epoch": 0.3272777319020094,
+      "grad_norm": 1.7047001699287403,
+      "learning_rate": 1.5712210078155495e-05,
+      "loss": 1.0294,
+      "step": 2378
+    },
+    {
+      "epoch": 0.3274153592072667,
+      "grad_norm": 1.8001029425852324,
+      "learning_rate": 1.570855087344253e-05,
+      "loss": 1.0833,
+      "step": 2379
+    },
+    {
+      "epoch": 0.32755298651252407,
+      "grad_norm": 1.6810697964791235,
+      "learning_rate": 1.570489053451865e-05,
+      "loss": 1.025,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3276906138177815,
+      "grad_norm": 1.541819086969105,
+      "learning_rate": 1.570122906211111e-05,
+      "loss": 1.0122,
+      "step": 2381
+    },
+    {
+      "epoch": 0.3278282411230388,
+      "grad_norm": 1.7618509939448488,
+      "learning_rate": 1.56975664569474e-05,
+      "loss": 1.0294,
+      "step": 2382
+    },
+    {
+      "epoch": 0.32796586842829617,
+      "grad_norm": 1.8598685173753144,
+      "learning_rate": 1.569390271975522e-05,
+      "loss": 1.0559,
+      "step": 2383
+    },
+    {
+      "epoch": 0.3281034957335535,
+      "grad_norm": 1.7747954467334333,
+      "learning_rate": 1.5690237851262514e-05,
+      "loss": 1.011,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3282411230388109,
+      "grad_norm": 1.6487990070930296,
+      "learning_rate": 1.568657185219744e-05,
+      "loss": 1.1034,
+      "step": 2385
+    },
+    {
+      "epoch": 0.32837875034406827,
+      "grad_norm": 2.0826714417480723,
+      "learning_rate": 1.5682904723288375e-05,
+      "loss": 1.1685,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3285163776493256,
+      "grad_norm": 1.468675794508899,
+      "learning_rate": 1.5679236465263938e-05,
+      "loss": 1.0781,
+      "step": 2387
+    },
+    {
+      "epoch": 0.32865400495458297,
+      "grad_norm": 1.832051478677324,
+      "learning_rate": 1.5675567078852953e-05,
+      "loss": 1.0866,
+      "step": 2388
+    },
+    {
+      "epoch": 0.32879163225984037,
+      "grad_norm": 1.5904825662952686,
+      "learning_rate": 1.567189656478448e-05,
+      "loss": 1.0368,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3289292595650977,
+      "grad_norm": 1.788046188361262,
+      "learning_rate": 1.56682249237878e-05,
+      "loss": 1.1599,
+      "step": 2390
+    },
+    {
+      "epoch": 0.32906688687035507,
+      "grad_norm": 1.752651508661983,
+      "learning_rate": 1.5664552156592412e-05,
+      "loss": 1.1091,
+      "step": 2391
+    },
+    {
+      "epoch": 0.3292045141756124,
+      "grad_norm": 2.050158250314157,
+      "learning_rate": 1.5660878263928054e-05,
+      "loss": 1.2802,
+      "step": 2392
+    },
+    {
+      "epoch": 0.3293421414808698,
+      "grad_norm": 1.866939139595996,
+      "learning_rate": 1.565720324652467e-05,
+      "loss": 1.1539,
+      "step": 2393
+    },
+    {
+      "epoch": 0.32947976878612717,
+      "grad_norm": 2.016290378236832,
+      "learning_rate": 1.5653527105112444e-05,
+      "loss": 1.0084,
+      "step": 2394
+    },
+    {
+      "epoch": 0.3296173960913845,
+      "grad_norm": 1.6525591375001598,
+      "learning_rate": 1.564984984042177e-05,
+      "loss": 1.095,
+      "step": 2395
+    },
+    {
+      "epoch": 0.32975502339664187,
+      "grad_norm": 1.7085894056865552,
+      "learning_rate": 1.564617145318327e-05,
+      "loss": 1.0396,
+      "step": 2396
+    },
+    {
+      "epoch": 0.32989265070189927,
+      "grad_norm": 1.8732548474485025,
+      "learning_rate": 1.5642491944127792e-05,
+      "loss": 1.1343,
+      "step": 2397
+    },
+    {
+      "epoch": 0.3300302780071566,
+      "grad_norm": 1.6349048433226607,
+      "learning_rate": 1.56388113139864e-05,
+      "loss": 1.0825,
+      "step": 2398
+    },
+    {
+      "epoch": 0.33016790531241397,
+      "grad_norm": 1.6646857406750188,
+      "learning_rate": 1.5635129563490384e-05,
+      "loss": 0.9959,
+      "step": 2399
+    },
+    {
+      "epoch": 0.33030553261767137,
+      "grad_norm": 1.6914573105875277,
+      "learning_rate": 1.563144669337126e-05,
+      "loss": 1.0849,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3304431599229287,
+      "grad_norm": 1.629293871225552,
+      "learning_rate": 1.5627762704360773e-05,
+      "loss": 1.0536,
+      "step": 2401
+    },
+    {
+      "epoch": 0.33058078722818607,
+      "grad_norm": 1.5937219087124495,
+      "learning_rate": 1.5624077597190864e-05,
+      "loss": 0.9655,
+      "step": 2402
+    },
+    {
+      "epoch": 0.3307184145334434,
+      "grad_norm": 1.7139382459330028,
+      "learning_rate": 1.5620391372593723e-05,
+      "loss": 1.0231,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3308560418387008,
+      "grad_norm": 1.5747178348505184,
+      "learning_rate": 1.5616704031301754e-05,
+      "loss": 1.1209,
+      "step": 2404
+    },
+    {
+      "epoch": 0.33099366914395817,
+      "grad_norm": 1.7413968191712683,
+      "learning_rate": 1.561301557404758e-05,
+      "loss": 1.0927,
+      "step": 2405
+    },
+    {
+      "epoch": 0.3311312964492155,
+      "grad_norm": 1.5321825619302887,
+      "learning_rate": 1.560932600156404e-05,
+      "loss": 0.9832,
+      "step": 2406
+    },
+    {
+      "epoch": 0.33126892375447287,
+      "grad_norm": 1.5011250820933424,
+      "learning_rate": 1.5605635314584206e-05,
+      "loss": 0.9461,
+      "step": 2407
+    },
+    {
+      "epoch": 0.33140655105973027,
+      "grad_norm": 1.6193644567964949,
+      "learning_rate": 1.560194351384137e-05,
+      "loss": 1.0257,
+      "step": 2408
+    },
+    {
+      "epoch": 0.3315441783649876,
+      "grad_norm": 1.9510663808578694,
+      "learning_rate": 1.559825060006904e-05,
+      "loss": 1.1339,
+      "step": 2409
+    },
+    {
+      "epoch": 0.33168180567024497,
+      "grad_norm": 1.6187957467910705,
+      "learning_rate": 1.559455657400095e-05,
+      "loss": 1.0121,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3318194329755023,
+      "grad_norm": 1.7384094839684594,
+      "learning_rate": 1.559086143637105e-05,
+      "loss": 1.0791,
+      "step": 2411
+    },
+    {
+      "epoch": 0.3319570602807597,
+      "grad_norm": 1.6977530496280118,
+      "learning_rate": 1.5587165187913507e-05,
+      "loss": 1.0451,
+      "step": 2412
+    },
+    {
+      "epoch": 0.33209468758601707,
+      "grad_norm": 1.6574988208816277,
+      "learning_rate": 1.5583467829362726e-05,
+      "loss": 1.0969,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3322323148912744,
+      "grad_norm": 1.729339781897529,
+      "learning_rate": 1.5579769361453317e-05,
+      "loss": 1.0843,
+      "step": 2414
+    },
+    {
+      "epoch": 0.33236994219653176,
+      "grad_norm": 1.6367312443181354,
+      "learning_rate": 1.557606978492011e-05,
+      "loss": 1.0804,
+      "step": 2415
+    },
+    {
+      "epoch": 0.33250756950178917,
+      "grad_norm": 1.9953687940267166,
+      "learning_rate": 1.5572369100498168e-05,
+      "loss": 1.1108,
+      "step": 2416
+    },
+    {
+      "epoch": 0.3326451968070465,
+      "grad_norm": 1.6533863863126634,
+      "learning_rate": 1.5568667308922764e-05,
+      "loss": 0.9502,
+      "step": 2417
+    },
+    {
+      "epoch": 0.33278282411230387,
+      "grad_norm": 1.862532215211359,
+      "learning_rate": 1.5564964410929394e-05,
+      "loss": 0.9961,
+      "step": 2418
+    },
+    {
+      "epoch": 0.33292045141756127,
+      "grad_norm": 1.6158055581339776,
+      "learning_rate": 1.5561260407253766e-05,
+      "loss": 1.1259,
+      "step": 2419
+    },
+    {
+      "epoch": 0.3330580787228186,
+      "grad_norm": 1.7673305943936002,
+      "learning_rate": 1.5557555298631825e-05,
+      "loss": 1.0545,
+      "step": 2420
+    },
+    {
+      "epoch": 0.33319570602807597,
+      "grad_norm": 1.7774490993114553,
+      "learning_rate": 1.555384908579972e-05,
+      "loss": 1.0466,
+      "step": 2421
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.7167457695563906,
+      "learning_rate": 1.5550141769493824e-05,
+      "loss": 1.0655,
+      "step": 2422
+    },
+    {
+      "epoch": 0.3334709606385907,
+      "grad_norm": 1.713492288604055,
+      "learning_rate": 1.554643335045073e-05,
+      "loss": 1.1201,
+      "step": 2423
+    },
+    {
+      "epoch": 0.33360858794384807,
+      "grad_norm": 1.687585260046263,
+      "learning_rate": 1.5542723829407255e-05,
+      "loss": 1.0734,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3337462152491054,
+      "grad_norm": 1.5930877211628152,
+      "learning_rate": 1.553901320710043e-05,
+      "loss": 1.0656,
+      "step": 2425
+    },
+    {
+      "epoch": 0.33388384255436276,
+      "grad_norm": 1.5543665765904255,
+      "learning_rate": 1.5535301484267496e-05,
+      "loss": 1.0608,
+      "step": 2426
+    },
+    {
+      "epoch": 0.33402146985962017,
+      "grad_norm": 1.7395507619351236,
+      "learning_rate": 1.553158866164593e-05,
+      "loss": 1.0794,
+      "step": 2427
+    },
+    {
+      "epoch": 0.3341590971648775,
+      "grad_norm": 1.9131707660834933,
+      "learning_rate": 1.5527874739973414e-05,
+      "loss": 1.1832,
+      "step": 2428
+    },
+    {
+      "epoch": 0.33429672447013487,
+      "grad_norm": 1.8638755076397202,
+      "learning_rate": 1.5524159719987855e-05,
+      "loss": 1.1638,
+      "step": 2429
+    },
+    {
+      "epoch": 0.3344343517753922,
+      "grad_norm": 1.8195774500045223,
+      "learning_rate": 1.5520443602427378e-05,
+      "loss": 1.0615,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3345719790806496,
+      "grad_norm": 1.6895453351369925,
+      "learning_rate": 1.5516726388030324e-05,
+      "loss": 1.2294,
+      "step": 2431
+    },
+    {
+      "epoch": 0.33470960638590697,
+      "grad_norm": 1.6691357329445753,
+      "learning_rate": 1.5513008077535255e-05,
+      "loss": 1.0025,
+      "step": 2432
+    },
+    {
+      "epoch": 0.3348472336911643,
+      "grad_norm": 1.6786652353004003,
+      "learning_rate": 1.5509288671680943e-05,
+      "loss": 1.1524,
+      "step": 2433
+    },
+    {
+      "epoch": 0.33498486099642166,
+      "grad_norm": 1.8928764926416464,
+      "learning_rate": 1.5505568171206387e-05,
+      "loss": 1.1015,
+      "step": 2434
+    },
+    {
+      "epoch": 0.33512248830167907,
+      "grad_norm": 1.645657502854033,
+      "learning_rate": 1.5501846576850796e-05,
+      "loss": 0.9984,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3352601156069364,
+      "grad_norm": 1.6090894767446973,
+      "learning_rate": 1.5498123889353603e-05,
+      "loss": 1.0492,
+      "step": 2436
+    },
+    {
+      "epoch": 0.33539774291219376,
+      "grad_norm": 1.7309350338962743,
+      "learning_rate": 1.5494400109454457e-05,
+      "loss": 1.0897,
+      "step": 2437
+    },
+    {
+      "epoch": 0.33553537021745117,
+      "grad_norm": 1.6156997037707437,
+      "learning_rate": 1.5490675237893217e-05,
+      "loss": 1.0903,
+      "step": 2438
+    },
+    {
+      "epoch": 0.3356729975227085,
+      "grad_norm": 1.577663058698449,
+      "learning_rate": 1.5486949275409967e-05,
+      "loss": 1.094,
+      "step": 2439
+    },
+    {
+      "epoch": 0.33581062482796586,
+      "grad_norm": 1.6178247978478184,
+      "learning_rate": 1.5483222222745004e-05,
+      "loss": 1.1846,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3359482521332232,
+      "grad_norm": 1.734992083328551,
+      "learning_rate": 1.5479494080638844e-05,
+      "loss": 1.0756,
+      "step": 2441
+    },
+    {
+      "epoch": 0.3360858794384806,
+      "grad_norm": 1.6004037925793275,
+      "learning_rate": 1.5475764849832215e-05,
+      "loss": 1.1094,
+      "step": 2442
+    },
+    {
+      "epoch": 0.33622350674373797,
+      "grad_norm": 1.5854947704566433,
+      "learning_rate": 1.5472034531066066e-05,
+      "loss": 1.084,
+      "step": 2443
+    },
+    {
+      "epoch": 0.3363611340489953,
+      "grad_norm": 1.6229880296735704,
+      "learning_rate": 1.546830312508156e-05,
+      "loss": 1.0798,
+      "step": 2444
+    },
+    {
+      "epoch": 0.33649876135425266,
+      "grad_norm": 1.7011715757128885,
+      "learning_rate": 1.5464570632620074e-05,
+      "loss": 1.1194,
+      "step": 2445
+    },
+    {
+      "epoch": 0.33663638865951007,
+      "grad_norm": 1.7828829837357527,
+      "learning_rate": 1.5460837054423206e-05,
+      "loss": 1.1322,
+      "step": 2446
+    },
+    {
+      "epoch": 0.3367740159647674,
+      "grad_norm": 1.7688893173722586,
+      "learning_rate": 1.5457102391232765e-05,
+      "loss": 1.0994,
+      "step": 2447
+    },
+    {
+      "epoch": 0.33691164327002476,
+      "grad_norm": 1.921710652768241,
+      "learning_rate": 1.545336664379078e-05,
+      "loss": 1.0077,
+      "step": 2448
+    },
+    {
+      "epoch": 0.3370492705752821,
+      "grad_norm": 1.6195241372579625,
+      "learning_rate": 1.5449629812839492e-05,
+      "loss": 1.0665,
+      "step": 2449
+    },
+    {
+      "epoch": 0.3371868978805395,
+      "grad_norm": 1.8714779759415912,
+      "learning_rate": 1.5445891899121356e-05,
+      "loss": 1.116,
+      "step": 2450
+    },
+    {
+      "epoch": 0.33732452518579686,
+      "grad_norm": 2.114463897136919,
+      "learning_rate": 1.5442152903379044e-05,
+      "loss": 1.1124,
+      "step": 2451
+    },
+    {
+      "epoch": 0.3374621524910542,
+      "grad_norm": 1.8057340321699364,
+      "learning_rate": 1.5438412826355447e-05,
+      "loss": 0.9909,
+      "step": 2452
+    },
+    {
+      "epoch": 0.33759977979631156,
+      "grad_norm": 1.630156294633219,
+      "learning_rate": 1.5434671668793668e-05,
+      "loss": 1.1287,
+      "step": 2453
+    },
+    {
+      "epoch": 0.33773740710156896,
+      "grad_norm": 1.8606501298007794,
+      "learning_rate": 1.5430929431437022e-05,
+      "loss": 1.0822,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3378750344068263,
+      "grad_norm": 1.9247552770118723,
+      "learning_rate": 1.5427186115029035e-05,
+      "loss": 1.0672,
+      "step": 2455
+    },
+    {
+      "epoch": 0.33801266171208366,
+      "grad_norm": 1.619581788791508,
+      "learning_rate": 1.5423441720313458e-05,
+      "loss": 1.1313,
+      "step": 2456
+    },
+    {
+      "epoch": 0.33815028901734107,
+      "grad_norm": 1.8311744317021126,
+      "learning_rate": 1.5419696248034245e-05,
+      "loss": 1.147,
+      "step": 2457
+    },
+    {
+      "epoch": 0.3382879163225984,
+      "grad_norm": 1.7716751128694765,
+      "learning_rate": 1.5415949698935582e-05,
+      "loss": 1.1304,
+      "step": 2458
+    },
+    {
+      "epoch": 0.33842554362785576,
+      "grad_norm": 1.7367918135044305,
+      "learning_rate": 1.541220207376185e-05,
+      "loss": 1.1842,
+      "step": 2459
+    },
+    {
+      "epoch": 0.3385631709331131,
+      "grad_norm": 1.552135337234452,
+      "learning_rate": 1.5408453373257646e-05,
+      "loss": 1.1422,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3387007982383705,
+      "grad_norm": 1.695730937226073,
+      "learning_rate": 1.5404703598167793e-05,
+      "loss": 1.1555,
+      "step": 2461
+    },
+    {
+      "epoch": 0.33883842554362786,
+      "grad_norm": 1.7549605747132613,
+      "learning_rate": 1.5400952749237317e-05,
+      "loss": 1.056,
+      "step": 2462
+    },
+    {
+      "epoch": 0.3389760528488852,
+      "grad_norm": 1.6848924726584689,
+      "learning_rate": 1.5397200827211464e-05,
+      "loss": 1.0401,
+      "step": 2463
+    },
+    {
+      "epoch": 0.33911368015414256,
+      "grad_norm": 1.5820613804703938,
+      "learning_rate": 1.539344783283568e-05,
+      "loss": 1.0543,
+      "step": 2464
+    },
+    {
+      "epoch": 0.33925130745939996,
+      "grad_norm": 1.657923352467557,
+      "learning_rate": 1.5389693766855645e-05,
+      "loss": 1.1277,
+      "step": 2465
+    },
+    {
+      "epoch": 0.3393889347646573,
+      "grad_norm": 1.8526755296739283,
+      "learning_rate": 1.5385938630017233e-05,
+      "loss": 1.1207,
+      "step": 2466
+    },
+    {
+      "epoch": 0.33952656206991466,
+      "grad_norm": 1.9587946569834347,
+      "learning_rate": 1.5382182423066544e-05,
+      "loss": 1.0795,
+      "step": 2467
+    },
+    {
+      "epoch": 0.339664189375172,
+      "grad_norm": 1.7563768495694851,
+      "learning_rate": 1.5378425146749878e-05,
+      "loss": 1.1257,
+      "step": 2468
+    },
+    {
+      "epoch": 0.3398018166804294,
+      "grad_norm": 1.5979272248706566,
+      "learning_rate": 1.537466680181376e-05,
+      "loss": 1.0399,
+      "step": 2469
+    },
+    {
+      "epoch": 0.33993944398568676,
+      "grad_norm": 1.725949860239849,
+      "learning_rate": 1.5370907389004917e-05,
+      "loss": 1.066,
+      "step": 2470
+    },
+    {
+      "epoch": 0.3400770712909441,
+      "grad_norm": 1.7333051148939074,
+      "learning_rate": 1.5367146909070297e-05,
+      "loss": 1.0831,
+      "step": 2471
+    },
+    {
+      "epoch": 0.34021469859620146,
+      "grad_norm": 1.650125366692987,
+      "learning_rate": 1.5363385362757054e-05,
+      "loss": 1.0597,
+      "step": 2472
+    },
+    {
+      "epoch": 0.34035232590145886,
+      "grad_norm": 1.8601542136550744,
+      "learning_rate": 1.535962275081256e-05,
+      "loss": 1.1184,
+      "step": 2473
+    },
+    {
+      "epoch": 0.3404899532067162,
+      "grad_norm": 1.513563528141664,
+      "learning_rate": 1.535585907398439e-05,
+      "loss": 1.0694,
+      "step": 2474
+    },
+    {
+      "epoch": 0.34062758051197356,
+      "grad_norm": 1.935330741873839,
+      "learning_rate": 1.535209433302033e-05,
+      "loss": 1.1241,
+      "step": 2475
+    },
+    {
+      "epoch": 0.34076520781723096,
+      "grad_norm": 1.5305038879136965,
+      "learning_rate": 1.534832852866839e-05,
+      "loss": 1.0233,
+      "step": 2476
+    },
+    {
+      "epoch": 0.3409028351224883,
+      "grad_norm": 1.8473654832600042,
+      "learning_rate": 1.5344561661676784e-05,
+      "loss": 1.1906,
+      "step": 2477
+    },
+    {
+      "epoch": 0.34104046242774566,
+      "grad_norm": 1.7355399287021502,
+      "learning_rate": 1.5340793732793934e-05,
+      "loss": 1.0456,
+      "step": 2478
+    },
+    {
+      "epoch": 0.341178089733003,
+      "grad_norm": 1.6626922000839517,
+      "learning_rate": 1.5337024742768475e-05,
+      "loss": 1.091,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3413157170382604,
+      "grad_norm": 2.249891605964178,
+      "learning_rate": 1.533325469234926e-05,
+      "loss": 1.1074,
+      "step": 2480
+    },
+    {
+      "epoch": 0.34145334434351776,
+      "grad_norm": 1.7521846163621029,
+      "learning_rate": 1.532948358228534e-05,
+      "loss": 1.0633,
+      "step": 2481
+    },
+    {
+      "epoch": 0.3415909716487751,
+      "grad_norm": 1.7730734920016111,
+      "learning_rate": 1.5325711413325984e-05,
+      "loss": 1.1702,
+      "step": 2482
+    },
+    {
+      "epoch": 0.34172859895403246,
+      "grad_norm": 1.6453841226552313,
+      "learning_rate": 1.532193818622067e-05,
+      "loss": 1.1156,
+      "step": 2483
+    },
+    {
+      "epoch": 0.34186622625928986,
+      "grad_norm": 2.106371047174693,
+      "learning_rate": 1.531816390171909e-05,
+      "loss": 1.1078,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3420038535645472,
+      "grad_norm": 1.827561678162473,
+      "learning_rate": 1.5314388560571142e-05,
+      "loss": 1.1435,
+      "step": 2485
+    },
+    {
+      "epoch": 0.34214148086980456,
+      "grad_norm": 1.5199705189299015,
+      "learning_rate": 1.531061216352693e-05,
+      "loss": 1.0091,
+      "step": 2486
+    },
+    {
+      "epoch": 0.3422791081750619,
+      "grad_norm": 1.6368250744534398,
+      "learning_rate": 1.530683471133678e-05,
+      "loss": 1.1342,
+      "step": 2487
+    },
+    {
+      "epoch": 0.3424167354803193,
+      "grad_norm": 1.7421683346333066,
+      "learning_rate": 1.530305620475121e-05,
+      "loss": 1.0608,
+      "step": 2488
+    },
+    {
+      "epoch": 0.34255436278557666,
+      "grad_norm": 1.6906875509047727,
+      "learning_rate": 1.5299276644520965e-05,
+      "loss": 1.1095,
+      "step": 2489
+    },
+    {
+      "epoch": 0.342691990090834,
+      "grad_norm": 1.7278192034774242,
+      "learning_rate": 1.529549603139699e-05,
+      "loss": 1.0095,
+      "step": 2490
+    },
+    {
+      "epoch": 0.34282961739609136,
+      "grad_norm": 1.704972681801542,
+      "learning_rate": 1.5291714366130438e-05,
+      "loss": 1.0859,
+      "step": 2491
+    },
+    {
+      "epoch": 0.34296724470134876,
+      "grad_norm": 1.626092616987487,
+      "learning_rate": 1.528793164947268e-05,
+      "loss": 1.1043,
+      "step": 2492
+    },
+    {
+      "epoch": 0.3431048720066061,
+      "grad_norm": 1.6347340492101472,
+      "learning_rate": 1.5284147882175287e-05,
+      "loss": 1.0104,
+      "step": 2493
+    },
+    {
+      "epoch": 0.34324249931186346,
+      "grad_norm": 1.672062990253176,
+      "learning_rate": 1.5280363064990042e-05,
+      "loss": 1.0345,
+      "step": 2494
+    },
+    {
+      "epoch": 0.34338012661712086,
+      "grad_norm": 1.5211368415279691,
+      "learning_rate": 1.5276577198668938e-05,
+      "loss": 1.0606,
+      "step": 2495
+    },
+    {
+      "epoch": 0.3435177539223782,
+      "grad_norm": 1.8170614012947848,
+      "learning_rate": 1.527279028396417e-05,
+      "loss": 1.1359,
+      "step": 2496
+    },
+    {
+      "epoch": 0.34365538122763556,
+      "grad_norm": 1.6213105199832738,
+      "learning_rate": 1.5269002321628152e-05,
+      "loss": 1.0982,
+      "step": 2497
+    },
+    {
+      "epoch": 0.3437930085328929,
+      "grad_norm": 1.8097405865032545,
+      "learning_rate": 1.5265213312413496e-05,
+      "loss": 1.1621,
+      "step": 2498
+    },
+    {
+      "epoch": 0.3439306358381503,
+      "grad_norm": 1.6753605631925006,
+      "learning_rate": 1.5261423257073028e-05,
+      "loss": 0.9842,
+      "step": 2499
+    },
+    {
+      "epoch": 0.34406826314340766,
+      "grad_norm": 1.5707995998681727,
+      "learning_rate": 1.5257632156359779e-05,
+      "loss": 1.0309,
+      "step": 2500
+    },
+    {
+      "epoch": 0.344205890448665,
+      "grad_norm": 1.8243804521449938,
+      "learning_rate": 1.5253840011026992e-05,
+      "loss": 1.0871,
+      "step": 2501
+    },
+    {
+      "epoch": 0.34434351775392236,
+      "grad_norm": 1.6824051943343137,
+      "learning_rate": 1.525004682182811e-05,
+      "loss": 1.0783,
+      "step": 2502
+    },
+    {
+      "epoch": 0.34448114505917976,
+      "grad_norm": 1.4785025945144012,
+      "learning_rate": 1.524625258951679e-05,
+      "loss": 0.9843,
+      "step": 2503
+    },
+    {
+      "epoch": 0.3446187723644371,
+      "grad_norm": 1.6224852362356657,
+      "learning_rate": 1.5242457314846894e-05,
+      "loss": 1.1099,
+      "step": 2504
+    },
+    {
+      "epoch": 0.34475639966969446,
+      "grad_norm": 1.777819533118021,
+      "learning_rate": 1.5238660998572492e-05,
+      "loss": 1.1213,
+      "step": 2505
+    },
+    {
+      "epoch": 0.3448940269749518,
+      "grad_norm": 1.75688063359809,
+      "learning_rate": 1.5234863641447857e-05,
+      "loss": 1.0783,
+      "step": 2506
+    },
+    {
+      "epoch": 0.3450316542802092,
+      "grad_norm": 1.7134935829124425,
+      "learning_rate": 1.5231065244227474e-05,
+      "loss": 1.0425,
+      "step": 2507
+    },
+    {
+      "epoch": 0.34516928158546656,
+      "grad_norm": 1.5224715381314857,
+      "learning_rate": 1.5227265807666034e-05,
+      "loss": 1.0335,
+      "step": 2508
+    },
+    {
+      "epoch": 0.3453069088907239,
+      "grad_norm": 1.7547787564037145,
+      "learning_rate": 1.522346533251843e-05,
+      "loss": 1.0733,
+      "step": 2509
+    },
+    {
+      "epoch": 0.34544453619598126,
+      "grad_norm": 1.549281608191395,
+      "learning_rate": 1.5219663819539769e-05,
+      "loss": 1.0559,
+      "step": 2510
+    },
+    {
+      "epoch": 0.34558216350123866,
+      "grad_norm": 1.7693444597990737,
+      "learning_rate": 1.5215861269485354e-05,
+      "loss": 1.1178,
+      "step": 2511
+    },
+    {
+      "epoch": 0.345719790806496,
+      "grad_norm": 1.755433986221727,
+      "learning_rate": 1.5212057683110702e-05,
+      "loss": 1.1059,
+      "step": 2512
+    },
+    {
+      "epoch": 0.34585741811175336,
+      "grad_norm": 1.901930132102428,
+      "learning_rate": 1.5208253061171535e-05,
+      "loss": 1.0424,
+      "step": 2513
+    },
+    {
+      "epoch": 0.34599504541701076,
+      "grad_norm": 1.8840027268699295,
+      "learning_rate": 1.5204447404423777e-05,
+      "loss": 1.0293,
+      "step": 2514
+    },
+    {
+      "epoch": 0.3461326727222681,
+      "grad_norm": 1.8006990095952689,
+      "learning_rate": 1.5200640713623558e-05,
+      "loss": 1.1076,
+      "step": 2515
+    },
+    {
+      "epoch": 0.34627030002752546,
+      "grad_norm": 1.9684085740961244,
+      "learning_rate": 1.5196832989527222e-05,
+      "loss": 1.0896,
+      "step": 2516
+    },
+    {
+      "epoch": 0.3464079273327828,
+      "grad_norm": 1.7150343354270194,
+      "learning_rate": 1.519302423289131e-05,
+      "loss": 1.0977,
+      "step": 2517
+    },
+    {
+      "epoch": 0.3465455546380402,
+      "grad_norm": 1.7833580700983938,
+      "learning_rate": 1.5189214444472562e-05,
+      "loss": 1.0532,
+      "step": 2518
+    },
+    {
+      "epoch": 0.34668318194329756,
+      "grad_norm": 1.7419282758217778,
+      "learning_rate": 1.5185403625027941e-05,
+      "loss": 1.0419,
+      "step": 2519
+    },
+    {
+      "epoch": 0.3468208092485549,
+      "grad_norm": 1.6509194077351477,
+      "learning_rate": 1.5181591775314597e-05,
+      "loss": 1.156,
+      "step": 2520
+    },
+    {
+      "epoch": 0.34695843655381226,
+      "grad_norm": 1.820930368117441,
+      "learning_rate": 1.51777788960899e-05,
+      "loss": 1.0791,
+      "step": 2521
+    },
+    {
+      "epoch": 0.34709606385906966,
+      "grad_norm": 1.6329745895645094,
+      "learning_rate": 1.5173964988111406e-05,
+      "loss": 1.0651,
+      "step": 2522
+    },
+    {
+      "epoch": 0.347233691164327,
+      "grad_norm": 1.730348201865413,
+      "learning_rate": 1.5170150052136896e-05,
+      "loss": 1.0522,
+      "step": 2523
+    },
+    {
+      "epoch": 0.34737131846958436,
+      "grad_norm": 1.9121073863256095,
+      "learning_rate": 1.5166334088924343e-05,
+      "loss": 1.058,
+      "step": 2524
+    },
+    {
+      "epoch": 0.3475089457748417,
+      "grad_norm": 1.8026279514826422,
+      "learning_rate": 1.5162517099231923e-05,
+      "loss": 1.0271,
+      "step": 2525
+    },
+    {
+      "epoch": 0.3476465730800991,
+      "grad_norm": 1.6796648858758643,
+      "learning_rate": 1.5158699083818027e-05,
+      "loss": 1.1328,
+      "step": 2526
+    },
+    {
+      "epoch": 0.34778420038535646,
+      "grad_norm": 2.0075598928410034,
+      "learning_rate": 1.5154880043441233e-05,
+      "loss": 1.1005,
+      "step": 2527
+    },
+    {
+      "epoch": 0.3479218276906138,
+      "grad_norm": 1.7869783621036,
+      "learning_rate": 1.5151059978860341e-05,
+      "loss": 1.0783,
+      "step": 2528
+    },
+    {
+      "epoch": 0.34805945499587115,
+      "grad_norm": 2.0584947623837295,
+      "learning_rate": 1.514723889083434e-05,
+      "loss": 1.1623,
+      "step": 2529
+    },
+    {
+      "epoch": 0.34819708230112856,
+      "grad_norm": 1.670771794433263,
+      "learning_rate": 1.514341678012243e-05,
+      "loss": 1.1976,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3483347096063859,
+      "grad_norm": 1.7201049901615104,
+      "learning_rate": 1.513959364748401e-05,
+      "loss": 1.088,
+      "step": 2531
+    },
+    {
+      "epoch": 0.34847233691164325,
+      "grad_norm": 2.0129916412525124,
+      "learning_rate": 1.5135769493678686e-05,
+      "loss": 1.047,
+      "step": 2532
+    },
+    {
+      "epoch": 0.34860996421690066,
+      "grad_norm": 1.7914746768355863,
+      "learning_rate": 1.5131944319466269e-05,
+      "loss": 1.1291,
+      "step": 2533
+    },
+    {
+      "epoch": 0.348747591522158,
+      "grad_norm": 1.807065567567135,
+      "learning_rate": 1.5128118125606759e-05,
+      "loss": 1.1191,
+      "step": 2534
+    },
+    {
+      "epoch": 0.34888521882741536,
+      "grad_norm": 1.5033738128006784,
+      "learning_rate": 1.5124290912860375e-05,
+      "loss": 1.14,
+      "step": 2535
+    },
+    {
+      "epoch": 0.3490228461326727,
+      "grad_norm": 1.804583645489471,
+      "learning_rate": 1.5120462681987534e-05,
+      "loss": 1.1359,
+      "step": 2536
+    },
+    {
+      "epoch": 0.3491604734379301,
+      "grad_norm": 1.7350917328773792,
+      "learning_rate": 1.5116633433748846e-05,
+      "loss": 1.0613,
+      "step": 2537
+    },
+    {
+      "epoch": 0.34929810074318746,
+      "grad_norm": 1.6981971491006336,
+      "learning_rate": 1.5112803168905133e-05,
+      "loss": 1.2286,
+      "step": 2538
+    },
+    {
+      "epoch": 0.3494357280484448,
+      "grad_norm": 1.9266151869593289,
+      "learning_rate": 1.5108971888217418e-05,
+      "loss": 1.1776,
+      "step": 2539
+    },
+    {
+      "epoch": 0.34957335535370215,
+      "grad_norm": 1.6882229497071304,
+      "learning_rate": 1.5105139592446923e-05,
+      "loss": 1.1066,
+      "step": 2540
+    },
+    {
+      "epoch": 0.34971098265895956,
+      "grad_norm": 2.073497695663884,
+      "learning_rate": 1.5101306282355076e-05,
+      "loss": 1.1714,
+      "step": 2541
+    },
+    {
+      "epoch": 0.3498486099642169,
+      "grad_norm": 1.7834935414058426,
+      "learning_rate": 1.5097471958703494e-05,
+      "loss": 1.1002,
+      "step": 2542
+    },
+    {
+      "epoch": 0.34998623726947425,
+      "grad_norm": 1.696892440493395,
+      "learning_rate": 1.5093636622254012e-05,
+      "loss": 1.0941,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3501238645747316,
+      "grad_norm": 2.011586114652374,
+      "learning_rate": 1.508980027376866e-05,
+      "loss": 1.0039,
+      "step": 2544
+    },
+    {
+      "epoch": 0.350261491879989,
+      "grad_norm": 1.7505746457800913,
+      "learning_rate": 1.5085962914009667e-05,
+      "loss": 1.1157,
+      "step": 2545
+    },
+    {
+      "epoch": 0.35039911918524635,
+      "grad_norm": 1.7380368411938478,
+      "learning_rate": 1.5082124543739456e-05,
+      "loss": 1.1612,
+      "step": 2546
+    },
+    {
+      "epoch": 0.3505367464905037,
+      "grad_norm": 1.886370876527671,
+      "learning_rate": 1.5078285163720669e-05,
+      "loss": 1.0449,
+      "step": 2547
+    },
+    {
+      "epoch": 0.35067437379576105,
+      "grad_norm": 1.6572261346667083,
+      "learning_rate": 1.5074444774716135e-05,
+      "loss": 1.0851,
+      "step": 2548
+    },
+    {
+      "epoch": 0.35081200110101846,
+      "grad_norm": 1.7824365555614288,
+      "learning_rate": 1.5070603377488885e-05,
+      "loss": 1.0881,
+      "step": 2549
+    },
+    {
+      "epoch": 0.3509496284062758,
+      "grad_norm": 1.5846972852212469,
+      "learning_rate": 1.5066760972802153e-05,
+      "loss": 1.0211,
+      "step": 2550
+    },
+    {
+      "epoch": 0.35108725571153315,
+      "grad_norm": 1.72820750993155,
+      "learning_rate": 1.5062917561419373e-05,
+      "loss": 1.0832,
+      "step": 2551
+    },
+    {
+      "epoch": 0.35122488301679056,
+      "grad_norm": 1.6371857352389667,
+      "learning_rate": 1.5059073144104174e-05,
+      "loss": 1.1442,
+      "step": 2552
+    },
+    {
+      "epoch": 0.3513625103220479,
+      "grad_norm": 1.6580463820512277,
+      "learning_rate": 1.5055227721620394e-05,
+      "loss": 1.0606,
+      "step": 2553
+    },
+    {
+      "epoch": 0.35150013762730525,
+      "grad_norm": 1.8394583108849039,
+      "learning_rate": 1.5051381294732064e-05,
+      "loss": 1.0887,
+      "step": 2554
+    },
+    {
+      "epoch": 0.3516377649325626,
+      "grad_norm": 1.8443317433163753,
+      "learning_rate": 1.5047533864203416e-05,
+      "loss": 1.093,
+      "step": 2555
+    },
+    {
+      "epoch": 0.35177539223782,
+      "grad_norm": 2.1114191504614577,
+      "learning_rate": 1.5043685430798884e-05,
+      "loss": 1.0976,
+      "step": 2556
+    },
+    {
+      "epoch": 0.35191301954307735,
+      "grad_norm": 1.6514430053152194,
+      "learning_rate": 1.5039835995283097e-05,
+      "loss": 1.0536,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3520506468483347,
+      "grad_norm": 1.70573544046439,
+      "learning_rate": 1.5035985558420887e-05,
+      "loss": 0.9519,
+      "step": 2558
+    },
+    {
+      "epoch": 0.35218827415359205,
+      "grad_norm": 1.7038618020890826,
+      "learning_rate": 1.503213412097728e-05,
+      "loss": 1.0284,
+      "step": 2559
+    },
+    {
+      "epoch": 0.35232590145884946,
+      "grad_norm": 2.165752814952019,
+      "learning_rate": 1.5028281683717505e-05,
+      "loss": 1.0882,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3524635287641068,
+      "grad_norm": 1.7832828119324977,
+      "learning_rate": 1.5024428247406992e-05,
+      "loss": 1.1454,
+      "step": 2561
+    },
+    {
+      "epoch": 0.35260115606936415,
+      "grad_norm": 1.6488418047142295,
+      "learning_rate": 1.5020573812811359e-05,
+      "loss": 0.9611,
+      "step": 2562
+    },
+    {
+      "epoch": 0.3527387833746215,
+      "grad_norm": 1.6937210807718899,
+      "learning_rate": 1.5016718380696436e-05,
+      "loss": 1.076,
+      "step": 2563
+    },
+    {
+      "epoch": 0.3528764106798789,
+      "grad_norm": 1.9066844014119584,
+      "learning_rate": 1.5012861951828241e-05,
+      "loss": 1.1385,
+      "step": 2564
+    },
+    {
+      "epoch": 0.35301403798513625,
+      "grad_norm": 1.9330330387088697,
+      "learning_rate": 1.5009004526972999e-05,
+      "loss": 1.1068,
+      "step": 2565
+    },
+    {
+      "epoch": 0.3531516652903936,
+      "grad_norm": 1.7702317599563946,
+      "learning_rate": 1.5005146106897117e-05,
+      "loss": 1.1374,
+      "step": 2566
+    },
+    {
+      "epoch": 0.35328929259565095,
+      "grad_norm": 1.8792050362701709,
+      "learning_rate": 1.5001286692367221e-05,
+      "loss": 1.0887,
+      "step": 2567
+    },
+    {
+      "epoch": 0.35342691990090835,
+      "grad_norm": 1.5790249603244053,
+      "learning_rate": 1.4997426284150119e-05,
+      "loss": 1.1345,
+      "step": 2568
+    },
+    {
+      "epoch": 0.3535645472061657,
+      "grad_norm": 1.7901694071333325,
+      "learning_rate": 1.4993564883012816e-05,
+      "loss": 1.1591,
+      "step": 2569
+    },
+    {
+      "epoch": 0.35370217451142305,
+      "grad_norm": 1.6620783060249198,
+      "learning_rate": 1.498970248972253e-05,
+      "loss": 1.1259,
+      "step": 2570
+    },
+    {
+      "epoch": 0.35383980181668045,
+      "grad_norm": 1.8469942506964478,
+      "learning_rate": 1.4985839105046663e-05,
+      "loss": 1.1577,
+      "step": 2571
+    },
+    {
+      "epoch": 0.3539774291219378,
+      "grad_norm": 1.7822492089898467,
+      "learning_rate": 1.498197472975281e-05,
+      "loss": 1.0662,
+      "step": 2572
+    },
+    {
+      "epoch": 0.35411505642719515,
+      "grad_norm": 1.8669572595714818,
+      "learning_rate": 1.4978109364608777e-05,
+      "loss": 0.9972,
+      "step": 2573
+    },
+    {
+      "epoch": 0.3542526837324525,
+      "grad_norm": 2.0359710317843147,
+      "learning_rate": 1.4974243010382554e-05,
+      "loss": 1.0561,
+      "step": 2574
+    },
+    {
+      "epoch": 0.3543903110377099,
+      "grad_norm": 1.5176717488190814,
+      "learning_rate": 1.4970375667842336e-05,
+      "loss": 1.0406,
+      "step": 2575
+    },
+    {
+      "epoch": 0.35452793834296725,
+      "grad_norm": 1.6240185999513308,
+      "learning_rate": 1.4966507337756511e-05,
+      "loss": 1.0974,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3546655656482246,
+      "grad_norm": 1.7107033626694317,
+      "learning_rate": 1.4962638020893659e-05,
+      "loss": 1.105,
+      "step": 2577
+    },
+    {
+      "epoch": 0.35480319295348195,
+      "grad_norm": 1.6202668149389496,
+      "learning_rate": 1.4958767718022565e-05,
+      "loss": 1.0086,
+      "step": 2578
+    },
+    {
+      "epoch": 0.35494082025873935,
+      "grad_norm": 1.8435129321440153,
+      "learning_rate": 1.4954896429912205e-05,
+      "loss": 1.1273,
+      "step": 2579
+    },
+    {
+      "epoch": 0.3550784475639967,
+      "grad_norm": 1.943671577885704,
+      "learning_rate": 1.4951024157331749e-05,
+      "loss": 1.1715,
+      "step": 2580
+    },
+    {
+      "epoch": 0.35521607486925405,
+      "grad_norm": 1.634725265708131,
+      "learning_rate": 1.4947150901050564e-05,
+      "loss": 1.0844,
+      "step": 2581
+    },
+    {
+      "epoch": 0.3553537021745114,
+      "grad_norm": 1.4985648750613396,
+      "learning_rate": 1.4943276661838215e-05,
+      "loss": 1.0894,
+      "step": 2582
+    },
+    {
+      "epoch": 0.3554913294797688,
+      "grad_norm": 1.710513708096674,
+      "learning_rate": 1.4939401440464458e-05,
+      "loss": 1.0518,
+      "step": 2583
+    },
+    {
+      "epoch": 0.35562895678502615,
+      "grad_norm": 1.5630120020357812,
+      "learning_rate": 1.4935525237699249e-05,
+      "loss": 1.0395,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3557665840902835,
+      "grad_norm": 1.6414020471874435,
+      "learning_rate": 1.4931648054312736e-05,
+      "loss": 1.0952,
+      "step": 2585
+    },
+    {
+      "epoch": 0.35590421139554085,
+      "grad_norm": 1.673189722639924,
+      "learning_rate": 1.4927769891075259e-05,
+      "loss": 1.113,
+      "step": 2586
+    },
+    {
+      "epoch": 0.35604183870079825,
+      "grad_norm": 1.5703468713576496,
+      "learning_rate": 1.492389074875736e-05,
+      "loss": 1.1234,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3561794660060556,
+      "grad_norm": 1.7437648144420366,
+      "learning_rate": 1.4920010628129769e-05,
+      "loss": 1.0545,
+      "step": 2588
+    },
+    {
+      "epoch": 0.35631709331131295,
+      "grad_norm": 1.6271766876597584,
+      "learning_rate": 1.4916129529963415e-05,
+      "loss": 1.1056,
+      "step": 2589
+    },
+    {
+      "epoch": 0.35645472061657035,
+      "grad_norm": 1.4760560745621767,
+      "learning_rate": 1.4912247455029417e-05,
+      "loss": 1.0833,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3565923479218277,
+      "grad_norm": 1.5875325526078738,
+      "learning_rate": 1.4908364404099093e-05,
+      "loss": 1.0419,
+      "step": 2591
+    },
+    {
+      "epoch": 0.35672997522708505,
+      "grad_norm": 1.6065080917246648,
+      "learning_rate": 1.490448037794395e-05,
+      "loss": 1.1092,
+      "step": 2592
+    },
+    {
+      "epoch": 0.3568676025323424,
+      "grad_norm": 1.5974413367525926,
+      "learning_rate": 1.4900595377335686e-05,
+      "loss": 1.0983,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3570052298375998,
+      "grad_norm": 1.795231911589888,
+      "learning_rate": 1.4896709403046205e-05,
+      "loss": 1.1178,
+      "step": 2594
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 1.7840239803686166,
+      "learning_rate": 1.4892822455847595e-05,
+      "loss": 1.0901,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3572804844481145,
+      "grad_norm": 1.6946752760523136,
+      "learning_rate": 1.4888934536512141e-05,
+      "loss": 1.0563,
+      "step": 2596
+    },
+    {
+      "epoch": 0.35741811175337185,
+      "grad_norm": 1.5866023653977352,
+      "learning_rate": 1.4885045645812316e-05,
+      "loss": 1.0079,
+      "step": 2597
+    },
+    {
+      "epoch": 0.35755573905862925,
+      "grad_norm": 1.6744673647122899,
+      "learning_rate": 1.4881155784520789e-05,
+      "loss": 1.0824,
+      "step": 2598
+    },
+    {
+      "epoch": 0.3576933663638866,
+      "grad_norm": 1.665295169616684,
+      "learning_rate": 1.4877264953410421e-05,
+      "loss": 1.0389,
+      "step": 2599
+    },
+    {
+      "epoch": 0.35783099366914395,
+      "grad_norm": 1.684963794945681,
+      "learning_rate": 1.4873373153254271e-05,
+      "loss": 1.0438,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3579686209744013,
+      "grad_norm": 1.7841742658221174,
+      "learning_rate": 1.4869480384825585e-05,
+      "loss": 1.1275,
+      "step": 2601
+    },
+    {
+      "epoch": 0.3581062482796587,
+      "grad_norm": 1.9192726378346312,
+      "learning_rate": 1.4865586648897803e-05,
+      "loss": 1.2261,
+      "step": 2602
+    },
+    {
+      "epoch": 0.35824387558491605,
+      "grad_norm": 1.7642299636910759,
+      "learning_rate": 1.4861691946244553e-05,
+      "loss": 1.0499,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3583815028901734,
+      "grad_norm": 1.8940935537328831,
+      "learning_rate": 1.4857796277639666e-05,
+      "loss": 1.1499,
+      "step": 2604
+    },
+    {
+      "epoch": 0.35851913019543075,
+      "grad_norm": 1.871691132191385,
+      "learning_rate": 1.4853899643857154e-05,
+      "loss": 1.039,
+      "step": 2605
+    },
+    {
+      "epoch": 0.35865675750068815,
+      "grad_norm": 1.878289580016883,
+      "learning_rate": 1.4850002045671228e-05,
+      "loss": 1.1044,
+      "step": 2606
+    },
+    {
+      "epoch": 0.3587943848059455,
+      "grad_norm": 1.724030358282651,
+      "learning_rate": 1.484610348385628e-05,
+      "loss": 1.0647,
+      "step": 2607
+    },
+    {
+      "epoch": 0.35893201211120285,
+      "grad_norm": 1.635979180951212,
+      "learning_rate": 1.4842203959186908e-05,
+      "loss": 1.0735,
+      "step": 2608
+    },
+    {
+      "epoch": 0.35906963941646025,
+      "grad_norm": 1.5928935356850182,
+      "learning_rate": 1.4838303472437891e-05,
+      "loss": 1.0673,
+      "step": 2609
+    },
+    {
+      "epoch": 0.3592072667217176,
+      "grad_norm": 1.5206395878957337,
+      "learning_rate": 1.4834402024384205e-05,
+      "loss": 0.9957,
+      "step": 2610
+    },
+    {
+      "epoch": 0.35934489402697495,
+      "grad_norm": 1.4998054938655745,
+      "learning_rate": 1.483049961580101e-05,
+      "loss": 0.9433,
+      "step": 2611
+    },
+    {
+      "epoch": 0.3594825213322323,
+      "grad_norm": 1.531581674235957,
+      "learning_rate": 1.4826596247463668e-05,
+      "loss": 1.0538,
+      "step": 2612
+    },
+    {
+      "epoch": 0.3596201486374897,
+      "grad_norm": 1.7723770086920334,
+      "learning_rate": 1.4822691920147718e-05,
+      "loss": 1.0298,
+      "step": 2613
+    },
+    {
+      "epoch": 0.35975777594274705,
+      "grad_norm": 1.5916166160419218,
+      "learning_rate": 1.4818786634628898e-05,
+      "loss": 1.1158,
+      "step": 2614
+    },
+    {
+      "epoch": 0.3598954032480044,
+      "grad_norm": 1.887287108507014,
+      "learning_rate": 1.4814880391683135e-05,
+      "loss": 1.1768,
+      "step": 2615
+    },
+    {
+      "epoch": 0.36003303055326175,
+      "grad_norm": 1.7066316915388042,
+      "learning_rate": 1.481097319208655e-05,
+      "loss": 1.0815,
+      "step": 2616
+    },
+    {
+      "epoch": 0.36017065785851915,
+      "grad_norm": 1.957956831013734,
+      "learning_rate": 1.4807065036615442e-05,
+      "loss": 1.1028,
+      "step": 2617
+    },
+    {
+      "epoch": 0.3603082851637765,
+      "grad_norm": 1.7748079524560387,
+      "learning_rate": 1.4803155926046318e-05,
+      "loss": 1.1503,
+      "step": 2618
+    },
+    {
+      "epoch": 0.36044591246903385,
+      "grad_norm": 1.5239821242872063,
+      "learning_rate": 1.4799245861155855e-05,
+      "loss": 1.1124,
+      "step": 2619
+    },
+    {
+      "epoch": 0.3605835397742912,
+      "grad_norm": 1.7541778112931568,
+      "learning_rate": 1.4795334842720936e-05,
+      "loss": 0.9751,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3607211670795486,
+      "grad_norm": 1.6525513042127815,
+      "learning_rate": 1.4791422871518626e-05,
+      "loss": 1.059,
+      "step": 2621
+    },
+    {
+      "epoch": 0.36085879438480595,
+      "grad_norm": 1.830757201850349,
+      "learning_rate": 1.4787509948326177e-05,
+      "loss": 1.1646,
+      "step": 2622
+    },
+    {
+      "epoch": 0.3609964216900633,
+      "grad_norm": 1.6778658624253295,
+      "learning_rate": 1.4783596073921036e-05,
+      "loss": 1.0811,
+      "step": 2623
+    },
+    {
+      "epoch": 0.36113404899532064,
+      "grad_norm": 1.7810316392226975,
+      "learning_rate": 1.4779681249080833e-05,
+      "loss": 1.0626,
+      "step": 2624
+    },
+    {
+      "epoch": 0.36127167630057805,
+      "grad_norm": 1.9080938471413562,
+      "learning_rate": 1.4775765474583396e-05,
+      "loss": 1.0933,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3614093036058354,
+      "grad_norm": 1.6805930538822762,
+      "learning_rate": 1.477184875120673e-05,
+      "loss": 1.008,
+      "step": 2626
+    },
+    {
+      "epoch": 0.36154693091109275,
+      "grad_norm": 1.6703255710796603,
+      "learning_rate": 1.4767931079729039e-05,
+      "loss": 1.0727,
+      "step": 2627
+    },
+    {
+      "epoch": 0.36168455821635015,
+      "grad_norm": 1.6431299437256073,
+      "learning_rate": 1.4764012460928707e-05,
+      "loss": 1.0769,
+      "step": 2628
+    },
+    {
+      "epoch": 0.3618221855216075,
+      "grad_norm": 1.6090280267592296,
+      "learning_rate": 1.4760092895584313e-05,
+      "loss": 1.0503,
+      "step": 2629
+    },
+    {
+      "epoch": 0.36195981282686485,
+      "grad_norm": 1.7262636610324742,
+      "learning_rate": 1.4756172384474615e-05,
+      "loss": 1.1083,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3620974401321222,
+      "grad_norm": 1.573042589490139,
+      "learning_rate": 1.475225092837857e-05,
+      "loss": 1.0369,
+      "step": 2631
+    },
+    {
+      "epoch": 0.3622350674373796,
+      "grad_norm": 2.0414580062299796,
+      "learning_rate": 1.4748328528075319e-05,
+      "loss": 1.0732,
+      "step": 2632
+    },
+    {
+      "epoch": 0.36237269474263695,
+      "grad_norm": 1.6210084710729677,
+      "learning_rate": 1.4744405184344186e-05,
+      "loss": 1.0047,
+      "step": 2633
+    },
+    {
+      "epoch": 0.3625103220478943,
+      "grad_norm": 1.5762925021621306,
+      "learning_rate": 1.4740480897964689e-05,
+      "loss": 1.0223,
+      "step": 2634
+    },
+    {
+      "epoch": 0.36264794935315164,
+      "grad_norm": 1.7067236136661312,
+      "learning_rate": 1.4736555669716526e-05,
+      "loss": 1.1422,
+      "step": 2635
+    },
+    {
+      "epoch": 0.36278557665840905,
+      "grad_norm": 1.7583200549687568,
+      "learning_rate": 1.4732629500379587e-05,
+      "loss": 1.0644,
+      "step": 2636
+    },
+    {
+      "epoch": 0.3629232039636664,
+      "grad_norm": 1.644533056988231,
+      "learning_rate": 1.472870239073395e-05,
+      "loss": 0.9271,
+      "step": 2637
+    },
+    {
+      "epoch": 0.36306083126892374,
+      "grad_norm": 1.8224029414932235,
+      "learning_rate": 1.4724774341559875e-05,
+      "loss": 1.0446,
+      "step": 2638
+    },
+    {
+      "epoch": 0.3631984585741811,
+      "grad_norm": 1.5900859912806318,
+      "learning_rate": 1.4720845353637815e-05,
+      "loss": 1.0504,
+      "step": 2639
+    },
+    {
+      "epoch": 0.3633360858794385,
+      "grad_norm": 1.7662328517331134,
+      "learning_rate": 1.4716915427748402e-05,
+      "loss": 1.151,
+      "step": 2640
+    },
+    {
+      "epoch": 0.36347371318469585,
+      "grad_norm": 1.4890588213570892,
+      "learning_rate": 1.4712984564672464e-05,
+      "loss": 1.0398,
+      "step": 2641
+    },
+    {
+      "epoch": 0.3636113404899532,
+      "grad_norm": 1.7067636092422127,
+      "learning_rate": 1.4709052765191006e-05,
+      "loss": 1.0662,
+      "step": 2642
+    },
+    {
+      "epoch": 0.36374896779521054,
+      "grad_norm": 1.8491287827961467,
+      "learning_rate": 1.4705120030085223e-05,
+      "loss": 1.0845,
+      "step": 2643
+    },
+    {
+      "epoch": 0.36388659510046795,
+      "grad_norm": 1.5339910308990317,
+      "learning_rate": 1.4701186360136497e-05,
+      "loss": 0.9987,
+      "step": 2644
+    },
+    {
+      "epoch": 0.3640242224057253,
+      "grad_norm": 2.1504107599847564,
+      "learning_rate": 1.4697251756126397e-05,
+      "loss": 1.2341,
+      "step": 2645
+    },
+    {
+      "epoch": 0.36416184971098264,
+      "grad_norm": 1.717175606320573,
+      "learning_rate": 1.4693316218836663e-05,
+      "loss": 1.0928,
+      "step": 2646
+    },
+    {
+      "epoch": 0.36429947701624005,
+      "grad_norm": 1.8049648056289809,
+      "learning_rate": 1.4689379749049246e-05,
+      "loss": 1.0742,
+      "step": 2647
+    },
+    {
+      "epoch": 0.3644371043214974,
+      "grad_norm": 1.589659055025965,
+      "learning_rate": 1.4685442347546261e-05,
+      "loss": 1.0916,
+      "step": 2648
+    },
+    {
+      "epoch": 0.36457473162675474,
+      "grad_norm": 1.6608028096883132,
+      "learning_rate": 1.4681504015110022e-05,
+      "loss": 1.0492,
+      "step": 2649
+    },
+    {
+      "epoch": 0.3647123589320121,
+      "grad_norm": 1.7923432516085307,
+      "learning_rate": 1.4677564752523017e-05,
+      "loss": 1.009,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3648499862372695,
+      "grad_norm": 1.6802524359420248,
+      "learning_rate": 1.4673624560567923e-05,
+      "loss": 1.0115,
+      "step": 2651
+    },
+    {
+      "epoch": 0.36498761354252685,
+      "grad_norm": 1.9459191122265007,
+      "learning_rate": 1.4669683440027605e-05,
+      "loss": 1.155,
+      "step": 2652
+    },
+    {
+      "epoch": 0.3651252408477842,
+      "grad_norm": 1.655905159050617,
+      "learning_rate": 1.4665741391685107e-05,
+      "loss": 1.0135,
+      "step": 2653
+    },
+    {
+      "epoch": 0.36526286815304154,
+      "grad_norm": 1.7859311097230404,
+      "learning_rate": 1.4661798416323663e-05,
+      "loss": 1.0839,
+      "step": 2654
+    },
+    {
+      "epoch": 0.36540049545829895,
+      "grad_norm": 1.8265440261483294,
+      "learning_rate": 1.4657854514726684e-05,
+      "loss": 1.0556,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3655381227635563,
+      "grad_norm": 1.7259038858999736,
+      "learning_rate": 1.4653909687677774e-05,
+      "loss": 1.047,
+      "step": 2656
+    },
+    {
+      "epoch": 0.36567575006881364,
+      "grad_norm": 1.9938766398304408,
+      "learning_rate": 1.4649963935960715e-05,
+      "loss": 1.2317,
+      "step": 2657
+    },
+    {
+      "epoch": 0.365813377374071,
+      "grad_norm": 1.6548834275409612,
+      "learning_rate": 1.4646017260359471e-05,
+      "loss": 1.0858,
+      "step": 2658
+    },
+    {
+      "epoch": 0.3659510046793284,
+      "grad_norm": 2.0066448541230515,
+      "learning_rate": 1.4642069661658196e-05,
+      "loss": 1.0508,
+      "step": 2659
+    },
+    {
+      "epoch": 0.36608863198458574,
+      "grad_norm": 1.8659325501224993,
+      "learning_rate": 1.4638121140641222e-05,
+      "loss": 1.1169,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3662262592898431,
+      "grad_norm": 1.631262011542562,
+      "learning_rate": 1.4634171698093068e-05,
+      "loss": 1.1514,
+      "step": 2661
+    },
+    {
+      "epoch": 0.36636388659510044,
+      "grad_norm": 1.4092484058028352,
+      "learning_rate": 1.4630221334798429e-05,
+      "loss": 1.1173,
+      "step": 2662
+    },
+    {
+      "epoch": 0.36650151390035784,
+      "grad_norm": 1.6424629596241145,
+      "learning_rate": 1.4626270051542192e-05,
+      "loss": 1.0597,
+      "step": 2663
+    },
+    {
+      "epoch": 0.3666391412056152,
+      "grad_norm": 1.8546990843944755,
+      "learning_rate": 1.4622317849109424e-05,
+      "loss": 1.0112,
+      "step": 2664
+    },
+    {
+      "epoch": 0.36677676851087254,
+      "grad_norm": 1.6589268606594818,
+      "learning_rate": 1.4618364728285373e-05,
+      "loss": 1.0415,
+      "step": 2665
+    },
+    {
+      "epoch": 0.36691439581612995,
+      "grad_norm": 1.5620239924072246,
+      "learning_rate": 1.4614410689855466e-05,
+      "loss": 1.0426,
+      "step": 2666
+    },
+    {
+      "epoch": 0.3670520231213873,
+      "grad_norm": 1.7015526917494908,
+      "learning_rate": 1.4610455734605323e-05,
+      "loss": 1.025,
+      "step": 2667
+    },
+    {
+      "epoch": 0.36718965042664464,
+      "grad_norm": 1.7539882110488587,
+      "learning_rate": 1.4606499863320734e-05,
+      "loss": 1.0856,
+      "step": 2668
+    },
+    {
+      "epoch": 0.367327277731902,
+      "grad_norm": 1.581280048335631,
+      "learning_rate": 1.460254307678768e-05,
+      "loss": 1.0486,
+      "step": 2669
+    },
+    {
+      "epoch": 0.3674649050371594,
+      "grad_norm": 1.667483074016372,
+      "learning_rate": 1.4598585375792315e-05,
+      "loss": 1.0789,
+      "step": 2670
+    },
+    {
+      "epoch": 0.36760253234241674,
+      "grad_norm": 1.755703790455354,
+      "learning_rate": 1.4594626761120983e-05,
+      "loss": 1.1694,
+      "step": 2671
+    },
+    {
+      "epoch": 0.3677401596476741,
+      "grad_norm": 1.8226280823619685,
+      "learning_rate": 1.459066723356021e-05,
+      "loss": 1.1164,
+      "step": 2672
+    },
+    {
+      "epoch": 0.36787778695293144,
+      "grad_norm": 1.5699803951107298,
+      "learning_rate": 1.4586706793896697e-05,
+      "loss": 1.0819,
+      "step": 2673
+    },
+    {
+      "epoch": 0.36801541425818884,
+      "grad_norm": 1.736392590840252,
+      "learning_rate": 1.4582745442917328e-05,
+      "loss": 1.0896,
+      "step": 2674
+    },
+    {
+      "epoch": 0.3681530415634462,
+      "grad_norm": 1.5171324061137426,
+      "learning_rate": 1.457878318140917e-05,
+      "loss": 1.0916,
+      "step": 2675
+    },
+    {
+      "epoch": 0.36829066886870354,
+      "grad_norm": 1.536076214682531,
+      "learning_rate": 1.4574820010159474e-05,
+      "loss": 1.0559,
+      "step": 2676
+    },
+    {
+      "epoch": 0.3684282961739609,
+      "grad_norm": 1.778050361389024,
+      "learning_rate": 1.4570855929955663e-05,
+      "loss": 1.0534,
+      "step": 2677
+    },
+    {
+      "epoch": 0.3685659234792183,
+      "grad_norm": 1.6377152929715204,
+      "learning_rate": 1.4566890941585345e-05,
+      "loss": 1.0626,
+      "step": 2678
+    },
+    {
+      "epoch": 0.36870355078447564,
+      "grad_norm": 1.5399938482312343,
+      "learning_rate": 1.4562925045836311e-05,
+      "loss": 1.0261,
+      "step": 2679
+    },
+    {
+      "epoch": 0.368841178089733,
+      "grad_norm": 1.6147069572120216,
+      "learning_rate": 1.4558958243496533e-05,
+      "loss": 0.9829,
+      "step": 2680
+    },
+    {
+      "epoch": 0.36897880539499034,
+      "grad_norm": 1.7144424287514364,
+      "learning_rate": 1.4554990535354155e-05,
+      "loss": 1.0636,
+      "step": 2681
+    },
+    {
+      "epoch": 0.36911643270024774,
+      "grad_norm": 1.5566125722011692,
+      "learning_rate": 1.4551021922197508e-05,
+      "loss": 0.9682,
+      "step": 2682
+    },
+    {
+      "epoch": 0.3692540600055051,
+      "grad_norm": 1.6788545482416568,
+      "learning_rate": 1.4547052404815103e-05,
+      "loss": 1.0782,
+      "step": 2683
+    },
+    {
+      "epoch": 0.36939168731076244,
+      "grad_norm": 1.6675383647995132,
+      "learning_rate": 1.4543081983995628e-05,
+      "loss": 0.9954,
+      "step": 2684
+    },
+    {
+      "epoch": 0.36952931461601984,
+      "grad_norm": 1.9066348937060653,
+      "learning_rate": 1.4539110660527951e-05,
+      "loss": 1.0843,
+      "step": 2685
+    },
+    {
+      "epoch": 0.3696669419212772,
+      "grad_norm": 1.6736938174409826,
+      "learning_rate": 1.4535138435201117e-05,
+      "loss": 0.9869,
+      "step": 2686
+    },
+    {
+      "epoch": 0.36980456922653454,
+      "grad_norm": 1.8308988350622064,
+      "learning_rate": 1.4531165308804355e-05,
+      "loss": 1.05,
+      "step": 2687
+    },
+    {
+      "epoch": 0.3699421965317919,
+      "grad_norm": 1.6582123061784981,
+      "learning_rate": 1.4527191282127074e-05,
+      "loss": 1.0798,
+      "step": 2688
+    },
+    {
+      "epoch": 0.3700798238370493,
+      "grad_norm": 1.7301101651851614,
+      "learning_rate": 1.4523216355958854e-05,
+      "loss": 1.0827,
+      "step": 2689
+    },
+    {
+      "epoch": 0.37021745114230664,
+      "grad_norm": 1.760665834637587,
+      "learning_rate": 1.451924053108946e-05,
+      "loss": 1.1413,
+      "step": 2690
+    },
+    {
+      "epoch": 0.370355078447564,
+      "grad_norm": 1.7619600069021069,
+      "learning_rate": 1.4515263808308837e-05,
+      "loss": 1.004,
+      "step": 2691
+    },
+    {
+      "epoch": 0.37049270575282134,
+      "grad_norm": 1.614914668217245,
+      "learning_rate": 1.4511286188407098e-05,
+      "loss": 1.114,
+      "step": 2692
+    },
+    {
+      "epoch": 0.37063033305807874,
+      "grad_norm": 1.7318183652847423,
+      "learning_rate": 1.450730767217455e-05,
+      "loss": 1.0979,
+      "step": 2693
+    },
+    {
+      "epoch": 0.3707679603633361,
+      "grad_norm": 1.7269469890493767,
+      "learning_rate": 1.4503328260401664e-05,
+      "loss": 1.1091,
+      "step": 2694
+    },
+    {
+      "epoch": 0.37090558766859344,
+      "grad_norm": 1.7036114883754234,
+      "learning_rate": 1.4499347953879097e-05,
+      "loss": 1.0276,
+      "step": 2695
+    },
+    {
+      "epoch": 0.3710432149738508,
+      "grad_norm": 1.5626689541007812,
+      "learning_rate": 1.4495366753397682e-05,
+      "loss": 1.0595,
+      "step": 2696
+    },
+    {
+      "epoch": 0.3711808422791082,
+      "grad_norm": 1.699743699958119,
+      "learning_rate": 1.4491384659748426e-05,
+      "loss": 1.075,
+      "step": 2697
+    },
+    {
+      "epoch": 0.37131846958436554,
+      "grad_norm": 1.8542358243470682,
+      "learning_rate": 1.4487401673722519e-05,
+      "loss": 1.0732,
+      "step": 2698
+    },
+    {
+      "epoch": 0.3714560968896229,
+      "grad_norm": 2.117585076097301,
+      "learning_rate": 1.4483417796111325e-05,
+      "loss": 1.0227,
+      "step": 2699
+    },
+    {
+      "epoch": 0.37159372419488024,
+      "grad_norm": 1.5997842502538744,
+      "learning_rate": 1.4479433027706386e-05,
+      "loss": 1.0447,
+      "step": 2700
+    },
+    {
+      "epoch": 0.37173135150013764,
+      "grad_norm": 1.679174527896832,
+      "learning_rate": 1.4475447369299422e-05,
+      "loss": 1.0579,
+      "step": 2701
+    },
+    {
+      "epoch": 0.371868978805395,
+      "grad_norm": 1.7639032047294902,
+      "learning_rate": 1.4471460821682328e-05,
+      "loss": 1.0975,
+      "step": 2702
+    },
+    {
+      "epoch": 0.37200660611065234,
+      "grad_norm": 1.683605772726475,
+      "learning_rate": 1.4467473385647175e-05,
+      "loss": 1.0886,
+      "step": 2703
+    },
+    {
+      "epoch": 0.37214423341590974,
+      "grad_norm": 1.6505623083120022,
+      "learning_rate": 1.4463485061986214e-05,
+      "loss": 1.001,
+      "step": 2704
+    },
+    {
+      "epoch": 0.3722818607211671,
+      "grad_norm": 1.9515320732039045,
+      "learning_rate": 1.4459495851491872e-05,
+      "loss": 1.0655,
+      "step": 2705
+    },
+    {
+      "epoch": 0.37241948802642444,
+      "grad_norm": 1.943275562019624,
+      "learning_rate": 1.4455505754956745e-05,
+      "loss": 1.1811,
+      "step": 2706
+    },
+    {
+      "epoch": 0.3725571153316818,
+      "grad_norm": 1.7815349334314485,
+      "learning_rate": 1.4451514773173614e-05,
+      "loss": 1.0753,
+      "step": 2707
+    },
+    {
+      "epoch": 0.3726947426369392,
+      "grad_norm": 1.5888572858352863,
+      "learning_rate": 1.4447522906935434e-05,
+      "loss": 1.0838,
+      "step": 2708
+    },
+    {
+      "epoch": 0.37283236994219654,
+      "grad_norm": 1.6564166331856693,
+      "learning_rate": 1.4443530157035333e-05,
+      "loss": 1.0038,
+      "step": 2709
+    },
+    {
+      "epoch": 0.3729699972474539,
+      "grad_norm": 1.5347079640620014,
+      "learning_rate": 1.4439536524266612e-05,
+      "loss": 1.0009,
+      "step": 2710
+    },
+    {
+      "epoch": 0.37310762455271124,
+      "grad_norm": 1.7732853316036703,
+      "learning_rate": 1.4435542009422756e-05,
+      "loss": 1.1285,
+      "step": 2711
+    },
+    {
+      "epoch": 0.37324525185796864,
+      "grad_norm": 1.64511553735762,
+      "learning_rate": 1.4431546613297417e-05,
+      "loss": 1.1137,
+      "step": 2712
+    },
+    {
+      "epoch": 0.373382879163226,
+      "grad_norm": 1.6552405033050148,
+      "learning_rate": 1.442755033668443e-05,
+      "loss": 1.053,
+      "step": 2713
+    },
+    {
+      "epoch": 0.37352050646848334,
+      "grad_norm": 1.6197243745606467,
+      "learning_rate": 1.4423553180377796e-05,
+      "loss": 1.1038,
+      "step": 2714
+    },
+    {
+      "epoch": 0.3736581337737407,
+      "grad_norm": 1.7433856046695835,
+      "learning_rate": 1.4419555145171698e-05,
+      "loss": 1.1624,
+      "step": 2715
+    },
+    {
+      "epoch": 0.3737957610789981,
+      "grad_norm": 1.578169137719525,
+      "learning_rate": 1.4415556231860488e-05,
+      "loss": 1.0756,
+      "step": 2716
+    },
+    {
+      "epoch": 0.37393338838425544,
+      "grad_norm": 1.801972526690994,
+      "learning_rate": 1.4411556441238697e-05,
+      "loss": 1.106,
+      "step": 2717
+    },
+    {
+      "epoch": 0.3740710156895128,
+      "grad_norm": 1.6397145112477818,
+      "learning_rate": 1.440755577410103e-05,
+      "loss": 1.0424,
+      "step": 2718
+    },
+    {
+      "epoch": 0.37420864299477014,
+      "grad_norm": 1.815560730118826,
+      "learning_rate": 1.4403554231242364e-05,
+      "loss": 1.0831,
+      "step": 2719
+    },
+    {
+      "epoch": 0.37434627030002754,
+      "grad_norm": 1.6838961062581774,
+      "learning_rate": 1.4399551813457748e-05,
+      "loss": 1.0109,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3744838976052849,
+      "grad_norm": 1.8267033249057691,
+      "learning_rate": 1.4395548521542413e-05,
+      "loss": 1.0455,
+      "step": 2721
+    },
+    {
+      "epoch": 0.37462152491054224,
+      "grad_norm": 1.6141987353417335,
+      "learning_rate": 1.4391544356291753e-05,
+      "loss": 1.1115,
+      "step": 2722
+    },
+    {
+      "epoch": 0.37475915221579964,
+      "grad_norm": 1.6274481251911723,
+      "learning_rate": 1.4387539318501347e-05,
+      "loss": 1.0239,
+      "step": 2723
+    },
+    {
+      "epoch": 0.374896779521057,
+      "grad_norm": 1.831046792432533,
+      "learning_rate": 1.4383533408966934e-05,
+      "loss": 1.1183,
+      "step": 2724
+    },
+    {
+      "epoch": 0.37503440682631434,
+      "grad_norm": 1.5506975667502614,
+      "learning_rate": 1.4379526628484437e-05,
+      "loss": 1.0708,
+      "step": 2725
+    },
+    {
+      "epoch": 0.3751720341315717,
+      "grad_norm": 1.746287815930841,
+      "learning_rate": 1.4375518977849946e-05,
+      "loss": 0.9987,
+      "step": 2726
+    },
+    {
+      "epoch": 0.3753096614368291,
+      "grad_norm": 1.829548509442392,
+      "learning_rate": 1.4371510457859731e-05,
+      "loss": 1.1004,
+      "step": 2727
+    },
+    {
+      "epoch": 0.37544728874208644,
+      "grad_norm": 1.6795355007672779,
+      "learning_rate": 1.4367501069310228e-05,
+      "loss": 1.1069,
+      "step": 2728
+    },
+    {
+      "epoch": 0.3755849160473438,
+      "grad_norm": 1.7684451442028153,
+      "learning_rate": 1.4363490812998047e-05,
+      "loss": 1.0346,
+      "step": 2729
+    },
+    {
+      "epoch": 0.37572254335260113,
+      "grad_norm": 2.003673527078075,
+      "learning_rate": 1.435947968971997e-05,
+      "loss": 1.0333,
+      "step": 2730
+    },
+    {
+      "epoch": 0.37586017065785854,
+      "grad_norm": 1.7532826668826607,
+      "learning_rate": 1.4355467700272954e-05,
+      "loss": 1.0591,
+      "step": 2731
+    },
+    {
+      "epoch": 0.3759977979631159,
+      "grad_norm": 1.898964115619472,
+      "learning_rate": 1.4351454845454124e-05,
+      "loss": 1.1188,
+      "step": 2732
+    },
+    {
+      "epoch": 0.37613542526837324,
+      "grad_norm": 1.6873958809582708,
+      "learning_rate": 1.4347441126060781e-05,
+      "loss": 1.1205,
+      "step": 2733
+    },
+    {
+      "epoch": 0.3762730525736306,
+      "grad_norm": 1.8417897535980432,
+      "learning_rate": 1.4343426542890399e-05,
+      "loss": 1.043,
+      "step": 2734
+    },
+    {
+      "epoch": 0.376410679878888,
+      "grad_norm": 1.737020769548892,
+      "learning_rate": 1.4339411096740616e-05,
+      "loss": 1.0801,
+      "step": 2735
+    },
+    {
+      "epoch": 0.37654830718414534,
+      "grad_norm": 1.8300389670846584,
+      "learning_rate": 1.433539478840925e-05,
+      "loss": 1.1405,
+      "step": 2736
+    },
+    {
+      "epoch": 0.3766859344894027,
+      "grad_norm": 1.7352009229303242,
+      "learning_rate": 1.4331377618694285e-05,
+      "loss": 1.0375,
+      "step": 2737
+    },
+    {
+      "epoch": 0.37682356179466003,
+      "grad_norm": 1.6667345961905546,
+      "learning_rate": 1.4327359588393873e-05,
+      "loss": 1.0458,
+      "step": 2738
+    },
+    {
+      "epoch": 0.37696118909991744,
+      "grad_norm": 1.7307121135951675,
+      "learning_rate": 1.4323340698306352e-05,
+      "loss": 1.0267,
+      "step": 2739
+    },
+    {
+      "epoch": 0.3770988164051748,
+      "grad_norm": 1.7501165795674134,
+      "learning_rate": 1.4319320949230208e-05,
+      "loss": 1.0677,
+      "step": 2740
+    },
+    {
+      "epoch": 0.37723644371043213,
+      "grad_norm": 1.690952823609567,
+      "learning_rate": 1.4315300341964123e-05,
+      "loss": 1.0815,
+      "step": 2741
+    },
+    {
+      "epoch": 0.37737407101568954,
+      "grad_norm": 1.6648265684785413,
+      "learning_rate": 1.4311278877306924e-05,
+      "loss": 1.0714,
+      "step": 2742
+    },
+    {
+      "epoch": 0.3775116983209469,
+      "grad_norm": 1.6880798781078672,
+      "learning_rate": 1.4307256556057632e-05,
+      "loss": 1.0125,
+      "step": 2743
+    },
+    {
+      "epoch": 0.37764932562620424,
+      "grad_norm": 1.7954920754417787,
+      "learning_rate": 1.430323337901542e-05,
+      "loss": 0.9859,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3777869529314616,
+      "grad_norm": 1.480765059646714,
+      "learning_rate": 1.4299209346979639e-05,
+      "loss": 1.0889,
+      "step": 2745
+    },
+    {
+      "epoch": 0.377924580236719,
+      "grad_norm": 1.7492060848309057,
+      "learning_rate": 1.4295184460749814e-05,
+      "loss": 1.0927,
+      "step": 2746
+    },
+    {
+      "epoch": 0.37806220754197634,
+      "grad_norm": 1.5518978766837375,
+      "learning_rate": 1.4291158721125626e-05,
+      "loss": 1.1561,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3781998348472337,
+      "grad_norm": 1.7417085673571024,
+      "learning_rate": 1.4287132128906942e-05,
+      "loss": 1.0612,
+      "step": 2748
+    },
+    {
+      "epoch": 0.37833746215249103,
+      "grad_norm": 1.7318995094554988,
+      "learning_rate": 1.428310468489379e-05,
+      "loss": 1.0131,
+      "step": 2749
+    },
+    {
+      "epoch": 0.37847508945774844,
+      "grad_norm": 1.6985233594886642,
+      "learning_rate": 1.4279076389886367e-05,
+      "loss": 1.0654,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3786127167630058,
+      "grad_norm": 1.8428072274348712,
+      "learning_rate": 1.4275047244685038e-05,
+      "loss": 1.0293,
+      "step": 2751
+    },
+    {
+      "epoch": 0.37875034406826313,
+      "grad_norm": 1.7001588773882064,
+      "learning_rate": 1.427101725009034e-05,
+      "loss": 1.0942,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3788879713735205,
+      "grad_norm": 1.608190275162232,
+      "learning_rate": 1.4266986406902979e-05,
+      "loss": 1.0662,
+      "step": 2753
+    },
+    {
+      "epoch": 0.3790255986787779,
+      "grad_norm": 1.8930520339351489,
+      "learning_rate": 1.4262954715923827e-05,
+      "loss": 1.1336,
+      "step": 2754
+    },
+    {
+      "epoch": 0.37916322598403523,
+      "grad_norm": 1.7215958220729648,
+      "learning_rate": 1.4258922177953926e-05,
+      "loss": 1.0686,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3793008532892926,
+      "grad_norm": 1.5098061738235085,
+      "learning_rate": 1.425488879379449e-05,
+      "loss": 1.023,
+      "step": 2756
+    },
+    {
+      "epoch": 0.37943848059454993,
+      "grad_norm": 1.8727323405998264,
+      "learning_rate": 1.425085456424689e-05,
+      "loss": 1.1194,
+      "step": 2757
+    },
+    {
+      "epoch": 0.37957610789980734,
+      "grad_norm": 1.4334432737626632,
+      "learning_rate": 1.4246819490112678e-05,
+      "loss": 1.0426,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3797137352050647,
+      "grad_norm": 1.8418156007421456,
+      "learning_rate": 1.4242783572193565e-05,
+      "loss": 1.083,
+      "step": 2759
+    },
+    {
+      "epoch": 0.37985136251032203,
+      "grad_norm": 1.7126767918685382,
+      "learning_rate": 1.4238746811291435e-05,
+      "loss": 1.1568,
+      "step": 2760
+    },
+    {
+      "epoch": 0.37998898981557944,
+      "grad_norm": 1.630191889150529,
+      "learning_rate": 1.4234709208208334e-05,
+      "loss": 1.058,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3801266171208368,
+      "grad_norm": 1.8541846666799495,
+      "learning_rate": 1.4230670763746482e-05,
+      "loss": 1.1455,
+      "step": 2762
+    },
+    {
+      "epoch": 0.38026424442609413,
+      "grad_norm": 1.4366581174352853,
+      "learning_rate": 1.422663147870826e-05,
+      "loss": 1.1075,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3804018717313515,
+      "grad_norm": 1.6530930004666824,
+      "learning_rate": 1.4222591353896215e-05,
+      "loss": 1.0199,
+      "step": 2764
+    },
+    {
+      "epoch": 0.3805394990366089,
+      "grad_norm": 1.825800254316283,
+      "learning_rate": 1.4218550390113075e-05,
+      "loss": 1.0724,
+      "step": 2765
+    },
+    {
+      "epoch": 0.38067712634186623,
+      "grad_norm": 1.5661996117201764,
+      "learning_rate": 1.4214508588161716e-05,
+      "loss": 1.0288,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3808147536471236,
+      "grad_norm": 1.7386178695867043,
+      "learning_rate": 1.4210465948845191e-05,
+      "loss": 1.1111,
+      "step": 2767
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 1.787246169988825,
+      "learning_rate": 1.4206422472966716e-05,
+      "loss": 1.0421,
+      "step": 2768
+    },
+    {
+      "epoch": 0.38109000825763834,
+      "grad_norm": 1.8136275569864353,
+      "learning_rate": 1.4202378161329678e-05,
+      "loss": 1.0939,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3812276355628957,
+      "grad_norm": 1.6482131482543099,
+      "learning_rate": 1.4198333014737619e-05,
+      "loss": 1.1085,
+      "step": 2770
+    },
+    {
+      "epoch": 0.38136526286815303,
+      "grad_norm": 1.9265813662398645,
+      "learning_rate": 1.4194287033994263e-05,
+      "loss": 1.0522,
+      "step": 2771
+    },
+    {
+      "epoch": 0.3815028901734104,
+      "grad_norm": 1.5519420199826888,
+      "learning_rate": 1.4190240219903486e-05,
+      "loss": 1.0724,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3816405174786678,
+      "grad_norm": 1.5263635566223055,
+      "learning_rate": 1.4186192573269335e-05,
+      "loss": 1.0685,
+      "step": 2773
+    },
+    {
+      "epoch": 0.38177814478392513,
+      "grad_norm": 1.4515778113407662,
+      "learning_rate": 1.4182144094896021e-05,
+      "loss": 1.0317,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3819157720891825,
+      "grad_norm": 1.7729726555936651,
+      "learning_rate": 1.4178094785587926e-05,
+      "loss": 1.0744,
+      "step": 2775
+    },
+    {
+      "epoch": 0.38205339939443983,
+      "grad_norm": 1.6459140430801447,
+      "learning_rate": 1.4174044646149587e-05,
+      "loss": 1.0432,
+      "step": 2776
+    },
+    {
+      "epoch": 0.38219102669969723,
+      "grad_norm": 1.6442565067801325,
+      "learning_rate": 1.4169993677385713e-05,
+      "loss": 1.0059,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3823286540049546,
+      "grad_norm": 1.5137037647737381,
+      "learning_rate": 1.4165941880101178e-05,
+      "loss": 1.0244,
+      "step": 2778
+    },
+    {
+      "epoch": 0.38246628131021193,
+      "grad_norm": 1.63695155077094,
+      "learning_rate": 1.4161889255101012e-05,
+      "loss": 1.0324,
+      "step": 2779
+    },
+    {
+      "epoch": 0.38260390861546933,
+      "grad_norm": 1.6789451655603265,
+      "learning_rate": 1.4157835803190425e-05,
+      "loss": 1.0423,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3827415359207267,
+      "grad_norm": 1.858996737574047,
+      "learning_rate": 1.4153781525174779e-05,
+      "loss": 1.1771,
+      "step": 2781
+    },
+    {
+      "epoch": 0.38287916322598403,
+      "grad_norm": 1.677502836194795,
+      "learning_rate": 1.4149726421859601e-05,
+      "loss": 1.0935,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3830167905312414,
+      "grad_norm": 1.663905831881494,
+      "learning_rate": 1.4145670494050588e-05,
+      "loss": 1.1535,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3831544178364988,
+      "grad_norm": 1.6475914981956006,
+      "learning_rate": 1.4141613742553597e-05,
+      "loss": 1.1272,
+      "step": 2784
+    },
+    {
+      "epoch": 0.38329204514175613,
+      "grad_norm": 1.5985873426894874,
+      "learning_rate": 1.4137556168174646e-05,
+      "loss": 1.0963,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3834296724470135,
+      "grad_norm": 1.5856430519814781,
+      "learning_rate": 1.4133497771719922e-05,
+      "loss": 1.02,
+      "step": 2786
+    },
+    {
+      "epoch": 0.38356729975227083,
+      "grad_norm": 1.6160498941334656,
+      "learning_rate": 1.4129438553995771e-05,
+      "loss": 1.06,
+      "step": 2787
+    },
+    {
+      "epoch": 0.38370492705752823,
+      "grad_norm": 1.659003100708784,
+      "learning_rate": 1.4125378515808706e-05,
+      "loss": 1.025,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3838425543627856,
+      "grad_norm": 1.5017961570536498,
+      "learning_rate": 1.4121317657965401e-05,
+      "loss": 1.0426,
+      "step": 2789
+    },
+    {
+      "epoch": 0.38398018166804293,
+      "grad_norm": 1.684299096133013,
+      "learning_rate": 1.4117255981272694e-05,
+      "loss": 1.0491,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3841178089733003,
+      "grad_norm": 1.708681892320337,
+      "learning_rate": 1.4113193486537578e-05,
+      "loss": 0.993,
+      "step": 2791
+    },
+    {
+      "epoch": 0.3842554362785577,
+      "grad_norm": 1.748382927508248,
+      "learning_rate": 1.4109130174567223e-05,
+      "loss": 1.0647,
+      "step": 2792
+    },
+    {
+      "epoch": 0.38439306358381503,
+      "grad_norm": 1.7476688161576805,
+      "learning_rate": 1.4105066046168947e-05,
+      "loss": 1.1161,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3845306908890724,
+      "grad_norm": 1.6827361516617716,
+      "learning_rate": 1.4101001102150243e-05,
+      "loss": 1.091,
+      "step": 2794
+    },
+    {
+      "epoch": 0.38466831819432973,
+      "grad_norm": 1.9505138657357013,
+      "learning_rate": 1.4096935343318752e-05,
+      "loss": 0.9877,
+      "step": 2795
+    },
+    {
+      "epoch": 0.38480594549958713,
+      "grad_norm": 1.702925962477622,
+      "learning_rate": 1.4092868770482293e-05,
+      "loss": 1.0363,
+      "step": 2796
+    },
+    {
+      "epoch": 0.3849435728048445,
+      "grad_norm": 1.8165739328261106,
+      "learning_rate": 1.4088801384448836e-05,
+      "loss": 0.9556,
+      "step": 2797
+    },
+    {
+      "epoch": 0.38508120011010183,
+      "grad_norm": 1.5071955945288082,
+      "learning_rate": 1.4084733186026512e-05,
+      "loss": 1.0316,
+      "step": 2798
+    },
+    {
+      "epoch": 0.38521882741535923,
+      "grad_norm": 1.798564467955146,
+      "learning_rate": 1.4080664176023615e-05,
+      "loss": 1.0828,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3853564547206166,
+      "grad_norm": 1.552309023558474,
+      "learning_rate": 1.407659435524861e-05,
+      "loss": 1.0185,
+      "step": 2800
+    },
+    {
+      "epoch": 0.38549408202587393,
+      "grad_norm": 1.6442194645085841,
+      "learning_rate": 1.4072523724510103e-05,
+      "loss": 1.0636,
+      "step": 2801
+    },
+    {
+      "epoch": 0.3856317093311313,
+      "grad_norm": 1.7542165789738038,
+      "learning_rate": 1.4068452284616883e-05,
+      "loss": 1.0829,
+      "step": 2802
+    },
+    {
+      "epoch": 0.3857693366363887,
+      "grad_norm": 1.4896378073129424,
+      "learning_rate": 1.4064380036377886e-05,
+      "loss": 1.0564,
+      "step": 2803
+    },
+    {
+      "epoch": 0.38590696394164603,
+      "grad_norm": 2.158109244246876,
+      "learning_rate": 1.4060306980602206e-05,
+      "loss": 1.0862,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3860445912469034,
+      "grad_norm": 1.673310575139339,
+      "learning_rate": 1.4056233118099112e-05,
+      "loss": 1.0747,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3861822185521607,
+      "grad_norm": 1.798520536664166,
+      "learning_rate": 1.4052158449678022e-05,
+      "loss": 1.0931,
+      "step": 2806
+    },
+    {
+      "epoch": 0.38631984585741813,
+      "grad_norm": 1.6241185375328566,
+      "learning_rate": 1.4048082976148513e-05,
+      "loss": 1.066,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3864574731626755,
+      "grad_norm": 1.8280721655476513,
+      "learning_rate": 1.4044006698320331e-05,
+      "loss": 1.0475,
+      "step": 2808
+    },
+    {
+      "epoch": 0.38659510046793283,
+      "grad_norm": 1.9089347048420024,
+      "learning_rate": 1.4039929617003373e-05,
+      "loss": 1.1459,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3867327277731902,
+      "grad_norm": 1.7673652037219538,
+      "learning_rate": 1.4035851733007704e-05,
+      "loss": 1.0582,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3868703550784476,
+      "grad_norm": 1.8197453011808504,
+      "learning_rate": 1.4031773047143536e-05,
+      "loss": 1.038,
+      "step": 2811
+    },
+    {
+      "epoch": 0.38700798238370493,
+      "grad_norm": 1.6569567074615181,
+      "learning_rate": 1.4027693560221254e-05,
+      "loss": 1.0937,
+      "step": 2812
+    },
+    {
+      "epoch": 0.3871456096889623,
+      "grad_norm": 1.5368502764739944,
+      "learning_rate": 1.4023613273051395e-05,
+      "loss": 1.0425,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3872832369942196,
+      "grad_norm": 1.8066807303576407,
+      "learning_rate": 1.4019532186444658e-05,
+      "loss": 1.1182,
+      "step": 2814
+    },
+    {
+      "epoch": 0.38742086429947703,
+      "grad_norm": 1.6474487668411018,
+      "learning_rate": 1.4015450301211895e-05,
+      "loss": 0.9853,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3875584916047344,
+      "grad_norm": 1.777816044740927,
+      "learning_rate": 1.4011367618164122e-05,
+      "loss": 1.0731,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3876961189099917,
+      "grad_norm": 1.62206762910681,
+      "learning_rate": 1.4007284138112513e-05,
+      "loss": 1.1097,
+      "step": 2817
+    },
+    {
+      "epoch": 0.38783374621524913,
+      "grad_norm": 1.7892548416883558,
+      "learning_rate": 1.4003199861868401e-05,
+      "loss": 1.1401,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3879713735205065,
+      "grad_norm": 1.7139921662755788,
+      "learning_rate": 1.3999114790243274e-05,
+      "loss": 0.9782,
+      "step": 2819
+    },
+    {
+      "epoch": 0.38810900082576383,
+      "grad_norm": 1.718055842959684,
+      "learning_rate": 1.3995028924048777e-05,
+      "loss": 1.0256,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3882466281310212,
+      "grad_norm": 1.720717608258826,
+      "learning_rate": 1.3990942264096725e-05,
+      "loss": 1.0292,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3883842554362786,
+      "grad_norm": 1.7284268173774238,
+      "learning_rate": 1.3986854811199072e-05,
+      "loss": 1.0631,
+      "step": 2822
+    },
+    {
+      "epoch": 0.38852188274153593,
+      "grad_norm": 1.95007539523667,
+      "learning_rate": 1.398276656616794e-05,
+      "loss": 1.0337,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3886595100467933,
+      "grad_norm": 1.9907854529267541,
+      "learning_rate": 1.3978677529815611e-05,
+      "loss": 1.1622,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3887971373520506,
+      "grad_norm": 1.5345290874743542,
+      "learning_rate": 1.3974587702954521e-05,
+      "loss": 1.0629,
+      "step": 2825
+    },
+    {
+      "epoch": 0.38893476465730803,
+      "grad_norm": 1.7447338064578646,
+      "learning_rate": 1.3970497086397257e-05,
+      "loss": 1.1048,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3890723919625654,
+      "grad_norm": 1.7753211043368142,
+      "learning_rate": 1.3966405680956574e-05,
+      "loss": 1.1087,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3892100192678227,
+      "grad_norm": 1.9171438986714817,
+      "learning_rate": 1.3962313487445376e-05,
+      "loss": 1.0227,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3893476465730801,
+      "grad_norm": 1.8582329931354122,
+      "learning_rate": 1.3958220506676724e-05,
+      "loss": 1.0738,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3894852738783375,
+      "grad_norm": 1.8762924712824558,
+      "learning_rate": 1.3954126739463844e-05,
+      "loss": 1.0044,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3896229011835948,
+      "grad_norm": 1.6099223896567645,
+      "learning_rate": 1.3950032186620104e-05,
+      "loss": 1.0781,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3897605284888522,
+      "grad_norm": 1.7431850336065398,
+      "learning_rate": 1.394593684895904e-05,
+      "loss": 1.0416,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3898981557941095,
+      "grad_norm": 1.906355352269881,
+      "learning_rate": 1.3941840727294337e-05,
+      "loss": 1.0314,
+      "step": 2833
+    },
+    {
+      "epoch": 0.39003578309936693,
+      "grad_norm": 1.8631715242906592,
+      "learning_rate": 1.3937743822439842e-05,
+      "loss": 1.0864,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3901734104046243,
+      "grad_norm": 1.533966672051595,
+      "learning_rate": 1.3933646135209546e-05,
+      "loss": 1.0437,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3903110377098816,
+      "grad_norm": 1.68284061308331,
+      "learning_rate": 1.3929547666417613e-05,
+      "loss": 0.9868,
+      "step": 2836
+    },
+    {
+      "epoch": 0.39044866501513903,
+      "grad_norm": 1.8823601913690837,
+      "learning_rate": 1.392544841687835e-05,
+      "loss": 1.1677,
+      "step": 2837
+    },
+    {
+      "epoch": 0.3905862923203964,
+      "grad_norm": 1.8206803882110318,
+      "learning_rate": 1.3921348387406215e-05,
+      "loss": 1.1914,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3907239196256537,
+      "grad_norm": 1.7498191749786491,
+      "learning_rate": 1.3917247578815836e-05,
+      "loss": 0.9928,
+      "step": 2839
+    },
+    {
+      "epoch": 0.3908615469309111,
+      "grad_norm": 1.6427953986338186,
+      "learning_rate": 1.3913145991921984e-05,
+      "loss": 1.0799,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3909991742361685,
+      "grad_norm": 1.798814535947167,
+      "learning_rate": 1.390904362753959e-05,
+      "loss": 1.0663,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3911368015414258,
+      "grad_norm": 1.7844607398746606,
+      "learning_rate": 1.3904940486483731e-05,
+      "loss": 1.0661,
+      "step": 2842
+    },
+    {
+      "epoch": 0.3912744288466832,
+      "grad_norm": 1.5503262467525056,
+      "learning_rate": 1.3900836569569654e-05,
+      "loss": 1.0856,
+      "step": 2843
+    },
+    {
+      "epoch": 0.3914120561519405,
+      "grad_norm": 1.7220886086413048,
+      "learning_rate": 1.3896731877612746e-05,
+      "loss": 1.1131,
+      "step": 2844
+    },
+    {
+      "epoch": 0.39154968345719793,
+      "grad_norm": 1.7428470809137904,
+      "learning_rate": 1.3892626411428554e-05,
+      "loss": 1.0326,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3916873107624553,
+      "grad_norm": 1.699844517017266,
+      "learning_rate": 1.388852017183278e-05,
+      "loss": 1.1668,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3918249380677126,
+      "grad_norm": 1.7018204503426606,
+      "learning_rate": 1.3884413159641276e-05,
+      "loss": 0.9984,
+      "step": 2847
+    },
+    {
+      "epoch": 0.39196256537297,
+      "grad_norm": 1.5879125555569387,
+      "learning_rate": 1.388030537567005e-05,
+      "loss": 1.0915,
+      "step": 2848
+    },
+    {
+      "epoch": 0.3921001926782274,
+      "grad_norm": 1.7598326042216341,
+      "learning_rate": 1.3876196820735258e-05,
+      "loss": 1.1225,
+      "step": 2849
+    },
+    {
+      "epoch": 0.3922378199834847,
+      "grad_norm": 1.8430338253447007,
+      "learning_rate": 1.3872087495653216e-05,
+      "loss": 0.9922,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3923754472887421,
+      "grad_norm": 1.6795102267220028,
+      "learning_rate": 1.3867977401240395e-05,
+      "loss": 0.9937,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3925130745939994,
+      "grad_norm": 1.6419010654053106,
+      "learning_rate": 1.386386653831341e-05,
+      "loss": 1.0787,
+      "step": 2852
+    },
+    {
+      "epoch": 0.3926507018992568,
+      "grad_norm": 1.8449651725340714,
+      "learning_rate": 1.3859754907689037e-05,
+      "loss": 1.1243,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3927883292045142,
+      "grad_norm": 1.8227072709007053,
+      "learning_rate": 1.3855642510184197e-05,
+      "loss": 1.1647,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3929259565097715,
+      "grad_norm": 1.6951511136911261,
+      "learning_rate": 1.3851529346615969e-05,
+      "loss": 1.1153,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3930635838150289,
+      "grad_norm": 1.6486258556385747,
+      "learning_rate": 1.384741541780158e-05,
+      "loss": 1.0801,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3932012111202863,
+      "grad_norm": 1.5608581347387593,
+      "learning_rate": 1.384330072455841e-05,
+      "loss": 1.0506,
+      "step": 2857
+    },
+    {
+      "epoch": 0.3933388384255436,
+      "grad_norm": 1.5342299328141695,
+      "learning_rate": 1.3839185267704e-05,
+      "loss": 1.186,
+      "step": 2858
+    },
+    {
+      "epoch": 0.393476465730801,
+      "grad_norm": 1.883069411284481,
+      "learning_rate": 1.3835069048056027e-05,
+      "loss": 1.092,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3936140930360584,
+      "grad_norm": 1.6602338518165858,
+      "learning_rate": 1.383095206643233e-05,
+      "loss": 1.0748,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3937517203413157,
+      "grad_norm": 1.6679655508122826,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 1.0826,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3938893476465731,
+      "grad_norm": 1.5221761686812818,
+      "learning_rate": 1.382271582052987e-05,
+      "loss": 1.0586,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3940269749518304,
+      "grad_norm": 1.6676628833365728,
+      "learning_rate": 1.3818596557887538e-05,
+      "loss": 1.0066,
+      "step": 2863
+    },
+    {
+      "epoch": 0.3941646022570878,
+      "grad_norm": 1.7534234475718389,
+      "learning_rate": 1.3814476536542338e-05,
+      "loss": 1.0889,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3943022295623452,
+      "grad_norm": 1.7149844273103154,
+      "learning_rate": 1.3810355757312867e-05,
+      "loss": 1.1446,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3944398568676025,
+      "grad_norm": 1.7858387283062693,
+      "learning_rate": 1.3806234221017862e-05,
+      "loss": 1.1562,
+      "step": 2866
+    },
+    {
+      "epoch": 0.39457748417285987,
+      "grad_norm": 1.5342536894772782,
+      "learning_rate": 1.380211192847622e-05,
+      "loss": 0.9946,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3947151114781173,
+      "grad_norm": 1.7411735606762555,
+      "learning_rate": 1.3797988880506986e-05,
+      "loss": 1.1148,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3948527387833746,
+      "grad_norm": 1.6039530433622324,
+      "learning_rate": 1.3793865077929346e-05,
+      "loss": 1.0024,
+      "step": 2869
+    },
+    {
+      "epoch": 0.394990366088632,
+      "grad_norm": 1.6544798462390056,
+      "learning_rate": 1.3789740521562652e-05,
+      "loss": 1.0579,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3951279933938893,
+      "grad_norm": 1.7570593401247045,
+      "learning_rate": 1.3785615212226389e-05,
+      "loss": 1.0591,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3952656206991467,
+      "grad_norm": 1.6666572847355376,
+      "learning_rate": 1.3781489150740204e-05,
+      "loss": 1.1048,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3954032480044041,
+      "grad_norm": 1.7827409850274374,
+      "learning_rate": 1.3777362337923885e-05,
+      "loss": 1.165,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3955408753096614,
+      "grad_norm": 1.545152690941674,
+      "learning_rate": 1.377323477459738e-05,
+      "loss": 0.9919,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3956785026149188,
+      "grad_norm": 1.7048045676093664,
+      "learning_rate": 1.3769106461580772e-05,
+      "loss": 1.0002,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3958161299201762,
+      "grad_norm": 1.814804636308928,
+      "learning_rate": 1.3764977399694304e-05,
+      "loss": 1.094,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3959537572254335,
+      "grad_norm": 1.6752898008018287,
+      "learning_rate": 1.3760847589758366e-05,
+      "loss": 1.0587,
+      "step": 2877
+    },
+    {
+      "epoch": 0.39609138453069087,
+      "grad_norm": 1.5721318119200358,
+      "learning_rate": 1.3756717032593489e-05,
+      "loss": 1.0333,
+      "step": 2878
+    },
+    {
+      "epoch": 0.3962290118359483,
+      "grad_norm": 1.6842544581612495,
+      "learning_rate": 1.3752585729020365e-05,
+      "loss": 1.0405,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3963666391412056,
+      "grad_norm": 1.5532006936194194,
+      "learning_rate": 1.3748453679859821e-05,
+      "loss": 1.0246,
+      "step": 2880
+    },
+    {
+      "epoch": 0.39650426644646297,
+      "grad_norm": 1.7696234744606163,
+      "learning_rate": 1.374432088593284e-05,
+      "loss": 0.9818,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3966418937517203,
+      "grad_norm": 1.7039161233958249,
+      "learning_rate": 1.3740187348060557e-05,
+      "loss": 1.0232,
+      "step": 2882
+    },
+    {
+      "epoch": 0.3967795210569777,
+      "grad_norm": 1.9230784395840668,
+      "learning_rate": 1.3736053067064246e-05,
+      "loss": 1.0808,
+      "step": 2883
+    },
+    {
+      "epoch": 0.3969171483622351,
+      "grad_norm": 1.651788627701754,
+      "learning_rate": 1.3731918043765329e-05,
+      "loss": 1.1031,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3970547756674924,
+      "grad_norm": 1.8443374497330645,
+      "learning_rate": 1.3727782278985382e-05,
+      "loss": 1.0575,
+      "step": 2885
+    },
+    {
+      "epoch": 0.39719240297274977,
+      "grad_norm": 1.9267220972916554,
+      "learning_rate": 1.3723645773546125e-05,
+      "loss": 1.0319,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3973300302780072,
+      "grad_norm": 1.575605507767993,
+      "learning_rate": 1.3719508528269424e-05,
+      "loss": 1.1246,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3974676575832645,
+      "grad_norm": 1.7378645749534758,
+      "learning_rate": 1.3715370543977292e-05,
+      "loss": 1.1407,
+      "step": 2888
+    },
+    {
+      "epoch": 0.39760528488852187,
+      "grad_norm": 1.4076332525841564,
+      "learning_rate": 1.3711231821491888e-05,
+      "loss": 1.0175,
+      "step": 2889
+    },
+    {
+      "epoch": 0.3977429121937792,
+      "grad_norm": 1.7092797267752737,
+      "learning_rate": 1.3707092361635526e-05,
+      "loss": 1.0677,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3978805394990366,
+      "grad_norm": 1.794996705940687,
+      "learning_rate": 1.3702952165230654e-05,
+      "loss": 1.0169,
+      "step": 2891
+    },
+    {
+      "epoch": 0.39801816680429397,
+      "grad_norm": 2.1478551322446564,
+      "learning_rate": 1.3698811233099876e-05,
+      "loss": 1.0925,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3981557941095513,
+      "grad_norm": 1.7449298274190403,
+      "learning_rate": 1.3694669566065938e-05,
+      "loss": 1.1363,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3982934214148087,
+      "grad_norm": 1.5593449199054368,
+      "learning_rate": 1.369052716495173e-05,
+      "loss": 1.1202,
+      "step": 2894
+    },
+    {
+      "epoch": 0.39843104872006607,
+      "grad_norm": 1.7208057786252413,
+      "learning_rate": 1.3686384030580294e-05,
+      "loss": 1.0608,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3985686760253234,
+      "grad_norm": 1.6013761813703682,
+      "learning_rate": 1.3682240163774812e-05,
+      "loss": 1.0052,
+      "step": 2896
+    },
+    {
+      "epoch": 0.39870630333058077,
+      "grad_norm": 1.7241850077159837,
+      "learning_rate": 1.367809556535861e-05,
+      "loss": 1.1633,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3988439306358382,
+      "grad_norm": 1.610377276405802,
+      "learning_rate": 1.367395023615517e-05,
+      "loss": 1.0717,
+      "step": 2898
+    },
+    {
+      "epoch": 0.3989815579410955,
+      "grad_norm": 1.7386583866651428,
+      "learning_rate": 1.366980417698811e-05,
+      "loss": 1.0244,
+      "step": 2899
+    },
+    {
+      "epoch": 0.39911918524635287,
+      "grad_norm": 1.6193890920420864,
+      "learning_rate": 1.3665657388681193e-05,
+      "loss": 1.0505,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3992568125516102,
+      "grad_norm": 1.7538389364710647,
+      "learning_rate": 1.366150987205833e-05,
+      "loss": 1.0679,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3993944398568676,
+      "grad_norm": 1.8223811632920925,
+      "learning_rate": 1.3657361627943577e-05,
+      "loss": 1.0505,
+      "step": 2902
+    },
+    {
+      "epoch": 0.39953206716212497,
+      "grad_norm": 1.8029120655484407,
+      "learning_rate": 1.3653212657161132e-05,
+      "loss": 1.0221,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3996696944673823,
+      "grad_norm": 1.6313830854432692,
+      "learning_rate": 1.3649062960535337e-05,
+      "loss": 1.0719,
+      "step": 2904
+    },
+    {
+      "epoch": 0.39980732177263967,
+      "grad_norm": 1.7618613597229262,
+      "learning_rate": 1.3644912538890684e-05,
+      "loss": 1.0888,
+      "step": 2905
+    },
+    {
+      "epoch": 0.39994494907789707,
+      "grad_norm": 1.9273894032861831,
+      "learning_rate": 1.3640761393051802e-05,
+      "loss": 1.091,
+      "step": 2906
+    },
+    {
+      "epoch": 0.4000825763831544,
+      "grad_norm": 1.7302772703895435,
+      "learning_rate": 1.363660952384347e-05,
+      "loss": 1.0516,
+      "step": 2907
+    },
+    {
+      "epoch": 0.40022020368841177,
+      "grad_norm": 1.8374161141176648,
+      "learning_rate": 1.3632456932090603e-05,
+      "loss": 1.126,
+      "step": 2908
+    },
+    {
+      "epoch": 0.4003578309936691,
+      "grad_norm": 1.7679970121187711,
+      "learning_rate": 1.3628303618618268e-05,
+      "loss": 1.0424,
+      "step": 2909
+    },
+    {
+      "epoch": 0.4004954582989265,
+      "grad_norm": 1.6725689106453228,
+      "learning_rate": 1.362414958425167e-05,
+      "loss": 1.0469,
+      "step": 2910
+    },
+    {
+      "epoch": 0.40063308560418387,
+      "grad_norm": 1.4677857318855299,
+      "learning_rate": 1.3619994829816156e-05,
+      "loss": 1.0221,
+      "step": 2911
+    },
+    {
+      "epoch": 0.4007707129094412,
+      "grad_norm": 1.74044141967409,
+      "learning_rate": 1.3615839356137223e-05,
+      "loss": 1.0353,
+      "step": 2912
+    },
+    {
+      "epoch": 0.4009083402146986,
+      "grad_norm": 1.616495983270934,
+      "learning_rate": 1.3611683164040508e-05,
+      "loss": 1.0609,
+      "step": 2913
+    },
+    {
+      "epoch": 0.40104596751995597,
+      "grad_norm": 1.849905448626749,
+      "learning_rate": 1.3607526254351784e-05,
+      "loss": 1.0425,
+      "step": 2914
+    },
+    {
+      "epoch": 0.4011835948252133,
+      "grad_norm": 1.883323270016199,
+      "learning_rate": 1.3603368627896976e-05,
+      "loss": 1.076,
+      "step": 2915
+    },
+    {
+      "epoch": 0.40132122213047067,
+      "grad_norm": 1.7738991997171931,
+      "learning_rate": 1.3599210285502147e-05,
+      "loss": 1.1664,
+      "step": 2916
+    },
+    {
+      "epoch": 0.40145884943572807,
+      "grad_norm": 1.6770990968514194,
+      "learning_rate": 1.3595051227993501e-05,
+      "loss": 0.9878,
+      "step": 2917
+    },
+    {
+      "epoch": 0.4015964767409854,
+      "grad_norm": 1.663971931688104,
+      "learning_rate": 1.3590891456197388e-05,
+      "loss": 1.0427,
+      "step": 2918
+    },
+    {
+      "epoch": 0.40173410404624277,
+      "grad_norm": 1.6391955229400554,
+      "learning_rate": 1.3586730970940293e-05,
+      "loss": 1.0024,
+      "step": 2919
+    },
+    {
+      "epoch": 0.4018717313515001,
+      "grad_norm": 1.8829275289563066,
+      "learning_rate": 1.3582569773048854e-05,
+      "loss": 1.2202,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4020093586567575,
+      "grad_norm": 1.6346589300519356,
+      "learning_rate": 1.3578407863349842e-05,
+      "loss": 1.0416,
+      "step": 2921
+    },
+    {
+      "epoch": 0.40214698596201487,
+      "grad_norm": 1.708553678187109,
+      "learning_rate": 1.357424524267017e-05,
+      "loss": 1.0308,
+      "step": 2922
+    },
+    {
+      "epoch": 0.4022846132672722,
+      "grad_norm": 1.9109937814325682,
+      "learning_rate": 1.3570081911836894e-05,
+      "loss": 1.025,
+      "step": 2923
+    },
+    {
+      "epoch": 0.40242224057252957,
+      "grad_norm": 1.5382565204248728,
+      "learning_rate": 1.3565917871677211e-05,
+      "loss": 1.0553,
+      "step": 2924
+    },
+    {
+      "epoch": 0.40255986787778697,
+      "grad_norm": 1.8484275796550542,
+      "learning_rate": 1.3561753123018462e-05,
+      "loss": 1.0685,
+      "step": 2925
+    },
+    {
+      "epoch": 0.4026974951830443,
+      "grad_norm": 1.649688172587731,
+      "learning_rate": 1.355758766668812e-05,
+      "loss": 1.0792,
+      "step": 2926
+    },
+    {
+      "epoch": 0.40283512248830167,
+      "grad_norm": 1.6912593499396404,
+      "learning_rate": 1.3553421503513806e-05,
+      "loss": 1.0631,
+      "step": 2927
+    },
+    {
+      "epoch": 0.402972749793559,
+      "grad_norm": 1.67003605456862,
+      "learning_rate": 1.3549254634323279e-05,
+      "loss": 1.0499,
+      "step": 2928
+    },
+    {
+      "epoch": 0.4031103770988164,
+      "grad_norm": 1.6654101034054862,
+      "learning_rate": 1.3545087059944442e-05,
+      "loss": 1.016,
+      "step": 2929
+    },
+    {
+      "epoch": 0.40324800440407377,
+      "grad_norm": 1.592044707580495,
+      "learning_rate": 1.3540918781205331e-05,
+      "loss": 1.0278,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4033856317093311,
+      "grad_norm": 1.7772042715619683,
+      "learning_rate": 1.3536749798934132e-05,
+      "loss": 1.0066,
+      "step": 2931
+    },
+    {
+      "epoch": 0.4035232590145885,
+      "grad_norm": 1.717851486816849,
+      "learning_rate": 1.3532580113959159e-05,
+      "loss": 1.1452,
+      "step": 2932
+    },
+    {
+      "epoch": 0.40366088631984587,
+      "grad_norm": 1.5819134390681329,
+      "learning_rate": 1.3528409727108873e-05,
+      "loss": 1.0167,
+      "step": 2933
+    },
+    {
+      "epoch": 0.4037985136251032,
+      "grad_norm": 1.7277949395740602,
+      "learning_rate": 1.3524238639211874e-05,
+      "loss": 1.0547,
+      "step": 2934
+    },
+    {
+      "epoch": 0.40393614093036057,
+      "grad_norm": 1.6402884657368522,
+      "learning_rate": 1.3520066851096898e-05,
+      "loss": 1.0417,
+      "step": 2935
+    },
+    {
+      "epoch": 0.40407376823561797,
+      "grad_norm": 1.8412595152047828,
+      "learning_rate": 1.3515894363592824e-05,
+      "loss": 1.155,
+      "step": 2936
+    },
+    {
+      "epoch": 0.4042113955408753,
+      "grad_norm": 2.0261770614225227,
+      "learning_rate": 1.3511721177528665e-05,
+      "loss": 1.0873,
+      "step": 2937
+    },
+    {
+      "epoch": 0.40434902284613267,
+      "grad_norm": 1.6612690271088073,
+      "learning_rate": 1.350754729373358e-05,
+      "loss": 1.1014,
+      "step": 2938
+    },
+    {
+      "epoch": 0.40448665015139,
+      "grad_norm": 2.1248230024037045,
+      "learning_rate": 1.350337271303686e-05,
+      "loss": 1.1304,
+      "step": 2939
+    },
+    {
+      "epoch": 0.4046242774566474,
+      "grad_norm": 1.6384645908469302,
+      "learning_rate": 1.349919743626794e-05,
+      "loss": 0.9833,
+      "step": 2940
+    },
+    {
+      "epoch": 0.40476190476190477,
+      "grad_norm": 1.6490128907932469,
+      "learning_rate": 1.3495021464256386e-05,
+      "loss": 1.14,
+      "step": 2941
+    },
+    {
+      "epoch": 0.4048995320671621,
+      "grad_norm": 1.6955289787977417,
+      "learning_rate": 1.3490844797831909e-05,
+      "loss": 1.0314,
+      "step": 2942
+    },
+    {
+      "epoch": 0.40503715937241946,
+      "grad_norm": 1.6726967460242488,
+      "learning_rate": 1.3486667437824355e-05,
+      "loss": 1.1179,
+      "step": 2943
+    },
+    {
+      "epoch": 0.40517478667767687,
+      "grad_norm": 1.61266506903315,
+      "learning_rate": 1.3482489385063705e-05,
+      "loss": 1.0058,
+      "step": 2944
+    },
+    {
+      "epoch": 0.4053124139829342,
+      "grad_norm": 1.8697777580991963,
+      "learning_rate": 1.347831064038009e-05,
+      "loss": 1.048,
+      "step": 2945
+    },
+    {
+      "epoch": 0.40545004128819156,
+      "grad_norm": 1.7893113857754643,
+      "learning_rate": 1.347413120460376e-05,
+      "loss": 1.0407,
+      "step": 2946
+    },
+    {
+      "epoch": 0.4055876685934489,
+      "grad_norm": 1.8562431868668634,
+      "learning_rate": 1.3469951078565117e-05,
+      "loss": 1.1318,
+      "step": 2947
+    },
+    {
+      "epoch": 0.4057252958987063,
+      "grad_norm": 1.660447148156247,
+      "learning_rate": 1.3465770263094693e-05,
+      "loss": 1.093,
+      "step": 2948
+    },
+    {
+      "epoch": 0.40586292320396367,
+      "grad_norm": 1.4902625436286638,
+      "learning_rate": 1.3461588759023162e-05,
+      "loss": 0.9832,
+      "step": 2949
+    },
+    {
+      "epoch": 0.406000550509221,
+      "grad_norm": 1.7912142935157125,
+      "learning_rate": 1.3457406567181324e-05,
+      "loss": 1.0651,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4061381778144784,
+      "grad_norm": 1.8015793746845836,
+      "learning_rate": 1.345322368840013e-05,
+      "loss": 0.9973,
+      "step": 2951
+    },
+    {
+      "epoch": 0.40627580511973577,
+      "grad_norm": 1.5879596196966237,
+      "learning_rate": 1.3449040123510657e-05,
+      "loss": 0.9815,
+      "step": 2952
+    },
+    {
+      "epoch": 0.4064134324249931,
+      "grad_norm": 1.928342325399757,
+      "learning_rate": 1.3444855873344124e-05,
+      "loss": 1.0279,
+      "step": 2953
+    },
+    {
+      "epoch": 0.40655105973025046,
+      "grad_norm": 1.8576430868480196,
+      "learning_rate": 1.3440670938731885e-05,
+      "loss": 1.0443,
+      "step": 2954
+    },
+    {
+      "epoch": 0.40668868703550787,
+      "grad_norm": 1.7652417908105098,
+      "learning_rate": 1.3436485320505427e-05,
+      "loss": 1.0239,
+      "step": 2955
+    },
+    {
+      "epoch": 0.4068263143407652,
+      "grad_norm": 1.8326917126914215,
+      "learning_rate": 1.343229901949638e-05,
+      "loss": 0.999,
+      "step": 2956
+    },
+    {
+      "epoch": 0.40696394164602256,
+      "grad_norm": 1.8697678271772764,
+      "learning_rate": 1.3428112036536497e-05,
+      "loss": 1.0753,
+      "step": 2957
+    },
+    {
+      "epoch": 0.4071015689512799,
+      "grad_norm": 1.759726808773196,
+      "learning_rate": 1.342392437245768e-05,
+      "loss": 1.0102,
+      "step": 2958
+    },
+    {
+      "epoch": 0.4072391962565373,
+      "grad_norm": 1.821670304566112,
+      "learning_rate": 1.3419736028091954e-05,
+      "loss": 0.9508,
+      "step": 2959
+    },
+    {
+      "epoch": 0.40737682356179467,
+      "grad_norm": 1.548223703164834,
+      "learning_rate": 1.3415547004271495e-05,
+      "loss": 1.0168,
+      "step": 2960
+    },
+    {
+      "epoch": 0.407514450867052,
+      "grad_norm": 1.842970019560257,
+      "learning_rate": 1.3411357301828597e-05,
+      "loss": 1.1346,
+      "step": 2961
+    },
+    {
+      "epoch": 0.40765207817230936,
+      "grad_norm": 1.7018190367740607,
+      "learning_rate": 1.3407166921595702e-05,
+      "loss": 1.0758,
+      "step": 2962
+    },
+    {
+      "epoch": 0.40778970547756677,
+      "grad_norm": 1.8811613874178925,
+      "learning_rate": 1.3402975864405378e-05,
+      "loss": 1.0393,
+      "step": 2963
+    },
+    {
+      "epoch": 0.4079273327828241,
+      "grad_norm": 1.709992365270689,
+      "learning_rate": 1.3398784131090331e-05,
+      "loss": 1.0105,
+      "step": 2964
+    },
+    {
+      "epoch": 0.40806496008808146,
+      "grad_norm": 1.6666909188666281,
+      "learning_rate": 1.3394591722483402e-05,
+      "loss": 1.068,
+      "step": 2965
+    },
+    {
+      "epoch": 0.4082025873933388,
+      "grad_norm": 1.7007821737303124,
+      "learning_rate": 1.3390398639417565e-05,
+      "loss": 1.0446,
+      "step": 2966
+    },
+    {
+      "epoch": 0.4083402146985962,
+      "grad_norm": 1.6047127039838578,
+      "learning_rate": 1.3386204882725928e-05,
+      "loss": 1.1704,
+      "step": 2967
+    },
+    {
+      "epoch": 0.40847784200385356,
+      "grad_norm": 1.7739982716949938,
+      "learning_rate": 1.3382010453241732e-05,
+      "loss": 1.1318,
+      "step": 2968
+    },
+    {
+      "epoch": 0.4086154693091109,
+      "grad_norm": 1.765811288206812,
+      "learning_rate": 1.3377815351798353e-05,
+      "loss": 1.0588,
+      "step": 2969
+    },
+    {
+      "epoch": 0.4087530966143683,
+      "grad_norm": 1.9224393278730878,
+      "learning_rate": 1.3373619579229304e-05,
+      "loss": 1.0796,
+      "step": 2970
+    },
+    {
+      "epoch": 0.40889072391962566,
+      "grad_norm": 1.8729112994103922,
+      "learning_rate": 1.3369423136368223e-05,
+      "loss": 1.1182,
+      "step": 2971
+    },
+    {
+      "epoch": 0.409028351224883,
+      "grad_norm": 1.8584839263124386,
+      "learning_rate": 1.3365226024048885e-05,
+      "loss": 1.0673,
+      "step": 2972
+    },
+    {
+      "epoch": 0.40916597853014036,
+      "grad_norm": 1.6380778844088122,
+      "learning_rate": 1.3361028243105203e-05,
+      "loss": 1.0226,
+      "step": 2973
+    },
+    {
+      "epoch": 0.40930360583539777,
+      "grad_norm": 1.6704536266598462,
+      "learning_rate": 1.3356829794371215e-05,
+      "loss": 1.141,
+      "step": 2974
+    },
+    {
+      "epoch": 0.4094412331406551,
+      "grad_norm": 2.109994885415102,
+      "learning_rate": 1.3352630678681095e-05,
+      "loss": 0.9887,
+      "step": 2975
+    },
+    {
+      "epoch": 0.40957886044591246,
+      "grad_norm": 1.6412817649522309,
+      "learning_rate": 1.3348430896869155e-05,
+      "loss": 1.0638,
+      "step": 2976
+    },
+    {
+      "epoch": 0.4097164877511698,
+      "grad_norm": 1.6881577342044665,
+      "learning_rate": 1.334423044976983e-05,
+      "loss": 1.1079,
+      "step": 2977
+    },
+    {
+      "epoch": 0.4098541150564272,
+      "grad_norm": 1.8595392528409433,
+      "learning_rate": 1.3340029338217693e-05,
+      "loss": 1.138,
+      "step": 2978
+    },
+    {
+      "epoch": 0.40999174236168456,
+      "grad_norm": 1.7842628613493872,
+      "learning_rate": 1.3335827563047446e-05,
+      "loss": 1.1872,
+      "step": 2979
+    },
+    {
+      "epoch": 0.4101293696669419,
+      "grad_norm": 1.6875242484948774,
+      "learning_rate": 1.3331625125093925e-05,
+      "loss": 0.9986,
+      "step": 2980
+    },
+    {
+      "epoch": 0.41026699697219926,
+      "grad_norm": 1.6459299499613893,
+      "learning_rate": 1.3327422025192095e-05,
+      "loss": 1.0929,
+      "step": 2981
+    },
+    {
+      "epoch": 0.41040462427745666,
+      "grad_norm": 1.8001413612098554,
+      "learning_rate": 1.3323218264177058e-05,
+      "loss": 1.1045,
+      "step": 2982
+    },
+    {
+      "epoch": 0.410542251582714,
+      "grad_norm": 1.6681634669086285,
+      "learning_rate": 1.331901384288404e-05,
+      "loss": 1.0364,
+      "step": 2983
+    },
+    {
+      "epoch": 0.41067987888797136,
+      "grad_norm": 1.8334633338239053,
+      "learning_rate": 1.3314808762148406e-05,
+      "loss": 1.0853,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4108175061932287,
+      "grad_norm": 1.8698260083817937,
+      "learning_rate": 1.3310603022805646e-05,
+      "loss": 1.1129,
+      "step": 2985
+    },
+    {
+      "epoch": 0.4109551334984861,
+      "grad_norm": 1.7163843005138193,
+      "learning_rate": 1.330639662569138e-05,
+      "loss": 1.0077,
+      "step": 2986
+    },
+    {
+      "epoch": 0.41109276080374346,
+      "grad_norm": 1.6802320794184566,
+      "learning_rate": 1.3302189571641366e-05,
+      "loss": 1.0256,
+      "step": 2987
+    },
+    {
+      "epoch": 0.4112303881090008,
+      "grad_norm": 1.7918702046657091,
+      "learning_rate": 1.3297981861491485e-05,
+      "loss": 1.0599,
+      "step": 2988
+    },
+    {
+      "epoch": 0.4113680154142582,
+      "grad_norm": 1.602094742710218,
+      "learning_rate": 1.3293773496077753e-05,
+      "loss": 1.0624,
+      "step": 2989
+    },
+    {
+      "epoch": 0.41150564271951556,
+      "grad_norm": 1.7846450937860054,
+      "learning_rate": 1.3289564476236311e-05,
+      "loss": 1.0882,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4116432700247729,
+      "grad_norm": 1.689622120044213,
+      "learning_rate": 1.328535480280344e-05,
+      "loss": 1.1801,
+      "step": 2991
+    },
+    {
+      "epoch": 0.41178089733003026,
+      "grad_norm": 1.7913570278271993,
+      "learning_rate": 1.3281144476615544e-05,
+      "loss": 1.0672,
+      "step": 2992
+    },
+    {
+      "epoch": 0.41191852463528766,
+      "grad_norm": 1.804477441428126,
+      "learning_rate": 1.327693349850915e-05,
+      "loss": 1.0626,
+      "step": 2993
+    },
+    {
+      "epoch": 0.412056151940545,
+      "grad_norm": 1.7246027880223527,
+      "learning_rate": 1.3272721869320929e-05,
+      "loss": 1.06,
+      "step": 2994
+    },
+    {
+      "epoch": 0.41219377924580236,
+      "grad_norm": 1.6350279679985293,
+      "learning_rate": 1.3268509589887671e-05,
+      "loss": 1.0573,
+      "step": 2995
+    },
+    {
+      "epoch": 0.4123314065510597,
+      "grad_norm": 1.6410190286345985,
+      "learning_rate": 1.3264296661046296e-05,
+      "loss": 1.0321,
+      "step": 2996
+    },
+    {
+      "epoch": 0.4124690338563171,
+      "grad_norm": 1.630683855486899,
+      "learning_rate": 1.3260083083633859e-05,
+      "loss": 1.078,
+      "step": 2997
+    },
+    {
+      "epoch": 0.41260666116157446,
+      "grad_norm": 1.6009085224037036,
+      "learning_rate": 1.325586885848754e-05,
+      "loss": 1.0214,
+      "step": 2998
+    },
+    {
+      "epoch": 0.4127442884668318,
+      "grad_norm": 1.793651585201714,
+      "learning_rate": 1.3251653986444649e-05,
+      "loss": 1.0719,
+      "step": 2999
+    },
+    {
+      "epoch": 0.41288191577208916,
+      "grad_norm": 1.7008889364397932,
+      "learning_rate": 1.3247438468342618e-05,
+      "loss": 1.0707,
+      "step": 3000
+    },
+    {
+      "epoch": 0.41301954307734656,
+      "grad_norm": 1.7144088971110383,
+      "learning_rate": 1.3243222305019015e-05,
+      "loss": 1.0366,
+      "step": 3001
+    },
+    {
+      "epoch": 0.4131571703826039,
+      "grad_norm": 1.6882872602981118,
+      "learning_rate": 1.3239005497311536e-05,
+      "loss": 0.967,
+      "step": 3002
+    },
+    {
+      "epoch": 0.41329479768786126,
+      "grad_norm": 1.6745126043975482,
+      "learning_rate": 1.3234788046058002e-05,
+      "loss": 1.1176,
+      "step": 3003
+    },
+    {
+      "epoch": 0.4134324249931186,
+      "grad_norm": 1.7006847342767089,
+      "learning_rate": 1.3230569952096363e-05,
+      "loss": 1.1309,
+      "step": 3004
+    },
+    {
+      "epoch": 0.413570052298376,
+      "grad_norm": 1.6259542321981322,
+      "learning_rate": 1.3226351216264695e-05,
+      "loss": 1.0389,
+      "step": 3005
+    },
+    {
+      "epoch": 0.41370767960363336,
+      "grad_norm": 1.686149145164333,
+      "learning_rate": 1.3222131839401204e-05,
+      "loss": 1.0479,
+      "step": 3006
+    },
+    {
+      "epoch": 0.4138453069088907,
+      "grad_norm": 1.7500062757682815,
+      "learning_rate": 1.3217911822344225e-05,
+      "loss": 1.0127,
+      "step": 3007
+    },
+    {
+      "epoch": 0.4139829342141481,
+      "grad_norm": 1.644974802227951,
+      "learning_rate": 1.3213691165932212e-05,
+      "loss": 1.0784,
+      "step": 3008
+    },
+    {
+      "epoch": 0.41412056151940546,
+      "grad_norm": 1.9036043564665068,
+      "learning_rate": 1.3209469871003759e-05,
+      "loss": 1.0577,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4142581888246628,
+      "grad_norm": 1.9253326810196307,
+      "learning_rate": 1.3205247938397574e-05,
+      "loss": 1.1272,
+      "step": 3010
+    },
+    {
+      "epoch": 0.41439581612992016,
+      "grad_norm": 1.725321886211019,
+      "learning_rate": 1.3201025368952498e-05,
+      "loss": 1.0535,
+      "step": 3011
+    },
+    {
+      "epoch": 0.41453344343517756,
+      "grad_norm": 1.7146791929559169,
+      "learning_rate": 1.3196802163507499e-05,
+      "loss": 1.0353,
+      "step": 3012
+    },
+    {
+      "epoch": 0.4146710707404349,
+      "grad_norm": 1.5358563686175672,
+      "learning_rate": 1.319257832290167e-05,
+      "loss": 1.0667,
+      "step": 3013
+    },
+    {
+      "epoch": 0.41480869804569226,
+      "grad_norm": 1.6672023937682923,
+      "learning_rate": 1.3188353847974228e-05,
+      "loss": 1.0331,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4149463253509496,
+      "grad_norm": 2.399074475840285,
+      "learning_rate": 1.3184128739564522e-05,
+      "loss": 1.11,
+      "step": 3015
+    },
+    {
+      "epoch": 0.415083952656207,
+      "grad_norm": 1.6848279638269332,
+      "learning_rate": 1.3179902998512021e-05,
+      "loss": 1.0201,
+      "step": 3016
+    },
+    {
+      "epoch": 0.41522157996146436,
+      "grad_norm": 1.8561403327176902,
+      "learning_rate": 1.3175676625656322e-05,
+      "loss": 1.0567,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4153592072667217,
+      "grad_norm": 1.6315369802411375,
+      "learning_rate": 1.3171449621837149e-05,
+      "loss": 1.105,
+      "step": 3018
+    },
+    {
+      "epoch": 0.41549683457197906,
+      "grad_norm": 1.7978487799811917,
+      "learning_rate": 1.3167221987894348e-05,
+      "loss": 1.0138,
+      "step": 3019
+    },
+    {
+      "epoch": 0.41563446187723646,
+      "grad_norm": 1.64728262221785,
+      "learning_rate": 1.316299372466789e-05,
+      "loss": 1.0908,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4157720891824938,
+      "grad_norm": 1.7760277538958098,
+      "learning_rate": 1.3158764832997877e-05,
+      "loss": 1.1453,
+      "step": 3021
+    },
+    {
+      "epoch": 0.41590971648775116,
+      "grad_norm": 1.7750347628021885,
+      "learning_rate": 1.315453531372453e-05,
+      "loss": 1.0228,
+      "step": 3022
+    },
+    {
+      "epoch": 0.4160473437930085,
+      "grad_norm": 1.9618108569242174,
+      "learning_rate": 1.31503051676882e-05,
+      "loss": 1.1407,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4161849710982659,
+      "grad_norm": 1.640611334139154,
+      "learning_rate": 1.3146074395729353e-05,
+      "loss": 1.075,
+      "step": 3024
+    },
+    {
+      "epoch": 0.41632259840352326,
+      "grad_norm": 1.7739567699435685,
+      "learning_rate": 1.3141842998688589e-05,
+      "loss": 1.1552,
+      "step": 3025
+    },
+    {
+      "epoch": 0.4164602257087806,
+      "grad_norm": 1.8208954570265325,
+      "learning_rate": 1.3137610977406632e-05,
+      "loss": 1.0867,
+      "step": 3026
+    },
+    {
+      "epoch": 0.416597853014038,
+      "grad_norm": 1.5918058194722966,
+      "learning_rate": 1.3133378332724318e-05,
+      "loss": 1.0667,
+      "step": 3027
+    },
+    {
+      "epoch": 0.41673548031929536,
+      "grad_norm": 1.5935519973891266,
+      "learning_rate": 1.3129145065482622e-05,
+      "loss": 0.9561,
+      "step": 3028
+    },
+    {
+      "epoch": 0.4168731076245527,
+      "grad_norm": 1.7914409754412541,
+      "learning_rate": 1.3124911176522637e-05,
+      "loss": 1.0361,
+      "step": 3029
+    },
+    {
+      "epoch": 0.41701073492981006,
+      "grad_norm": 1.6585946385018528,
+      "learning_rate": 1.3120676666685576e-05,
+      "loss": 1.1673,
+      "step": 3030
+    },
+    {
+      "epoch": 0.41714836223506746,
+      "grad_norm": 1.4574205058884744,
+      "learning_rate": 1.311644153681278e-05,
+      "loss": 1.0251,
+      "step": 3031
+    },
+    {
+      "epoch": 0.4172859895403248,
+      "grad_norm": 1.6170242599783526,
+      "learning_rate": 1.3112205787745706e-05,
+      "loss": 0.9872,
+      "step": 3032
+    },
+    {
+      "epoch": 0.41742361684558216,
+      "grad_norm": 1.571178139899867,
+      "learning_rate": 1.3107969420325945e-05,
+      "loss": 1.0358,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4175612441508395,
+      "grad_norm": 1.6305116645574207,
+      "learning_rate": 1.3103732435395205e-05,
+      "loss": 0.9876,
+      "step": 3034
+    },
+    {
+      "epoch": 0.4176988714560969,
+      "grad_norm": 1.7678608616559375,
+      "learning_rate": 1.3099494833795314e-05,
+      "loss": 1.0423,
+      "step": 3035
+    },
+    {
+      "epoch": 0.41783649876135426,
+      "grad_norm": 1.714414158370946,
+      "learning_rate": 1.3095256616368228e-05,
+      "loss": 1.0253,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4179741260666116,
+      "grad_norm": 1.972336932522941,
+      "learning_rate": 1.3091017783956018e-05,
+      "loss": 1.0415,
+      "step": 3037
+    },
+    {
+      "epoch": 0.41811175337186895,
+      "grad_norm": 1.562178277884062,
+      "learning_rate": 1.308677833740089e-05,
+      "loss": 1.0806,
+      "step": 3038
+    },
+    {
+      "epoch": 0.41824938067712636,
+      "grad_norm": 1.8926683508220141,
+      "learning_rate": 1.3082538277545161e-05,
+      "loss": 1.0586,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4183870079823837,
+      "grad_norm": 1.8309892195362432,
+      "learning_rate": 1.3078297605231267e-05,
+      "loss": 1.0923,
+      "step": 3040
+    },
+    {
+      "epoch": 0.41852463528764106,
+      "grad_norm": 1.6887500776382363,
+      "learning_rate": 1.307405632130178e-05,
+      "loss": 1.05,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4186622625928984,
+      "grad_norm": 1.7806125117237206,
+      "learning_rate": 1.306981442659938e-05,
+      "loss": 1.1288,
+      "step": 3042
+    },
+    {
+      "epoch": 0.4187998898981558,
+      "grad_norm": 1.7910770335784474,
+      "learning_rate": 1.3065571921966875e-05,
+      "loss": 1.0756,
+      "step": 3043
+    },
+    {
+      "epoch": 0.41893751720341316,
+      "grad_norm": 1.838712411557385,
+      "learning_rate": 1.3061328808247191e-05,
+      "loss": 1.1604,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4190751445086705,
+      "grad_norm": 1.7197796200414144,
+      "learning_rate": 1.3057085086283378e-05,
+      "loss": 1.0795,
+      "step": 3045
+    },
+    {
+      "epoch": 0.4192127718139279,
+      "grad_norm": 1.7018948218861754,
+      "learning_rate": 1.3052840756918607e-05,
+      "loss": 1.2068,
+      "step": 3046
+    },
+    {
+      "epoch": 0.41935039911918526,
+      "grad_norm": 1.6568255903403415,
+      "learning_rate": 1.3048595820996166e-05,
+      "loss": 1.04,
+      "step": 3047
+    },
+    {
+      "epoch": 0.4194880264244426,
+      "grad_norm": 1.6799032575971349,
+      "learning_rate": 1.3044350279359464e-05,
+      "loss": 1.0302,
+      "step": 3048
+    },
+    {
+      "epoch": 0.41962565372969995,
+      "grad_norm": 1.6821346953822536,
+      "learning_rate": 1.3040104132852036e-05,
+      "loss": 1.0525,
+      "step": 3049
+    },
+    {
+      "epoch": 0.41976328103495736,
+      "grad_norm": 1.7865195732332562,
+      "learning_rate": 1.3035857382317532e-05,
+      "loss": 1.0928,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4199009083402147,
+      "grad_norm": 1.4638278026668325,
+      "learning_rate": 1.303161002859972e-05,
+      "loss": 1.0447,
+      "step": 3051
+    },
+    {
+      "epoch": 0.42003853564547206,
+      "grad_norm": 1.7002449982730747,
+      "learning_rate": 1.3027362072542493e-05,
+      "loss": 1.0182,
+      "step": 3052
+    },
+    {
+      "epoch": 0.4201761629507294,
+      "grad_norm": 1.6080529102479715,
+      "learning_rate": 1.3023113514989865e-05,
+      "loss": 1.0538,
+      "step": 3053
+    },
+    {
+      "epoch": 0.4203137902559868,
+      "grad_norm": 1.7625827635816944,
+      "learning_rate": 1.3018864356785961e-05,
+      "loss": 1.0532,
+      "step": 3054
+    },
+    {
+      "epoch": 0.42045141756124416,
+      "grad_norm": 1.7576122640538192,
+      "learning_rate": 1.3014614598775035e-05,
+      "loss": 1.0429,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4205890448665015,
+      "grad_norm": 1.6339978276592457,
+      "learning_rate": 1.3010364241801453e-05,
+      "loss": 1.0963,
+      "step": 3056
+    },
+    {
+      "epoch": 0.42072667217175885,
+      "grad_norm": 1.6950136685311903,
+      "learning_rate": 1.3006113286709705e-05,
+      "loss": 1.0627,
+      "step": 3057
+    },
+    {
+      "epoch": 0.42086429947701626,
+      "grad_norm": 1.5560320034254649,
+      "learning_rate": 1.3001861734344394e-05,
+      "loss": 1.058,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4210019267822736,
+      "grad_norm": 1.7062727742830026,
+      "learning_rate": 1.2997609585550248e-05,
+      "loss": 1.0159,
+      "step": 3059
+    },
+    {
+      "epoch": 0.42113955408753095,
+      "grad_norm": 1.745887472527452,
+      "learning_rate": 1.299335684117211e-05,
+      "loss": 1.0592,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4212771813927883,
+      "grad_norm": 1.6818398345153933,
+      "learning_rate": 1.2989103502054943e-05,
+      "loss": 1.118,
+      "step": 3061
+    },
+    {
+      "epoch": 0.4214148086980457,
+      "grad_norm": 1.827710372264519,
+      "learning_rate": 1.298484956904383e-05,
+      "loss": 1.0067,
+      "step": 3062
+    },
+    {
+      "epoch": 0.42155243600330305,
+      "grad_norm": 1.8441029517937946,
+      "learning_rate": 1.2980595042983961e-05,
+      "loss": 1.1122,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4216900633085604,
+      "grad_norm": 1.6289015601743708,
+      "learning_rate": 1.2976339924720657e-05,
+      "loss": 1.0885,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4218276906138178,
+      "grad_norm": 1.5713109877966056,
+      "learning_rate": 1.2972084215099353e-05,
+      "loss": 1.038,
+      "step": 3065
+    },
+    {
+      "epoch": 0.42196531791907516,
+      "grad_norm": 1.525857752531764,
+      "learning_rate": 1.2967827914965598e-05,
+      "loss": 1.0143,
+      "step": 3066
+    },
+    {
+      "epoch": 0.4221029452243325,
+      "grad_norm": 1.6058407106345536,
+      "learning_rate": 1.2963571025165062e-05,
+      "loss": 1.1011,
+      "step": 3067
+    },
+    {
+      "epoch": 0.42224057252958985,
+      "grad_norm": 1.8189877180181475,
+      "learning_rate": 1.2959313546543529e-05,
+      "loss": 1.0948,
+      "step": 3068
+    },
+    {
+      "epoch": 0.42237819983484726,
+      "grad_norm": 1.579561229311487,
+      "learning_rate": 1.2955055479946906e-05,
+      "loss": 1.0679,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4225158271401046,
+      "grad_norm": 2.0246197722684447,
+      "learning_rate": 1.2950796826221206e-05,
+      "loss": 1.2288,
+      "step": 3070
+    },
+    {
+      "epoch": 0.42265345444536195,
+      "grad_norm": 1.8903250652978796,
+      "learning_rate": 1.2946537586212568e-05,
+      "loss": 1.0872,
+      "step": 3071
+    },
+    {
+      "epoch": 0.4227910817506193,
+      "grad_norm": 1.720880088606622,
+      "learning_rate": 1.294227776076725e-05,
+      "loss": 0.9987,
+      "step": 3072
+    },
+    {
+      "epoch": 0.4229287090558767,
+      "grad_norm": 1.9539893777382384,
+      "learning_rate": 1.2938017350731614e-05,
+      "loss": 1.0599,
+      "step": 3073
+    },
+    {
+      "epoch": 0.42306633636113405,
+      "grad_norm": 1.7244222840933896,
+      "learning_rate": 1.2933756356952145e-05,
+      "loss": 1.0882,
+      "step": 3074
+    },
+    {
+      "epoch": 0.4232039636663914,
+      "grad_norm": 1.7286828272122046,
+      "learning_rate": 1.2929494780275449e-05,
+      "loss": 1.1122,
+      "step": 3075
+    },
+    {
+      "epoch": 0.42334159097164875,
+      "grad_norm": 1.6237419450437434,
+      "learning_rate": 1.2925232621548242e-05,
+      "loss": 1.0258,
+      "step": 3076
+    },
+    {
+      "epoch": 0.42347921827690616,
+      "grad_norm": 1.5627382435020283,
+      "learning_rate": 1.292096988161735e-05,
+      "loss": 1.0188,
+      "step": 3077
+    },
+    {
+      "epoch": 0.4236168455821635,
+      "grad_norm": 1.6123659145228613,
+      "learning_rate": 1.2916706561329733e-05,
+      "loss": 1.0319,
+      "step": 3078
+    },
+    {
+      "epoch": 0.42375447288742085,
+      "grad_norm": 1.7700479727198817,
+      "learning_rate": 1.2912442661532446e-05,
+      "loss": 1.0838,
+      "step": 3079
+    },
+    {
+      "epoch": 0.4238921001926782,
+      "grad_norm": 1.752295399227553,
+      "learning_rate": 1.2908178183072666e-05,
+      "loss": 1.0925,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4240297274979356,
+      "grad_norm": 1.7457267540198458,
+      "learning_rate": 1.2903913126797692e-05,
+      "loss": 1.0508,
+      "step": 3081
+    },
+    {
+      "epoch": 0.42416735480319295,
+      "grad_norm": 1.5320268243290713,
+      "learning_rate": 1.2899647493554927e-05,
+      "loss": 0.9961,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4243049821084503,
+      "grad_norm": 1.5532457792908723,
+      "learning_rate": 1.2895381284191898e-05,
+      "loss": 1.0869,
+      "step": 3083
+    },
+    {
+      "epoch": 0.4244426094137077,
+      "grad_norm": 1.7727562876627874,
+      "learning_rate": 1.2891114499556243e-05,
+      "loss": 0.9572,
+      "step": 3084
+    },
+    {
+      "epoch": 0.42458023671896505,
+      "grad_norm": 1.5036474824967767,
+      "learning_rate": 1.288684714049571e-05,
+      "loss": 1.0621,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4247178640242224,
+      "grad_norm": 1.6808772078439156,
+      "learning_rate": 1.288257920785817e-05,
+      "loss": 0.9915,
+      "step": 3086
+    },
+    {
+      "epoch": 0.42485549132947975,
+      "grad_norm": 1.5788369949490637,
+      "learning_rate": 1.2878310702491597e-05,
+      "loss": 1.1019,
+      "step": 3087
+    },
+    {
+      "epoch": 0.42499311863473715,
+      "grad_norm": 1.7304591414068993,
+      "learning_rate": 1.2874041625244087e-05,
+      "loss": 1.1207,
+      "step": 3088
+    },
+    {
+      "epoch": 0.4251307459399945,
+      "grad_norm": 1.788340679239401,
+      "learning_rate": 1.286977197696385e-05,
+      "loss": 1.2451,
+      "step": 3089
+    },
+    {
+      "epoch": 0.42526837324525185,
+      "grad_norm": 1.8970870132144813,
+      "learning_rate": 1.2865501758499202e-05,
+      "loss": 1.0715,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4254060005505092,
+      "grad_norm": 1.597682991506191,
+      "learning_rate": 1.2861230970698579e-05,
+      "loss": 1.0074,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4255436278557666,
+      "grad_norm": 1.7772810202750515,
+      "learning_rate": 1.2856959614410527e-05,
+      "loss": 1.007,
+      "step": 3092
+    },
+    {
+      "epoch": 0.42568125516102395,
+      "grad_norm": 1.820221520661318,
+      "learning_rate": 1.2852687690483709e-05,
+      "loss": 1.0239,
+      "step": 3093
+    },
+    {
+      "epoch": 0.4258188824662813,
+      "grad_norm": 1.8300867184214384,
+      "learning_rate": 1.2848415199766896e-05,
+      "loss": 1.1073,
+      "step": 3094
+    },
+    {
+      "epoch": 0.42595650977153865,
+      "grad_norm": 1.9000215652436345,
+      "learning_rate": 1.2844142143108976e-05,
+      "loss": 1.1761,
+      "step": 3095
+    },
+    {
+      "epoch": 0.42609413707679605,
+      "grad_norm": 1.7602665475561385,
+      "learning_rate": 1.283986852135894e-05,
+      "loss": 1.085,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4262317643820534,
+      "grad_norm": 1.4883104177985613,
+      "learning_rate": 1.2835594335365904e-05,
+      "loss": 1.0512,
+      "step": 3097
+    },
+    {
+      "epoch": 0.42636939168731075,
+      "grad_norm": 1.7071133665255267,
+      "learning_rate": 1.2831319585979088e-05,
+      "loss": 0.9966,
+      "step": 3098
+    },
+    {
+      "epoch": 0.4265070189925681,
+      "grad_norm": 1.776204696723145,
+      "learning_rate": 1.2827044274047825e-05,
+      "loss": 1.0764,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4266446462978255,
+      "grad_norm": 1.725909259019703,
+      "learning_rate": 1.2822768400421564e-05,
+      "loss": 1.1123,
+      "step": 3100
+    },
+    {
+      "epoch": 0.42678227360308285,
+      "grad_norm": 1.5073815134995774,
+      "learning_rate": 1.2818491965949862e-05,
+      "loss": 1.0791,
+      "step": 3101
+    },
+    {
+      "epoch": 0.4269199009083402,
+      "grad_norm": 1.5743084088660422,
+      "learning_rate": 1.2814214971482389e-05,
+      "loss": 1.0221,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4270575282135976,
+      "grad_norm": 1.5566053543231022,
+      "learning_rate": 1.280993741786892e-05,
+      "loss": 1.0166,
+      "step": 3103
+    },
+    {
+      "epoch": 0.42719515551885495,
+      "grad_norm": 1.804668527474538,
+      "learning_rate": 1.2805659305959352e-05,
+      "loss": 1.0647,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4273327828241123,
+      "grad_norm": 1.6486344133142343,
+      "learning_rate": 1.2801380636603686e-05,
+      "loss": 1.0884,
+      "step": 3105
+    },
+    {
+      "epoch": 0.42747041012936965,
+      "grad_norm": 1.531564978705585,
+      "learning_rate": 1.2797101410652033e-05,
+      "loss": 1.0703,
+      "step": 3106
+    },
+    {
+      "epoch": 0.42760803743462705,
+      "grad_norm": 1.7568537503522137,
+      "learning_rate": 1.279282162895462e-05,
+      "loss": 1.0662,
+      "step": 3107
+    },
+    {
+      "epoch": 0.4277456647398844,
+      "grad_norm": 1.6489692997287986,
+      "learning_rate": 1.2788541292361774e-05,
+      "loss": 1.0562,
+      "step": 3108
+    },
+    {
+      "epoch": 0.42788329204514175,
+      "grad_norm": 1.6027806389338546,
+      "learning_rate": 1.2784260401723947e-05,
+      "loss": 1.037,
+      "step": 3109
+    },
+    {
+      "epoch": 0.4280209193503991,
+      "grad_norm": 1.5865446336100815,
+      "learning_rate": 1.2779978957891692e-05,
+      "loss": 1.0713,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4281585466556565,
+      "grad_norm": 1.6586520929151236,
+      "learning_rate": 1.2775696961715671e-05,
+      "loss": 0.9449,
+      "step": 3111
+    },
+    {
+      "epoch": 0.42829617396091385,
+      "grad_norm": 1.7463607641763577,
+      "learning_rate": 1.2771414414046658e-05,
+      "loss": 1.0658,
+      "step": 3112
+    },
+    {
+      "epoch": 0.4284338012661712,
+      "grad_norm": 1.9651552457368968,
+      "learning_rate": 1.276713131573554e-05,
+      "loss": 1.0476,
+      "step": 3113
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.6249884113305426,
+      "learning_rate": 1.2762847667633305e-05,
+      "loss": 1.0193,
+      "step": 3114
+    },
+    {
+      "epoch": 0.42870905587668595,
+      "grad_norm": 1.6616066471384072,
+      "learning_rate": 1.2758563470591057e-05,
+      "loss": 1.0358,
+      "step": 3115
+    },
+    {
+      "epoch": 0.4288466831819433,
+      "grad_norm": 1.7948450094151074,
+      "learning_rate": 1.2754278725460008e-05,
+      "loss": 1.1198,
+      "step": 3116
+    },
+    {
+      "epoch": 0.42898431048720065,
+      "grad_norm": 1.6728406241485596,
+      "learning_rate": 1.2749993433091479e-05,
+      "loss": 1.0511,
+      "step": 3117
+    },
+    {
+      "epoch": 0.429121937792458,
+      "grad_norm": 1.702515640176959,
+      "learning_rate": 1.27457075943369e-05,
+      "loss": 1.1458,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4292595650977154,
+      "grad_norm": 1.8574280686784768,
+      "learning_rate": 1.2741421210047803e-05,
+      "loss": 0.969,
+      "step": 3119
+    },
+    {
+      "epoch": 0.42939719240297275,
+      "grad_norm": 1.7999194187457404,
+      "learning_rate": 1.273713428107584e-05,
+      "loss": 1.0494,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4295348197082301,
+      "grad_norm": 1.7416193703403509,
+      "learning_rate": 1.2732846808272762e-05,
+      "loss": 1.1068,
+      "step": 3121
+    },
+    {
+      "epoch": 0.4296724470134875,
+      "grad_norm": 1.643127031413654,
+      "learning_rate": 1.2728558792490433e-05,
+      "loss": 1.0124,
+      "step": 3122
+    },
+    {
+      "epoch": 0.42981007431874485,
+      "grad_norm": 1.8549211720104217,
+      "learning_rate": 1.272427023458082e-05,
+      "loss": 1.0371,
+      "step": 3123
+    },
+    {
+      "epoch": 0.4299477016240022,
+      "grad_norm": 1.85952636340522,
+      "learning_rate": 1.2719981135396002e-05,
+      "loss": 1.0825,
+      "step": 3124
+    },
+    {
+      "epoch": 0.43008532892925955,
+      "grad_norm": 1.7360753216908393,
+      "learning_rate": 1.2715691495788167e-05,
+      "loss": 1.0565,
+      "step": 3125
+    },
+    {
+      "epoch": 0.43022295623451695,
+      "grad_norm": 1.6938326389465288,
+      "learning_rate": 1.2711401316609609e-05,
+      "loss": 1.0943,
+      "step": 3126
+    },
+    {
+      "epoch": 0.4303605835397743,
+      "grad_norm": 1.8172848863364526,
+      "learning_rate": 1.270711059871272e-05,
+      "loss": 1.0686,
+      "step": 3127
+    },
+    {
+      "epoch": 0.43049821084503165,
+      "grad_norm": 1.6048678965865795,
+      "learning_rate": 1.270281934295001e-05,
+      "loss": 0.9601,
+      "step": 3128
+    },
+    {
+      "epoch": 0.430635838150289,
+      "grad_norm": 1.6098443949117176,
+      "learning_rate": 1.2698527550174096e-05,
+      "loss": 1.0907,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4307734654555464,
+      "grad_norm": 1.9938263775233502,
+      "learning_rate": 1.2694235221237695e-05,
+      "loss": 1.0383,
+      "step": 3130
+    },
+    {
+      "epoch": 0.43091109276080375,
+      "grad_norm": 1.646435798285442,
+      "learning_rate": 1.2689942356993637e-05,
+      "loss": 0.9769,
+      "step": 3131
+    },
+    {
+      "epoch": 0.4310487200660611,
+      "grad_norm": 1.8229832346215644,
+      "learning_rate": 1.2685648958294854e-05,
+      "loss": 1.155,
+      "step": 3132
+    },
+    {
+      "epoch": 0.43118634737131845,
+      "grad_norm": 1.687962810813789,
+      "learning_rate": 1.2681355025994386e-05,
+      "loss": 1.0828,
+      "step": 3133
+    },
+    {
+      "epoch": 0.43132397467657585,
+      "grad_norm": 1.8961903710574464,
+      "learning_rate": 1.2677060560945377e-05,
+      "loss": 1.074,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4314616019818332,
+      "grad_norm": 1.543217766170686,
+      "learning_rate": 1.2672765564001082e-05,
+      "loss": 0.9539,
+      "step": 3135
+    },
+    {
+      "epoch": 0.43159922928709055,
+      "grad_norm": 1.7379587821564906,
+      "learning_rate": 1.2668470036014852e-05,
+      "loss": 1.0339,
+      "step": 3136
+    },
+    {
+      "epoch": 0.4317368565923479,
+      "grad_norm": 1.7511599167144067,
+      "learning_rate": 1.2664173977840155e-05,
+      "loss": 1.1296,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4318744838976053,
+      "grad_norm": 1.6327434234842504,
+      "learning_rate": 1.2659877390330557e-05,
+      "loss": 1.0541,
+      "step": 3138
+    },
+    {
+      "epoch": 0.43201211120286265,
+      "grad_norm": 1.7944906594830923,
+      "learning_rate": 1.265558027433973e-05,
+      "loss": 1.141,
+      "step": 3139
+    },
+    {
+      "epoch": 0.43214973850812,
+      "grad_norm": 1.7984560448078661,
+      "learning_rate": 1.2651282630721454e-05,
+      "loss": 1.1396,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4322873658133774,
+      "grad_norm": 1.6224796215661335,
+      "learning_rate": 1.2646984460329612e-05,
+      "loss": 1.0961,
+      "step": 3141
+    },
+    {
+      "epoch": 0.43242499311863475,
+      "grad_norm": 1.542697735749195,
+      "learning_rate": 1.2642685764018195e-05,
+      "loss": 1.0846,
+      "step": 3142
+    },
+    {
+      "epoch": 0.4325626204238921,
+      "grad_norm": 1.7516892322251574,
+      "learning_rate": 1.2638386542641286e-05,
+      "loss": 0.9619,
+      "step": 3143
+    },
+    {
+      "epoch": 0.43270024772914945,
+      "grad_norm": 1.6781228288564856,
+      "learning_rate": 1.2634086797053089e-05,
+      "loss": 1.0081,
+      "step": 3144
+    },
+    {
+      "epoch": 0.43283787503440685,
+      "grad_norm": 1.8321574667939375,
+      "learning_rate": 1.2629786528107901e-05,
+      "loss": 1.0235,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4329755023396642,
+      "grad_norm": 1.6262013509718087,
+      "learning_rate": 1.2625485736660126e-05,
+      "loss": 0.9767,
+      "step": 3146
+    },
+    {
+      "epoch": 0.43311312964492155,
+      "grad_norm": 1.6357209504785235,
+      "learning_rate": 1.2621184423564275e-05,
+      "loss": 0.9934,
+      "step": 3147
+    },
+    {
+      "epoch": 0.4332507569501789,
+      "grad_norm": 1.5489993817368388,
+      "learning_rate": 1.261688258967496e-05,
+      "loss": 1.0205,
+      "step": 3148
+    },
+    {
+      "epoch": 0.4333883842554363,
+      "grad_norm": 1.7006853093592256,
+      "learning_rate": 1.2612580235846895e-05,
+      "loss": 0.997,
+      "step": 3149
+    },
+    {
+      "epoch": 0.43352601156069365,
+      "grad_norm": 1.5811003569160225,
+      "learning_rate": 1.2608277362934901e-05,
+      "loss": 1.027,
+      "step": 3150
+    },
+    {
+      "epoch": 0.433663638865951,
+      "grad_norm": 1.7177391956181791,
+      "learning_rate": 1.2603973971793897e-05,
+      "loss": 1.0833,
+      "step": 3151
+    },
+    {
+      "epoch": 0.43380126617120834,
+      "grad_norm": 1.8362146115801377,
+      "learning_rate": 1.2599670063278911e-05,
+      "loss": 1.038,
+      "step": 3152
+    },
+    {
+      "epoch": 0.43393889347646575,
+      "grad_norm": 1.6568752419241706,
+      "learning_rate": 1.2595365638245068e-05,
+      "loss": 1.0115,
+      "step": 3153
+    },
+    {
+      "epoch": 0.4340765207817231,
+      "grad_norm": 1.7787602054076295,
+      "learning_rate": 1.2591060697547596e-05,
+      "loss": 1.0963,
+      "step": 3154
+    },
+    {
+      "epoch": 0.43421414808698044,
+      "grad_norm": 1.71318078175494,
+      "learning_rate": 1.2586755242041835e-05,
+      "loss": 1.1773,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4343517753922378,
+      "grad_norm": 1.6298724444752155,
+      "learning_rate": 1.2582449272583218e-05,
+      "loss": 1.0541,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4344894026974952,
+      "grad_norm": 1.5946830347426724,
+      "learning_rate": 1.2578142790027275e-05,
+      "loss": 1.0508,
+      "step": 3157
+    },
+    {
+      "epoch": 0.43462703000275255,
+      "grad_norm": 1.8644424747640742,
+      "learning_rate": 1.2573835795229656e-05,
+      "loss": 1.0091,
+      "step": 3158
+    },
+    {
+      "epoch": 0.4347646573080099,
+      "grad_norm": 1.7629854196859822,
+      "learning_rate": 1.2569528289046096e-05,
+      "loss": 1.0989,
+      "step": 3159
+    },
+    {
+      "epoch": 0.4349022846132673,
+      "grad_norm": 1.7849150933978042,
+      "learning_rate": 1.2565220272332435e-05,
+      "loss": 1.0707,
+      "step": 3160
+    },
+    {
+      "epoch": 0.43503991191852465,
+      "grad_norm": 1.6640070034228307,
+      "learning_rate": 1.2560911745944617e-05,
+      "loss": 1.1032,
+      "step": 3161
+    },
+    {
+      "epoch": 0.435177539223782,
+      "grad_norm": 1.886310816136597,
+      "learning_rate": 1.2556602710738695e-05,
+      "loss": 1.0434,
+      "step": 3162
+    },
+    {
+      "epoch": 0.43531516652903934,
+      "grad_norm": 1.792270173244599,
+      "learning_rate": 1.255229316757081e-05,
+      "loss": 1.0184,
+      "step": 3163
+    },
+    {
+      "epoch": 0.43545279383429675,
+      "grad_norm": 1.5772509109304522,
+      "learning_rate": 1.2547983117297208e-05,
+      "loss": 1.0575,
+      "step": 3164
+    },
+    {
+      "epoch": 0.4355904211395541,
+      "grad_norm": 1.7610745759451705,
+      "learning_rate": 1.2543672560774241e-05,
+      "loss": 0.9908,
+      "step": 3165
+    },
+    {
+      "epoch": 0.43572804844481144,
+      "grad_norm": 1.6190827877993157,
+      "learning_rate": 1.2539361498858354e-05,
+      "loss": 1.071,
+      "step": 3166
+    },
+    {
+      "epoch": 0.4358656757500688,
+      "grad_norm": 1.557941579429338,
+      "learning_rate": 1.2535049932406098e-05,
+      "loss": 0.9981,
+      "step": 3167
+    },
+    {
+      "epoch": 0.4360033030553262,
+      "grad_norm": 1.6404936552889935,
+      "learning_rate": 1.253073786227412e-05,
+      "loss": 1.0329,
+      "step": 3168
+    },
+    {
+      "epoch": 0.43614093036058355,
+      "grad_norm": 1.5588288919919873,
+      "learning_rate": 1.2526425289319173e-05,
+      "loss": 1.0018,
+      "step": 3169
+    },
+    {
+      "epoch": 0.4362785576658409,
+      "grad_norm": 1.6500157198685181,
+      "learning_rate": 1.25221122143981e-05,
+      "loss": 0.9798,
+      "step": 3170
+    },
+    {
+      "epoch": 0.43641618497109824,
+      "grad_norm": 1.6880322636698197,
+      "learning_rate": 1.2517798638367858e-05,
+      "loss": 1.0596,
+      "step": 3171
+    },
+    {
+      "epoch": 0.43655381227635565,
+      "grad_norm": 1.6648805509143199,
+      "learning_rate": 1.251348456208549e-05,
+      "loss": 1.0727,
+      "step": 3172
+    },
+    {
+      "epoch": 0.436691439581613,
+      "grad_norm": 1.7659517570629897,
+      "learning_rate": 1.2509169986408145e-05,
+      "loss": 1.0299,
+      "step": 3173
+    },
+    {
+      "epoch": 0.43682906688687034,
+      "grad_norm": 1.7395465184471834,
+      "learning_rate": 1.2504854912193075e-05,
+      "loss": 1.0318,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4369666941921277,
+      "grad_norm": 1.6029235128883876,
+      "learning_rate": 1.2500539340297614e-05,
+      "loss": 0.9996,
+      "step": 3175
+    },
+    {
+      "epoch": 0.4371043214973851,
+      "grad_norm": 1.6627560283267624,
+      "learning_rate": 1.2496223271579219e-05,
+      "loss": 1.0411,
+      "step": 3176
+    },
+    {
+      "epoch": 0.43724194880264244,
+      "grad_norm": 1.6977250298361346,
+      "learning_rate": 1.2491906706895427e-05,
+      "loss": 0.9759,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4373795761078998,
+      "grad_norm": 1.7833305712647323,
+      "learning_rate": 1.2487589647103887e-05,
+      "loss": 1.0375,
+      "step": 3178
+    },
+    {
+      "epoch": 0.4375172034131572,
+      "grad_norm": 1.849835517983004,
+      "learning_rate": 1.2483272093062332e-05,
+      "loss": 1.0026,
+      "step": 3179
+    },
+    {
+      "epoch": 0.43765483071841454,
+      "grad_norm": 1.7608875779329944,
+      "learning_rate": 1.2478954045628606e-05,
+      "loss": 1.0985,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4377924580236719,
+      "grad_norm": 1.7544916347044048,
+      "learning_rate": 1.2474635505660645e-05,
+      "loss": 1.0367,
+      "step": 3181
+    },
+    {
+      "epoch": 0.43793008532892924,
+      "grad_norm": 1.55109606450465,
+      "learning_rate": 1.2470316474016482e-05,
+      "loss": 0.9968,
+      "step": 3182
+    },
+    {
+      "epoch": 0.43806771263418665,
+      "grad_norm": 1.9658896178337446,
+      "learning_rate": 1.246599695155425e-05,
+      "loss": 1.0278,
+      "step": 3183
+    },
+    {
+      "epoch": 0.438205339939444,
+      "grad_norm": 1.7328909496489404,
+      "learning_rate": 1.2461676939132179e-05,
+      "loss": 1.0244,
+      "step": 3184
+    },
+    {
+      "epoch": 0.43834296724470134,
+      "grad_norm": 1.6622757547894675,
+      "learning_rate": 1.2457356437608596e-05,
+      "loss": 0.9542,
+      "step": 3185
+    },
+    {
+      "epoch": 0.4384805945499587,
+      "grad_norm": 1.5325297556261395,
+      "learning_rate": 1.245303544784193e-05,
+      "loss": 1.0552,
+      "step": 3186
+    },
+    {
+      "epoch": 0.4386182218552161,
+      "grad_norm": 1.6754719636753852,
+      "learning_rate": 1.2448713970690696e-05,
+      "loss": 1.0384,
+      "step": 3187
+    },
+    {
+      "epoch": 0.43875584916047344,
+      "grad_norm": 2.086655518413788,
+      "learning_rate": 1.2444392007013518e-05,
+      "loss": 1.0947,
+      "step": 3188
+    },
+    {
+      "epoch": 0.4388934764657308,
+      "grad_norm": 1.700161558159017,
+      "learning_rate": 1.2440069557669106e-05,
+      "loss": 1.0756,
+      "step": 3189
+    },
+    {
+      "epoch": 0.43903110377098814,
+      "grad_norm": 1.9117027265303874,
+      "learning_rate": 1.2435746623516276e-05,
+      "loss": 1.0411,
+      "step": 3190
+    },
+    {
+      "epoch": 0.43916873107624554,
+      "grad_norm": 1.6254447132741074,
+      "learning_rate": 1.2431423205413933e-05,
+      "loss": 0.9813,
+      "step": 3191
+    },
+    {
+      "epoch": 0.4393063583815029,
+      "grad_norm": 1.6843316275313973,
+      "learning_rate": 1.2427099304221079e-05,
+      "loss": 1.0611,
+      "step": 3192
+    },
+    {
+      "epoch": 0.43944398568676024,
+      "grad_norm": 1.524002664379864,
+      "learning_rate": 1.2422774920796818e-05,
+      "loss": 1.0109,
+      "step": 3193
+    },
+    {
+      "epoch": 0.4395816129920176,
+      "grad_norm": 1.7493421633312474,
+      "learning_rate": 1.2418450056000346e-05,
+      "loss": 1.1647,
+      "step": 3194
+    },
+    {
+      "epoch": 0.439719240297275,
+      "grad_norm": 1.6230740957301961,
+      "learning_rate": 1.241412471069095e-05,
+      "loss": 1.0323,
+      "step": 3195
+    },
+    {
+      "epoch": 0.43985686760253234,
+      "grad_norm": 1.5636832552465756,
+      "learning_rate": 1.2409798885728023e-05,
+      "loss": 1.0169,
+      "step": 3196
+    },
+    {
+      "epoch": 0.4399944949077897,
+      "grad_norm": 1.636831351027792,
+      "learning_rate": 1.2405472581971042e-05,
+      "loss": 1.1052,
+      "step": 3197
+    },
+    {
+      "epoch": 0.4401321222130471,
+      "grad_norm": 1.9941306692535827,
+      "learning_rate": 1.2401145800279588e-05,
+      "loss": 1.0797,
+      "step": 3198
+    },
+    {
+      "epoch": 0.44026974951830444,
+      "grad_norm": 1.6933492278073254,
+      "learning_rate": 1.239681854151333e-05,
+      "loss": 1.0639,
+      "step": 3199
+    },
+    {
+      "epoch": 0.4404073768235618,
+      "grad_norm": 1.761482654922105,
+      "learning_rate": 1.2392490806532038e-05,
+      "loss": 1.0718,
+      "step": 3200
+    },
+    {
+      "epoch": 0.44054500412881914,
+      "grad_norm": 1.5820362891530688,
+      "learning_rate": 1.238816259619557e-05,
+      "loss": 1.0551,
+      "step": 3201
+    },
+    {
+      "epoch": 0.44068263143407654,
+      "grad_norm": 1.674842342949645,
+      "learning_rate": 1.2383833911363885e-05,
+      "loss": 0.9517,
+      "step": 3202
+    },
+    {
+      "epoch": 0.4408202587393339,
+      "grad_norm": 1.737347397302642,
+      "learning_rate": 1.2379504752897034e-05,
+      "loss": 1.0854,
+      "step": 3203
+    },
+    {
+      "epoch": 0.44095788604459124,
+      "grad_norm": 1.7246422831833141,
+      "learning_rate": 1.2375175121655161e-05,
+      "loss": 0.9239,
+      "step": 3204
+    },
+    {
+      "epoch": 0.4410955133498486,
+      "grad_norm": 1.733967508578194,
+      "learning_rate": 1.2370845018498503e-05,
+      "loss": 1.0965,
+      "step": 3205
+    },
+    {
+      "epoch": 0.441233140655106,
+      "grad_norm": 1.6078103889529902,
+      "learning_rate": 1.2366514444287396e-05,
+      "loss": 1.0201,
+      "step": 3206
+    },
+    {
+      "epoch": 0.44137076796036334,
+      "grad_norm": 1.7381363979598115,
+      "learning_rate": 1.2362183399882262e-05,
+      "loss": 1.1206,
+      "step": 3207
+    },
+    {
+      "epoch": 0.4415083952656207,
+      "grad_norm": 1.7496823738127036,
+      "learning_rate": 1.2357851886143616e-05,
+      "loss": 1.0512,
+      "step": 3208
+    },
+    {
+      "epoch": 0.44164602257087804,
+      "grad_norm": 1.7271127489763485,
+      "learning_rate": 1.235351990393208e-05,
+      "loss": 1.0508,
+      "step": 3209
+    },
+    {
+      "epoch": 0.44178364987613544,
+      "grad_norm": 1.611808272457047,
+      "learning_rate": 1.2349187454108354e-05,
+      "loss": 1.0606,
+      "step": 3210
+    },
+    {
+      "epoch": 0.4419212771813928,
+      "grad_norm": 1.776021783150459,
+      "learning_rate": 1.234485453753324e-05,
+      "loss": 1.0235,
+      "step": 3211
+    },
+    {
+      "epoch": 0.44205890448665014,
+      "grad_norm": 1.76542122779535,
+      "learning_rate": 1.2340521155067623e-05,
+      "loss": 0.9886,
+      "step": 3212
+    },
+    {
+      "epoch": 0.4421965317919075,
+      "grad_norm": 1.7284881609504283,
+      "learning_rate": 1.2336187307572493e-05,
+      "loss": 1.0095,
+      "step": 3213
+    },
+    {
+      "epoch": 0.4423341590971649,
+      "grad_norm": 1.8033416942477851,
+      "learning_rate": 1.2331852995908927e-05,
+      "loss": 1.0488,
+      "step": 3214
+    },
+    {
+      "epoch": 0.44247178640242224,
+      "grad_norm": 1.7772731380019744,
+      "learning_rate": 1.2327518220938086e-05,
+      "loss": 1.0057,
+      "step": 3215
+    },
+    {
+      "epoch": 0.4426094137076796,
+      "grad_norm": 1.648924059394994,
+      "learning_rate": 1.2323182983521236e-05,
+      "loss": 0.9851,
+      "step": 3216
+    },
+    {
+      "epoch": 0.442747041012937,
+      "grad_norm": 1.837133707876487,
+      "learning_rate": 1.2318847284519726e-05,
+      "loss": 1.127,
+      "step": 3217
+    },
+    {
+      "epoch": 0.44288466831819434,
+      "grad_norm": 1.6841317459593383,
+      "learning_rate": 1.2314511124795003e-05,
+      "loss": 1.0502,
+      "step": 3218
+    },
+    {
+      "epoch": 0.4430222956234517,
+      "grad_norm": 1.657717780390998,
+      "learning_rate": 1.2310174505208604e-05,
+      "loss": 1.0316,
+      "step": 3219
+    },
+    {
+      "epoch": 0.44315992292870904,
+      "grad_norm": 1.8515675456237604,
+      "learning_rate": 1.2305837426622153e-05,
+      "loss": 1.0282,
+      "step": 3220
+    },
+    {
+      "epoch": 0.44329755023396644,
+      "grad_norm": 1.6111843970120465,
+      "learning_rate": 1.2301499889897368e-05,
+      "loss": 1.0707,
+      "step": 3221
+    },
+    {
+      "epoch": 0.4434351775392238,
+      "grad_norm": 1.650082239657082,
+      "learning_rate": 1.2297161895896062e-05,
+      "loss": 1.1191,
+      "step": 3222
+    },
+    {
+      "epoch": 0.44357280484448114,
+      "grad_norm": 1.8678479527451124,
+      "learning_rate": 1.2292823445480126e-05,
+      "loss": 1.0852,
+      "step": 3223
+    },
+    {
+      "epoch": 0.4437104321497385,
+      "grad_norm": 1.6235834596643877,
+      "learning_rate": 1.228848453951156e-05,
+      "loss": 0.9877,
+      "step": 3224
+    },
+    {
+      "epoch": 0.4438480594549959,
+      "grad_norm": 2.1832705376431036,
+      "learning_rate": 1.2284145178852445e-05,
+      "loss": 1.1397,
+      "step": 3225
+    },
+    {
+      "epoch": 0.44398568676025324,
+      "grad_norm": 1.7026121411926158,
+      "learning_rate": 1.2279805364364945e-05,
+      "loss": 1.043,
+      "step": 3226
+    },
+    {
+      "epoch": 0.4441233140655106,
+      "grad_norm": 1.6998623698786843,
+      "learning_rate": 1.2275465096911327e-05,
+      "loss": 1.038,
+      "step": 3227
+    },
+    {
+      "epoch": 0.44426094137076794,
+      "grad_norm": 1.8479290861957356,
+      "learning_rate": 1.2271124377353945e-05,
+      "loss": 1.1217,
+      "step": 3228
+    },
+    {
+      "epoch": 0.44439856867602534,
+      "grad_norm": 1.8607164462218913,
+      "learning_rate": 1.2266783206555235e-05,
+      "loss": 1.2002,
+      "step": 3229
+    },
+    {
+      "epoch": 0.4445361959812827,
+      "grad_norm": 1.650611949226281,
+      "learning_rate": 1.2262441585377731e-05,
+      "loss": 1.0317,
+      "step": 3230
+    },
+    {
+      "epoch": 0.44467382328654004,
+      "grad_norm": 1.5279001064836084,
+      "learning_rate": 1.2258099514684052e-05,
+      "loss": 0.9919,
+      "step": 3231
+    },
+    {
+      "epoch": 0.4448114505917974,
+      "grad_norm": 1.6281976127874973,
+      "learning_rate": 1.225375699533691e-05,
+      "loss": 0.95,
+      "step": 3232
+    },
+    {
+      "epoch": 0.4449490778970548,
+      "grad_norm": 1.6345456989676252,
+      "learning_rate": 1.2249414028199104e-05,
+      "loss": 0.9755,
+      "step": 3233
+    },
+    {
+      "epoch": 0.44508670520231214,
+      "grad_norm": 1.6395157354413419,
+      "learning_rate": 1.2245070614133522e-05,
+      "loss": 1.0486,
+      "step": 3234
+    },
+    {
+      "epoch": 0.4452243325075695,
+      "grad_norm": 1.7404472833699915,
+      "learning_rate": 1.224072675400314e-05,
+      "loss": 1.0204,
+      "step": 3235
+    },
+    {
+      "epoch": 0.4453619598128269,
+      "grad_norm": 1.701781681798363,
+      "learning_rate": 1.2236382448671027e-05,
+      "loss": 1.0122,
+      "step": 3236
+    },
+    {
+      "epoch": 0.44549958711808424,
+      "grad_norm": 1.560882896897504,
+      "learning_rate": 1.2232037699000332e-05,
+      "loss": 1.0201,
+      "step": 3237
+    },
+    {
+      "epoch": 0.4456372144233416,
+      "grad_norm": 1.5766243079034652,
+      "learning_rate": 1.2227692505854303e-05,
+      "loss": 1.0768,
+      "step": 3238
+    },
+    {
+      "epoch": 0.44577484172859894,
+      "grad_norm": 1.7290501288363898,
+      "learning_rate": 1.222334687009626e-05,
+      "loss": 1.0398,
+      "step": 3239
+    },
+    {
+      "epoch": 0.44591246903385634,
+      "grad_norm": 1.6275044121561326,
+      "learning_rate": 1.2219000792589633e-05,
+      "loss": 0.9727,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4460500963391137,
+      "grad_norm": 1.7118732517388435,
+      "learning_rate": 1.2214654274197922e-05,
+      "loss": 0.9847,
+      "step": 3241
+    },
+    {
+      "epoch": 0.44618772364437104,
+      "grad_norm": 1.6170963787510475,
+      "learning_rate": 1.2210307315784721e-05,
+      "loss": 1.0,
+      "step": 3242
+    },
+    {
+      "epoch": 0.4463253509496284,
+      "grad_norm": 1.620718718673401,
+      "learning_rate": 1.2205959918213714e-05,
+      "loss": 1.0055,
+      "step": 3243
+    },
+    {
+      "epoch": 0.4464629782548858,
+      "grad_norm": 1.589082499762054,
+      "learning_rate": 1.2201612082348666e-05,
+      "loss": 1.0485,
+      "step": 3244
+    },
+    {
+      "epoch": 0.44660060556014314,
+      "grad_norm": 1.8054830779124131,
+      "learning_rate": 1.2197263809053435e-05,
+      "loss": 1.0317,
+      "step": 3245
+    },
+    {
+      "epoch": 0.4467382328654005,
+      "grad_norm": 1.8013118709417002,
+      "learning_rate": 1.2192915099191964e-05,
+      "loss": 1.1345,
+      "step": 3246
+    },
+    {
+      "epoch": 0.44687586017065783,
+      "grad_norm": 1.8993159126135937,
+      "learning_rate": 1.2188565953628278e-05,
+      "loss": 1.1772,
+      "step": 3247
+    },
+    {
+      "epoch": 0.44701348747591524,
+      "grad_norm": 1.6425834295678363,
+      "learning_rate": 1.2184216373226494e-05,
+      "loss": 1.0359,
+      "step": 3248
+    },
+    {
+      "epoch": 0.4471511147811726,
+      "grad_norm": 1.749221294967735,
+      "learning_rate": 1.2179866358850818e-05,
+      "loss": 1.0139,
+      "step": 3249
+    },
+    {
+      "epoch": 0.44728874208642994,
+      "grad_norm": 1.8998095837523783,
+      "learning_rate": 1.2175515911365535e-05,
+      "loss": 1.097,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4474263693916873,
+      "grad_norm": 1.5855957774891942,
+      "learning_rate": 1.217116503163502e-05,
+      "loss": 0.9543,
+      "step": 3251
+    },
+    {
+      "epoch": 0.4475639966969447,
+      "grad_norm": 1.6167812546582958,
+      "learning_rate": 1.2166813720523734e-05,
+      "loss": 0.99,
+      "step": 3252
+    },
+    {
+      "epoch": 0.44770162400220204,
+      "grad_norm": 1.6844539503387257,
+      "learning_rate": 1.216246197889622e-05,
+      "loss": 1.0674,
+      "step": 3253
+    },
+    {
+      "epoch": 0.4478392513074594,
+      "grad_norm": 1.6531401848885043,
+      "learning_rate": 1.2158109807617114e-05,
+      "loss": 0.9557,
+      "step": 3254
+    },
+    {
+      "epoch": 0.4479768786127168,
+      "grad_norm": 1.6572737161849602,
+      "learning_rate": 1.2153757207551129e-05,
+      "loss": 1.0477,
+      "step": 3255
+    },
+    {
+      "epoch": 0.44811450591797414,
+      "grad_norm": 1.9124510820055534,
+      "learning_rate": 1.2149404179563069e-05,
+      "loss": 1.1294,
+      "step": 3256
+    },
+    {
+      "epoch": 0.4482521332232315,
+      "grad_norm": 1.6534364697395338,
+      "learning_rate": 1.2145050724517819e-05,
+      "loss": 1.0068,
+      "step": 3257
+    },
+    {
+      "epoch": 0.44838976052848883,
+      "grad_norm": 1.962284770379394,
+      "learning_rate": 1.2140696843280355e-05,
+      "loss": 1.1436,
+      "step": 3258
+    },
+    {
+      "epoch": 0.44852738783374624,
+      "grad_norm": 1.683058512893367,
+      "learning_rate": 1.2136342536715731e-05,
+      "loss": 1.0346,
+      "step": 3259
+    },
+    {
+      "epoch": 0.4486650151390036,
+      "grad_norm": 1.6929402925395962,
+      "learning_rate": 1.2131987805689089e-05,
+      "loss": 0.9749,
+      "step": 3260
+    },
+    {
+      "epoch": 0.44880264244426094,
+      "grad_norm": 1.7051742034293074,
+      "learning_rate": 1.2127632651065652e-05,
+      "loss": 1.0327,
+      "step": 3261
+    },
+    {
+      "epoch": 0.4489402697495183,
+      "grad_norm": 1.613591598627337,
+      "learning_rate": 1.2123277073710733e-05,
+      "loss": 1.0038,
+      "step": 3262
+    },
+    {
+      "epoch": 0.4490778970547757,
+      "grad_norm": 1.7745919001342014,
+      "learning_rate": 1.211892107448972e-05,
+      "loss": 1.1127,
+      "step": 3263
+    },
+    {
+      "epoch": 0.44921552436003304,
+      "grad_norm": 1.8376395575154916,
+      "learning_rate": 1.2114564654268097e-05,
+      "loss": 1.0368,
+      "step": 3264
+    },
+    {
+      "epoch": 0.4493531516652904,
+      "grad_norm": 1.825012022181292,
+      "learning_rate": 1.2110207813911424e-05,
+      "loss": 1.0819,
+      "step": 3265
+    },
+    {
+      "epoch": 0.44949077897054773,
+      "grad_norm": 1.5116503982193217,
+      "learning_rate": 1.210585055428534e-05,
+      "loss": 1.0371,
+      "step": 3266
+    },
+    {
+      "epoch": 0.44962840627580514,
+      "grad_norm": 1.643664894329445,
+      "learning_rate": 1.2101492876255581e-05,
+      "loss": 1.0227,
+      "step": 3267
+    },
+    {
+      "epoch": 0.4497660335810625,
+      "grad_norm": 1.6713361797160597,
+      "learning_rate": 1.2097134780687954e-05,
+      "loss": 1.0699,
+      "step": 3268
+    },
+    {
+      "epoch": 0.44990366088631983,
+      "grad_norm": 1.8169204619927837,
+      "learning_rate": 1.209277626844835e-05,
+      "loss": 0.9811,
+      "step": 3269
+    },
+    {
+      "epoch": 0.4500412881915772,
+      "grad_norm": 1.6525341918012835,
+      "learning_rate": 1.2088417340402749e-05,
+      "loss": 1.0237,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4501789154968346,
+      "grad_norm": 1.757007263730867,
+      "learning_rate": 1.208405799741721e-05,
+      "loss": 1.0882,
+      "step": 3271
+    },
+    {
+      "epoch": 0.45031654280209193,
+      "grad_norm": 1.8662296996129175,
+      "learning_rate": 1.2079698240357876e-05,
+      "loss": 1.0269,
+      "step": 3272
+    },
+    {
+      "epoch": 0.4504541701073493,
+      "grad_norm": 1.7294007337929087,
+      "learning_rate": 1.2075338070090967e-05,
+      "loss": 1.0011,
+      "step": 3273
+    },
+    {
+      "epoch": 0.4505917974126067,
+      "grad_norm": 1.7574734200970403,
+      "learning_rate": 1.2070977487482793e-05,
+      "loss": 1.1127,
+      "step": 3274
+    },
+    {
+      "epoch": 0.45072942471786404,
+      "grad_norm": 1.8981318739917472,
+      "learning_rate": 1.2066616493399744e-05,
+      "loss": 1.0285,
+      "step": 3275
+    },
+    {
+      "epoch": 0.4508670520231214,
+      "grad_norm": 1.523934879663697,
+      "learning_rate": 1.2062255088708282e-05,
+      "loss": 0.986,
+      "step": 3276
+    },
+    {
+      "epoch": 0.45100467932837873,
+      "grad_norm": 1.6938745946870397,
+      "learning_rate": 1.2057893274274965e-05,
+      "loss": 1.1151,
+      "step": 3277
+    },
+    {
+      "epoch": 0.45114230663363614,
+      "grad_norm": 1.6222238094214725,
+      "learning_rate": 1.2053531050966422e-05,
+      "loss": 1.0518,
+      "step": 3278
+    },
+    {
+      "epoch": 0.4512799339388935,
+      "grad_norm": 1.769985464183407,
+      "learning_rate": 1.204916841964937e-05,
+      "loss": 1.0843,
+      "step": 3279
+    },
+    {
+      "epoch": 0.45141756124415083,
+      "grad_norm": 1.6932381614669898,
+      "learning_rate": 1.2044805381190602e-05,
+      "loss": 0.9843,
+      "step": 3280
+    },
+    {
+      "epoch": 0.4515551885494082,
+      "grad_norm": 1.646395215440809,
+      "learning_rate": 1.2040441936456998e-05,
+      "loss": 1.0624,
+      "step": 3281
+    },
+    {
+      "epoch": 0.4516928158546656,
+      "grad_norm": 1.9077963678666485,
+      "learning_rate": 1.203607808631551e-05,
+      "loss": 1.0897,
+      "step": 3282
+    },
+    {
+      "epoch": 0.45183044315992293,
+      "grad_norm": 1.4679761450586144,
+      "learning_rate": 1.2031713831633178e-05,
+      "loss": 1.0368,
+      "step": 3283
+    },
+    {
+      "epoch": 0.4519680704651803,
+      "grad_norm": 1.5926135535533286,
+      "learning_rate": 1.2027349173277118e-05,
+      "loss": 1.0364,
+      "step": 3284
+    },
+    {
+      "epoch": 0.45210569777043763,
+      "grad_norm": 1.494292963672221,
+      "learning_rate": 1.202298411211453e-05,
+      "loss": 1.0076,
+      "step": 3285
+    },
+    {
+      "epoch": 0.45224332507569504,
+      "grad_norm": 1.6864107443958332,
+      "learning_rate": 1.201861864901269e-05,
+      "loss": 1.0986,
+      "step": 3286
+    },
+    {
+      "epoch": 0.4523809523809524,
+      "grad_norm": 1.8805721551700996,
+      "learning_rate": 1.2014252784838955e-05,
+      "loss": 1.1704,
+      "step": 3287
+    },
+    {
+      "epoch": 0.45251857968620973,
+      "grad_norm": 1.853903433378315,
+      "learning_rate": 1.2009886520460765e-05,
+      "loss": 1.1493,
+      "step": 3288
+    },
+    {
+      "epoch": 0.4526562069914671,
+      "grad_norm": 1.586150426027311,
+      "learning_rate": 1.2005519856745638e-05,
+      "loss": 1.1387,
+      "step": 3289
+    },
+    {
+      "epoch": 0.4527938342967245,
+      "grad_norm": 1.9058522499999693,
+      "learning_rate": 1.2001152794561164e-05,
+      "loss": 1.157,
+      "step": 3290
+    },
+    {
+      "epoch": 0.45293146160198183,
+      "grad_norm": 1.739575629903443,
+      "learning_rate": 1.1996785334775023e-05,
+      "loss": 1.1242,
+      "step": 3291
+    },
+    {
+      "epoch": 0.4530690889072392,
+      "grad_norm": 1.6240616286441973,
+      "learning_rate": 1.1992417478254972e-05,
+      "loss": 1.0479,
+      "step": 3292
+    },
+    {
+      "epoch": 0.4532067162124966,
+      "grad_norm": 1.6226496643111894,
+      "learning_rate": 1.1988049225868839e-05,
+      "loss": 1.0278,
+      "step": 3293
+    },
+    {
+      "epoch": 0.45334434351775393,
+      "grad_norm": 1.7352197338209585,
+      "learning_rate": 1.1983680578484535e-05,
+      "loss": 1.0205,
+      "step": 3294
+    },
+    {
+      "epoch": 0.4534819708230113,
+      "grad_norm": 1.459082502483015,
+      "learning_rate": 1.197931153697006e-05,
+      "loss": 1.0409,
+      "step": 3295
+    },
+    {
+      "epoch": 0.45361959812826863,
+      "grad_norm": 1.683616513197909,
+      "learning_rate": 1.1974942102193472e-05,
+      "loss": 1.0829,
+      "step": 3296
+    },
+    {
+      "epoch": 0.45375722543352603,
+      "grad_norm": 1.5487414148744094,
+      "learning_rate": 1.1970572275022925e-05,
+      "loss": 1.0594,
+      "step": 3297
+    },
+    {
+      "epoch": 0.4538948527387834,
+      "grad_norm": 1.7601979137626567,
+      "learning_rate": 1.1966202056326638e-05,
+      "loss": 1.0983,
+      "step": 3298
+    },
+    {
+      "epoch": 0.45403248004404073,
+      "grad_norm": 1.697816890378664,
+      "learning_rate": 1.1961831446972921e-05,
+      "loss": 1.0527,
+      "step": 3299
+    },
+    {
+      "epoch": 0.4541701073492981,
+      "grad_norm": 1.8005918305171795,
+      "learning_rate": 1.1957460447830147e-05,
+      "loss": 1.0426,
+      "step": 3300
+    },
+    {
+      "epoch": 0.4543077346545555,
+      "grad_norm": 1.6480841493777627,
+      "learning_rate": 1.1953089059766778e-05,
+      "loss": 1.0703,
+      "step": 3301
+    },
+    {
+      "epoch": 0.45444536195981283,
+      "grad_norm": 1.7281580576427387,
+      "learning_rate": 1.1948717283651347e-05,
+      "loss": 1.0769,
+      "step": 3302
+    },
+    {
+      "epoch": 0.4545829892650702,
+      "grad_norm": 1.7986794938202983,
+      "learning_rate": 1.1944345120352467e-05,
+      "loss": 1.1647,
+      "step": 3303
+    },
+    {
+      "epoch": 0.45472061657032753,
+      "grad_norm": 1.919631116432969,
+      "learning_rate": 1.1939972570738827e-05,
+      "loss": 1.1783,
+      "step": 3304
+    },
+    {
+      "epoch": 0.45485824387558493,
+      "grad_norm": 1.6448367853200376,
+      "learning_rate": 1.1935599635679194e-05,
+      "loss": 1.0931,
+      "step": 3305
+    },
+    {
+      "epoch": 0.4549958711808423,
+      "grad_norm": 1.702070872290541,
+      "learning_rate": 1.1931226316042409e-05,
+      "loss": 1.0966,
+      "step": 3306
+    },
+    {
+      "epoch": 0.45513349848609963,
+      "grad_norm": 1.790366601900612,
+      "learning_rate": 1.1926852612697386e-05,
+      "loss": 1.0871,
+      "step": 3307
+    },
+    {
+      "epoch": 0.455271125791357,
+      "grad_norm": 1.7029154995010782,
+      "learning_rate": 1.1922478526513128e-05,
+      "loss": 1.1683,
+      "step": 3308
+    },
+    {
+      "epoch": 0.4554087530966144,
+      "grad_norm": 1.7941516183797472,
+      "learning_rate": 1.1918104058358701e-05,
+      "loss": 1.0944,
+      "step": 3309
+    },
+    {
+      "epoch": 0.45554638040187173,
+      "grad_norm": 1.6624531880032065,
+      "learning_rate": 1.1913729209103253e-05,
+      "loss": 1.0,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4556840077071291,
+      "grad_norm": 1.4266514133697288,
+      "learning_rate": 1.1909353979616009e-05,
+      "loss": 1.0173,
+      "step": 3311
+    },
+    {
+      "epoch": 0.4558216350123865,
+      "grad_norm": 1.785067928838574,
+      "learning_rate": 1.1904978370766267e-05,
+      "loss": 1.1155,
+      "step": 3312
+    },
+    {
+      "epoch": 0.45595926231764383,
+      "grad_norm": 1.7162788001945006,
+      "learning_rate": 1.1900602383423397e-05,
+      "loss": 1.1032,
+      "step": 3313
+    },
+    {
+      "epoch": 0.4560968896229012,
+      "grad_norm": 1.8060119790215499,
+      "learning_rate": 1.189622601845685e-05,
+      "loss": 1.0749,
+      "step": 3314
+    },
+    {
+      "epoch": 0.45623451692815853,
+      "grad_norm": 1.6414380294504327,
+      "learning_rate": 1.1891849276736151e-05,
+      "loss": 1.0991,
+      "step": 3315
+    },
+    {
+      "epoch": 0.45637214423341593,
+      "grad_norm": 1.690254162608865,
+      "learning_rate": 1.1887472159130895e-05,
+      "loss": 1.1257,
+      "step": 3316
+    },
+    {
+      "epoch": 0.4565097715386733,
+      "grad_norm": 1.8865705721325123,
+      "learning_rate": 1.188309466651076e-05,
+      "loss": 1.1003,
+      "step": 3317
+    },
+    {
+      "epoch": 0.45664739884393063,
+      "grad_norm": 1.895780507482217,
+      "learning_rate": 1.187871679974549e-05,
+      "loss": 1.1299,
+      "step": 3318
+    },
+    {
+      "epoch": 0.456785026149188,
+      "grad_norm": 1.998893476419672,
+      "learning_rate": 1.1874338559704909e-05,
+      "loss": 1.0949,
+      "step": 3319
+    },
+    {
+      "epoch": 0.4569226534544454,
+      "grad_norm": 1.5010609951701976,
+      "learning_rate": 1.1869959947258913e-05,
+      "loss": 1.1002,
+      "step": 3320
+    },
+    {
+      "epoch": 0.45706028075970273,
+      "grad_norm": 1.925106912353012,
+      "learning_rate": 1.1865580963277474e-05,
+      "loss": 1.2157,
+      "step": 3321
+    },
+    {
+      "epoch": 0.4571979080649601,
+      "grad_norm": 1.7281530797737379,
+      "learning_rate": 1.1861201608630632e-05,
+      "loss": 1.1124,
+      "step": 3322
+    },
+    {
+      "epoch": 0.4573355353702174,
+      "grad_norm": 1.640029272841053,
+      "learning_rate": 1.185682188418851e-05,
+      "loss": 1.0297,
+      "step": 3323
+    },
+    {
+      "epoch": 0.45747316267547483,
+      "grad_norm": 1.6208714907666426,
+      "learning_rate": 1.1852441790821297e-05,
+      "loss": 1.0323,
+      "step": 3324
+    },
+    {
+      "epoch": 0.4576107899807322,
+      "grad_norm": 1.6224707938246572,
+      "learning_rate": 1.1848061329399252e-05,
+      "loss": 1.0842,
+      "step": 3325
+    },
+    {
+      "epoch": 0.45774841728598953,
+      "grad_norm": 1.7832473801855395,
+      "learning_rate": 1.1843680500792723e-05,
+      "loss": 0.9929,
+      "step": 3326
+    },
+    {
+      "epoch": 0.4578860445912469,
+      "grad_norm": 1.602852670317834,
+      "learning_rate": 1.1839299305872114e-05,
+      "loss": 0.9024,
+      "step": 3327
+    },
+    {
+      "epoch": 0.4580236718965043,
+      "grad_norm": 1.7475116456703832,
+      "learning_rate": 1.1834917745507911e-05,
+      "loss": 1.107,
+      "step": 3328
+    },
+    {
+      "epoch": 0.45816129920176163,
+      "grad_norm": 1.6957599958994916,
+      "learning_rate": 1.183053582057067e-05,
+      "loss": 1.0747,
+      "step": 3329
+    },
+    {
+      "epoch": 0.458298926507019,
+      "grad_norm": 1.8365986028534491,
+      "learning_rate": 1.1826153531931018e-05,
+      "loss": 1.0982,
+      "step": 3330
+    },
+    {
+      "epoch": 0.4584365538122764,
+      "grad_norm": 1.7364651026027218,
+      "learning_rate": 1.1821770880459655e-05,
+      "loss": 1.0513,
+      "step": 3331
+    },
+    {
+      "epoch": 0.45857418111753373,
+      "grad_norm": 1.834959931027001,
+      "learning_rate": 1.1817387867027357e-05,
+      "loss": 1.0597,
+      "step": 3332
+    },
+    {
+      "epoch": 0.4587118084227911,
+      "grad_norm": 1.9666407652609499,
+      "learning_rate": 1.1813004492504967e-05,
+      "loss": 1.0657,
+      "step": 3333
+    },
+    {
+      "epoch": 0.4588494357280484,
+      "grad_norm": 1.6711532701038987,
+      "learning_rate": 1.1808620757763403e-05,
+      "loss": 1.0397,
+      "step": 3334
+    },
+    {
+      "epoch": 0.45898706303330583,
+      "grad_norm": 1.681638195922923,
+      "learning_rate": 1.180423666367365e-05,
+      "loss": 0.9255,
+      "step": 3335
+    },
+    {
+      "epoch": 0.4591246903385632,
+      "grad_norm": 1.6853010186423814,
+      "learning_rate": 1.1799852211106768e-05,
+      "loss": 1.0736,
+      "step": 3336
+    },
+    {
+      "epoch": 0.45926231764382053,
+      "grad_norm": 1.7557304107832628,
+      "learning_rate": 1.1795467400933891e-05,
+      "loss": 1.001,
+      "step": 3337
+    },
+    {
+      "epoch": 0.4593999449490779,
+      "grad_norm": 1.713567640099152,
+      "learning_rate": 1.1791082234026218e-05,
+      "loss": 1.02,
+      "step": 3338
+    },
+    {
+      "epoch": 0.4595375722543353,
+      "grad_norm": 1.8168813446596255,
+      "learning_rate": 1.1786696711255024e-05,
+      "loss": 1.1123,
+      "step": 3339
+    },
+    {
+      "epoch": 0.45967519955959263,
+      "grad_norm": 1.702815453978539,
+      "learning_rate": 1.1782310833491648e-05,
+      "loss": 0.9685,
+      "step": 3340
+    },
+    {
+      "epoch": 0.45981282686485,
+      "grad_norm": 1.6684501250686197,
+      "learning_rate": 1.1777924601607507e-05,
+      "loss": 0.9983,
+      "step": 3341
+    },
+    {
+      "epoch": 0.4599504541701073,
+      "grad_norm": 1.8863530593523867,
+      "learning_rate": 1.1773538016474085e-05,
+      "loss": 1.1947,
+      "step": 3342
+    },
+    {
+      "epoch": 0.46008808147536473,
+      "grad_norm": 1.7037891330902808,
+      "learning_rate": 1.1769151078962936e-05,
+      "loss": 1.0483,
+      "step": 3343
+    },
+    {
+      "epoch": 0.4602257087806221,
+      "grad_norm": 1.8045849099551563,
+      "learning_rate": 1.1764763789945686e-05,
+      "loss": 1.0213,
+      "step": 3344
+    },
+    {
+      "epoch": 0.4603633360858794,
+      "grad_norm": 1.7849923401604424,
+      "learning_rate": 1.1760376150294026e-05,
+      "loss": 0.9848,
+      "step": 3345
+    },
+    {
+      "epoch": 0.4605009633911368,
+      "grad_norm": 1.8904809312620632,
+      "learning_rate": 1.1755988160879723e-05,
+      "loss": 1.0098,
+      "step": 3346
+    },
+    {
+      "epoch": 0.4606385906963942,
+      "grad_norm": 1.790989122054825,
+      "learning_rate": 1.175159982257461e-05,
+      "loss": 1.0134,
+      "step": 3347
+    },
+    {
+      "epoch": 0.4607762180016515,
+      "grad_norm": 1.9736653760979912,
+      "learning_rate": 1.174721113625059e-05,
+      "loss": 1.1192,
+      "step": 3348
+    },
+    {
+      "epoch": 0.4609138453069089,
+      "grad_norm": 1.7632300782153767,
+      "learning_rate": 1.1742822102779632e-05,
+      "loss": 1.0644,
+      "step": 3349
+    },
+    {
+      "epoch": 0.4610514726121663,
+      "grad_norm": 1.8224722920191472,
+      "learning_rate": 1.1738432723033778e-05,
+      "loss": 1.0307,
+      "step": 3350
+    },
+    {
+      "epoch": 0.46118909991742363,
+      "grad_norm": 1.7387714339318252,
+      "learning_rate": 1.173404299788514e-05,
+      "loss": 1.055,
+      "step": 3351
+    },
+    {
+      "epoch": 0.461326727222681,
+      "grad_norm": 1.625768817429624,
+      "learning_rate": 1.1729652928205894e-05,
+      "loss": 1.0091,
+      "step": 3352
+    },
+    {
+      "epoch": 0.4614643545279383,
+      "grad_norm": 1.5866465824220013,
+      "learning_rate": 1.172526251486829e-05,
+      "loss": 1.106,
+      "step": 3353
+    },
+    {
+      "epoch": 0.46160198183319573,
+      "grad_norm": 1.772910385935587,
+      "learning_rate": 1.1720871758744639e-05,
+      "loss": 1.0737,
+      "step": 3354
+    },
+    {
+      "epoch": 0.4617396091384531,
+      "grad_norm": 1.7727922271345744,
+      "learning_rate": 1.1716480660707326e-05,
+      "loss": 0.982,
+      "step": 3355
+    },
+    {
+      "epoch": 0.4618772364437104,
+      "grad_norm": 1.7444197284708047,
+      "learning_rate": 1.1712089221628797e-05,
+      "loss": 1.0881,
+      "step": 3356
+    },
+    {
+      "epoch": 0.4620148637489678,
+      "grad_norm": 1.7351443967909226,
+      "learning_rate": 1.1707697442381581e-05,
+      "loss": 1.1493,
+      "step": 3357
+    },
+    {
+      "epoch": 0.4621524910542252,
+      "grad_norm": 1.695813585179952,
+      "learning_rate": 1.1703305323838257e-05,
+      "loss": 1.0356,
+      "step": 3358
+    },
+    {
+      "epoch": 0.4622901183594825,
+      "grad_norm": 1.9009294636616443,
+      "learning_rate": 1.169891286687148e-05,
+      "loss": 1.0604,
+      "step": 3359
+    },
+    {
+      "epoch": 0.4624277456647399,
+      "grad_norm": 1.5970016688460025,
+      "learning_rate": 1.1694520072353977e-05,
+      "loss": 1.1069,
+      "step": 3360
+    },
+    {
+      "epoch": 0.4625653729699972,
+      "grad_norm": 1.6844374968728684,
+      "learning_rate": 1.1690126941158526e-05,
+      "loss": 1.042,
+      "step": 3361
+    },
+    {
+      "epoch": 0.46270300027525463,
+      "grad_norm": 1.6726448337287079,
+      "learning_rate": 1.168573347415799e-05,
+      "loss": 1.0164,
+      "step": 3362
+    },
+    {
+      "epoch": 0.462840627580512,
+      "grad_norm": 1.6783630302453567,
+      "learning_rate": 1.1681339672225287e-05,
+      "loss": 1.0263,
+      "step": 3363
+    },
+    {
+      "epoch": 0.4629782548857693,
+      "grad_norm": 1.6010225956441717,
+      "learning_rate": 1.1676945536233404e-05,
+      "loss": 1.1162,
+      "step": 3364
+    },
+    {
+      "epoch": 0.4631158821910267,
+      "grad_norm": 1.9479987567053825,
+      "learning_rate": 1.1672551067055397e-05,
+      "loss": 1.0369,
+      "step": 3365
+    },
+    {
+      "epoch": 0.4632535094962841,
+      "grad_norm": 1.9946391623919404,
+      "learning_rate": 1.1668156265564395e-05,
+      "loss": 1.0509,
+      "step": 3366
+    },
+    {
+      "epoch": 0.4633911368015414,
+      "grad_norm": 1.7654512266198432,
+      "learning_rate": 1.166376113263357e-05,
+      "loss": 1.0193,
+      "step": 3367
+    },
+    {
+      "epoch": 0.4635287641067988,
+      "grad_norm": 1.671398476523646,
+      "learning_rate": 1.1659365669136187e-05,
+      "loss": 1.0309,
+      "step": 3368
+    },
+    {
+      "epoch": 0.4636663914120562,
+      "grad_norm": 1.683970298323594,
+      "learning_rate": 1.165496987594556e-05,
+      "loss": 1.0473,
+      "step": 3369
+    },
+    {
+      "epoch": 0.4638040187173135,
+      "grad_norm": 1.7217534447874112,
+      "learning_rate": 1.1650573753935067e-05,
+      "loss": 1.0535,
+      "step": 3370
+    },
+    {
+      "epoch": 0.4639416460225709,
+      "grad_norm": 1.7075654309394153,
+      "learning_rate": 1.1646177303978167e-05,
+      "loss": 1.0059,
+      "step": 3371
+    },
+    {
+      "epoch": 0.4640792733278282,
+      "grad_norm": 1.6761899959669933,
+      "learning_rate": 1.164178052694837e-05,
+      "loss": 1.0365,
+      "step": 3372
+    },
+    {
+      "epoch": 0.4642169006330856,
+      "grad_norm": 1.6937540134544466,
+      "learning_rate": 1.1637383423719255e-05,
+      "loss": 1.0402,
+      "step": 3373
+    },
+    {
+      "epoch": 0.464354527938343,
+      "grad_norm": 1.6037158561675933,
+      "learning_rate": 1.163298599516447e-05,
+      "loss": 1.1022,
+      "step": 3374
+    },
+    {
+      "epoch": 0.4644921552436003,
+      "grad_norm": 1.8083468429130618,
+      "learning_rate": 1.1628588242157719e-05,
+      "loss": 1.1337,
+      "step": 3375
+    },
+    {
+      "epoch": 0.4646297825488577,
+      "grad_norm": 1.7244388191904836,
+      "learning_rate": 1.1624190165572777e-05,
+      "loss": 1.0884,
+      "step": 3376
+    },
+    {
+      "epoch": 0.4647674098541151,
+      "grad_norm": 1.6416748204742255,
+      "learning_rate": 1.1619791766283481e-05,
+      "loss": 1.0445,
+      "step": 3377
+    },
+    {
+      "epoch": 0.4649050371593724,
+      "grad_norm": 1.5745217862949055,
+      "learning_rate": 1.1615393045163734e-05,
+      "loss": 1.0214,
+      "step": 3378
+    },
+    {
+      "epoch": 0.4650426644646298,
+      "grad_norm": 1.6715018007330258,
+      "learning_rate": 1.1610994003087504e-05,
+      "loss": 1.1757,
+      "step": 3379
+    },
+    {
+      "epoch": 0.4651802917698871,
+      "grad_norm": 1.6877155258332566,
+      "learning_rate": 1.1606594640928815e-05,
+      "loss": 1.0116,
+      "step": 3380
+    },
+    {
+      "epoch": 0.4653179190751445,
+      "grad_norm": 1.65126399078884,
+      "learning_rate": 1.1602194959561761e-05,
+      "loss": 1.0235,
+      "step": 3381
+    },
+    {
+      "epoch": 0.4654555463804019,
+      "grad_norm": 1.6992252131651249,
+      "learning_rate": 1.1597794959860505e-05,
+      "loss": 1.0953,
+      "step": 3382
+    },
+    {
+      "epoch": 0.4655931736856592,
+      "grad_norm": 1.7955430817009295,
+      "learning_rate": 1.1593394642699258e-05,
+      "loss": 1.1016,
+      "step": 3383
+    },
+    {
+      "epoch": 0.46573080099091657,
+      "grad_norm": 1.5867435642542689,
+      "learning_rate": 1.1588994008952307e-05,
+      "loss": 1.0749,
+      "step": 3384
+    },
+    {
+      "epoch": 0.465868428296174,
+      "grad_norm": 1.5699320398519483,
+      "learning_rate": 1.1584593059494e-05,
+      "loss": 0.9877,
+      "step": 3385
+    },
+    {
+      "epoch": 0.4660060556014313,
+      "grad_norm": 1.5605191647765546,
+      "learning_rate": 1.158019179519874e-05,
+      "loss": 0.9822,
+      "step": 3386
+    },
+    {
+      "epoch": 0.46614368290668867,
+      "grad_norm": 1.7702375449651875,
+      "learning_rate": 1.1575790216941e-05,
+      "loss": 1.0368,
+      "step": 3387
+    },
+    {
+      "epoch": 0.4662813102119461,
+      "grad_norm": 1.6651215668810446,
+      "learning_rate": 1.1571388325595319e-05,
+      "loss": 1.0527,
+      "step": 3388
+    },
+    {
+      "epoch": 0.4664189375172034,
+      "grad_norm": 1.7320979204382785,
+      "learning_rate": 1.1566986122036283e-05,
+      "loss": 1.1581,
+      "step": 3389
+    },
+    {
+      "epoch": 0.4665565648224608,
+      "grad_norm": 1.7421398836450757,
+      "learning_rate": 1.1562583607138556e-05,
+      "loss": 1.0984,
+      "step": 3390
+    },
+    {
+      "epoch": 0.4666941921277181,
+      "grad_norm": 1.5684890151003559,
+      "learning_rate": 1.1558180781776856e-05,
+      "loss": 1.0758,
+      "step": 3391
+    },
+    {
+      "epoch": 0.4668318194329755,
+      "grad_norm": 1.7065370494719578,
+      "learning_rate": 1.1553777646825963e-05,
+      "loss": 1.0616,
+      "step": 3392
+    },
+    {
+      "epoch": 0.4669694467382329,
+      "grad_norm": 1.6955138726608856,
+      "learning_rate": 1.154937420316072e-05,
+      "loss": 1.0649,
+      "step": 3393
+    },
+    {
+      "epoch": 0.4671070740434902,
+      "grad_norm": 1.5949273381469316,
+      "learning_rate": 1.1544970451656035e-05,
+      "loss": 1.0605,
+      "step": 3394
+    },
+    {
+      "epoch": 0.46724470134874757,
+      "grad_norm": 1.722080743848845,
+      "learning_rate": 1.1540566393186868e-05,
+      "loss": 1.0546,
+      "step": 3395
+    },
+    {
+      "epoch": 0.467382328654005,
+      "grad_norm": 1.8149888496462947,
+      "learning_rate": 1.1536162028628242e-05,
+      "loss": 1.0735,
+      "step": 3396
+    },
+    {
+      "epoch": 0.4675199559592623,
+      "grad_norm": 1.5560808303868876,
+      "learning_rate": 1.1531757358855253e-05,
+      "loss": 1.0552,
+      "step": 3397
+    },
+    {
+      "epoch": 0.46765758326451967,
+      "grad_norm": 1.6820785392575153,
+      "learning_rate": 1.1527352384743047e-05,
+      "loss": 0.9632,
+      "step": 3398
+    },
+    {
+      "epoch": 0.467795210569777,
+      "grad_norm": 1.9090923868548884,
+      "learning_rate": 1.1522947107166825e-05,
+      "loss": 1.0176,
+      "step": 3399
+    },
+    {
+      "epoch": 0.4679328378750344,
+      "grad_norm": 1.7521702061489202,
+      "learning_rate": 1.151854152700186e-05,
+      "loss": 1.017,
+      "step": 3400
+    },
+    {
+      "epoch": 0.4680704651802918,
+      "grad_norm": 1.9316621401191454,
+      "learning_rate": 1.151413564512348e-05,
+      "loss": 1.1116,
+      "step": 3401
+    },
+    {
+      "epoch": 0.4682080924855491,
+      "grad_norm": 1.7498879378255123,
+      "learning_rate": 1.1509729462407075e-05,
+      "loss": 1.0817,
+      "step": 3402
+    },
+    {
+      "epoch": 0.46834571979080647,
+      "grad_norm": 1.645633367704172,
+      "learning_rate": 1.150532297972809e-05,
+      "loss": 1.0759,
+      "step": 3403
+    },
+    {
+      "epoch": 0.4684833470960639,
+      "grad_norm": 1.799876086541932,
+      "learning_rate": 1.1500916197962036e-05,
+      "loss": 1.1274,
+      "step": 3404
+    },
+    {
+      "epoch": 0.4686209744013212,
+      "grad_norm": 1.6577174883923727,
+      "learning_rate": 1.1496509117984478e-05,
+      "loss": 1.0407,
+      "step": 3405
+    },
+    {
+      "epoch": 0.46875860170657857,
+      "grad_norm": 1.5964320157311451,
+      "learning_rate": 1.1492101740671044e-05,
+      "loss": 1.0631,
+      "step": 3406
+    },
+    {
+      "epoch": 0.468896229011836,
+      "grad_norm": 1.9886570476945782,
+      "learning_rate": 1.1487694066897419e-05,
+      "loss": 1.0381,
+      "step": 3407
+    },
+    {
+      "epoch": 0.4690338563170933,
+      "grad_norm": 1.9360353433469417,
+      "learning_rate": 1.1483286097539348e-05,
+      "loss": 1.0262,
+      "step": 3408
+    },
+    {
+      "epoch": 0.46917148362235067,
+      "grad_norm": 1.6249804209024834,
+      "learning_rate": 1.147887783347263e-05,
+      "loss": 1.047,
+      "step": 3409
+    },
+    {
+      "epoch": 0.469309110927608,
+      "grad_norm": 1.6494391231711798,
+      "learning_rate": 1.1474469275573135e-05,
+      "loss": 1.0139,
+      "step": 3410
+    },
+    {
+      "epoch": 0.4694467382328654,
+      "grad_norm": 1.6812721140218512,
+      "learning_rate": 1.1470060424716773e-05,
+      "loss": 0.9263,
+      "step": 3411
+    },
+    {
+      "epoch": 0.46958436553812277,
+      "grad_norm": 1.7196591614099348,
+      "learning_rate": 1.1465651281779529e-05,
+      "loss": 1.0832,
+      "step": 3412
+    },
+    {
+      "epoch": 0.4697219928433801,
+      "grad_norm": 1.9418529126139144,
+      "learning_rate": 1.146124184763744e-05,
+      "loss": 1.0232,
+      "step": 3413
+    },
+    {
+      "epoch": 0.46985962014863747,
+      "grad_norm": 1.690681684076394,
+      "learning_rate": 1.14568321231666e-05,
+      "loss": 0.9815,
+      "step": 3414
+    },
+    {
+      "epoch": 0.4699972474538949,
+      "grad_norm": 1.7913465004547984,
+      "learning_rate": 1.1452422109243155e-05,
+      "loss": 1.0586,
+      "step": 3415
+    },
+    {
+      "epoch": 0.4701348747591522,
+      "grad_norm": 1.7517903052122032,
+      "learning_rate": 1.144801180674332e-05,
+      "loss": 1.0372,
+      "step": 3416
+    },
+    {
+      "epoch": 0.47027250206440957,
+      "grad_norm": 1.8409775017330277,
+      "learning_rate": 1.1443601216543359e-05,
+      "loss": 1.1062,
+      "step": 3417
+    },
+    {
+      "epoch": 0.4704101293696669,
+      "grad_norm": 1.7501999706791864,
+      "learning_rate": 1.1439190339519597e-05,
+      "loss": 1.0958,
+      "step": 3418
+    },
+    {
+      "epoch": 0.4705477566749243,
+      "grad_norm": 1.848297711350787,
+      "learning_rate": 1.1434779176548414e-05,
+      "loss": 1.0542,
+      "step": 3419
+    },
+    {
+      "epoch": 0.47068538398018167,
+      "grad_norm": 1.6400020609396726,
+      "learning_rate": 1.1430367728506251e-05,
+      "loss": 1.0036,
+      "step": 3420
+    },
+    {
+      "epoch": 0.470823011285439,
+      "grad_norm": 1.87957309623098,
+      "learning_rate": 1.1425955996269597e-05,
+      "loss": 1.0706,
+      "step": 3421
+    },
+    {
+      "epoch": 0.47096063859069637,
+      "grad_norm": 1.7924968076770076,
+      "learning_rate": 1.1421543980715009e-05,
+      "loss": 1.1028,
+      "step": 3422
+    },
+    {
+      "epoch": 0.47109826589595377,
+      "grad_norm": 1.9061538869423025,
+      "learning_rate": 1.1417131682719088e-05,
+      "loss": 1.0631,
+      "step": 3423
+    },
+    {
+      "epoch": 0.4712358932012111,
+      "grad_norm": 1.7290513556324654,
+      "learning_rate": 1.1412719103158503e-05,
+      "loss": 1.0742,
+      "step": 3424
+    },
+    {
+      "epoch": 0.47137352050646847,
+      "grad_norm": 1.878145688288672,
+      "learning_rate": 1.140830624290997e-05,
+      "loss": 1.079,
+      "step": 3425
+    },
+    {
+      "epoch": 0.4715111478117259,
+      "grad_norm": 1.7697409137465385,
+      "learning_rate": 1.1403893102850263e-05,
+      "loss": 1.0036,
+      "step": 3426
+    },
+    {
+      "epoch": 0.4716487751169832,
+      "grad_norm": 1.6463625722631452,
+      "learning_rate": 1.1399479683856211e-05,
+      "loss": 0.9507,
+      "step": 3427
+    },
+    {
+      "epoch": 0.47178640242224057,
+      "grad_norm": 1.7485552427334696,
+      "learning_rate": 1.1395065986804705e-05,
+      "loss": 1.0057,
+      "step": 3428
+    },
+    {
+      "epoch": 0.4719240297274979,
+      "grad_norm": 1.804487396361699,
+      "learning_rate": 1.1390652012572684e-05,
+      "loss": 1.0873,
+      "step": 3429
+    },
+    {
+      "epoch": 0.4720616570327553,
+      "grad_norm": 1.7211790340482807,
+      "learning_rate": 1.1386237762037147e-05,
+      "loss": 0.9663,
+      "step": 3430
+    },
+    {
+      "epoch": 0.47219928433801267,
+      "grad_norm": 1.9535106987321724,
+      "learning_rate": 1.138182323607514e-05,
+      "loss": 1.113,
+      "step": 3431
+    },
+    {
+      "epoch": 0.47233691164327,
+      "grad_norm": 2.016452622312912,
+      "learning_rate": 1.137740843556377e-05,
+      "loss": 1.1021,
+      "step": 3432
+    },
+    {
+      "epoch": 0.47247453894852737,
+      "grad_norm": 1.8736887109904863,
+      "learning_rate": 1.1372993361380199e-05,
+      "loss": 1.0398,
+      "step": 3433
+    },
+    {
+      "epoch": 0.47261216625378477,
+      "grad_norm": 2.0374779129456417,
+      "learning_rate": 1.1368578014401638e-05,
+      "loss": 1.0328,
+      "step": 3434
+    },
+    {
+      "epoch": 0.4727497935590421,
+      "grad_norm": 1.7800147182598518,
+      "learning_rate": 1.1364162395505365e-05,
+      "loss": 1.0623,
+      "step": 3435
+    },
+    {
+      "epoch": 0.47288742086429947,
+      "grad_norm": 2.035339219883238,
+      "learning_rate": 1.1359746505568694e-05,
+      "loss": 1.0733,
+      "step": 3436
+    },
+    {
+      "epoch": 0.4730250481695568,
+      "grad_norm": 1.7423984305643005,
+      "learning_rate": 1.1355330345469006e-05,
+      "loss": 1.0837,
+      "step": 3437
+    },
+    {
+      "epoch": 0.4731626754748142,
+      "grad_norm": 1.7753299306354817,
+      "learning_rate": 1.1350913916083733e-05,
+      "loss": 1.006,
+      "step": 3438
+    },
+    {
+      "epoch": 0.47330030278007157,
+      "grad_norm": 1.9189751182574306,
+      "learning_rate": 1.1346497218290354e-05,
+      "loss": 1.067,
+      "step": 3439
+    },
+    {
+      "epoch": 0.4734379300853289,
+      "grad_norm": 1.6247065491550436,
+      "learning_rate": 1.134208025296641e-05,
+      "loss": 1.014,
+      "step": 3440
+    },
+    {
+      "epoch": 0.47357555739058627,
+      "grad_norm": 2.032874036597277,
+      "learning_rate": 1.133766302098949e-05,
+      "loss": 1.125,
+      "step": 3441
+    },
+    {
+      "epoch": 0.47371318469584367,
+      "grad_norm": 1.7252935866372037,
+      "learning_rate": 1.1333245523237239e-05,
+      "loss": 1.0058,
+      "step": 3442
+    },
+    {
+      "epoch": 0.473850812001101,
+      "grad_norm": 1.6383320135625083,
+      "learning_rate": 1.1328827760587353e-05,
+      "loss": 1.0339,
+      "step": 3443
+    },
+    {
+      "epoch": 0.47398843930635837,
+      "grad_norm": 1.8619014040152602,
+      "learning_rate": 1.1324409733917581e-05,
+      "loss": 0.9868,
+      "step": 3444
+    },
+    {
+      "epoch": 0.47412606661161577,
+      "grad_norm": 1.7512250477107274,
+      "learning_rate": 1.1319991444105722e-05,
+      "loss": 1.1194,
+      "step": 3445
+    },
+    {
+      "epoch": 0.4742636939168731,
+      "grad_norm": 1.7105090113238877,
+      "learning_rate": 1.1315572892029635e-05,
+      "loss": 1.1826,
+      "step": 3446
+    },
+    {
+      "epoch": 0.47440132122213047,
+      "grad_norm": 1.7033523678449303,
+      "learning_rate": 1.1311154078567224e-05,
+      "loss": 1.1011,
+      "step": 3447
+    },
+    {
+      "epoch": 0.4745389485273878,
+      "grad_norm": 1.7664853235559337,
+      "learning_rate": 1.130673500459644e-05,
+      "loss": 1.0836,
+      "step": 3448
+    },
+    {
+      "epoch": 0.4746765758326452,
+      "grad_norm": 1.6825428252960772,
+      "learning_rate": 1.13023156709953e-05,
+      "loss": 1.1703,
+      "step": 3449
+    },
+    {
+      "epoch": 0.47481420313790257,
+      "grad_norm": 1.5872166021080731,
+      "learning_rate": 1.1297896078641864e-05,
+      "loss": 1.117,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4749518304431599,
+      "grad_norm": 1.6726504994266929,
+      "learning_rate": 1.1293476228414242e-05,
+      "loss": 1.0637,
+      "step": 3451
+    },
+    {
+      "epoch": 0.47508945774841727,
+      "grad_norm": 2.335500491252527,
+      "learning_rate": 1.12890561211906e-05,
+      "loss": 1.061,
+      "step": 3452
+    },
+    {
+      "epoch": 0.47522708505367467,
+      "grad_norm": 1.7906724471640978,
+      "learning_rate": 1.1284635757849151e-05,
+      "loss": 1.0241,
+      "step": 3453
+    },
+    {
+      "epoch": 0.475364712358932,
+      "grad_norm": 1.8597762926035852,
+      "learning_rate": 1.1280215139268163e-05,
+      "loss": 0.9982,
+      "step": 3454
+    },
+    {
+      "epoch": 0.47550233966418937,
+      "grad_norm": 1.7953922989735378,
+      "learning_rate": 1.127579426632595e-05,
+      "loss": 1.0546,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4756399669694467,
+      "grad_norm": 1.4927847270143364,
+      "learning_rate": 1.127137313990088e-05,
+      "loss": 1.0,
+      "step": 3456
+    },
+    {
+      "epoch": 0.4757775942747041,
+      "grad_norm": 1.8142620792748545,
+      "learning_rate": 1.1266951760871372e-05,
+      "loss": 1.1491,
+      "step": 3457
+    },
+    {
+      "epoch": 0.47591522157996147,
+      "grad_norm": 1.9029938595405198,
+      "learning_rate": 1.1262530130115887e-05,
+      "loss": 1.0352,
+      "step": 3458
+    },
+    {
+      "epoch": 0.4760528488852188,
+      "grad_norm": 1.7609993143946812,
+      "learning_rate": 1.1258108248512951e-05,
+      "loss": 1.0354,
+      "step": 3459
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 1.789998098092864,
+      "learning_rate": 1.1253686116941129e-05,
+      "loss": 0.9582,
+      "step": 3460
+    },
+    {
+      "epoch": 0.47632810349573357,
+      "grad_norm": 1.8523481631060819,
+      "learning_rate": 1.1249263736279035e-05,
+      "loss": 1.2396,
+      "step": 3461
+    },
+    {
+      "epoch": 0.4764657308009909,
+      "grad_norm": 1.7705459937680121,
+      "learning_rate": 1.124484110740534e-05,
+      "loss": 1.0169,
+      "step": 3462
+    },
+    {
+      "epoch": 0.47660335810624826,
+      "grad_norm": 1.5814134703884635,
+      "learning_rate": 1.1240418231198753e-05,
+      "loss": 1.0413,
+      "step": 3463
+    },
+    {
+      "epoch": 0.47674098541150567,
+      "grad_norm": 1.7999615523373897,
+      "learning_rate": 1.1235995108538045e-05,
+      "loss": 1.0507,
+      "step": 3464
+    },
+    {
+      "epoch": 0.476878612716763,
+      "grad_norm": 2.2425542359294135,
+      "learning_rate": 1.1231571740302025e-05,
+      "loss": 1.058,
+      "step": 3465
+    },
+    {
+      "epoch": 0.47701624002202037,
+      "grad_norm": 1.5572219730913666,
+      "learning_rate": 1.122714812736956e-05,
+      "loss": 1.0715,
+      "step": 3466
+    },
+    {
+      "epoch": 0.4771538673272777,
+      "grad_norm": 1.6446105254722247,
+      "learning_rate": 1.1222724270619563e-05,
+      "loss": 1.0847,
+      "step": 3467
+    },
+    {
+      "epoch": 0.4772914946325351,
+      "grad_norm": 1.8424139326415117,
+      "learning_rate": 1.1218300170930987e-05,
+      "loss": 1.0688,
+      "step": 3468
+    },
+    {
+      "epoch": 0.47742912193779247,
+      "grad_norm": 1.696170750012753,
+      "learning_rate": 1.1213875829182845e-05,
+      "loss": 1.0131,
+      "step": 3469
+    },
+    {
+      "epoch": 0.4775667492430498,
+      "grad_norm": 1.6668185977659102,
+      "learning_rate": 1.1209451246254193e-05,
+      "loss": 1.0555,
+      "step": 3470
+    },
+    {
+      "epoch": 0.47770437654830716,
+      "grad_norm": 1.6368871789609842,
+      "learning_rate": 1.1205026423024132e-05,
+      "loss": 1.0083,
+      "step": 3471
+    },
+    {
+      "epoch": 0.47784200385356457,
+      "grad_norm": 1.8113324600535936,
+      "learning_rate": 1.1200601360371816e-05,
+      "loss": 1.098,
+      "step": 3472
+    },
+    {
+      "epoch": 0.4779796311588219,
+      "grad_norm": 1.6570845094441033,
+      "learning_rate": 1.1196176059176446e-05,
+      "loss": 1.0723,
+      "step": 3473
+    },
+    {
+      "epoch": 0.47811725846407926,
+      "grad_norm": 1.793862760962084,
+      "learning_rate": 1.1191750520317264e-05,
+      "loss": 1.0552,
+      "step": 3474
+    },
+    {
+      "epoch": 0.4782548857693366,
+      "grad_norm": 1.5711821602784453,
+      "learning_rate": 1.1187324744673569e-05,
+      "loss": 1.0843,
+      "step": 3475
+    },
+    {
+      "epoch": 0.478392513074594,
+      "grad_norm": 1.8213396747572819,
+      "learning_rate": 1.1182898733124699e-05,
+      "loss": 1.0545,
+      "step": 3476
+    },
+    {
+      "epoch": 0.47853014037985137,
+      "grad_norm": 1.6779529333384606,
+      "learning_rate": 1.1178472486550043e-05,
+      "loss": 0.9788,
+      "step": 3477
+    },
+    {
+      "epoch": 0.4786677676851087,
+      "grad_norm": 1.8718176639067832,
+      "learning_rate": 1.1174046005829038e-05,
+      "loss": 1.0809,
+      "step": 3478
+    },
+    {
+      "epoch": 0.47880539499036606,
+      "grad_norm": 1.6851169197608131,
+      "learning_rate": 1.116961929184116e-05,
+      "loss": 0.9601,
+      "step": 3479
+    },
+    {
+      "epoch": 0.47894302229562347,
+      "grad_norm": 1.538320518285011,
+      "learning_rate": 1.1165192345465941e-05,
+      "loss": 1.0137,
+      "step": 3480
+    },
+    {
+      "epoch": 0.4790806496008808,
+      "grad_norm": 1.747098512522145,
+      "learning_rate": 1.1160765167582947e-05,
+      "loss": 1.0023,
+      "step": 3481
+    },
+    {
+      "epoch": 0.47921827690613816,
+      "grad_norm": 1.8676841228507637,
+      "learning_rate": 1.115633775907181e-05,
+      "loss": 0.9468,
+      "step": 3482
+    },
+    {
+      "epoch": 0.47935590421139557,
+      "grad_norm": 1.7011510029279595,
+      "learning_rate": 1.1151910120812185e-05,
+      "loss": 1.0207,
+      "step": 3483
+    },
+    {
+      "epoch": 0.4794935315166529,
+      "grad_norm": 1.7519314114814235,
+      "learning_rate": 1.1147482253683788e-05,
+      "loss": 1.0562,
+      "step": 3484
+    },
+    {
+      "epoch": 0.47963115882191026,
+      "grad_norm": 1.6128264537500054,
+      "learning_rate": 1.1143054158566376e-05,
+      "loss": 0.9831,
+      "step": 3485
+    },
+    {
+      "epoch": 0.4797687861271676,
+      "grad_norm": 1.7753107798096068,
+      "learning_rate": 1.1138625836339746e-05,
+      "loss": 1.0524,
+      "step": 3486
+    },
+    {
+      "epoch": 0.479906413432425,
+      "grad_norm": 1.664649603429524,
+      "learning_rate": 1.1134197287883752e-05,
+      "loss": 1.1971,
+      "step": 3487
+    },
+    {
+      "epoch": 0.48004404073768236,
+      "grad_norm": 1.5558478448324111,
+      "learning_rate": 1.112976851407828e-05,
+      "loss": 1.0491,
+      "step": 3488
+    },
+    {
+      "epoch": 0.4801816680429397,
+      "grad_norm": 1.8130513660564276,
+      "learning_rate": 1.1125339515803271e-05,
+      "loss": 1.0594,
+      "step": 3489
+    },
+    {
+      "epoch": 0.48031929534819706,
+      "grad_norm": 1.523778412805424,
+      "learning_rate": 1.1120910293938703e-05,
+      "loss": 1.0539,
+      "step": 3490
+    },
+    {
+      "epoch": 0.48045692265345447,
+      "grad_norm": 1.945464037068747,
+      "learning_rate": 1.1116480849364603e-05,
+      "loss": 1.1278,
+      "step": 3491
+    },
+    {
+      "epoch": 0.4805945499587118,
+      "grad_norm": 1.56474393624702,
+      "learning_rate": 1.1112051182961042e-05,
+      "loss": 1.0121,
+      "step": 3492
+    },
+    {
+      "epoch": 0.48073217726396916,
+      "grad_norm": 1.6116149806382005,
+      "learning_rate": 1.1107621295608133e-05,
+      "loss": 1.1033,
+      "step": 3493
+    },
+    {
+      "epoch": 0.4808698045692265,
+      "grad_norm": 1.6622348856836084,
+      "learning_rate": 1.1103191188186038e-05,
+      "loss": 1.1077,
+      "step": 3494
+    },
+    {
+      "epoch": 0.4810074318744839,
+      "grad_norm": 1.744583728989335,
+      "learning_rate": 1.1098760861574951e-05,
+      "loss": 1.0788,
+      "step": 3495
+    },
+    {
+      "epoch": 0.48114505917974126,
+      "grad_norm": 1.6344367802786446,
+      "learning_rate": 1.1094330316655119e-05,
+      "loss": 1.0135,
+      "step": 3496
+    },
+    {
+      "epoch": 0.4812826864849986,
+      "grad_norm": 1.580084652148695,
+      "learning_rate": 1.1089899554306835e-05,
+      "loss": 1.0447,
+      "step": 3497
+    },
+    {
+      "epoch": 0.48142031379025596,
+      "grad_norm": 1.7172256726484543,
+      "learning_rate": 1.108546857541043e-05,
+      "loss": 1.0163,
+      "step": 3498
+    },
+    {
+      "epoch": 0.48155794109551336,
+      "grad_norm": 1.5833836865657778,
+      "learning_rate": 1.1081037380846276e-05,
+      "loss": 1.0574,
+      "step": 3499
+    },
+    {
+      "epoch": 0.4816955684007707,
+      "grad_norm": 1.6859862996145802,
+      "learning_rate": 1.1076605971494793e-05,
+      "loss": 1.1456,
+      "step": 3500
+    },
+    {
+      "epoch": 0.48183319570602806,
+      "grad_norm": 1.6299681597553304,
+      "learning_rate": 1.107217434823644e-05,
+      "loss": 1.0395,
+      "step": 3501
+    },
+    {
+      "epoch": 0.48197082301128547,
+      "grad_norm": 1.8394087068023697,
+      "learning_rate": 1.106774251195172e-05,
+      "loss": 1.0919,
+      "step": 3502
+    },
+    {
+      "epoch": 0.4821084503165428,
+      "grad_norm": 1.6338034410174067,
+      "learning_rate": 1.106331046352118e-05,
+      "loss": 1.1217,
+      "step": 3503
+    },
+    {
+      "epoch": 0.48224607762180016,
+      "grad_norm": 1.698479071616687,
+      "learning_rate": 1.1058878203825407e-05,
+      "loss": 1.0515,
+      "step": 3504
+    },
+    {
+      "epoch": 0.4823837049270575,
+      "grad_norm": 1.6692086408282805,
+      "learning_rate": 1.1054445733745025e-05,
+      "loss": 0.9861,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4825213322323149,
+      "grad_norm": 1.6743437880744207,
+      "learning_rate": 1.1050013054160713e-05,
+      "loss": 1.1255,
+      "step": 3506
+    },
+    {
+      "epoch": 0.48265895953757226,
+      "grad_norm": 1.7687771360209608,
+      "learning_rate": 1.1045580165953179e-05,
+      "loss": 1.0873,
+      "step": 3507
+    },
+    {
+      "epoch": 0.4827965868428296,
+      "grad_norm": 1.5505344116454078,
+      "learning_rate": 1.104114707000318e-05,
+      "loss": 1.0077,
+      "step": 3508
+    },
+    {
+      "epoch": 0.48293421414808696,
+      "grad_norm": 1.5418368343185167,
+      "learning_rate": 1.103671376719151e-05,
+      "loss": 1.0689,
+      "step": 3509
+    },
+    {
+      "epoch": 0.48307184145334436,
+      "grad_norm": 1.6344129576227135,
+      "learning_rate": 1.1032280258399006e-05,
+      "loss": 1.0506,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4832094687586017,
+      "grad_norm": 1.6459098985265208,
+      "learning_rate": 1.1027846544506542e-05,
+      "loss": 0.9854,
+      "step": 3511
+    },
+    {
+      "epoch": 0.48334709606385906,
+      "grad_norm": 1.6926117614057368,
+      "learning_rate": 1.102341262639504e-05,
+      "loss": 0.9909,
+      "step": 3512
+    },
+    {
+      "epoch": 0.4834847233691164,
+      "grad_norm": 1.6341847012839852,
+      "learning_rate": 1.101897850494546e-05,
+      "loss": 1.0329,
+      "step": 3513
+    },
+    {
+      "epoch": 0.4836223506743738,
+      "grad_norm": 1.7087022370602607,
+      "learning_rate": 1.1014544181038798e-05,
+      "loss": 0.9712,
+      "step": 3514
+    },
+    {
+      "epoch": 0.48375997797963116,
+      "grad_norm": 1.7346152230224559,
+      "learning_rate": 1.1010109655556094e-05,
+      "loss": 1.0887,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4838976052848885,
+      "grad_norm": 1.7555438649286805,
+      "learning_rate": 1.100567492937843e-05,
+      "loss": 1.0306,
+      "step": 3516
+    },
+    {
+      "epoch": 0.48403523259014586,
+      "grad_norm": 1.7303283324437906,
+      "learning_rate": 1.1001240003386924e-05,
+      "loss": 1.1079,
+      "step": 3517
+    },
+    {
+      "epoch": 0.48417285989540326,
+      "grad_norm": 1.5644133689385946,
+      "learning_rate": 1.0996804878462735e-05,
+      "loss": 1.0728,
+      "step": 3518
+    },
+    {
+      "epoch": 0.4843104872006606,
+      "grad_norm": 1.8264423942066474,
+      "learning_rate": 1.099236955548706e-05,
+      "loss": 1.0379,
+      "step": 3519
+    },
+    {
+      "epoch": 0.48444811450591796,
+      "grad_norm": 1.7265951567237316,
+      "learning_rate": 1.0987934035341138e-05,
+      "loss": 0.992,
+      "step": 3520
+    },
+    {
+      "epoch": 0.48458574181117536,
+      "grad_norm": 2.024122042919884,
+      "learning_rate": 1.0983498318906245e-05,
+      "loss": 1.0994,
+      "step": 3521
+    },
+    {
+      "epoch": 0.4847233691164327,
+      "grad_norm": 1.6472654440087442,
+      "learning_rate": 1.0979062407063701e-05,
+      "loss": 1.0589,
+      "step": 3522
+    },
+    {
+      "epoch": 0.48486099642169006,
+      "grad_norm": 1.7069636837887578,
+      "learning_rate": 1.097462630069486e-05,
+      "loss": 1.1063,
+      "step": 3523
+    },
+    {
+      "epoch": 0.4849986237269474,
+      "grad_norm": 1.6846746197961866,
+      "learning_rate": 1.097019000068111e-05,
+      "loss": 1.0683,
+      "step": 3524
+    },
+    {
+      "epoch": 0.4851362510322048,
+      "grad_norm": 1.5930571762617842,
+      "learning_rate": 1.0965753507903894e-05,
+      "loss": 0.9549,
+      "step": 3525
+    },
+    {
+      "epoch": 0.48527387833746216,
+      "grad_norm": 1.5248622019745357,
+      "learning_rate": 1.0961316823244672e-05,
+      "loss": 1.0529,
+      "step": 3526
+    },
+    {
+      "epoch": 0.4854115056427195,
+      "grad_norm": 1.751758760245993,
+      "learning_rate": 1.0956879947584954e-05,
+      "loss": 0.9967,
+      "step": 3527
+    },
+    {
+      "epoch": 0.48554913294797686,
+      "grad_norm": 1.7160198382755134,
+      "learning_rate": 1.0952442881806291e-05,
+      "loss": 1.0415,
+      "step": 3528
+    },
+    {
+      "epoch": 0.48568676025323426,
+      "grad_norm": 1.8182142133453385,
+      "learning_rate": 1.0948005626790266e-05,
+      "loss": 1.0591,
+      "step": 3529
+    },
+    {
+      "epoch": 0.4858243875584916,
+      "grad_norm": 1.8036628384516902,
+      "learning_rate": 1.0943568183418499e-05,
+      "loss": 0.9981,
+      "step": 3530
+    },
+    {
+      "epoch": 0.48596201486374896,
+      "grad_norm": 1.6956712328530221,
+      "learning_rate": 1.0939130552572648e-05,
+      "loss": 1.0465,
+      "step": 3531
+    },
+    {
+      "epoch": 0.4860996421690063,
+      "grad_norm": 1.9483672952401996,
+      "learning_rate": 1.0934692735134413e-05,
+      "loss": 1.0801,
+      "step": 3532
+    },
+    {
+      "epoch": 0.4862372694742637,
+      "grad_norm": 1.5872547128618153,
+      "learning_rate": 1.0930254731985526e-05,
+      "loss": 1.0565,
+      "step": 3533
+    },
+    {
+      "epoch": 0.48637489677952106,
+      "grad_norm": 1.7210338012273994,
+      "learning_rate": 1.092581654400776e-05,
+      "loss": 1.0056,
+      "step": 3534
+    },
+    {
+      "epoch": 0.4865125240847784,
+      "grad_norm": 1.6637425662780916,
+      "learning_rate": 1.0921378172082915e-05,
+      "loss": 1.0566,
+      "step": 3535
+    },
+    {
+      "epoch": 0.48665015139003576,
+      "grad_norm": 1.9408092110193347,
+      "learning_rate": 1.091693961709284e-05,
+      "loss": 1.0515,
+      "step": 3536
+    },
+    {
+      "epoch": 0.48678777869529316,
+      "grad_norm": 1.743589855088594,
+      "learning_rate": 1.0912500879919416e-05,
+      "loss": 1.1286,
+      "step": 3537
+    },
+    {
+      "epoch": 0.4869254060005505,
+      "grad_norm": 1.7108384842812852,
+      "learning_rate": 1.0908061961444559e-05,
+      "loss": 1.0454,
+      "step": 3538
+    },
+    {
+      "epoch": 0.48706303330580786,
+      "grad_norm": 1.7834015400923728,
+      "learning_rate": 1.090362286255022e-05,
+      "loss": 0.9996,
+      "step": 3539
+    },
+    {
+      "epoch": 0.48720066061106526,
+      "grad_norm": 1.6258761384419342,
+      "learning_rate": 1.0899183584118387e-05,
+      "loss": 1.0489,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4873382879163226,
+      "grad_norm": 1.8075868724474395,
+      "learning_rate": 1.0894744127031084e-05,
+      "loss": 1.0045,
+      "step": 3541
+    },
+    {
+      "epoch": 0.48747591522157996,
+      "grad_norm": 1.6588153561627,
+      "learning_rate": 1.0890304492170373e-05,
+      "loss": 1.0444,
+      "step": 3542
+    },
+    {
+      "epoch": 0.4876135425268373,
+      "grad_norm": 1.5877071243322192,
+      "learning_rate": 1.0885864680418343e-05,
+      "loss": 1.0361,
+      "step": 3543
+    },
+    {
+      "epoch": 0.4877511698320947,
+      "grad_norm": 1.8180082646382312,
+      "learning_rate": 1.088142469265713e-05,
+      "loss": 1.0328,
+      "step": 3544
+    },
+    {
+      "epoch": 0.48788879713735206,
+      "grad_norm": 1.9133564700784407,
+      "learning_rate": 1.0876984529768895e-05,
+      "loss": 1.0465,
+      "step": 3545
+    },
+    {
+      "epoch": 0.4880264244426094,
+      "grad_norm": 1.6525276133072366,
+      "learning_rate": 1.0872544192635836e-05,
+      "loss": 1.0492,
+      "step": 3546
+    },
+    {
+      "epoch": 0.48816405174786676,
+      "grad_norm": 1.7712878232281224,
+      "learning_rate": 1.0868103682140191e-05,
+      "loss": 1.0153,
+      "step": 3547
+    },
+    {
+      "epoch": 0.48830167905312416,
+      "grad_norm": 1.7745035085087129,
+      "learning_rate": 1.086366299916423e-05,
+      "loss": 1.0928,
+      "step": 3548
+    },
+    {
+      "epoch": 0.4884393063583815,
+      "grad_norm": 1.6005132836849814,
+      "learning_rate": 1.0859222144590252e-05,
+      "loss": 1.047,
+      "step": 3549
+    },
+    {
+      "epoch": 0.48857693366363886,
+      "grad_norm": 1.8217660920924623,
+      "learning_rate": 1.0854781119300596e-05,
+      "loss": 1.1032,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4887145609688962,
+      "grad_norm": 1.777210319809963,
+      "learning_rate": 1.085033992417763e-05,
+      "loss": 1.0592,
+      "step": 3551
+    },
+    {
+      "epoch": 0.4888521882741536,
+      "grad_norm": 1.6165252974250917,
+      "learning_rate": 1.084589856010376e-05,
+      "loss": 1.0842,
+      "step": 3552
+    },
+    {
+      "epoch": 0.48898981557941096,
+      "grad_norm": 1.4999912340679995,
+      "learning_rate": 1.0841457027961429e-05,
+      "loss": 0.997,
+      "step": 3553
+    },
+    {
+      "epoch": 0.4891274428846683,
+      "grad_norm": 1.8106229842277155,
+      "learning_rate": 1.0837015328633102e-05,
+      "loss": 1.0678,
+      "step": 3554
+    },
+    {
+      "epoch": 0.48926507018992565,
+      "grad_norm": 1.901233887151323,
+      "learning_rate": 1.0832573463001289e-05,
+      "loss": 1.0901,
+      "step": 3555
+    },
+    {
+      "epoch": 0.48940269749518306,
+      "grad_norm": 1.6277103884309303,
+      "learning_rate": 1.0828131431948523e-05,
+      "loss": 1.0862,
+      "step": 3556
+    },
+    {
+      "epoch": 0.4895403248004404,
+      "grad_norm": 1.6314947151289938,
+      "learning_rate": 1.0823689236357379e-05,
+      "loss": 0.9439,
+      "step": 3557
+    },
+    {
+      "epoch": 0.48967795210569776,
+      "grad_norm": 1.6237013771517066,
+      "learning_rate": 1.081924687711046e-05,
+      "loss": 1.0498,
+      "step": 3558
+    },
+    {
+      "epoch": 0.48981557941095516,
+      "grad_norm": 1.7363682642100624,
+      "learning_rate": 1.0814804355090399e-05,
+      "loss": 1.1205,
+      "step": 3559
+    },
+    {
+      "epoch": 0.4899532067162125,
+      "grad_norm": 1.541585670538475,
+      "learning_rate": 1.0810361671179867e-05,
+      "loss": 0.9952,
+      "step": 3560
+    },
+    {
+      "epoch": 0.49009083402146986,
+      "grad_norm": 1.650520818593143,
+      "learning_rate": 1.0805918826261563e-05,
+      "loss": 1.072,
+      "step": 3561
+    },
+    {
+      "epoch": 0.4902284613267272,
+      "grad_norm": 1.609325540762437,
+      "learning_rate": 1.0801475821218219e-05,
+      "loss": 0.982,
+      "step": 3562
+    },
+    {
+      "epoch": 0.4903660886319846,
+      "grad_norm": 1.6053868007785101,
+      "learning_rate": 1.0797032656932602e-05,
+      "loss": 1.0525,
+      "step": 3563
+    },
+    {
+      "epoch": 0.49050371593724196,
+      "grad_norm": 1.6045610581867689,
+      "learning_rate": 1.0792589334287506e-05,
+      "loss": 1.1038,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4906413432424993,
+      "grad_norm": 1.6326746237203646,
+      "learning_rate": 1.078814585416576e-05,
+      "loss": 1.0913,
+      "step": 3565
+    },
+    {
+      "epoch": 0.49077897054775665,
+      "grad_norm": 1.5337801014644312,
+      "learning_rate": 1.0783702217450219e-05,
+      "loss": 0.9896,
+      "step": 3566
+    },
+    {
+      "epoch": 0.49091659785301406,
+      "grad_norm": 1.6948687633743464,
+      "learning_rate": 1.0779258425023773e-05,
+      "loss": 1.0409,
+      "step": 3567
+    },
+    {
+      "epoch": 0.4910542251582714,
+      "grad_norm": 1.6231738021872701,
+      "learning_rate": 1.0774814477769348e-05,
+      "loss": 1.008,
+      "step": 3568
+    },
+    {
+      "epoch": 0.49119185246352876,
+      "grad_norm": 1.5716834903100676,
+      "learning_rate": 1.0770370376569892e-05,
+      "loss": 0.9943,
+      "step": 3569
+    },
+    {
+      "epoch": 0.4913294797687861,
+      "grad_norm": 1.722729925653256,
+      "learning_rate": 1.0765926122308387e-05,
+      "loss": 1.1153,
+      "step": 3570
+    },
+    {
+      "epoch": 0.4914671070740435,
+      "grad_norm": 1.7875398513381686,
+      "learning_rate": 1.0761481715867847e-05,
+      "loss": 1.0571,
+      "step": 3571
+    },
+    {
+      "epoch": 0.49160473437930086,
+      "grad_norm": 2.2108313271017193,
+      "learning_rate": 1.0757037158131314e-05,
+      "loss": 1.1396,
+      "step": 3572
+    },
+    {
+      "epoch": 0.4917423616845582,
+      "grad_norm": 1.9209424684229361,
+      "learning_rate": 1.0752592449981858e-05,
+      "loss": 1.0804,
+      "step": 3573
+    },
+    {
+      "epoch": 0.49187998898981555,
+      "grad_norm": 1.668238514864507,
+      "learning_rate": 1.0748147592302584e-05,
+      "loss": 1.0199,
+      "step": 3574
+    },
+    {
+      "epoch": 0.49201761629507296,
+      "grad_norm": 1.9882455416410778,
+      "learning_rate": 1.0743702585976628e-05,
+      "loss": 1.0336,
+      "step": 3575
+    },
+    {
+      "epoch": 0.4921552436003303,
+      "grad_norm": 1.7027484847816787,
+      "learning_rate": 1.0739257431887144e-05,
+      "loss": 1.0194,
+      "step": 3576
+    },
+    {
+      "epoch": 0.49229287090558765,
+      "grad_norm": 1.8411809651606363,
+      "learning_rate": 1.0734812130917331e-05,
+      "loss": 1.104,
+      "step": 3577
+    },
+    {
+      "epoch": 0.49243049821084506,
+      "grad_norm": 1.8707777196196447,
+      "learning_rate": 1.0730366683950406e-05,
+      "loss": 0.9991,
+      "step": 3578
+    },
+    {
+      "epoch": 0.4925681255161024,
+      "grad_norm": 1.5934745086991755,
+      "learning_rate": 1.0725921091869616e-05,
+      "loss": 1.0111,
+      "step": 3579
+    },
+    {
+      "epoch": 0.49270575282135975,
+      "grad_norm": 1.6936956222628783,
+      "learning_rate": 1.0721475355558243e-05,
+      "loss": 1.0579,
+      "step": 3580
+    },
+    {
+      "epoch": 0.4928433801266171,
+      "grad_norm": 1.6433330423273211,
+      "learning_rate": 1.0717029475899593e-05,
+      "loss": 1.1173,
+      "step": 3581
+    },
+    {
+      "epoch": 0.4929810074318745,
+      "grad_norm": 1.4375624327102545,
+      "learning_rate": 1.0712583453777002e-05,
+      "loss": 0.9984,
+      "step": 3582
+    },
+    {
+      "epoch": 0.49311863473713186,
+      "grad_norm": 1.7047067038360502,
+      "learning_rate": 1.070813729007383e-05,
+      "loss": 0.9384,
+      "step": 3583
+    },
+    {
+      "epoch": 0.4932562620423892,
+      "grad_norm": 1.9033695426736614,
+      "learning_rate": 1.0703690985673474e-05,
+      "loss": 0.9702,
+      "step": 3584
+    },
+    {
+      "epoch": 0.49339388934764655,
+      "grad_norm": 1.6028695622701497,
+      "learning_rate": 1.069924454145935e-05,
+      "loss": 1.0602,
+      "step": 3585
+    },
+    {
+      "epoch": 0.49353151665290396,
+      "grad_norm": 1.5792029918591102,
+      "learning_rate": 1.069479795831491e-05,
+      "loss": 1.0214,
+      "step": 3586
+    },
+    {
+      "epoch": 0.4936691439581613,
+      "grad_norm": 1.7874895050963413,
+      "learning_rate": 1.0690351237123626e-05,
+      "loss": 0.9222,
+      "step": 3587
+    },
+    {
+      "epoch": 0.49380677126341865,
+      "grad_norm": 1.7626879257229984,
+      "learning_rate": 1.0685904378768998e-05,
+      "loss": 1.0143,
+      "step": 3588
+    },
+    {
+      "epoch": 0.493944398568676,
+      "grad_norm": 1.8953974613526487,
+      "learning_rate": 1.0681457384134564e-05,
+      "loss": 1.088,
+      "step": 3589
+    },
+    {
+      "epoch": 0.4940820258739334,
+      "grad_norm": 1.787501322781793,
+      "learning_rate": 1.0677010254103872e-05,
+      "loss": 1.081,
+      "step": 3590
+    },
+    {
+      "epoch": 0.49421965317919075,
+      "grad_norm": 1.7094726927793387,
+      "learning_rate": 1.0672562989560511e-05,
+      "loss": 1.0665,
+      "step": 3591
+    },
+    {
+      "epoch": 0.4943572804844481,
+      "grad_norm": 1.7366339545950398,
+      "learning_rate": 1.066811559138809e-05,
+      "loss": 1.0937,
+      "step": 3592
+    },
+    {
+      "epoch": 0.49449490778970545,
+      "grad_norm": 1.5584257757784012,
+      "learning_rate": 1.0663668060470246e-05,
+      "loss": 0.9538,
+      "step": 3593
+    },
+    {
+      "epoch": 0.49463253509496286,
+      "grad_norm": 1.637444794043393,
+      "learning_rate": 1.0659220397690643e-05,
+      "loss": 1.0215,
+      "step": 3594
+    },
+    {
+      "epoch": 0.4947701624002202,
+      "grad_norm": 1.790669341553123,
+      "learning_rate": 1.0654772603932972e-05,
+      "loss": 1.0808,
+      "step": 3595
+    },
+    {
+      "epoch": 0.49490778970547755,
+      "grad_norm": 1.747385323715335,
+      "learning_rate": 1.0650324680080948e-05,
+      "loss": 1.1295,
+      "step": 3596
+    },
+    {
+      "epoch": 0.49504541701073496,
+      "grad_norm": 1.8021376377240488,
+      "learning_rate": 1.064587662701831e-05,
+      "loss": 1.0112,
+      "step": 3597
+    },
+    {
+      "epoch": 0.4951830443159923,
+      "grad_norm": 1.7306591852080317,
+      "learning_rate": 1.0641428445628827e-05,
+      "loss": 1.1666,
+      "step": 3598
+    },
+    {
+      "epoch": 0.49532067162124965,
+      "grad_norm": 1.7837803004465724,
+      "learning_rate": 1.0636980136796294e-05,
+      "loss": 1.1454,
+      "step": 3599
+    },
+    {
+      "epoch": 0.495458298926507,
+      "grad_norm": 1.637576550159746,
+      "learning_rate": 1.0632531701404524e-05,
+      "loss": 1.0093,
+      "step": 3600
+    },
+    {
+      "epoch": 0.4955959262317644,
+      "grad_norm": 1.721484005039665,
+      "learning_rate": 1.0628083140337368e-05,
+      "loss": 1.0718,
+      "step": 3601
+    },
+    {
+      "epoch": 0.49573355353702175,
+      "grad_norm": 1.6340611641789315,
+      "learning_rate": 1.0623634454478686e-05,
+      "loss": 1.0315,
+      "step": 3602
+    },
+    {
+      "epoch": 0.4958711808422791,
+      "grad_norm": 1.4875693776699317,
+      "learning_rate": 1.0619185644712377e-05,
+      "loss": 0.9972,
+      "step": 3603
+    },
+    {
+      "epoch": 0.49600880814753645,
+      "grad_norm": 1.536553497918394,
+      "learning_rate": 1.0614736711922356e-05,
+      "loss": 1.0261,
+      "step": 3604
+    },
+    {
+      "epoch": 0.49614643545279385,
+      "grad_norm": 1.6635709470715365,
+      "learning_rate": 1.0610287656992564e-05,
+      "loss": 1.0309,
+      "step": 3605
+    },
+    {
+      "epoch": 0.4962840627580512,
+      "grad_norm": 1.674577456774639,
+      "learning_rate": 1.0605838480806967e-05,
+      "loss": 0.984,
+      "step": 3606
+    },
+    {
+      "epoch": 0.49642169006330855,
+      "grad_norm": 1.773679587939119,
+      "learning_rate": 1.060138918424956e-05,
+      "loss": 1.0404,
+      "step": 3607
+    },
+    {
+      "epoch": 0.4965593173685659,
+      "grad_norm": 1.7963060828139978,
+      "learning_rate": 1.0596939768204353e-05,
+      "loss": 1.0476,
+      "step": 3608
+    },
+    {
+      "epoch": 0.4966969446738233,
+      "grad_norm": 1.5995070628774557,
+      "learning_rate": 1.0592490233555383e-05,
+      "loss": 1.0898,
+      "step": 3609
+    },
+    {
+      "epoch": 0.49683457197908065,
+      "grad_norm": 2.0680028649885878,
+      "learning_rate": 1.0588040581186716e-05,
+      "loss": 0.9877,
+      "step": 3610
+    },
+    {
+      "epoch": 0.496972199284338,
+      "grad_norm": 1.5483714623871836,
+      "learning_rate": 1.0583590811982437e-05,
+      "loss": 1.0199,
+      "step": 3611
+    },
+    {
+      "epoch": 0.49710982658959535,
+      "grad_norm": 1.63478872273824,
+      "learning_rate": 1.057914092682665e-05,
+      "loss": 1.0808,
+      "step": 3612
+    },
+    {
+      "epoch": 0.49724745389485275,
+      "grad_norm": 1.5937401841631034,
+      "learning_rate": 1.0574690926603488e-05,
+      "loss": 1.0132,
+      "step": 3613
+    },
+    {
+      "epoch": 0.4973850812001101,
+      "grad_norm": 1.725287176899473,
+      "learning_rate": 1.0570240812197103e-05,
+      "loss": 1.1035,
+      "step": 3614
+    },
+    {
+      "epoch": 0.49752270850536745,
+      "grad_norm": 1.698308149921419,
+      "learning_rate": 1.0565790584491677e-05,
+      "loss": 1.0335,
+      "step": 3615
+    },
+    {
+      "epoch": 0.49766033581062485,
+      "grad_norm": 1.6124922349424369,
+      "learning_rate": 1.0561340244371407e-05,
+      "loss": 1.0846,
+      "step": 3616
+    },
+    {
+      "epoch": 0.4977979631158822,
+      "grad_norm": 1.8394423116958303,
+      "learning_rate": 1.0556889792720516e-05,
+      "loss": 0.9412,
+      "step": 3617
+    },
+    {
+      "epoch": 0.49793559042113955,
+      "grad_norm": 1.9080921434643454,
+      "learning_rate": 1.0552439230423244e-05,
+      "loss": 1.001,
+      "step": 3618
+    },
+    {
+      "epoch": 0.4980732177263969,
+      "grad_norm": 1.810735264072154,
+      "learning_rate": 1.0547988558363858e-05,
+      "loss": 1.0359,
+      "step": 3619
+    },
+    {
+      "epoch": 0.4982108450316543,
+      "grad_norm": 1.6328037385370924,
+      "learning_rate": 1.0543537777426649e-05,
+      "loss": 1.054,
+      "step": 3620
+    },
+    {
+      "epoch": 0.49834847233691165,
+      "grad_norm": 1.4789925779419533,
+      "learning_rate": 1.0539086888495922e-05,
+      "loss": 1.0309,
+      "step": 3621
+    },
+    {
+      "epoch": 0.498486099642169,
+      "grad_norm": 1.8183255533610405,
+      "learning_rate": 1.053463589245601e-05,
+      "loss": 1.077,
+      "step": 3622
+    },
+    {
+      "epoch": 0.49862372694742635,
+      "grad_norm": 1.617203455947018,
+      "learning_rate": 1.0530184790191267e-05,
+      "loss": 1.0888,
+      "step": 3623
+    },
+    {
+      "epoch": 0.49876135425268375,
+      "grad_norm": 1.632080188536248,
+      "learning_rate": 1.0525733582586061e-05,
+      "loss": 1.0153,
+      "step": 3624
+    },
+    {
+      "epoch": 0.4988989815579411,
+      "grad_norm": 1.8022685267978615,
+      "learning_rate": 1.052128227052479e-05,
+      "loss": 1.1296,
+      "step": 3625
+    },
+    {
+      "epoch": 0.49903660886319845,
+      "grad_norm": 1.8534712718244832,
+      "learning_rate": 1.0516830854891866e-05,
+      "loss": 1.144,
+      "step": 3626
+    },
+    {
+      "epoch": 0.4991742361684558,
+      "grad_norm": 1.6285972283508823,
+      "learning_rate": 1.0512379336571725e-05,
+      "loss": 1.0246,
+      "step": 3627
+    },
+    {
+      "epoch": 0.4993118634737132,
+      "grad_norm": 2.093019253537683,
+      "learning_rate": 1.0507927716448827e-05,
+      "loss": 1.1888,
+      "step": 3628
+    },
+    {
+      "epoch": 0.49944949077897055,
+      "grad_norm": 1.570896400760891,
+      "learning_rate": 1.050347599540764e-05,
+      "loss": 0.9462,
+      "step": 3629
+    },
+    {
+      "epoch": 0.4995871180842279,
+      "grad_norm": 1.6907771311454358,
+      "learning_rate": 1.0499024174332667e-05,
+      "loss": 1.0303,
+      "step": 3630
+    },
+    {
+      "epoch": 0.49972474538948525,
+      "grad_norm": 1.9629858081703044,
+      "learning_rate": 1.0494572254108422e-05,
+      "loss": 0.9924,
+      "step": 3631
+    },
+    {
+      "epoch": 0.49986237269474265,
+      "grad_norm": 1.736487979873794,
+      "learning_rate": 1.0490120235619439e-05,
+      "loss": 1.0266,
+      "step": 3632
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.866803282577337,
+      "learning_rate": 1.0485668119750273e-05,
+      "loss": 1.0246,
+      "step": 3633
+    },
+    {
+      "epoch": 0.5001376273052573,
+      "grad_norm": 1.803134982184472,
+      "learning_rate": 1.0481215907385503e-05,
+      "loss": 1.0757,
+      "step": 3634
+    },
+    {
+      "epoch": 0.5002752546105147,
+      "grad_norm": 1.818991274709966,
+      "learning_rate": 1.0476763599409719e-05,
+      "loss": 1.1309,
+      "step": 3635
+    },
+    {
+      "epoch": 0.500412881915772,
+      "grad_norm": 2.0056917166173824,
+      "learning_rate": 1.0472311196707532e-05,
+      "loss": 1.0962,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5005505092210295,
+      "grad_norm": 1.7948790292155148,
+      "learning_rate": 1.0467858700163575e-05,
+      "loss": 1.0443,
+      "step": 3637
+    },
+    {
+      "epoch": 0.5006881365262869,
+      "grad_norm": 1.620657912612342,
+      "learning_rate": 1.0463406110662502e-05,
+      "loss": 1.0692,
+      "step": 3638
+    },
+    {
+      "epoch": 0.5008257638315442,
+      "grad_norm": 1.6784728545621452,
+      "learning_rate": 1.0458953429088975e-05,
+      "loss": 1.0731,
+      "step": 3639
+    },
+    {
+      "epoch": 0.5009633911368016,
+      "grad_norm": 1.8236351897225729,
+      "learning_rate": 1.0454500656327683e-05,
+      "loss": 1.0534,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5011010184420589,
+      "grad_norm": 1.6964682727433784,
+      "learning_rate": 1.0450047793263337e-05,
+      "loss": 1.0647,
+      "step": 3641
+    },
+    {
+      "epoch": 0.5012386457473162,
+      "grad_norm": 1.8288714228239622,
+      "learning_rate": 1.044559484078065e-05,
+      "loss": 0.9893,
+      "step": 3642
+    },
+    {
+      "epoch": 0.5013762730525736,
+      "grad_norm": 1.6805213707480573,
+      "learning_rate": 1.0441141799764371e-05,
+      "loss": 1.0351,
+      "step": 3643
+    },
+    {
+      "epoch": 0.5015139003578309,
+      "grad_norm": 1.7384175133945252,
+      "learning_rate": 1.0436688671099252e-05,
+      "loss": 0.9991,
+      "step": 3644
+    },
+    {
+      "epoch": 0.5016515276630884,
+      "grad_norm": 1.6194220832537984,
+      "learning_rate": 1.0432235455670072e-05,
+      "loss": 0.9723,
+      "step": 3645
+    },
+    {
+      "epoch": 0.5017891549683458,
+      "grad_norm": 1.7205513483578003,
+      "learning_rate": 1.0427782154361625e-05,
+      "loss": 1.0988,
+      "step": 3646
+    },
+    {
+      "epoch": 0.5019267822736031,
+      "grad_norm": 1.6024933545225586,
+      "learning_rate": 1.042332876805872e-05,
+      "loss": 1.1487,
+      "step": 3647
+    },
+    {
+      "epoch": 0.5020644095788604,
+      "grad_norm": 1.6459542067469202,
+      "learning_rate": 1.0418875297646183e-05,
+      "loss": 1.0045,
+      "step": 3648
+    },
+    {
+      "epoch": 0.5022020368841178,
+      "grad_norm": 1.7443434476130137,
+      "learning_rate": 1.0414421744008857e-05,
+      "loss": 1.0653,
+      "step": 3649
+    },
+    {
+      "epoch": 0.5023396641893751,
+      "grad_norm": 1.6814560218440744,
+      "learning_rate": 1.0409968108031605e-05,
+      "loss": 1.0421,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5024772914946325,
+      "grad_norm": 1.6464292491425856,
+      "learning_rate": 1.0405514390599303e-05,
+      "loss": 1.0209,
+      "step": 3651
+    },
+    {
+      "epoch": 0.5026149187998898,
+      "grad_norm": 1.7979900805457167,
+      "learning_rate": 1.0401060592596838e-05,
+      "loss": 1.0851,
+      "step": 3652
+    },
+    {
+      "epoch": 0.5027525461051473,
+      "grad_norm": 2.1100172281340193,
+      "learning_rate": 1.0396606714909127e-05,
+      "loss": 1.0334,
+      "step": 3653
+    },
+    {
+      "epoch": 0.5028901734104047,
+      "grad_norm": 1.9233464480399822,
+      "learning_rate": 1.039215275842109e-05,
+      "loss": 1.1319,
+      "step": 3654
+    },
+    {
+      "epoch": 0.503027800715662,
+      "grad_norm": 1.6498657479995467,
+      "learning_rate": 1.0387698724017668e-05,
+      "loss": 1.0685,
+      "step": 3655
+    },
+    {
+      "epoch": 0.5031654280209193,
+      "grad_norm": 1.8524281908235674,
+      "learning_rate": 1.0383244612583815e-05,
+      "loss": 1.1121,
+      "step": 3656
+    },
+    {
+      "epoch": 0.5033030553261767,
+      "grad_norm": 1.5900931978272586,
+      "learning_rate": 1.0378790425004506e-05,
+      "loss": 1.0318,
+      "step": 3657
+    },
+    {
+      "epoch": 0.503440682631434,
+      "grad_norm": 1.606794452020487,
+      "learning_rate": 1.0374336162164722e-05,
+      "loss": 1.001,
+      "step": 3658
+    },
+    {
+      "epoch": 0.5035783099366914,
+      "grad_norm": 1.765166752234181,
+      "learning_rate": 1.0369881824949469e-05,
+      "loss": 1.0793,
+      "step": 3659
+    },
+    {
+      "epoch": 0.5037159372419489,
+      "grad_norm": 1.8339332660364427,
+      "learning_rate": 1.0365427414243756e-05,
+      "loss": 1.0924,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5038535645472062,
+      "grad_norm": 1.612316780324593,
+      "learning_rate": 1.0360972930932622e-05,
+      "loss": 1.1001,
+      "step": 3661
+    },
+    {
+      "epoch": 0.5039911918524635,
+      "grad_norm": 1.551479507101269,
+      "learning_rate": 1.0356518375901109e-05,
+      "loss": 1.0443,
+      "step": 3662
+    },
+    {
+      "epoch": 0.5041288191577209,
+      "grad_norm": 1.7203104169229857,
+      "learning_rate": 1.035206375003427e-05,
+      "loss": 1.046,
+      "step": 3663
+    },
+    {
+      "epoch": 0.5042664464629782,
+      "grad_norm": 1.6533651930309448,
+      "learning_rate": 1.0347609054217187e-05,
+      "loss": 1.0369,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5044040737682356,
+      "grad_norm": 1.956336430699767,
+      "learning_rate": 1.0343154289334942e-05,
+      "loss": 1.06,
+      "step": 3665
+    },
+    {
+      "epoch": 0.5045417010734929,
+      "grad_norm": 1.6841171227539313,
+      "learning_rate": 1.0338699456272639e-05,
+      "loss": 1.0529,
+      "step": 3666
+    },
+    {
+      "epoch": 0.5046793283787503,
+      "grad_norm": 1.7529286707237446,
+      "learning_rate": 1.0334244555915388e-05,
+      "loss": 1.0199,
+      "step": 3667
+    },
+    {
+      "epoch": 0.5048169556840078,
+      "grad_norm": 1.5963719489699255,
+      "learning_rate": 1.032978958914832e-05,
+      "loss": 1.0522,
+      "step": 3668
+    },
+    {
+      "epoch": 0.5049545829892651,
+      "grad_norm": 1.6442938121182555,
+      "learning_rate": 1.0325334556856572e-05,
+      "loss": 1.0516,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5050922102945224,
+      "grad_norm": 1.6496228970468823,
+      "learning_rate": 1.0320879459925306e-05,
+      "loss": 1.0674,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5052298375997798,
+      "grad_norm": 1.7738652646046575,
+      "learning_rate": 1.031642429923968e-05,
+      "loss": 1.1196,
+      "step": 3671
+    },
+    {
+      "epoch": 0.5053674649050371,
+      "grad_norm": 1.7512180479554356,
+      "learning_rate": 1.031196907568488e-05,
+      "loss": 1.0533,
+      "step": 3672
+    },
+    {
+      "epoch": 0.5055050922102945,
+      "grad_norm": 1.6690540317633202,
+      "learning_rate": 1.0307513790146092e-05,
+      "loss": 0.9972,
+      "step": 3673
+    },
+    {
+      "epoch": 0.5056427195155518,
+      "grad_norm": 1.6251525794378752,
+      "learning_rate": 1.0303058443508524e-05,
+      "loss": 1.0118,
+      "step": 3674
+    },
+    {
+      "epoch": 0.5057803468208093,
+      "grad_norm": 1.7283757766951422,
+      "learning_rate": 1.0298603036657394e-05,
+      "loss": 1.0931,
+      "step": 3675
+    },
+    {
+      "epoch": 0.5059179741260666,
+      "grad_norm": 1.7087621949750793,
+      "learning_rate": 1.0294147570477924e-05,
+      "loss": 0.9911,
+      "step": 3676
+    },
+    {
+      "epoch": 0.506055601431324,
+      "grad_norm": 1.6233840094476675,
+      "learning_rate": 1.0289692045855362e-05,
+      "loss": 1.0266,
+      "step": 3677
+    },
+    {
+      "epoch": 0.5061932287365813,
+      "grad_norm": 1.82417963132422,
+      "learning_rate": 1.0285236463674952e-05,
+      "loss": 0.9795,
+      "step": 3678
+    },
+    {
+      "epoch": 0.5063308560418387,
+      "grad_norm": 2.5928922843401794,
+      "learning_rate": 1.0280780824821965e-05,
+      "loss": 1.0613,
+      "step": 3679
+    },
+    {
+      "epoch": 0.506468483347096,
+      "grad_norm": 1.7288797356142398,
+      "learning_rate": 1.027632513018167e-05,
+      "loss": 1.1681,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5066061106523534,
+      "grad_norm": 1.7681771958526726,
+      "learning_rate": 1.0271869380639355e-05,
+      "loss": 1.0144,
+      "step": 3681
+    },
+    {
+      "epoch": 0.5067437379576107,
+      "grad_norm": 1.6423966704100725,
+      "learning_rate": 1.0267413577080315e-05,
+      "loss": 1.0661,
+      "step": 3682
+    },
+    {
+      "epoch": 0.5068813652628682,
+      "grad_norm": 1.570715117590332,
+      "learning_rate": 1.0262957720389857e-05,
+      "loss": 1.0323,
+      "step": 3683
+    },
+    {
+      "epoch": 0.5070189925681255,
+      "grad_norm": 1.6558910770797217,
+      "learning_rate": 1.02585018114533e-05,
+      "loss": 1.0297,
+      "step": 3684
+    },
+    {
+      "epoch": 0.5071566198733829,
+      "grad_norm": 1.7295672012698569,
+      "learning_rate": 1.025404585115597e-05,
+      "loss": 1.0854,
+      "step": 3685
+    },
+    {
+      "epoch": 0.5072942471786402,
+      "grad_norm": 1.5846774849157363,
+      "learning_rate": 1.0249589840383211e-05,
+      "loss": 1.0253,
+      "step": 3686
+    },
+    {
+      "epoch": 0.5074318744838976,
+      "grad_norm": 1.7786646903483323,
+      "learning_rate": 1.0245133780020364e-05,
+      "loss": 1.0074,
+      "step": 3687
+    },
+    {
+      "epoch": 0.5075695017891549,
+      "grad_norm": 1.9360098635358838,
+      "learning_rate": 1.0240677670952792e-05,
+      "loss": 1.1628,
+      "step": 3688
+    },
+    {
+      "epoch": 0.5077071290944123,
+      "grad_norm": 1.6677157386515384,
+      "learning_rate": 1.023622151406586e-05,
+      "loss": 1.0564,
+      "step": 3689
+    },
+    {
+      "epoch": 0.5078447563996696,
+      "grad_norm": 2.0006921397486583,
+      "learning_rate": 1.0231765310244948e-05,
+      "loss": 1.1413,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5079823837049271,
+      "grad_norm": 1.543490019914018,
+      "learning_rate": 1.022730906037544e-05,
+      "loss": 1.0165,
+      "step": 3691
+    },
+    {
+      "epoch": 0.5081200110101844,
+      "grad_norm": 1.7960222886831883,
+      "learning_rate": 1.0222852765342732e-05,
+      "loss": 1.1044,
+      "step": 3692
+    },
+    {
+      "epoch": 0.5082576383154418,
+      "grad_norm": 1.6839564983728792,
+      "learning_rate": 1.021839642603223e-05,
+      "loss": 1.032,
+      "step": 3693
+    },
+    {
+      "epoch": 0.5083952656206991,
+      "grad_norm": 1.4523234570971106,
+      "learning_rate": 1.021394004332935e-05,
+      "loss": 1.0247,
+      "step": 3694
+    },
+    {
+      "epoch": 0.5085328929259565,
+      "grad_norm": 1.4614194327142296,
+      "learning_rate": 1.0209483618119513e-05,
+      "loss": 1.0572,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5086705202312138,
+      "grad_norm": 1.6763058207341053,
+      "learning_rate": 1.0205027151288147e-05,
+      "loss": 1.0782,
+      "step": 3696
+    },
+    {
+      "epoch": 0.5088081475364712,
+      "grad_norm": 1.8280593860327268,
+      "learning_rate": 1.0200570643720693e-05,
+      "loss": 1.0466,
+      "step": 3697
+    },
+    {
+      "epoch": 0.5089457748417286,
+      "grad_norm": 1.7538140873688142,
+      "learning_rate": 1.0196114096302597e-05,
+      "loss": 1.1403,
+      "step": 3698
+    },
+    {
+      "epoch": 0.509083402146986,
+      "grad_norm": 1.7658685185606402,
+      "learning_rate": 1.0191657509919315e-05,
+      "loss": 1.0587,
+      "step": 3699
+    },
+    {
+      "epoch": 0.5092210294522433,
+      "grad_norm": 1.7729097365408693,
+      "learning_rate": 1.0187200885456306e-05,
+      "loss": 1.0477,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5093586567575007,
+      "grad_norm": 1.820431589560232,
+      "learning_rate": 1.0182744223799046e-05,
+      "loss": 0.9971,
+      "step": 3701
+    },
+    {
+      "epoch": 0.509496284062758,
+      "grad_norm": 1.743844584621045,
+      "learning_rate": 1.0178287525833012e-05,
+      "loss": 1.0243,
+      "step": 3702
+    },
+    {
+      "epoch": 0.5096339113680154,
+      "grad_norm": 1.6903413632210242,
+      "learning_rate": 1.017383079244369e-05,
+      "loss": 1.0348,
+      "step": 3703
+    },
+    {
+      "epoch": 0.5097715386732727,
+      "grad_norm": 1.8460537715040826,
+      "learning_rate": 1.0169374024516562e-05,
+      "loss": 0.9714,
+      "step": 3704
+    },
+    {
+      "epoch": 0.5099091659785301,
+      "grad_norm": 1.5125088212257936,
+      "learning_rate": 1.0164917222937138e-05,
+      "loss": 0.9888,
+      "step": 3705
+    },
+    {
+      "epoch": 0.5100467932837875,
+      "grad_norm": 1.825005120629029,
+      "learning_rate": 1.0160460388590918e-05,
+      "loss": 0.9849,
+      "step": 3706
+    },
+    {
+      "epoch": 0.5101844205890449,
+      "grad_norm": 1.8582237919090017,
+      "learning_rate": 1.0156003522363413e-05,
+      "loss": 1.1005,
+      "step": 3707
+    },
+    {
+      "epoch": 0.5103220478943022,
+      "grad_norm": 1.7306031823587031,
+      "learning_rate": 1.0151546625140144e-05,
+      "loss": 1.0635,
+      "step": 3708
+    },
+    {
+      "epoch": 0.5104596751995596,
+      "grad_norm": 1.5780905729950525,
+      "learning_rate": 1.0147089697806638e-05,
+      "loss": 1.0444,
+      "step": 3709
+    },
+    {
+      "epoch": 0.5105973025048169,
+      "grad_norm": 1.7503046433238172,
+      "learning_rate": 1.0142632741248421e-05,
+      "loss": 1.0751,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5107349298100743,
+      "grad_norm": 1.656146099387431,
+      "learning_rate": 1.0138175756351032e-05,
+      "loss": 1.0329,
+      "step": 3711
+    },
+    {
+      "epoch": 0.5108725571153316,
+      "grad_norm": 1.7012765662785974,
+      "learning_rate": 1.0133718744000009e-05,
+      "loss": 0.9921,
+      "step": 3712
+    },
+    {
+      "epoch": 0.5110101844205891,
+      "grad_norm": 1.599312054472005,
+      "learning_rate": 1.0129261705080906e-05,
+      "loss": 0.9828,
+      "step": 3713
+    },
+    {
+      "epoch": 0.5111478117258464,
+      "grad_norm": 1.5802052450242108,
+      "learning_rate": 1.0124804640479268e-05,
+      "loss": 0.9721,
+      "step": 3714
+    },
+    {
+      "epoch": 0.5112854390311038,
+      "grad_norm": 1.5633795859045756,
+      "learning_rate": 1.0120347551080658e-05,
+      "loss": 1.0904,
+      "step": 3715
+    },
+    {
+      "epoch": 0.5114230663363611,
+      "grad_norm": 1.6904453085234274,
+      "learning_rate": 1.0115890437770634e-05,
+      "loss": 1.1297,
+      "step": 3716
+    },
+    {
+      "epoch": 0.5115606936416185,
+      "grad_norm": 1.7482556045701696,
+      "learning_rate": 1.011143330143477e-05,
+      "loss": 1.0572,
+      "step": 3717
+    },
+    {
+      "epoch": 0.5116983209468758,
+      "grad_norm": 1.639525588162019,
+      "learning_rate": 1.0106976142958635e-05,
+      "loss": 1.0319,
+      "step": 3718
+    },
+    {
+      "epoch": 0.5118359482521332,
+      "grad_norm": 1.7377113405788274,
+      "learning_rate": 1.0102518963227803e-05,
+      "loss": 1.089,
+      "step": 3719
+    },
+    {
+      "epoch": 0.5119735755573905,
+      "grad_norm": 1.7632065698167017,
+      "learning_rate": 1.0098061763127858e-05,
+      "loss": 1.0105,
+      "step": 3720
+    },
+    {
+      "epoch": 0.512111202862648,
+      "grad_norm": 1.9817509295490594,
+      "learning_rate": 1.009360454354438e-05,
+      "loss": 1.1615,
+      "step": 3721
+    },
+    {
+      "epoch": 0.5122488301679053,
+      "grad_norm": 2.155299597310412,
+      "learning_rate": 1.0089147305362963e-05,
+      "loss": 1.1497,
+      "step": 3722
+    },
+    {
+      "epoch": 0.5123864574731627,
+      "grad_norm": 1.7152715888780967,
+      "learning_rate": 1.0084690049469192e-05,
+      "loss": 1.0044,
+      "step": 3723
+    },
+    {
+      "epoch": 0.51252408477842,
+      "grad_norm": 1.7350698569933563,
+      "learning_rate": 1.0080232776748674e-05,
+      "loss": 1.0503,
+      "step": 3724
+    },
+    {
+      "epoch": 0.5126617120836774,
+      "grad_norm": 1.735764841018283,
+      "learning_rate": 1.0075775488086996e-05,
+      "loss": 0.9355,
+      "step": 3725
+    },
+    {
+      "epoch": 0.5127993393889347,
+      "grad_norm": 1.8867100601426237,
+      "learning_rate": 1.007131818436977e-05,
+      "loss": 1.0739,
+      "step": 3726
+    },
+    {
+      "epoch": 0.5129369666941921,
+      "grad_norm": 1.6994828166701672,
+      "learning_rate": 1.0066860866482595e-05,
+      "loss": 0.9865,
+      "step": 3727
+    },
+    {
+      "epoch": 0.5130745939994494,
+      "grad_norm": 1.5722621818538653,
+      "learning_rate": 1.006240353531108e-05,
+      "loss": 0.9691,
+      "step": 3728
+    },
+    {
+      "epoch": 0.5132122213047069,
+      "grad_norm": 1.773267881388477,
+      "learning_rate": 1.0057946191740839e-05,
+      "loss": 1.1118,
+      "step": 3729
+    },
+    {
+      "epoch": 0.5133498486099642,
+      "grad_norm": 1.6831046517602444,
+      "learning_rate": 1.0053488836657483e-05,
+      "loss": 0.9948,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5134874759152216,
+      "grad_norm": 1.7493845421698786,
+      "learning_rate": 1.0049031470946624e-05,
+      "loss": 1.0342,
+      "step": 3731
+    },
+    {
+      "epoch": 0.5136251032204789,
+      "grad_norm": 1.7496108141988185,
+      "learning_rate": 1.0044574095493884e-05,
+      "loss": 1.0449,
+      "step": 3732
+    },
+    {
+      "epoch": 0.5137627305257363,
+      "grad_norm": 1.7186088852375372,
+      "learning_rate": 1.0040116711184884e-05,
+      "loss": 0.9886,
+      "step": 3733
+    },
+    {
+      "epoch": 0.5139003578309936,
+      "grad_norm": 1.840951697227051,
+      "learning_rate": 1.0035659318905244e-05,
+      "loss": 0.9859,
+      "step": 3734
+    },
+    {
+      "epoch": 0.514037985136251,
+      "grad_norm": 1.5660185396231594,
+      "learning_rate": 1.0031201919540583e-05,
+      "loss": 0.9309,
+      "step": 3735
+    },
+    {
+      "epoch": 0.5141756124415084,
+      "grad_norm": 1.6383801012737529,
+      "learning_rate": 1.0026744513976527e-05,
+      "loss": 1.0711,
+      "step": 3736
+    },
+    {
+      "epoch": 0.5143132397467658,
+      "grad_norm": 1.6443499115211788,
+      "learning_rate": 1.0022287103098702e-05,
+      "loss": 0.9736,
+      "step": 3737
+    },
+    {
+      "epoch": 0.5144508670520231,
+      "grad_norm": 1.5188753468483267,
+      "learning_rate": 1.0017829687792732e-05,
+      "loss": 1.0048,
+      "step": 3738
+    },
+    {
+      "epoch": 0.5145884943572805,
+      "grad_norm": 1.7140193032228657,
+      "learning_rate": 1.0013372268944253e-05,
+      "loss": 1.0075,
+      "step": 3739
+    },
+    {
+      "epoch": 0.5147261216625378,
+      "grad_norm": 1.6243350415644062,
+      "learning_rate": 1.0008914847438884e-05,
+      "loss": 1.0515,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5148637489677952,
+      "grad_norm": 1.8033473776499138,
+      "learning_rate": 1.0004457424162258e-05,
+      "loss": 1.0539,
+      "step": 3741
+    },
+    {
+      "epoch": 0.5150013762730525,
+      "grad_norm": 1.6825516404366865,
+      "learning_rate": 1e-05,
+      "loss": 0.9723,
+      "step": 3742
+    },
+    {
+      "epoch": 0.5151390035783099,
+      "grad_norm": 1.704995417090172,
+      "learning_rate": 9.995542575837747e-06,
+      "loss": 1.0115,
+      "step": 3743
+    },
+    {
+      "epoch": 0.5152766308835673,
+      "grad_norm": 1.6369464660271422,
+      "learning_rate": 9.991085152561119e-06,
+      "loss": 1.0148,
+      "step": 3744
+    },
+    {
+      "epoch": 0.5154142581888247,
+      "grad_norm": 1.624171201480478,
+      "learning_rate": 9.986627731055752e-06,
+      "loss": 0.9912,
+      "step": 3745
+    },
+    {
+      "epoch": 0.515551885494082,
+      "grad_norm": 1.9238153794159611,
+      "learning_rate": 9.98217031220727e-06,
+      "loss": 1.0713,
+      "step": 3746
+    },
+    {
+      "epoch": 0.5156895127993394,
+      "grad_norm": 1.72421206296874,
+      "learning_rate": 9.977712896901303e-06,
+      "loss": 1.0526,
+      "step": 3747
+    },
+    {
+      "epoch": 0.5158271401045967,
+      "grad_norm": 1.8604156533553424,
+      "learning_rate": 9.973255486023477e-06,
+      "loss": 1.1955,
+      "step": 3748
+    },
+    {
+      "epoch": 0.5159647674098541,
+      "grad_norm": 1.676058822024568,
+      "learning_rate": 9.968798080459424e-06,
+      "loss": 1.0224,
+      "step": 3749
+    },
+    {
+      "epoch": 0.5161023947151114,
+      "grad_norm": 1.6284108733913727,
+      "learning_rate": 9.96434068109476e-06,
+      "loss": 1.0415,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5162400220203689,
+      "grad_norm": 1.89726297905937,
+      "learning_rate": 9.959883288815119e-06,
+      "loss": 1.015,
+      "step": 3751
+    },
+    {
+      "epoch": 0.5163776493256262,
+      "grad_norm": 1.7503170137033843,
+      "learning_rate": 9.955425904506114e-06,
+      "loss": 1.103,
+      "step": 3752
+    },
+    {
+      "epoch": 0.5165152766308836,
+      "grad_norm": 1.8725881109303228,
+      "learning_rate": 9.95096852905338e-06,
+      "loss": 1.005,
+      "step": 3753
+    },
+    {
+      "epoch": 0.5166529039361409,
+      "grad_norm": 1.6992899913800072,
+      "learning_rate": 9.94651116334252e-06,
+      "loss": 1.0641,
+      "step": 3754
+    },
+    {
+      "epoch": 0.5167905312413983,
+      "grad_norm": 1.582651809741018,
+      "learning_rate": 9.942053808259163e-06,
+      "loss": 1.0406,
+      "step": 3755
+    },
+    {
+      "epoch": 0.5169281585466556,
+      "grad_norm": 2.1302244688840317,
+      "learning_rate": 9.93759646468892e-06,
+      "loss": 1.0427,
+      "step": 3756
+    },
+    {
+      "epoch": 0.517065785851913,
+      "grad_norm": 1.6577359388648831,
+      "learning_rate": 9.933139133517409e-06,
+      "loss": 0.9762,
+      "step": 3757
+    },
+    {
+      "epoch": 0.5172034131571703,
+      "grad_norm": 1.796510940150195,
+      "learning_rate": 9.92868181563023e-06,
+      "loss": 1.0503,
+      "step": 3758
+    },
+    {
+      "epoch": 0.5173410404624278,
+      "grad_norm": 1.6269280324764142,
+      "learning_rate": 9.924224511913006e-06,
+      "loss": 1.0682,
+      "step": 3759
+    },
+    {
+      "epoch": 0.5174786677676851,
+      "grad_norm": 2.0315694171854175,
+      "learning_rate": 9.919767223251333e-06,
+      "loss": 1.1341,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5176162950729425,
+      "grad_norm": 1.7248910405210414,
+      "learning_rate": 9.91530995053081e-06,
+      "loss": 1.0222,
+      "step": 3761
+    },
+    {
+      "epoch": 0.5177539223781998,
+      "grad_norm": 1.5296155761976447,
+      "learning_rate": 9.910852694637043e-06,
+      "loss": 1.0625,
+      "step": 3762
+    },
+    {
+      "epoch": 0.5178915496834572,
+      "grad_norm": 1.6801950441859106,
+      "learning_rate": 9.906395456455622e-06,
+      "loss": 1.0309,
+      "step": 3763
+    },
+    {
+      "epoch": 0.5180291769887145,
+      "grad_norm": 1.767967122473914,
+      "learning_rate": 9.901938236872149e-06,
+      "loss": 1.0404,
+      "step": 3764
+    },
+    {
+      "epoch": 0.5181668042939719,
+      "grad_norm": 1.9789534892954859,
+      "learning_rate": 9.8974810367722e-06,
+      "loss": 1.0079,
+      "step": 3765
+    },
+    {
+      "epoch": 0.5183044315992292,
+      "grad_norm": 1.6968858918967822,
+      "learning_rate": 9.89302385704137e-06,
+      "loss": 1.0021,
+      "step": 3766
+    },
+    {
+      "epoch": 0.5184420589044867,
+      "grad_norm": 1.5794358422814134,
+      "learning_rate": 9.88856669856523e-06,
+      "loss": 0.989,
+      "step": 3767
+    },
+    {
+      "epoch": 0.518579686209744,
+      "grad_norm": 1.806437091633027,
+      "learning_rate": 9.884109562229368e-06,
+      "loss": 1.0371,
+      "step": 3768
+    },
+    {
+      "epoch": 0.5187173135150014,
+      "grad_norm": 1.7189994892870109,
+      "learning_rate": 9.879652448919344e-06,
+      "loss": 1.0357,
+      "step": 3769
+    },
+    {
+      "epoch": 0.5188549408202587,
+      "grad_norm": 1.8139837047636342,
+      "learning_rate": 9.875195359520734e-06,
+      "loss": 1.0272,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5189925681255161,
+      "grad_norm": 1.8320132643809295,
+      "learning_rate": 9.870738294919096e-06,
+      "loss": 1.0207,
+      "step": 3771
+    },
+    {
+      "epoch": 0.5191301954307734,
+      "grad_norm": 1.886779616600022,
+      "learning_rate": 9.866281255999993e-06,
+      "loss": 0.9705,
+      "step": 3772
+    },
+    {
+      "epoch": 0.5192678227360308,
+      "grad_norm": 1.6853487947782546,
+      "learning_rate": 9.861824243648968e-06,
+      "loss": 1.0485,
+      "step": 3773
+    },
+    {
+      "epoch": 0.5194054500412882,
+      "grad_norm": 1.6942470178959521,
+      "learning_rate": 9.857367258751582e-06,
+      "loss": 1.0498,
+      "step": 3774
+    },
+    {
+      "epoch": 0.5195430773465456,
+      "grad_norm": 1.8883628942388777,
+      "learning_rate": 9.852910302193367e-06,
+      "loss": 1.0257,
+      "step": 3775
+    },
+    {
+      "epoch": 0.5196807046518029,
+      "grad_norm": 1.698204080891631,
+      "learning_rate": 9.848453374859857e-06,
+      "loss": 1.1391,
+      "step": 3776
+    },
+    {
+      "epoch": 0.5198183319570603,
+      "grad_norm": 1.780949060296349,
+      "learning_rate": 9.84399647763659e-06,
+      "loss": 0.9374,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5199559592623176,
+      "grad_norm": 1.7159764787602076,
+      "learning_rate": 9.839539611409086e-06,
+      "loss": 1.0077,
+      "step": 3778
+    },
+    {
+      "epoch": 0.520093586567575,
+      "grad_norm": 1.7752053989592338,
+      "learning_rate": 9.835082777062868e-06,
+      "loss": 1.1353,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5202312138728323,
+      "grad_norm": 1.679182268597926,
+      "learning_rate": 9.83062597548344e-06,
+      "loss": 1.0714,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5203688411780897,
+      "grad_norm": 1.6947975609275099,
+      "learning_rate": 9.826169207556318e-06,
+      "loss": 1.0394,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5205064684833471,
+      "grad_norm": 1.6759159654407623,
+      "learning_rate": 9.82171247416699e-06,
+      "loss": 0.9814,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5206440957886045,
+      "grad_norm": 1.788822794581056,
+      "learning_rate": 9.817255776200957e-06,
+      "loss": 1.0709,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5207817230938618,
+      "grad_norm": 1.746786872320396,
+      "learning_rate": 9.812799114543695e-06,
+      "loss": 1.0359,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5209193503991192,
+      "grad_norm": 1.666359513302726,
+      "learning_rate": 9.808342490080689e-06,
+      "loss": 1.1051,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5210569777043765,
+      "grad_norm": 1.9743488023445492,
+      "learning_rate": 9.803885903697407e-06,
+      "loss": 0.9999,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5211946050096339,
+      "grad_norm": 1.4362515492111947,
+      "learning_rate": 9.79942935627931e-06,
+      "loss": 1.0175,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5213322323148912,
+      "grad_norm": 1.8622280027250133,
+      "learning_rate": 9.794972848711855e-06,
+      "loss": 0.9889,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5214698596201487,
+      "grad_norm": 1.880202556161118,
+      "learning_rate": 9.79051638188049e-06,
+      "loss": 1.0473,
+      "step": 3789
+    },
+    {
+      "epoch": 0.521607486925406,
+      "grad_norm": 1.935376240850844,
+      "learning_rate": 9.78605995667065e-06,
+      "loss": 1.0181,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5217451142306634,
+      "grad_norm": 1.6872372205994932,
+      "learning_rate": 9.781603573967771e-06,
+      "loss": 1.011,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5218827415359207,
+      "grad_norm": 1.5869172939716962,
+      "learning_rate": 9.777147234657271e-06,
+      "loss": 1.0231,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5220203688411781,
+      "grad_norm": 1.8067251515147564,
+      "learning_rate": 9.772690939624563e-06,
+      "loss": 1.0737,
+      "step": 3793
+    },
+    {
+      "epoch": 0.5221579961464354,
+      "grad_norm": 1.6464598640446624,
+      "learning_rate": 9.768234689755057e-06,
+      "loss": 0.993,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5222956234516928,
+      "grad_norm": 1.7486532209739112,
+      "learning_rate": 9.763778485934143e-06,
+      "loss": 1.0085,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5224332507569501,
+      "grad_norm": 1.8556316497350904,
+      "learning_rate": 9.759322329047214e-06,
+      "loss": 1.0327,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5225708780622076,
+      "grad_norm": 1.6525516351671823,
+      "learning_rate": 9.75486621997964e-06,
+      "loss": 1.0926,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5227085053674649,
+      "grad_norm": 1.6439850856264127,
+      "learning_rate": 9.750410159616794e-06,
+      "loss": 1.1132,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5228461326727223,
+      "grad_norm": 1.5584784145356112,
+      "learning_rate": 9.745954148844032e-06,
+      "loss": 1.0024,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5229837599779796,
+      "grad_norm": 1.8288025718733074,
+      "learning_rate": 9.741498188546703e-06,
+      "loss": 0.9761,
+      "step": 3800
+    },
+    {
+      "epoch": 0.523121387283237,
+      "grad_norm": 1.5838380640509493,
+      "learning_rate": 9.737042279610146e-06,
+      "loss": 1.0688,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5232590145884943,
+      "grad_norm": 1.6152130732351877,
+      "learning_rate": 9.732586422919689e-06,
+      "loss": 1.0151,
+      "step": 3802
+    },
+    {
+      "epoch": 0.5233966418937517,
+      "grad_norm": 1.5112628263593764,
+      "learning_rate": 9.728130619360647e-06,
+      "loss": 0.9623,
+      "step": 3803
+    },
+    {
+      "epoch": 0.523534269199009,
+      "grad_norm": 1.732071231659601,
+      "learning_rate": 9.723674869818332e-06,
+      "loss": 0.9979,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5236718965042665,
+      "grad_norm": 1.6913098661366945,
+      "learning_rate": 9.719219175178037e-06,
+      "loss": 0.994,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 1.7326210408760323,
+      "learning_rate": 9.71476353632505e-06,
+      "loss": 1.0873,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5239471511147812,
+      "grad_norm": 1.6004285950649455,
+      "learning_rate": 9.710307954144643e-06,
+      "loss": 1.0545,
+      "step": 3807
+    },
+    {
+      "epoch": 0.5240847784200385,
+      "grad_norm": 1.6560546504490241,
+      "learning_rate": 9.70585242952208e-06,
+      "loss": 1.0376,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5242224057252959,
+      "grad_norm": 1.7193129761288588,
+      "learning_rate": 9.701396963342611e-06,
+      "loss": 1.0495,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5243600330305532,
+      "grad_norm": 1.667947819133195,
+      "learning_rate": 9.696941556491477e-06,
+      "loss": 1.0057,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5244976603358106,
+      "grad_norm": 1.90895465679153,
+      "learning_rate": 9.692486209853911e-06,
+      "loss": 1.1257,
+      "step": 3811
+    },
+    {
+      "epoch": 0.524635287641068,
+      "grad_norm": 1.6730046156745004,
+      "learning_rate": 9.688030924315124e-06,
+      "loss": 1.0954,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5247729149463254,
+      "grad_norm": 1.8130463515578326,
+      "learning_rate": 9.683575700760323e-06,
+      "loss": 1.0986,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5249105422515827,
+      "grad_norm": 1.541011123419202,
+      "learning_rate": 9.679120540074697e-06,
+      "loss": 0.8903,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5250481695568401,
+      "grad_norm": 1.7049055466375391,
+      "learning_rate": 9.67466544314343e-06,
+      "loss": 1.0217,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5251857968620974,
+      "grad_norm": 1.6971724643081114,
+      "learning_rate": 9.670210410851684e-06,
+      "loss": 1.0272,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5253234241673548,
+      "grad_norm": 1.8837959227824357,
+      "learning_rate": 9.665755444084614e-06,
+      "loss": 1.0624,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5254610514726121,
+      "grad_norm": 1.7845727136599527,
+      "learning_rate": 9.661300543727363e-06,
+      "loss": 1.0018,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5255986787778695,
+      "grad_norm": 1.7056200203654783,
+      "learning_rate": 9.656845710665061e-06,
+      "loss": 1.0259,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5257363060831269,
+      "grad_norm": 1.7276654122842436,
+      "learning_rate": 9.652390945782813e-06,
+      "loss": 1.0123,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5258739333883843,
+      "grad_norm": 1.4996999017539958,
+      "learning_rate": 9.647936249965731e-06,
+      "loss": 1.0215,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5260115606936416,
+      "grad_norm": 1.6252114665201842,
+      "learning_rate": 9.643481624098893e-06,
+      "loss": 1.0444,
+      "step": 3822
+    },
+    {
+      "epoch": 0.526149187998899,
+      "grad_norm": 1.8939508955753133,
+      "learning_rate": 9.63902706906738e-06,
+      "loss": 1.1951,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5262868153041563,
+      "grad_norm": 1.6392983815275937,
+      "learning_rate": 9.634572585756247e-06,
+      "loss": 1.048,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5264244426094137,
+      "grad_norm": 1.837708054304148,
+      "learning_rate": 9.630118175050535e-06,
+      "loss": 0.9705,
+      "step": 3825
+    },
+    {
+      "epoch": 0.526562069914671,
+      "grad_norm": 1.753698623714684,
+      "learning_rate": 9.625663837835283e-06,
+      "loss": 1.0801,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5266996972199285,
+      "grad_norm": 1.6337344831551184,
+      "learning_rate": 9.621209574995497e-06,
+      "loss": 1.0135,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5268373245251858,
+      "grad_norm": 1.5601032232157375,
+      "learning_rate": 9.61675538741619e-06,
+      "loss": 1.0361,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5269749518304432,
+      "grad_norm": 1.8366964942383963,
+      "learning_rate": 9.612301275982335e-06,
+      "loss": 1.0835,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5271125791357005,
+      "grad_norm": 2.1299054437352525,
+      "learning_rate": 9.607847241578914e-06,
+      "loss": 1.0124,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5272502064409579,
+      "grad_norm": 1.7093939932872375,
+      "learning_rate": 9.603393285090874e-06,
+      "loss": 1.0807,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5273878337462152,
+      "grad_norm": 1.6674394246610071,
+      "learning_rate": 9.598939407403163e-06,
+      "loss": 1.0061,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5275254610514726,
+      "grad_norm": 1.5982903143499696,
+      "learning_rate": 9.594485609400698e-06,
+      "loss": 0.9418,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5276630883567299,
+      "grad_norm": 1.8152079823054406,
+      "learning_rate": 9.590031891968397e-06,
+      "loss": 1.0061,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5278007156619874,
+      "grad_norm": 1.6266171384650536,
+      "learning_rate": 9.585578255991143e-06,
+      "loss": 1.0738,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5279383429672447,
+      "grad_norm": 1.9092604004182847,
+      "learning_rate": 9.58112470235382e-06,
+      "loss": 0.9816,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5280759702725021,
+      "grad_norm": 2.046959008468833,
+      "learning_rate": 9.57667123194128e-06,
+      "loss": 1.1644,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5282135975777594,
+      "grad_norm": 1.697050879017924,
+      "learning_rate": 9.572217845638378e-06,
+      "loss": 1.0657,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5283512248830168,
+      "grad_norm": 1.482869175498403,
+      "learning_rate": 9.56776454432993e-06,
+      "loss": 0.9644,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5284888521882741,
+      "grad_norm": 1.65618306977903,
+      "learning_rate": 9.56331132890075e-06,
+      "loss": 1.0686,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5286264794935315,
+      "grad_norm": 1.7472309716485506,
+      "learning_rate": 9.558858200235635e-06,
+      "loss": 1.0017,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5287641067987888,
+      "grad_norm": 1.7006867970345885,
+      "learning_rate": 9.554405159219353e-06,
+      "loss": 1.0906,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5289017341040463,
+      "grad_norm": 1.8403493687227561,
+      "learning_rate": 9.549952206736668e-06,
+      "loss": 1.1752,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5290393614093036,
+      "grad_norm": 1.7239930064535378,
+      "learning_rate": 9.545499343672318e-06,
+      "loss": 1.0315,
+      "step": 3844
+    },
+    {
+      "epoch": 0.529176988714561,
+      "grad_norm": 1.9005594730355417,
+      "learning_rate": 9.54104657091103e-06,
+      "loss": 1.0045,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5293146160198183,
+      "grad_norm": 1.5552004223658424,
+      "learning_rate": 9.536593889337502e-06,
+      "loss": 1.0153,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5294522433250757,
+      "grad_norm": 1.89501673805388,
+      "learning_rate": 9.532141299836426e-06,
+      "loss": 1.1348,
+      "step": 3847
+    },
+    {
+      "epoch": 0.529589870630333,
+      "grad_norm": 1.7934443481546254,
+      "learning_rate": 9.527688803292472e-06,
+      "loss": 1.0594,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5297274979355904,
+      "grad_norm": 1.65381651631811,
+      "learning_rate": 9.523236400590284e-06,
+      "loss": 1.1535,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5298651252408478,
+      "grad_norm": 1.7849886985874959,
+      "learning_rate": 9.518784092614497e-06,
+      "loss": 0.9903,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5300027525461052,
+      "grad_norm": 1.7434102318373084,
+      "learning_rate": 9.514331880249728e-06,
+      "loss": 0.9623,
+      "step": 3851
+    },
+    {
+      "epoch": 0.5301403798513625,
+      "grad_norm": 1.7120951184504127,
+      "learning_rate": 9.509879764380563e-06,
+      "loss": 0.979,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5302780071566199,
+      "grad_norm": 1.6255376843054818,
+      "learning_rate": 9.505427745891582e-06,
+      "loss": 1.0198,
+      "step": 3853
+    },
+    {
+      "epoch": 0.5304156344618772,
+      "grad_norm": 1.7738051929558583,
+      "learning_rate": 9.500975825667333e-06,
+      "loss": 1.0102,
+      "step": 3854
+    },
+    {
+      "epoch": 0.5305532617671346,
+      "grad_norm": 1.605912422522909,
+      "learning_rate": 9.496524004592364e-06,
+      "loss": 0.9952,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5306908890723919,
+      "grad_norm": 1.638841243339728,
+      "learning_rate": 9.49207228355118e-06,
+      "loss": 0.9361,
+      "step": 3856
+    },
+    {
+      "epoch": 0.5308285163776493,
+      "grad_norm": 1.6066175502071258,
+      "learning_rate": 9.487620663428276e-06,
+      "loss": 1.0568,
+      "step": 3857
+    },
+    {
+      "epoch": 0.5309661436829067,
+      "grad_norm": 1.6546692713402367,
+      "learning_rate": 9.48316914510814e-06,
+      "loss": 1.0065,
+      "step": 3858
+    },
+    {
+      "epoch": 0.5311037709881641,
+      "grad_norm": 1.5744529875649458,
+      "learning_rate": 9.478717729475214e-06,
+      "loss": 1.0474,
+      "step": 3859
+    },
+    {
+      "epoch": 0.5312413982934214,
+      "grad_norm": 1.7758170501412414,
+      "learning_rate": 9.474266417413944e-06,
+      "loss": 1.0316,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5313790255986788,
+      "grad_norm": 1.6729868841164004,
+      "learning_rate": 9.469815209808738e-06,
+      "loss": 1.0092,
+      "step": 3861
+    },
+    {
+      "epoch": 0.5315166529039361,
+      "grad_norm": 1.5665269084764804,
+      "learning_rate": 9.465364107543991e-06,
+      "loss": 1.023,
+      "step": 3862
+    },
+    {
+      "epoch": 0.5316542802091935,
+      "grad_norm": 1.6297225665815323,
+      "learning_rate": 9.46091311150408e-06,
+      "loss": 1.0334,
+      "step": 3863
+    },
+    {
+      "epoch": 0.5317919075144508,
+      "grad_norm": 1.624633544965137,
+      "learning_rate": 9.456462222573354e-06,
+      "loss": 1.0832,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5319295348197083,
+      "grad_norm": 1.5305370778043634,
+      "learning_rate": 9.452011441636142e-06,
+      "loss": 0.9794,
+      "step": 3865
+    },
+    {
+      "epoch": 0.5320671621249656,
+      "grad_norm": 1.699367530579064,
+      "learning_rate": 9.44756076957676e-06,
+      "loss": 1.0278,
+      "step": 3866
+    },
+    {
+      "epoch": 0.532204789430223,
+      "grad_norm": 1.5978923902682878,
+      "learning_rate": 9.443110207279486e-06,
+      "loss": 1.051,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5323424167354803,
+      "grad_norm": 1.6452691714047998,
+      "learning_rate": 9.438659755628595e-06,
+      "loss": 1.0462,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5324800440407377,
+      "grad_norm": 1.6006061765582986,
+      "learning_rate": 9.434209415508322e-06,
+      "loss": 0.9676,
+      "step": 3869
+    },
+    {
+      "epoch": 0.532617671345995,
+      "grad_norm": 1.6581083042037867,
+      "learning_rate": 9.4297591878029e-06,
+      "loss": 1.0157,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5327552986512524,
+      "grad_norm": 1.6416976965233043,
+      "learning_rate": 9.425309073396519e-06,
+      "loss": 1.0774,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5328929259565097,
+      "grad_norm": 1.794811193612008,
+      "learning_rate": 9.420859073173355e-06,
+      "loss": 0.9986,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5330305532617672,
+      "grad_norm": 1.7253972789669378,
+      "learning_rate": 9.416409188017568e-06,
+      "loss": 1.0253,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5331681805670245,
+      "grad_norm": 1.5708128159296446,
+      "learning_rate": 9.411959418813286e-06,
+      "loss": 1.0938,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5333058078722819,
+      "grad_norm": 2.0736462514555547,
+      "learning_rate": 9.40750976644462e-06,
+      "loss": 0.9924,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5334434351775392,
+      "grad_norm": 1.6615070298183594,
+      "learning_rate": 9.40306023179565e-06,
+      "loss": 1.0301,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5335810624827966,
+      "grad_norm": 1.8777588434328962,
+      "learning_rate": 9.398610815750446e-06,
+      "loss": 0.9795,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5337186897880539,
+      "grad_norm": 1.5007003198464026,
+      "learning_rate": 9.394161519193035e-06,
+      "loss": 0.9139,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5338563170933113,
+      "grad_norm": 1.7134556342587728,
+      "learning_rate": 9.389712343007439e-06,
+      "loss": 1.0428,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5339939443985686,
+      "grad_norm": 1.7271494353068595,
+      "learning_rate": 9.385263288077645e-06,
+      "loss": 0.9963,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5341315717038261,
+      "grad_norm": 1.8380041169482233,
+      "learning_rate": 9.380814355287626e-06,
+      "loss": 0.968,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5342691990090834,
+      "grad_norm": 1.5918904838607604,
+      "learning_rate": 9.376365545521314e-06,
+      "loss": 1.0005,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5344068263143408,
+      "grad_norm": 1.7785911580799227,
+      "learning_rate": 9.371916859662635e-06,
+      "loss": 1.0918,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5345444536195981,
+      "grad_norm": 1.711909601874512,
+      "learning_rate": 9.367468298595474e-06,
+      "loss": 1.1005,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5346820809248555,
+      "grad_norm": 1.6937343320985883,
+      "learning_rate": 9.36301986320371e-06,
+      "loss": 0.9808,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5348197082301128,
+      "grad_norm": 1.5833376851184493,
+      "learning_rate": 9.358571554371174e-06,
+      "loss": 0.8992,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5349573355353702,
+      "grad_norm": 1.7887014569868183,
+      "learning_rate": 9.354123372981692e-06,
+      "loss": 0.9902,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5350949628406276,
+      "grad_norm": 2.0857967285007284,
+      "learning_rate": 9.349675319919057e-06,
+      "loss": 0.9584,
+      "step": 3888
+    },
+    {
+      "epoch": 0.535232590145885,
+      "grad_norm": 1.9131669645790712,
+      "learning_rate": 9.34522739606703e-06,
+      "loss": 1.1007,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5353702174511423,
+      "grad_norm": 1.8640943455233843,
+      "learning_rate": 9.34077960230936e-06,
+      "loss": 1.0409,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5355078447563997,
+      "grad_norm": 1.6482922248771166,
+      "learning_rate": 9.336331939529755e-06,
+      "loss": 1.0128,
+      "step": 3891
+    },
+    {
+      "epoch": 0.535645472061657,
+      "grad_norm": 1.6118289315008025,
+      "learning_rate": 9.331884408611915e-06,
+      "loss": 1.0664,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5357830993669144,
+      "grad_norm": 1.8266556287769047,
+      "learning_rate": 9.327437010439492e-06,
+      "loss": 0.9581,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5359207266721717,
+      "grad_norm": 1.5437768551855613,
+      "learning_rate": 9.322989745896131e-06,
+      "loss": 1.0562,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5360583539774291,
+      "grad_norm": 1.529280484862347,
+      "learning_rate": 9.318542615865437e-06,
+      "loss": 0.8906,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5361959812826865,
+      "grad_norm": 1.4687591709202528,
+      "learning_rate": 9.314095621231003e-06,
+      "loss": 0.9726,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5363336085879439,
+      "grad_norm": 1.9214992882949178,
+      "learning_rate": 9.309648762876376e-06,
+      "loss": 1.0279,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5364712358932012,
+      "grad_norm": 1.7351924040997424,
+      "learning_rate": 9.305202041685092e-06,
+      "loss": 0.937,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5366088631984586,
+      "grad_norm": 1.7687317033171404,
+      "learning_rate": 9.300755458540648e-06,
+      "loss": 0.9939,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5367464905037159,
+      "grad_norm": 1.5872159660138214,
+      "learning_rate": 9.29630901432653e-06,
+      "loss": 0.997,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5368841178089733,
+      "grad_norm": 1.7404538361882436,
+      "learning_rate": 9.29186270992617e-06,
+      "loss": 1.0434,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5370217451142306,
+      "grad_norm": 1.637431221175942,
+      "learning_rate": 9.287416546223e-06,
+      "loss": 1.1033,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5371593724194881,
+      "grad_norm": 1.6419849416190846,
+      "learning_rate": 9.282970524100412e-06,
+      "loss": 1.08,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5372969997247454,
+      "grad_norm": 1.7016117074662158,
+      "learning_rate": 9.278524644441758e-06,
+      "loss": 1.0184,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5374346270300028,
+      "grad_norm": 1.6932237338058265,
+      "learning_rate": 9.274078908130389e-06,
+      "loss": 1.0189,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5375722543352601,
+      "grad_norm": 1.5523720946995796,
+      "learning_rate": 9.269633316049597e-06,
+      "loss": 1.0553,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5377098816405175,
+      "grad_norm": 1.6741630595932375,
+      "learning_rate": 9.265187869082674e-06,
+      "loss": 1.0548,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5378475089457748,
+      "grad_norm": 1.5354530327789282,
+      "learning_rate": 9.260742568112857e-06,
+      "loss": 1.0171,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5379851362510322,
+      "grad_norm": 1.775194832420557,
+      "learning_rate": 9.256297414023376e-06,
+      "loss": 0.9707,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5381227635562895,
+      "grad_norm": 1.5134466892664977,
+      "learning_rate": 9.251852407697417e-06,
+      "loss": 1.0266,
+      "step": 3910
+    },
+    {
+      "epoch": 0.538260390861547,
+      "grad_norm": 1.797947051439264,
+      "learning_rate": 9.247407550018145e-06,
+      "loss": 1.0801,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5383980181668043,
+      "grad_norm": 1.8512508465850501,
+      "learning_rate": 9.242962841868688e-06,
+      "loss": 1.1125,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5385356454720617,
+      "grad_norm": 1.6270419428118281,
+      "learning_rate": 9.238518284132156e-06,
+      "loss": 1.0266,
+      "step": 3913
+    },
+    {
+      "epoch": 0.538673272777319,
+      "grad_norm": 2.026777366120372,
+      "learning_rate": 9.234073877691613e-06,
+      "loss": 1.1091,
+      "step": 3914
+    },
+    {
+      "epoch": 0.5388109000825764,
+      "grad_norm": 1.6851614827533623,
+      "learning_rate": 9.229629623430112e-06,
+      "loss": 1.0031,
+      "step": 3915
+    },
+    {
+      "epoch": 0.5389485273878337,
+      "grad_norm": 1.6286985112139007,
+      "learning_rate": 9.225185522230654e-06,
+      "loss": 1.0377,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5390861546930911,
+      "grad_norm": 1.827713262636613,
+      "learning_rate": 9.220741574976229e-06,
+      "loss": 1.0928,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5392237819983484,
+      "grad_norm": 1.8207472933077034,
+      "learning_rate": 9.216297782549784e-06,
+      "loss": 0.9845,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5393614093036059,
+      "grad_norm": 1.8617041462397839,
+      "learning_rate": 9.211854145834243e-06,
+      "loss": 1.1651,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5394990366088632,
+      "grad_norm": 1.8468211678927453,
+      "learning_rate": 9.207410665712499e-06,
+      "loss": 0.9784,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5396366639141206,
+      "grad_norm": 1.542069759394412,
+      "learning_rate": 9.2029673430674e-06,
+      "loss": 1.0396,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5397742912193779,
+      "grad_norm": 1.4505777641325768,
+      "learning_rate": 9.198524178781785e-06,
+      "loss": 0.9671,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5399119185246353,
+      "grad_norm": 1.6050589474619645,
+      "learning_rate": 9.19408117373844e-06,
+      "loss": 0.9645,
+      "step": 3923
+    },
+    {
+      "epoch": 0.5400495458298926,
+      "grad_norm": 1.8800979240650662,
+      "learning_rate": 9.189638328820138e-06,
+      "loss": 1.0226,
+      "step": 3924
+    },
+    {
+      "epoch": 0.54018717313515,
+      "grad_norm": 1.8980282072915977,
+      "learning_rate": 9.185195644909603e-06,
+      "loss": 1.0913,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5403248004404074,
+      "grad_norm": 1.6419855042391214,
+      "learning_rate": 9.180753122889542e-06,
+      "loss": 1.0352,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5404624277456648,
+      "grad_norm": 1.8408462225792674,
+      "learning_rate": 9.176310763642621e-06,
+      "loss": 1.0152,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5406000550509221,
+      "grad_norm": 1.629404883011823,
+      "learning_rate": 9.171868568051479e-06,
+      "loss": 1.0165,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5407376823561795,
+      "grad_norm": 1.5762819283854779,
+      "learning_rate": 9.167426536998713e-06,
+      "loss": 1.0764,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5408753096614368,
+      "grad_norm": 1.7235475583029523,
+      "learning_rate": 9.162984671366901e-06,
+      "loss": 0.9726,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5410129369666942,
+      "grad_norm": 1.5201467071012251,
+      "learning_rate": 9.158542972038573e-06,
+      "loss": 1.0861,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5411505642719515,
+      "grad_norm": 1.5836489834186152,
+      "learning_rate": 9.154101439896243e-06,
+      "loss": 1.0395,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5412881915772089,
+      "grad_norm": 1.5900271413266611,
+      "learning_rate": 9.149660075822373e-06,
+      "loss": 1.0234,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5414258188824663,
+      "grad_norm": 1.5888588231503058,
+      "learning_rate": 9.145218880699408e-06,
+      "loss": 0.9906,
+      "step": 3934
+    },
+    {
+      "epoch": 0.5415634461877237,
+      "grad_norm": 1.8325390179702046,
+      "learning_rate": 9.140777855409752e-06,
+      "loss": 0.9132,
+      "step": 3935
+    },
+    {
+      "epoch": 0.541701073492981,
+      "grad_norm": 1.6377214717238329,
+      "learning_rate": 9.136337000835773e-06,
+      "loss": 1.0307,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5418387007982384,
+      "grad_norm": 1.607856121647772,
+      "learning_rate": 9.131896317859812e-06,
+      "loss": 0.9973,
+      "step": 3937
+    },
+    {
+      "epoch": 0.5419763281034957,
+      "grad_norm": 1.9817131767073952,
+      "learning_rate": 9.127455807364165e-06,
+      "loss": 1.0445,
+      "step": 3938
+    },
+    {
+      "epoch": 0.5421139554087531,
+      "grad_norm": 1.5868201020121806,
+      "learning_rate": 9.123015470231111e-06,
+      "loss": 1.0087,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5422515827140104,
+      "grad_norm": 1.7277173434537934,
+      "learning_rate": 9.118575307342874e-06,
+      "loss": 1.0655,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5423892100192679,
+      "grad_norm": 2.0310904777546903,
+      "learning_rate": 9.114135319581659e-06,
+      "loss": 1.0148,
+      "step": 3941
+    },
+    {
+      "epoch": 0.5425268373245252,
+      "grad_norm": 1.9426031707748488,
+      "learning_rate": 9.109695507829628e-06,
+      "loss": 1.1568,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5426644646297826,
+      "grad_norm": 1.8031406907537484,
+      "learning_rate": 9.105255872968917e-06,
+      "loss": 0.9951,
+      "step": 3943
+    },
+    {
+      "epoch": 0.5428020919350399,
+      "grad_norm": 1.6136701055228408,
+      "learning_rate": 9.100816415881613e-06,
+      "loss": 1.0321,
+      "step": 3944
+    },
+    {
+      "epoch": 0.5429397192402973,
+      "grad_norm": 1.6153210574067458,
+      "learning_rate": 9.096377137449783e-06,
+      "loss": 1.034,
+      "step": 3945
+    },
+    {
+      "epoch": 0.5430773465455546,
+      "grad_norm": 1.6891714347457558,
+      "learning_rate": 9.091938038555441e-06,
+      "loss": 1.011,
+      "step": 3946
+    },
+    {
+      "epoch": 0.543214973850812,
+      "grad_norm": 1.6380296814974908,
+      "learning_rate": 9.087499120080586e-06,
+      "loss": 0.9859,
+      "step": 3947
+    },
+    {
+      "epoch": 0.5433526011560693,
+      "grad_norm": 1.6594119497567779,
+      "learning_rate": 9.083060382907161e-06,
+      "loss": 1.0138,
+      "step": 3948
+    },
+    {
+      "epoch": 0.5434902284613268,
+      "grad_norm": 1.6752641511851678,
+      "learning_rate": 9.078621827917086e-06,
+      "loss": 1.018,
+      "step": 3949
+    },
+    {
+      "epoch": 0.5436278557665841,
+      "grad_norm": 1.8886647851390153,
+      "learning_rate": 9.074183455992244e-06,
+      "loss": 1.0103,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5437654830718415,
+      "grad_norm": 1.767866556426858,
+      "learning_rate": 9.069745268014476e-06,
+      "loss": 1.0941,
+      "step": 3951
+    },
+    {
+      "epoch": 0.5439031103770988,
+      "grad_norm": 1.906685930012221,
+      "learning_rate": 9.065307264865592e-06,
+      "loss": 0.9813,
+      "step": 3952
+    },
+    {
+      "epoch": 0.5440407376823562,
+      "grad_norm": 1.8137655451651837,
+      "learning_rate": 9.060869447427353e-06,
+      "loss": 1.0682,
+      "step": 3953
+    },
+    {
+      "epoch": 0.5441783649876135,
+      "grad_norm": 1.6938215858836734,
+      "learning_rate": 9.056431816581506e-06,
+      "loss": 1.0512,
+      "step": 3954
+    },
+    {
+      "epoch": 0.5443159922928709,
+      "grad_norm": 1.7498041504404622,
+      "learning_rate": 9.051994373209738e-06,
+      "loss": 1.036,
+      "step": 3955
+    },
+    {
+      "epoch": 0.5444536195981282,
+      "grad_norm": 1.6251545894987196,
+      "learning_rate": 9.047557118193712e-06,
+      "loss": 1.0488,
+      "step": 3956
+    },
+    {
+      "epoch": 0.5445912469033857,
+      "grad_norm": 2.195265894990174,
+      "learning_rate": 9.043120052415049e-06,
+      "loss": 1.1594,
+      "step": 3957
+    },
+    {
+      "epoch": 0.544728874208643,
+      "grad_norm": 1.6321302368606516,
+      "learning_rate": 9.038683176755333e-06,
+      "loss": 0.9998,
+      "step": 3958
+    },
+    {
+      "epoch": 0.5448665015139004,
+      "grad_norm": 1.6100665080175312,
+      "learning_rate": 9.034246492096108e-06,
+      "loss": 1.0128,
+      "step": 3959
+    },
+    {
+      "epoch": 0.5450041288191577,
+      "grad_norm": 1.9271504717703556,
+      "learning_rate": 9.02980999931889e-06,
+      "loss": 1.0048,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5451417561244151,
+      "grad_norm": 1.693137169320187,
+      "learning_rate": 9.02537369930514e-06,
+      "loss": 1.0224,
+      "step": 3961
+    },
+    {
+      "epoch": 0.5452793834296724,
+      "grad_norm": 1.6776886025620386,
+      "learning_rate": 9.020937592936302e-06,
+      "loss": 1.0745,
+      "step": 3962
+    },
+    {
+      "epoch": 0.5454170107349298,
+      "grad_norm": 2.1728619659702506,
+      "learning_rate": 9.016501681093755e-06,
+      "loss": 1.1489,
+      "step": 3963
+    },
+    {
+      "epoch": 0.5455546380401872,
+      "grad_norm": 1.5943403502270752,
+      "learning_rate": 9.012065964658865e-06,
+      "loss": 1.0116,
+      "step": 3964
+    },
+    {
+      "epoch": 0.5456922653454446,
+      "grad_norm": 1.8256872249987628,
+      "learning_rate": 9.007630444512944e-06,
+      "loss": 1.0708,
+      "step": 3965
+    },
+    {
+      "epoch": 0.5458298926507019,
+      "grad_norm": 1.7966555703844447,
+      "learning_rate": 9.00319512153727e-06,
+      "loss": 0.9358,
+      "step": 3966
+    },
+    {
+      "epoch": 0.5459675199559593,
+      "grad_norm": 1.679911444438586,
+      "learning_rate": 8.998759996613083e-06,
+      "loss": 0.999,
+      "step": 3967
+    },
+    {
+      "epoch": 0.5461051472612166,
+      "grad_norm": 1.976250180889497,
+      "learning_rate": 8.994325070621572e-06,
+      "loss": 1.0064,
+      "step": 3968
+    },
+    {
+      "epoch": 0.546242774566474,
+      "grad_norm": 1.7529014872738853,
+      "learning_rate": 8.98989034444391e-06,
+      "loss": 0.9588,
+      "step": 3969
+    },
+    {
+      "epoch": 0.5463804018717313,
+      "grad_norm": 1.6986757017523486,
+      "learning_rate": 8.985455818961206e-06,
+      "loss": 1.076,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5465180291769887,
+      "grad_norm": 1.8171871229262688,
+      "learning_rate": 8.981021495054544e-06,
+      "loss": 0.9973,
+      "step": 3971
+    },
+    {
+      "epoch": 0.5466556564822461,
+      "grad_norm": 1.6591342846280948,
+      "learning_rate": 8.976587373604963e-06,
+      "loss": 1.0198,
+      "step": 3972
+    },
+    {
+      "epoch": 0.5467932837875035,
+      "grad_norm": 1.664420597783526,
+      "learning_rate": 8.972153455493462e-06,
+      "loss": 0.996,
+      "step": 3973
+    },
+    {
+      "epoch": 0.5469309110927608,
+      "grad_norm": 1.466244263978851,
+      "learning_rate": 8.967719741600997e-06,
+      "loss": 0.9687,
+      "step": 3974
+    },
+    {
+      "epoch": 0.5470685383980182,
+      "grad_norm": 1.63065074731512,
+      "learning_rate": 8.963286232808494e-06,
+      "loss": 1.1125,
+      "step": 3975
+    },
+    {
+      "epoch": 0.5472061657032755,
+      "grad_norm": 1.7816756175677586,
+      "learning_rate": 8.958852929996821e-06,
+      "loss": 1.0265,
+      "step": 3976
+    },
+    {
+      "epoch": 0.5473437930085329,
+      "grad_norm": 1.6621845010171312,
+      "learning_rate": 8.954419834046823e-06,
+      "loss": 1.0096,
+      "step": 3977
+    },
+    {
+      "epoch": 0.5474814203137902,
+      "grad_norm": 1.5516500163791924,
+      "learning_rate": 8.949986945839289e-06,
+      "loss": 1.0378,
+      "step": 3978
+    },
+    {
+      "epoch": 0.5476190476190477,
+      "grad_norm": 1.937129623332254,
+      "learning_rate": 8.945554266254978e-06,
+      "loss": 1.0714,
+      "step": 3979
+    },
+    {
+      "epoch": 0.547756674924305,
+      "grad_norm": 1.657712255431642,
+      "learning_rate": 8.941121796174596e-06,
+      "loss": 1.0442,
+      "step": 3980
+    },
+    {
+      "epoch": 0.5478943022295624,
+      "grad_norm": 1.6511210146796935,
+      "learning_rate": 8.936689536478822e-06,
+      "loss": 0.9567,
+      "step": 3981
+    },
+    {
+      "epoch": 0.5480319295348197,
+      "grad_norm": 1.7331073572877322,
+      "learning_rate": 8.93225748804828e-06,
+      "loss": 0.9835,
+      "step": 3982
+    },
+    {
+      "epoch": 0.5481695568400771,
+      "grad_norm": 1.6032467015442413,
+      "learning_rate": 8.927825651763563e-06,
+      "loss": 0.9593,
+      "step": 3983
+    },
+    {
+      "epoch": 0.5483071841453344,
+      "grad_norm": 1.8320194597201023,
+      "learning_rate": 8.923394028505212e-06,
+      "loss": 1.0053,
+      "step": 3984
+    },
+    {
+      "epoch": 0.5484448114505918,
+      "grad_norm": 1.6095440819565592,
+      "learning_rate": 8.918962619153727e-06,
+      "loss": 1.0478,
+      "step": 3985
+    },
+    {
+      "epoch": 0.5485824387558491,
+      "grad_norm": 1.8318282956182468,
+      "learning_rate": 8.914531424589575e-06,
+      "loss": 1.0706,
+      "step": 3986
+    },
+    {
+      "epoch": 0.5487200660611066,
+      "grad_norm": 1.7532991842675543,
+      "learning_rate": 8.910100445693168e-06,
+      "loss": 1.0921,
+      "step": 3987
+    },
+    {
+      "epoch": 0.5488576933663639,
+      "grad_norm": 1.722389035071202,
+      "learning_rate": 8.905669683344884e-06,
+      "loss": 1.0027,
+      "step": 3988
+    },
+    {
+      "epoch": 0.5489953206716213,
+      "grad_norm": 1.806594161119985,
+      "learning_rate": 8.901239138425054e-06,
+      "loss": 1.0553,
+      "step": 3989
+    },
+    {
+      "epoch": 0.5491329479768786,
+      "grad_norm": 1.9463313871694514,
+      "learning_rate": 8.896808811813969e-06,
+      "loss": 1.0043,
+      "step": 3990
+    },
+    {
+      "epoch": 0.549270575282136,
+      "grad_norm": 1.644955832287456,
+      "learning_rate": 8.892378704391867e-06,
+      "loss": 1.0799,
+      "step": 3991
+    },
+    {
+      "epoch": 0.5494082025873933,
+      "grad_norm": 1.6820277595725697,
+      "learning_rate": 8.88794881703896e-06,
+      "loss": 0.9991,
+      "step": 3992
+    },
+    {
+      "epoch": 0.5495458298926507,
+      "grad_norm": 1.6806249224646335,
+      "learning_rate": 8.883519150635398e-06,
+      "loss": 1.0465,
+      "step": 3993
+    },
+    {
+      "epoch": 0.549683457197908,
+      "grad_norm": 2.075053601257837,
+      "learning_rate": 8.8790897060613e-06,
+      "loss": 1.0677,
+      "step": 3994
+    },
+    {
+      "epoch": 0.5498210845031655,
+      "grad_norm": 1.932985751040996,
+      "learning_rate": 8.87466048419673e-06,
+      "loss": 0.9824,
+      "step": 3995
+    },
+    {
+      "epoch": 0.5499587118084228,
+      "grad_norm": 1.7139076546024135,
+      "learning_rate": 8.870231485921721e-06,
+      "loss": 1.0104,
+      "step": 3996
+    },
+    {
+      "epoch": 0.5500963391136802,
+      "grad_norm": 1.6743793269743235,
+      "learning_rate": 8.865802712116248e-06,
+      "loss": 0.9583,
+      "step": 3997
+    },
+    {
+      "epoch": 0.5502339664189375,
+      "grad_norm": 1.8631323122129777,
+      "learning_rate": 8.861374163660255e-06,
+      "loss": 1.103,
+      "step": 3998
+    },
+    {
+      "epoch": 0.5503715937241949,
+      "grad_norm": 1.8053725691802114,
+      "learning_rate": 8.856945841433629e-06,
+      "loss": 1.0138,
+      "step": 3999
+    },
+    {
+      "epoch": 0.5505092210294522,
+      "grad_norm": 1.7692770233224753,
+      "learning_rate": 8.852517746316214e-06,
+      "loss": 0.904,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5506468483347096,
+      "grad_norm": 2.3958745104845653,
+      "learning_rate": 8.84808987918782e-06,
+      "loss": 1.0869,
+      "step": 4001
+    },
+    {
+      "epoch": 0.550784475639967,
+      "grad_norm": 1.7069923624706607,
+      "learning_rate": 8.843662240928194e-06,
+      "loss": 0.965,
+      "step": 4002
+    },
+    {
+      "epoch": 0.5509221029452244,
+      "grad_norm": 1.8362772562101821,
+      "learning_rate": 8.839234832417055e-06,
+      "loss": 1.1439,
+      "step": 4003
+    },
+    {
+      "epoch": 0.5510597302504817,
+      "grad_norm": 1.6376008618183546,
+      "learning_rate": 8.834807654534062e-06,
+      "loss": 1.0178,
+      "step": 4004
+    },
+    {
+      "epoch": 0.5511973575557391,
+      "grad_norm": 1.8150890728629336,
+      "learning_rate": 8.830380708158844e-06,
+      "loss": 0.9878,
+      "step": 4005
+    },
+    {
+      "epoch": 0.5513349848609964,
+      "grad_norm": 1.8397883561360153,
+      "learning_rate": 8.825953994170966e-06,
+      "loss": 1.0777,
+      "step": 4006
+    },
+    {
+      "epoch": 0.5514726121662538,
+      "grad_norm": 1.7939942560549935,
+      "learning_rate": 8.821527513449959e-06,
+      "loss": 1.0434,
+      "step": 4007
+    },
+    {
+      "epoch": 0.5516102394715111,
+      "grad_norm": 1.7230518711173215,
+      "learning_rate": 8.817101266875303e-06,
+      "loss": 1.0185,
+      "step": 4008
+    },
+    {
+      "epoch": 0.5517478667767685,
+      "grad_norm": 1.7652588136734766,
+      "learning_rate": 8.812675255326435e-06,
+      "loss": 1.0446,
+      "step": 4009
+    },
+    {
+      "epoch": 0.5518854940820259,
+      "grad_norm": 1.969856149240467,
+      "learning_rate": 8.808249479682736e-06,
+      "loss": 0.968,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5520231213872833,
+      "grad_norm": 2.5667842507868297,
+      "learning_rate": 8.803823940823557e-06,
+      "loss": 0.9724,
+      "step": 4011
+    },
+    {
+      "epoch": 0.5521607486925406,
+      "grad_norm": 1.5113473973455593,
+      "learning_rate": 8.799398639628186e-06,
+      "loss": 1.0037,
+      "step": 4012
+    },
+    {
+      "epoch": 0.552298375997798,
+      "grad_norm": 1.810026846678308,
+      "learning_rate": 8.79497357697587e-06,
+      "loss": 0.8971,
+      "step": 4013
+    },
+    {
+      "epoch": 0.5524360033030553,
+      "grad_norm": 1.639963655110258,
+      "learning_rate": 8.790548753745808e-06,
+      "loss": 1.038,
+      "step": 4014
+    },
+    {
+      "epoch": 0.5525736306083127,
+      "grad_norm": 1.8695529145548873,
+      "learning_rate": 8.786124170817158e-06,
+      "loss": 1.0049,
+      "step": 4015
+    },
+    {
+      "epoch": 0.55271125791357,
+      "grad_norm": 1.746157513417196,
+      "learning_rate": 8.781699829069018e-06,
+      "loss": 0.9324,
+      "step": 4016
+    },
+    {
+      "epoch": 0.5528488852188275,
+      "grad_norm": 1.6869901825141598,
+      "learning_rate": 8.777275729380442e-06,
+      "loss": 0.977,
+      "step": 4017
+    },
+    {
+      "epoch": 0.5529865125240848,
+      "grad_norm": 1.965230850630973,
+      "learning_rate": 8.772851872630442e-06,
+      "loss": 1.0562,
+      "step": 4018
+    },
+    {
+      "epoch": 0.5531241398293422,
+      "grad_norm": 1.6411055931948964,
+      "learning_rate": 8.768428259697978e-06,
+      "loss": 1.1002,
+      "step": 4019
+    },
+    {
+      "epoch": 0.5532617671345995,
+      "grad_norm": 1.8347479565068978,
+      "learning_rate": 8.76400489146196e-06,
+      "loss": 0.9842,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5533993944398569,
+      "grad_norm": 1.6047107626068835,
+      "learning_rate": 8.759581768801248e-06,
+      "loss": 1.0443,
+      "step": 4021
+    },
+    {
+      "epoch": 0.5535370217451142,
+      "grad_norm": 1.6718159477573014,
+      "learning_rate": 8.755158892594665e-06,
+      "loss": 0.9755,
+      "step": 4022
+    },
+    {
+      "epoch": 0.5536746490503716,
+      "grad_norm": 1.54774555392231,
+      "learning_rate": 8.750736263720966e-06,
+      "loss": 1.0119,
+      "step": 4023
+    },
+    {
+      "epoch": 0.5538122763556289,
+      "grad_norm": 1.7992675325515233,
+      "learning_rate": 8.746313883058875e-06,
+      "loss": 1.0391,
+      "step": 4024
+    },
+    {
+      "epoch": 0.5539499036608864,
+      "grad_norm": 1.6449476719333789,
+      "learning_rate": 8.741891751487049e-06,
+      "loss": 1.0209,
+      "step": 4025
+    },
+    {
+      "epoch": 0.5540875309661437,
+      "grad_norm": 1.847939023220919,
+      "learning_rate": 8.737469869884115e-06,
+      "loss": 1.062,
+      "step": 4026
+    },
+    {
+      "epoch": 0.5542251582714011,
+      "grad_norm": 1.7278692815253256,
+      "learning_rate": 8.733048239128632e-06,
+      "loss": 1.1497,
+      "step": 4027
+    },
+    {
+      "epoch": 0.5543627855766584,
+      "grad_norm": 1.723829063014673,
+      "learning_rate": 8.728626860099123e-06,
+      "loss": 1.0605,
+      "step": 4028
+    },
+    {
+      "epoch": 0.5545004128819158,
+      "grad_norm": 1.7343319783835072,
+      "learning_rate": 8.72420573367405e-06,
+      "loss": 1.0697,
+      "step": 4029
+    },
+    {
+      "epoch": 0.5546380401871731,
+      "grad_norm": 1.6170309632616842,
+      "learning_rate": 8.71978486073184e-06,
+      "loss": 0.9294,
+      "step": 4030
+    },
+    {
+      "epoch": 0.5547756674924305,
+      "grad_norm": 1.677417727288273,
+      "learning_rate": 8.715364242150854e-06,
+      "loss": 0.9211,
+      "step": 4031
+    },
+    {
+      "epoch": 0.5549132947976878,
+      "grad_norm": 1.648150433954116,
+      "learning_rate": 8.710943878809404e-06,
+      "loss": 0.9992,
+      "step": 4032
+    },
+    {
+      "epoch": 0.5550509221029453,
+      "grad_norm": 1.8157585018859017,
+      "learning_rate": 8.706523771585763e-06,
+      "loss": 0.9549,
+      "step": 4033
+    },
+    {
+      "epoch": 0.5551885494082026,
+      "grad_norm": 1.5317489786835896,
+      "learning_rate": 8.702103921358141e-06,
+      "loss": 0.9875,
+      "step": 4034
+    },
+    {
+      "epoch": 0.55532617671346,
+      "grad_norm": 1.8266023298073102,
+      "learning_rate": 8.697684329004703e-06,
+      "loss": 1.0267,
+      "step": 4035
+    },
+    {
+      "epoch": 0.5554638040187173,
+      "grad_norm": 1.6517954445642442,
+      "learning_rate": 8.693264995403562e-06,
+      "loss": 1.0624,
+      "step": 4036
+    },
+    {
+      "epoch": 0.5556014313239747,
+      "grad_norm": 1.6643250004330805,
+      "learning_rate": 8.688845921432783e-06,
+      "loss": 1.0087,
+      "step": 4037
+    },
+    {
+      "epoch": 0.555739058629232,
+      "grad_norm": 1.82944166980534,
+      "learning_rate": 8.684427107970366e-06,
+      "loss": 1.0424,
+      "step": 4038
+    },
+    {
+      "epoch": 0.5558766859344894,
+      "grad_norm": 1.6016502000775719,
+      "learning_rate": 8.680008555894281e-06,
+      "loss": 0.9601,
+      "step": 4039
+    },
+    {
+      "epoch": 0.5560143132397468,
+      "grad_norm": 1.8162718506778595,
+      "learning_rate": 8.67559026608242e-06,
+      "loss": 1.0208,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5561519405450042,
+      "grad_norm": 1.7925171268797053,
+      "learning_rate": 8.67117223941265e-06,
+      "loss": 1.0698,
+      "step": 4041
+    },
+    {
+      "epoch": 0.5562895678502615,
+      "grad_norm": 1.9367795024546761,
+      "learning_rate": 8.666754476762763e-06,
+      "loss": 0.9788,
+      "step": 4042
+    },
+    {
+      "epoch": 0.5564271951555189,
+      "grad_norm": 1.721739609579454,
+      "learning_rate": 8.662336979010511e-06,
+      "loss": 0.9788,
+      "step": 4043
+    },
+    {
+      "epoch": 0.5565648224607762,
+      "grad_norm": 1.67068907990827,
+      "learning_rate": 8.65791974703359e-06,
+      "loss": 0.995,
+      "step": 4044
+    },
+    {
+      "epoch": 0.5567024497660336,
+      "grad_norm": 1.6841709022819278,
+      "learning_rate": 8.65350278170965e-06,
+      "loss": 0.9932,
+      "step": 4045
+    },
+    {
+      "epoch": 0.5568400770712909,
+      "grad_norm": 1.6726611603998105,
+      "learning_rate": 8.649086083916269e-06,
+      "loss": 1.0123,
+      "step": 4046
+    },
+    {
+      "epoch": 0.5569777043765483,
+      "grad_norm": 1.879855732792412,
+      "learning_rate": 8.644669654530995e-06,
+      "loss": 1.0081,
+      "step": 4047
+    },
+    {
+      "epoch": 0.5571153316818057,
+      "grad_norm": 4.559539425623358,
+      "learning_rate": 8.640253494431311e-06,
+      "loss": 1.0354,
+      "step": 4048
+    },
+    {
+      "epoch": 0.5572529589870631,
+      "grad_norm": 1.6266265189142677,
+      "learning_rate": 8.635837604494638e-06,
+      "loss": 0.9671,
+      "step": 4049
+    },
+    {
+      "epoch": 0.5573905862923204,
+      "grad_norm": 1.8928121903374728,
+      "learning_rate": 8.631421985598363e-06,
+      "loss": 0.9685,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5575282135975778,
+      "grad_norm": 1.65852396430693,
+      "learning_rate": 8.627006638619803e-06,
+      "loss": 1.0248,
+      "step": 4051
+    },
+    {
+      "epoch": 0.5576658409028351,
+      "grad_norm": 1.6967778889823284,
+      "learning_rate": 8.622591564436235e-06,
+      "loss": 0.9628,
+      "step": 4052
+    },
+    {
+      "epoch": 0.5578034682080925,
+      "grad_norm": 1.5656230937495763,
+      "learning_rate": 8.618176763924864e-06,
+      "loss": 1.0085,
+      "step": 4053
+    },
+    {
+      "epoch": 0.5579410955133498,
+      "grad_norm": 1.8058466656836607,
+      "learning_rate": 8.613762237962858e-06,
+      "loss": 1.0253,
+      "step": 4054
+    },
+    {
+      "epoch": 0.5580787228186073,
+      "grad_norm": 1.8885789266120219,
+      "learning_rate": 8.609347987427318e-06,
+      "loss": 1.0161,
+      "step": 4055
+    },
+    {
+      "epoch": 0.5582163501238646,
+      "grad_norm": 1.6454330256265008,
+      "learning_rate": 8.604934013195297e-06,
+      "loss": 0.9388,
+      "step": 4056
+    },
+    {
+      "epoch": 0.558353977429122,
+      "grad_norm": 1.6538748153710856,
+      "learning_rate": 8.60052031614379e-06,
+      "loss": 1.0463,
+      "step": 4057
+    },
+    {
+      "epoch": 0.5584916047343793,
+      "grad_norm": 1.7003689336770682,
+      "learning_rate": 8.59610689714974e-06,
+      "loss": 0.9918,
+      "step": 4058
+    },
+    {
+      "epoch": 0.5586292320396367,
+      "grad_norm": 1.8742696488359554,
+      "learning_rate": 8.591693757090034e-06,
+      "loss": 1.11,
+      "step": 4059
+    },
+    {
+      "epoch": 0.558766859344894,
+      "grad_norm": 1.61900153167684,
+      "learning_rate": 8.587280896841499e-06,
+      "loss": 0.9137,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5589044866501514,
+      "grad_norm": 1.6611891086280002,
+      "learning_rate": 8.582868317280911e-06,
+      "loss": 0.9523,
+      "step": 4061
+    },
+    {
+      "epoch": 0.5590421139554087,
+      "grad_norm": 1.9671402986853632,
+      "learning_rate": 8.578456019284994e-06,
+      "loss": 1.0609,
+      "step": 4062
+    },
+    {
+      "epoch": 0.5591797412606662,
+      "grad_norm": 1.862561616468973,
+      "learning_rate": 8.574044003730408e-06,
+      "loss": 0.978,
+      "step": 4063
+    },
+    {
+      "epoch": 0.5593173685659235,
+      "grad_norm": 1.6964521811301743,
+      "learning_rate": 8.569632271493752e-06,
+      "loss": 1.0476,
+      "step": 4064
+    },
+    {
+      "epoch": 0.5594549958711809,
+      "grad_norm": 1.6443698193151444,
+      "learning_rate": 8.56522082345159e-06,
+      "loss": 0.9877,
+      "step": 4065
+    },
+    {
+      "epoch": 0.5595926231764382,
+      "grad_norm": 1.6216029750041632,
+      "learning_rate": 8.560809660480404e-06,
+      "loss": 0.9407,
+      "step": 4066
+    },
+    {
+      "epoch": 0.5597302504816956,
+      "grad_norm": 2.102355596975558,
+      "learning_rate": 8.556398783456646e-06,
+      "loss": 1.0712,
+      "step": 4067
+    },
+    {
+      "epoch": 0.5598678777869529,
+      "grad_norm": 1.8415325131689604,
+      "learning_rate": 8.551988193256683e-06,
+      "loss": 1.0887,
+      "step": 4068
+    },
+    {
+      "epoch": 0.5600055050922103,
+      "grad_norm": 2.0849150967039405,
+      "learning_rate": 8.54757789075685e-06,
+      "loss": 1.0445,
+      "step": 4069
+    },
+    {
+      "epoch": 0.5601431323974676,
+      "grad_norm": 1.9241149133496436,
+      "learning_rate": 8.543167876833404e-06,
+      "loss": 1.0984,
+      "step": 4070
+    },
+    {
+      "epoch": 0.5602807597027251,
+      "grad_norm": 1.7858312818164144,
+      "learning_rate": 8.538758152362562e-06,
+      "loss": 1.0158,
+      "step": 4071
+    },
+    {
+      "epoch": 0.5604183870079824,
+      "grad_norm": 1.8552281674567,
+      "learning_rate": 8.534348718220471e-06,
+      "loss": 0.959,
+      "step": 4072
+    },
+    {
+      "epoch": 0.5605560143132398,
+      "grad_norm": 1.9698397455504308,
+      "learning_rate": 8.529939575283229e-06,
+      "loss": 0.9777,
+      "step": 4073
+    },
+    {
+      "epoch": 0.5606936416184971,
+      "grad_norm": 1.7615288978532848,
+      "learning_rate": 8.52553072442687e-06,
+      "loss": 1.0806,
+      "step": 4074
+    },
+    {
+      "epoch": 0.5608312689237545,
+      "grad_norm": 1.7076912862558042,
+      "learning_rate": 8.521122166527371e-06,
+      "loss": 0.9263,
+      "step": 4075
+    },
+    {
+      "epoch": 0.5609688962290118,
+      "grad_norm": 1.5597410190011463,
+      "learning_rate": 8.516713902460654e-06,
+      "loss": 1.0283,
+      "step": 4076
+    },
+    {
+      "epoch": 0.5611065235342692,
+      "grad_norm": 1.810549438251086,
+      "learning_rate": 8.512305933102584e-06,
+      "loss": 1.061,
+      "step": 4077
+    },
+    {
+      "epoch": 0.5612441508395266,
+      "grad_norm": 1.7597212102633768,
+      "learning_rate": 8.507898259328955e-06,
+      "loss": 1.0767,
+      "step": 4078
+    },
+    {
+      "epoch": 0.561381778144784,
+      "grad_norm": 1.7646516749701504,
+      "learning_rate": 8.503490882015523e-06,
+      "loss": 1.0038,
+      "step": 4079
+    },
+    {
+      "epoch": 0.5615194054500413,
+      "grad_norm": 1.6985309707618284,
+      "learning_rate": 8.499083802037966e-06,
+      "loss": 1.0404,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5616570327552987,
+      "grad_norm": 1.8058245361566534,
+      "learning_rate": 8.494677020271913e-06,
+      "loss": 0.9264,
+      "step": 4081
+    },
+    {
+      "epoch": 0.561794660060556,
+      "grad_norm": 1.726167554726961,
+      "learning_rate": 8.490270537592928e-06,
+      "loss": 0.9795,
+      "step": 4082
+    },
+    {
+      "epoch": 0.5619322873658134,
+      "grad_norm": 1.5351131897418806,
+      "learning_rate": 8.485864354876521e-06,
+      "loss": 1.0113,
+      "step": 4083
+    },
+    {
+      "epoch": 0.5620699146710707,
+      "grad_norm": 1.8562635186896628,
+      "learning_rate": 8.481458472998145e-06,
+      "loss": 0.9286,
+      "step": 4084
+    },
+    {
+      "epoch": 0.562207541976328,
+      "grad_norm": 1.5587993968423375,
+      "learning_rate": 8.477052892833177e-06,
+      "loss": 1.0345,
+      "step": 4085
+    },
+    {
+      "epoch": 0.5623451692815855,
+      "grad_norm": 1.6956688642396152,
+      "learning_rate": 8.47264761525696e-06,
+      "loss": 0.935,
+      "step": 4086
+    },
+    {
+      "epoch": 0.5624827965868429,
+      "grad_norm": 1.737043827251866,
+      "learning_rate": 8.468242641144748e-06,
+      "loss": 0.9354,
+      "step": 4087
+    },
+    {
+      "epoch": 0.5626204238921002,
+      "grad_norm": 2.359954573574087,
+      "learning_rate": 8.46383797137176e-06,
+      "loss": 1.0025,
+      "step": 4088
+    },
+    {
+      "epoch": 0.5627580511973576,
+      "grad_norm": 1.741436545414816,
+      "learning_rate": 8.459433606813137e-06,
+      "loss": 1.0246,
+      "step": 4089
+    },
+    {
+      "epoch": 0.5628956785026149,
+      "grad_norm": 2.11279603890354,
+      "learning_rate": 8.455029548343969e-06,
+      "loss": 1.1304,
+      "step": 4090
+    },
+    {
+      "epoch": 0.5630333058078723,
+      "grad_norm": 1.6667892461492047,
+      "learning_rate": 8.45062579683928e-06,
+      "loss": 1.0209,
+      "step": 4091
+    },
+    {
+      "epoch": 0.5631709331131296,
+      "grad_norm": 1.8375287889582337,
+      "learning_rate": 8.446222353174039e-06,
+      "loss": 1.0425,
+      "step": 4092
+    },
+    {
+      "epoch": 0.5633085604183871,
+      "grad_norm": 1.7491430393656549,
+      "learning_rate": 8.441819218223144e-06,
+      "loss": 1.0108,
+      "step": 4093
+    },
+    {
+      "epoch": 0.5634461877236444,
+      "grad_norm": 1.670382579953507,
+      "learning_rate": 8.437416392861445e-06,
+      "loss": 1.0904,
+      "step": 4094
+    },
+    {
+      "epoch": 0.5635838150289018,
+      "grad_norm": 1.8840698238412177,
+      "learning_rate": 8.433013877963722e-06,
+      "loss": 1.0433,
+      "step": 4095
+    },
+    {
+      "epoch": 0.5637214423341591,
+      "grad_norm": 2.2659205124774218,
+      "learning_rate": 8.428611674404686e-06,
+      "loss": 1.0398,
+      "step": 4096
+    },
+    {
+      "epoch": 0.5638590696394165,
+      "grad_norm": 1.824576646614936,
+      "learning_rate": 8.424209783059001e-06,
+      "loss": 1.0627,
+      "step": 4097
+    },
+    {
+      "epoch": 0.5639966969446738,
+      "grad_norm": 1.904794571862147,
+      "learning_rate": 8.419808204801261e-06,
+      "loss": 1.0601,
+      "step": 4098
+    },
+    {
+      "epoch": 0.5641343242499312,
+      "grad_norm": 1.6467001724313644,
+      "learning_rate": 8.415406940506005e-06,
+      "loss": 0.99,
+      "step": 4099
+    },
+    {
+      "epoch": 0.5642719515551885,
+      "grad_norm": 1.5494121172512694,
+      "learning_rate": 8.411005991047695e-06,
+      "loss": 0.9969,
+      "step": 4100
+    },
+    {
+      "epoch": 0.564409578860446,
+      "grad_norm": 1.6290232300187526,
+      "learning_rate": 8.406605357300747e-06,
+      "loss": 1.0118,
+      "step": 4101
+    },
+    {
+      "epoch": 0.5645472061657033,
+      "grad_norm": 1.7657400921338566,
+      "learning_rate": 8.402205040139498e-06,
+      "loss": 0.9674,
+      "step": 4102
+    },
+    {
+      "epoch": 0.5646848334709607,
+      "grad_norm": 1.585490905574864,
+      "learning_rate": 8.397805040438242e-06,
+      "loss": 1.0422,
+      "step": 4103
+    },
+    {
+      "epoch": 0.564822460776218,
+      "grad_norm": 1.7485420981466837,
+      "learning_rate": 8.393405359071188e-06,
+      "loss": 0.9111,
+      "step": 4104
+    },
+    {
+      "epoch": 0.5649600880814754,
+      "grad_norm": 1.80029257820778,
+      "learning_rate": 8.3890059969125e-06,
+      "loss": 0.9125,
+      "step": 4105
+    },
+    {
+      "epoch": 0.5650977153867327,
+      "grad_norm": 1.7879356397236177,
+      "learning_rate": 8.384606954836264e-06,
+      "loss": 1.0473,
+      "step": 4106
+    },
+    {
+      "epoch": 0.56523534269199,
+      "grad_norm": 1.7142478817277533,
+      "learning_rate": 8.38020823371652e-06,
+      "loss": 0.968,
+      "step": 4107
+    },
+    {
+      "epoch": 0.5653729699972474,
+      "grad_norm": 1.6927398443145103,
+      "learning_rate": 8.375809834427225e-06,
+      "loss": 0.9704,
+      "step": 4108
+    },
+    {
+      "epoch": 0.5655105973025049,
+      "grad_norm": 1.720216511232695,
+      "learning_rate": 8.371411757842285e-06,
+      "loss": 1.0429,
+      "step": 4109
+    },
+    {
+      "epoch": 0.5656482246077622,
+      "grad_norm": 1.837175632656309,
+      "learning_rate": 8.367014004835531e-06,
+      "loss": 1.0877,
+      "step": 4110
+    },
+    {
+      "epoch": 0.5657858519130196,
+      "grad_norm": 1.6763421175317594,
+      "learning_rate": 8.362616576280747e-06,
+      "loss": 0.981,
+      "step": 4111
+    },
+    {
+      "epoch": 0.5659234792182769,
+      "grad_norm": 1.9256625709035307,
+      "learning_rate": 8.358219473051632e-06,
+      "loss": 0.9022,
+      "step": 4112
+    },
+    {
+      "epoch": 0.5660611065235343,
+      "grad_norm": 1.69977757583853,
+      "learning_rate": 8.353822696021835e-06,
+      "loss": 0.9143,
+      "step": 4113
+    },
+    {
+      "epoch": 0.5661987338287916,
+      "grad_norm": 1.71099276646812,
+      "learning_rate": 8.349426246064936e-06,
+      "loss": 0.9974,
+      "step": 4114
+    },
+    {
+      "epoch": 0.566336361134049,
+      "grad_norm": 1.5619621400764025,
+      "learning_rate": 8.345030124054446e-06,
+      "loss": 0.9613,
+      "step": 4115
+    },
+    {
+      "epoch": 0.5664739884393064,
+      "grad_norm": 1.8272880564800456,
+      "learning_rate": 8.340634330863818e-06,
+      "loss": 1.0615,
+      "step": 4116
+    },
+    {
+      "epoch": 0.5666116157445638,
+      "grad_norm": 1.7163420357547579,
+      "learning_rate": 8.336238867366431e-06,
+      "loss": 1.0695,
+      "step": 4117
+    },
+    {
+      "epoch": 0.5667492430498211,
+      "grad_norm": 2.032561028652858,
+      "learning_rate": 8.331843734435612e-06,
+      "loss": 0.9495,
+      "step": 4118
+    },
+    {
+      "epoch": 0.5668868703550785,
+      "grad_norm": 2.0101585082255515,
+      "learning_rate": 8.327448932944601e-06,
+      "loss": 0.956,
+      "step": 4119
+    },
+    {
+      "epoch": 0.5670244976603358,
+      "grad_norm": 1.7803288666189858,
+      "learning_rate": 8.323054463766598e-06,
+      "loss": 0.9781,
+      "step": 4120
+    },
+    {
+      "epoch": 0.5671621249655932,
+      "grad_norm": 1.6629451221539324,
+      "learning_rate": 8.318660327774716e-06,
+      "loss": 1.024,
+      "step": 4121
+    },
+    {
+      "epoch": 0.5672997522708505,
+      "grad_norm": 1.824961916746767,
+      "learning_rate": 8.314266525842014e-06,
+      "loss": 1.0462,
+      "step": 4122
+    },
+    {
+      "epoch": 0.5674373795761078,
+      "grad_norm": 1.7673697102343662,
+      "learning_rate": 8.309873058841474e-06,
+      "loss": 1.0283,
+      "step": 4123
+    },
+    {
+      "epoch": 0.5675750068813653,
+      "grad_norm": 1.6882592944303085,
+      "learning_rate": 8.305479927646028e-06,
+      "loss": 1.0205,
+      "step": 4124
+    },
+    {
+      "epoch": 0.5677126341866227,
+      "grad_norm": 1.5938230552812813,
+      "learning_rate": 8.301087133128518e-06,
+      "loss": 0.9735,
+      "step": 4125
+    },
+    {
+      "epoch": 0.56785026149188,
+      "grad_norm": 1.6891376968182923,
+      "learning_rate": 8.296694676161745e-06,
+      "loss": 0.9967,
+      "step": 4126
+    },
+    {
+      "epoch": 0.5679878887971374,
+      "grad_norm": 1.7463496108187126,
+      "learning_rate": 8.292302557618422e-06,
+      "loss": 1.0411,
+      "step": 4127
+    },
+    {
+      "epoch": 0.5681255161023947,
+      "grad_norm": 2.142899485341186,
+      "learning_rate": 8.287910778371206e-06,
+      "loss": 1.0389,
+      "step": 4128
+    },
+    {
+      "epoch": 0.568263143407652,
+      "grad_norm": 1.9476377944016832,
+      "learning_rate": 8.28351933929268e-06,
+      "loss": 1.0298,
+      "step": 4129
+    },
+    {
+      "epoch": 0.5684007707129094,
+      "grad_norm": 1.6706843649527119,
+      "learning_rate": 8.279128241255365e-06,
+      "loss": 1.0298,
+      "step": 4130
+    },
+    {
+      "epoch": 0.5685383980181669,
+      "grad_norm": 1.8046434549279606,
+      "learning_rate": 8.274737485131715e-06,
+      "loss": 1.0265,
+      "step": 4131
+    },
+    {
+      "epoch": 0.5686760253234242,
+      "grad_norm": 1.6909768333276554,
+      "learning_rate": 8.270347071794107e-06,
+      "loss": 0.9827,
+      "step": 4132
+    },
+    {
+      "epoch": 0.5688136526286816,
+      "grad_norm": 1.794147547346384,
+      "learning_rate": 8.265957002114864e-06,
+      "loss": 0.9373,
+      "step": 4133
+    },
+    {
+      "epoch": 0.5689512799339389,
+      "grad_norm": 1.8415423873415095,
+      "learning_rate": 8.261567276966223e-06,
+      "loss": 0.9539,
+      "step": 4134
+    },
+    {
+      "epoch": 0.5690889072391963,
+      "grad_norm": 1.831705029545054,
+      "learning_rate": 8.257177897220371e-06,
+      "loss": 1.0131,
+      "step": 4135
+    },
+    {
+      "epoch": 0.5692265345444536,
+      "grad_norm": 1.818652899879206,
+      "learning_rate": 8.252788863749414e-06,
+      "loss": 0.9498,
+      "step": 4136
+    },
+    {
+      "epoch": 0.569364161849711,
+      "grad_norm": 1.641762885434339,
+      "learning_rate": 8.248400177425392e-06,
+      "loss": 1.022,
+      "step": 4137
+    },
+    {
+      "epoch": 0.5695017891549683,
+      "grad_norm": 1.78633652596997,
+      "learning_rate": 8.244011839120277e-06,
+      "loss": 1.0665,
+      "step": 4138
+    },
+    {
+      "epoch": 0.5696394164602258,
+      "grad_norm": 1.6074253916566155,
+      "learning_rate": 8.239623849705976e-06,
+      "loss": 1.0811,
+      "step": 4139
+    },
+    {
+      "epoch": 0.5697770437654831,
+      "grad_norm": 1.7452729177445854,
+      "learning_rate": 8.235236210054315e-06,
+      "loss": 0.9746,
+      "step": 4140
+    },
+    {
+      "epoch": 0.5699146710707405,
+      "grad_norm": 1.8319463797494118,
+      "learning_rate": 8.230848921037066e-06,
+      "loss": 0.9687,
+      "step": 4141
+    },
+    {
+      "epoch": 0.5700522983759978,
+      "grad_norm": 1.9204861996323552,
+      "learning_rate": 8.226461983525917e-06,
+      "loss": 0.9407,
+      "step": 4142
+    },
+    {
+      "epoch": 0.5701899256812551,
+      "grad_norm": 1.6694819668121021,
+      "learning_rate": 8.222075398392498e-06,
+      "loss": 0.9357,
+      "step": 4143
+    },
+    {
+      "epoch": 0.5703275529865125,
+      "grad_norm": 1.546947872661611,
+      "learning_rate": 8.217689166508357e-06,
+      "loss": 0.986,
+      "step": 4144
+    },
+    {
+      "epoch": 0.5704651802917698,
+      "grad_norm": 1.6841022492177011,
+      "learning_rate": 8.21330328874498e-06,
+      "loss": 0.9404,
+      "step": 4145
+    },
+    {
+      "epoch": 0.5706028075970272,
+      "grad_norm": 1.757863698683949,
+      "learning_rate": 8.208917765973787e-06,
+      "loss": 1.0384,
+      "step": 4146
+    },
+    {
+      "epoch": 0.5707404349022847,
+      "grad_norm": 1.5671479923831644,
+      "learning_rate": 8.204532599066112e-06,
+      "loss": 0.994,
+      "step": 4147
+    },
+    {
+      "epoch": 0.570878062207542,
+      "grad_norm": 1.7908940269865012,
+      "learning_rate": 8.200147788893238e-06,
+      "loss": 1.0381,
+      "step": 4148
+    },
+    {
+      "epoch": 0.5710156895127994,
+      "grad_norm": 1.5031083094303492,
+      "learning_rate": 8.195763336326354e-06,
+      "loss": 0.9687,
+      "step": 4149
+    },
+    {
+      "epoch": 0.5711533168180567,
+      "grad_norm": 1.647351778653535,
+      "learning_rate": 8.191379242236603e-06,
+      "loss": 0.9369,
+      "step": 4150
+    },
+    {
+      "epoch": 0.571290944123314,
+      "grad_norm": 1.6423571612387926,
+      "learning_rate": 8.186995507495036e-06,
+      "loss": 1.0494,
+      "step": 4151
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.6667378774333477,
+      "learning_rate": 8.182612132972646e-06,
+      "loss": 0.9383,
+      "step": 4152
+    },
+    {
+      "epoch": 0.5715661987338287,
+      "grad_norm": 1.747671069083027,
+      "learning_rate": 8.178229119540345e-06,
+      "loss": 1.0125,
+      "step": 4153
+    },
+    {
+      "epoch": 0.5717038260390862,
+      "grad_norm": 1.6848249564610762,
+      "learning_rate": 8.173846468068987e-06,
+      "loss": 0.9725,
+      "step": 4154
+    },
+    {
+      "epoch": 0.5718414533443436,
+      "grad_norm": 1.7014445639576605,
+      "learning_rate": 8.169464179429332e-06,
+      "loss": 1.0055,
+      "step": 4155
+    },
+    {
+      "epoch": 0.5719790806496009,
+      "grad_norm": 1.852314589150778,
+      "learning_rate": 8.16508225449209e-06,
+      "loss": 0.9877,
+      "step": 4156
+    },
+    {
+      "epoch": 0.5721167079548583,
+      "grad_norm": 3.3906933936873083,
+      "learning_rate": 8.160700694127887e-06,
+      "loss": 0.9692,
+      "step": 4157
+    },
+    {
+      "epoch": 0.5722543352601156,
+      "grad_norm": 1.8309410676235658,
+      "learning_rate": 8.15631949920728e-06,
+      "loss": 1.0929,
+      "step": 4158
+    },
+    {
+      "epoch": 0.572391962565373,
+      "grad_norm": 1.752374117749652,
+      "learning_rate": 8.15193867060075e-06,
+      "loss": 0.9239,
+      "step": 4159
+    },
+    {
+      "epoch": 0.5725295898706303,
+      "grad_norm": 2.071523066067662,
+      "learning_rate": 8.147558209178708e-06,
+      "loss": 0.9809,
+      "step": 4160
+    },
+    {
+      "epoch": 0.5726672171758876,
+      "grad_norm": 2.0002157567798,
+      "learning_rate": 8.143178115811495e-06,
+      "loss": 1.0785,
+      "step": 4161
+    },
+    {
+      "epoch": 0.5728048444811451,
+      "grad_norm": 1.870131482615628,
+      "learning_rate": 8.13879839136937e-06,
+      "loss": 1.0327,
+      "step": 4162
+    },
+    {
+      "epoch": 0.5729424717864025,
+      "grad_norm": 1.8950512765929906,
+      "learning_rate": 8.13441903672253e-06,
+      "loss": 0.9886,
+      "step": 4163
+    },
+    {
+      "epoch": 0.5730800990916598,
+      "grad_norm": 1.7654884561549298,
+      "learning_rate": 8.130040052741089e-06,
+      "loss": 1.0525,
+      "step": 4164
+    },
+    {
+      "epoch": 0.5732177263969171,
+      "grad_norm": 1.8178888660495405,
+      "learning_rate": 8.125661440295094e-06,
+      "loss": 1.0443,
+      "step": 4165
+    },
+    {
+      "epoch": 0.5733553537021745,
+      "grad_norm": 1.7033579224334903,
+      "learning_rate": 8.121283200254513e-06,
+      "loss": 1.0342,
+      "step": 4166
+    },
+    {
+      "epoch": 0.5734929810074318,
+      "grad_norm": 1.7209275288783732,
+      "learning_rate": 8.116905333489242e-06,
+      "loss": 0.9357,
+      "step": 4167
+    },
+    {
+      "epoch": 0.5736306083126892,
+      "grad_norm": 1.870786556336648,
+      "learning_rate": 8.112527840869105e-06,
+      "loss": 1.0286,
+      "step": 4168
+    },
+    {
+      "epoch": 0.5737682356179467,
+      "grad_norm": 1.7912247826211616,
+      "learning_rate": 8.108150723263852e-06,
+      "loss": 1.0459,
+      "step": 4169
+    },
+    {
+      "epoch": 0.573905862923204,
+      "grad_norm": 1.755538482703091,
+      "learning_rate": 8.103773981543151e-06,
+      "loss": 0.9169,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5740434902284614,
+      "grad_norm": 2.6369218349891885,
+      "learning_rate": 8.099397616576605e-06,
+      "loss": 1.0694,
+      "step": 4171
+    },
+    {
+      "epoch": 0.5741811175337187,
+      "grad_norm": 1.5730415568174594,
+      "learning_rate": 8.095021629233733e-06,
+      "loss": 0.9562,
+      "step": 4172
+    },
+    {
+      "epoch": 0.574318744838976,
+      "grad_norm": 1.5959210576067544,
+      "learning_rate": 8.090646020383993e-06,
+      "loss": 0.9689,
+      "step": 4173
+    },
+    {
+      "epoch": 0.5744563721442334,
+      "grad_norm": 1.717630115731798,
+      "learning_rate": 8.086270790896747e-06,
+      "loss": 0.986,
+      "step": 4174
+    },
+    {
+      "epoch": 0.5745939994494907,
+      "grad_norm": 1.7834953519071666,
+      "learning_rate": 8.0818959416413e-06,
+      "loss": 0.9045,
+      "step": 4175
+    },
+    {
+      "epoch": 0.5747316267547481,
+      "grad_norm": 1.8576152238066745,
+      "learning_rate": 8.077521473486877e-06,
+      "loss": 0.9996,
+      "step": 4176
+    },
+    {
+      "epoch": 0.5748692540600056,
+      "grad_norm": 1.7341241637037148,
+      "learning_rate": 8.073147387302616e-06,
+      "loss": 1.0178,
+      "step": 4177
+    },
+    {
+      "epoch": 0.5750068813652629,
+      "grad_norm": 1.6271349682690879,
+      "learning_rate": 8.068773683957598e-06,
+      "loss": 0.9513,
+      "step": 4178
+    },
+    {
+      "epoch": 0.5751445086705202,
+      "grad_norm": 1.9750022725718093,
+      "learning_rate": 8.064400364320809e-06,
+      "loss": 1.052,
+      "step": 4179
+    },
+    {
+      "epoch": 0.5752821359757776,
+      "grad_norm": 1.7249060856776555,
+      "learning_rate": 8.060027429261176e-06,
+      "loss": 1.0218,
+      "step": 4180
+    },
+    {
+      "epoch": 0.575419763281035,
+      "grad_norm": 1.7620318514282844,
+      "learning_rate": 8.055654879647535e-06,
+      "loss": 1.0455,
+      "step": 4181
+    },
+    {
+      "epoch": 0.5755573905862923,
+      "grad_norm": 1.6750596616229183,
+      "learning_rate": 8.051282716348655e-06,
+      "loss": 0.9941,
+      "step": 4182
+    },
+    {
+      "epoch": 0.5756950178915496,
+      "grad_norm": 1.6714530412314512,
+      "learning_rate": 8.046910940233224e-06,
+      "loss": 1.0569,
+      "step": 4183
+    },
+    {
+      "epoch": 0.575832645196807,
+      "grad_norm": 1.6481958595720827,
+      "learning_rate": 8.042539552169856e-06,
+      "loss": 0.9535,
+      "step": 4184
+    },
+    {
+      "epoch": 0.5759702725020645,
+      "grad_norm": 1.5239819349429768,
+      "learning_rate": 8.03816855302708e-06,
+      "loss": 0.9833,
+      "step": 4185
+    },
+    {
+      "epoch": 0.5761078998073218,
+      "grad_norm": 1.7220436770127434,
+      "learning_rate": 8.033797943673363e-06,
+      "loss": 1.0686,
+      "step": 4186
+    },
+    {
+      "epoch": 0.5762455271125791,
+      "grad_norm": 1.7925415287677329,
+      "learning_rate": 8.029427724977077e-06,
+      "loss": 1.0022,
+      "step": 4187
+    },
+    {
+      "epoch": 0.5763831544178365,
+      "grad_norm": 1.7053016051731829,
+      "learning_rate": 8.025057897806531e-06,
+      "loss": 0.9145,
+      "step": 4188
+    },
+    {
+      "epoch": 0.5765207817230938,
+      "grad_norm": 1.9748909127090817,
+      "learning_rate": 8.020688463029942e-06,
+      "loss": 0.998,
+      "step": 4189
+    },
+    {
+      "epoch": 0.5766584090283512,
+      "grad_norm": 2.051342883570065,
+      "learning_rate": 8.016319421515467e-06,
+      "loss": 0.9439,
+      "step": 4190
+    },
+    {
+      "epoch": 0.5767960363336085,
+      "grad_norm": 1.79927213892093,
+      "learning_rate": 8.011950774131167e-06,
+      "loss": 1.0192,
+      "step": 4191
+    },
+    {
+      "epoch": 0.576933663638866,
+      "grad_norm": 1.6074888927522777,
+      "learning_rate": 8.007582521745031e-06,
+      "loss": 0.9507,
+      "step": 4192
+    },
+    {
+      "epoch": 0.5770712909441233,
+      "grad_norm": 1.6705054548553586,
+      "learning_rate": 8.00321466522498e-06,
+      "loss": 0.995,
+      "step": 4193
+    },
+    {
+      "epoch": 0.5772089182493807,
+      "grad_norm": 1.6991705761238471,
+      "learning_rate": 7.998847205438838e-06,
+      "loss": 1.0305,
+      "step": 4194
+    },
+    {
+      "epoch": 0.577346545554638,
+      "grad_norm": 2.0318020481804138,
+      "learning_rate": 7.994480143254369e-06,
+      "loss": 1.0029,
+      "step": 4195
+    },
+    {
+      "epoch": 0.5774841728598954,
+      "grad_norm": 1.7024632485724989,
+      "learning_rate": 7.990113479539237e-06,
+      "loss": 0.9726,
+      "step": 4196
+    },
+    {
+      "epoch": 0.5776218001651527,
+      "grad_norm": 1.799983809221181,
+      "learning_rate": 7.985747215161048e-06,
+      "loss": 0.9961,
+      "step": 4197
+    },
+    {
+      "epoch": 0.5777594274704101,
+      "grad_norm": 1.7015472207834037,
+      "learning_rate": 7.981381350987313e-06,
+      "loss": 0.9778,
+      "step": 4198
+    },
+    {
+      "epoch": 0.5778970547756674,
+      "grad_norm": 1.560337422992669,
+      "learning_rate": 7.977015887885473e-06,
+      "loss": 0.9807,
+      "step": 4199
+    },
+    {
+      "epoch": 0.5780346820809249,
+      "grad_norm": 1.6587536028509373,
+      "learning_rate": 7.97265082672288e-06,
+      "loss": 0.9441,
+      "step": 4200
+    },
+    {
+      "epoch": 0.5781723093861822,
+      "grad_norm": 1.9037971739152533,
+      "learning_rate": 7.968286168366824e-06,
+      "loss": 1.0559,
+      "step": 4201
+    },
+    {
+      "epoch": 0.5783099366914396,
+      "grad_norm": 1.8890985826009206,
+      "learning_rate": 7.963921913684488e-06,
+      "loss": 0.9952,
+      "step": 4202
+    },
+    {
+      "epoch": 0.578447563996697,
+      "grad_norm": 1.6974743388793723,
+      "learning_rate": 7.959558063543005e-06,
+      "loss": 0.9835,
+      "step": 4203
+    },
+    {
+      "epoch": 0.5785851913019543,
+      "grad_norm": 1.669679861450674,
+      "learning_rate": 7.955194618809396e-06,
+      "loss": 0.9588,
+      "step": 4204
+    },
+    {
+      "epoch": 0.5787228186072116,
+      "grad_norm": 1.8994863316113748,
+      "learning_rate": 7.950831580350633e-06,
+      "loss": 1.0297,
+      "step": 4205
+    },
+    {
+      "epoch": 0.578860445912469,
+      "grad_norm": 1.650843678180431,
+      "learning_rate": 7.94646894903358e-06,
+      "loss": 1.0407,
+      "step": 4206
+    },
+    {
+      "epoch": 0.5789980732177264,
+      "grad_norm": 1.8180652299205629,
+      "learning_rate": 7.942106725725038e-06,
+      "loss": 1.0354,
+      "step": 4207
+    },
+    {
+      "epoch": 0.5791357005229838,
+      "grad_norm": 1.5641663870802276,
+      "learning_rate": 7.937744911291723e-06,
+      "loss": 0.9799,
+      "step": 4208
+    },
+    {
+      "epoch": 0.5792733278282411,
+      "grad_norm": 1.6109763841552938,
+      "learning_rate": 7.933383506600261e-06,
+      "loss": 1.0174,
+      "step": 4209
+    },
+    {
+      "epoch": 0.5794109551334985,
+      "grad_norm": 1.811855611098829,
+      "learning_rate": 7.92902251251721e-06,
+      "loss": 0.9682,
+      "step": 4210
+    },
+    {
+      "epoch": 0.5795485824387558,
+      "grad_norm": 1.895223018215018,
+      "learning_rate": 7.924661929909035e-06,
+      "loss": 1.0217,
+      "step": 4211
+    },
+    {
+      "epoch": 0.5796862097440132,
+      "grad_norm": 1.6468542091015417,
+      "learning_rate": 7.92030175964213e-06,
+      "loss": 1.0009,
+      "step": 4212
+    },
+    {
+      "epoch": 0.5798238370492705,
+      "grad_norm": 2.0673536864865696,
+      "learning_rate": 7.915942002582793e-06,
+      "loss": 0.9827,
+      "step": 4213
+    },
+    {
+      "epoch": 0.5799614643545279,
+      "grad_norm": 1.650148406061492,
+      "learning_rate": 7.911582659597253e-06,
+      "loss": 1.0068,
+      "step": 4214
+    },
+    {
+      "epoch": 0.5800990916597853,
+      "grad_norm": 2.0893035906691617,
+      "learning_rate": 7.90722373155165e-06,
+      "loss": 0.9624,
+      "step": 4215
+    },
+    {
+      "epoch": 0.5802367189650427,
+      "grad_norm": 1.8386097947898845,
+      "learning_rate": 7.90286521931205e-06,
+      "loss": 0.9638,
+      "step": 4216
+    },
+    {
+      "epoch": 0.5803743462703,
+      "grad_norm": 1.666757602230867,
+      "learning_rate": 7.898507123744419e-06,
+      "loss": 1.0621,
+      "step": 4217
+    },
+    {
+      "epoch": 0.5805119735755574,
+      "grad_norm": 1.9407347766851486,
+      "learning_rate": 7.894149445714661e-06,
+      "loss": 1.0124,
+      "step": 4218
+    },
+    {
+      "epoch": 0.5806496008808147,
+      "grad_norm": 1.922149174688596,
+      "learning_rate": 7.889792186088578e-06,
+      "loss": 1.0014,
+      "step": 4219
+    },
+    {
+      "epoch": 0.5807872281860721,
+      "grad_norm": 1.573704489778096,
+      "learning_rate": 7.885435345731905e-06,
+      "loss": 1.0084,
+      "step": 4220
+    },
+    {
+      "epoch": 0.5809248554913294,
+      "grad_norm": 1.7760277958691533,
+      "learning_rate": 7.88107892551028e-06,
+      "loss": 0.9841,
+      "step": 4221
+    },
+    {
+      "epoch": 0.5810624827965868,
+      "grad_norm": 1.634474543912554,
+      "learning_rate": 7.87672292628927e-06,
+      "loss": 0.9303,
+      "step": 4222
+    },
+    {
+      "epoch": 0.5812001101018442,
+      "grad_norm": 1.6689815214624326,
+      "learning_rate": 7.872367348934353e-06,
+      "loss": 1.0037,
+      "step": 4223
+    },
+    {
+      "epoch": 0.5813377374071016,
+      "grad_norm": 1.6955558305552014,
+      "learning_rate": 7.868012194310914e-06,
+      "loss": 1.0099,
+      "step": 4224
+    },
+    {
+      "epoch": 0.5814753647123589,
+      "grad_norm": 1.8226906694150415,
+      "learning_rate": 7.863657463284274e-06,
+      "loss": 1.0254,
+      "step": 4225
+    },
+    {
+      "epoch": 0.5816129920176163,
+      "grad_norm": 1.7254136983608634,
+      "learning_rate": 7.859303156719646e-06,
+      "loss": 1.0531,
+      "step": 4226
+    },
+    {
+      "epoch": 0.5817506193228736,
+      "grad_norm": 1.9269842478890744,
+      "learning_rate": 7.854949275482185e-06,
+      "loss": 1.0381,
+      "step": 4227
+    },
+    {
+      "epoch": 0.581888246628131,
+      "grad_norm": 1.7153618793183898,
+      "learning_rate": 7.850595820436933e-06,
+      "loss": 0.9376,
+      "step": 4228
+    },
+    {
+      "epoch": 0.5820258739333883,
+      "grad_norm": 1.7399612048476527,
+      "learning_rate": 7.846242792448873e-06,
+      "loss": 0.9228,
+      "step": 4229
+    },
+    {
+      "epoch": 0.5821635012386458,
+      "grad_norm": 2.0173364543126393,
+      "learning_rate": 7.84189019238289e-06,
+      "loss": 1.0168,
+      "step": 4230
+    },
+    {
+      "epoch": 0.5823011285439031,
+      "grad_norm": 1.7744042579516992,
+      "learning_rate": 7.837538021103783e-06,
+      "loss": 0.9436,
+      "step": 4231
+    },
+    {
+      "epoch": 0.5824387558491605,
+      "grad_norm": 1.938055025754311,
+      "learning_rate": 7.833186279476268e-06,
+      "loss": 1.1145,
+      "step": 4232
+    },
+    {
+      "epoch": 0.5825763831544178,
+      "grad_norm": 2.078467394147425,
+      "learning_rate": 7.828834968364984e-06,
+      "loss": 0.9641,
+      "step": 4233
+    },
+    {
+      "epoch": 0.5827140104596752,
+      "grad_norm": 1.7046173033468408,
+      "learning_rate": 7.824484088634467e-06,
+      "loss": 0.9885,
+      "step": 4234
+    },
+    {
+      "epoch": 0.5828516377649325,
+      "grad_norm": 1.8967074193608617,
+      "learning_rate": 7.820133641149186e-06,
+      "loss": 1.0123,
+      "step": 4235
+    },
+    {
+      "epoch": 0.5829892650701899,
+      "grad_norm": 1.6776998313675089,
+      "learning_rate": 7.815783626773506e-06,
+      "loss": 1.0483,
+      "step": 4236
+    },
+    {
+      "epoch": 0.5831268923754472,
+      "grad_norm": 1.5852957697810253,
+      "learning_rate": 7.811434046371724e-06,
+      "loss": 0.9784,
+      "step": 4237
+    },
+    {
+      "epoch": 0.5832645196807047,
+      "grad_norm": 1.868697205954871,
+      "learning_rate": 7.80708490080804e-06,
+      "loss": 1.0508,
+      "step": 4238
+    },
+    {
+      "epoch": 0.583402146985962,
+      "grad_norm": 1.627225740070741,
+      "learning_rate": 7.802736190946567e-06,
+      "loss": 0.9671,
+      "step": 4239
+    },
+    {
+      "epoch": 0.5835397742912194,
+      "grad_norm": 1.5673256140407603,
+      "learning_rate": 7.798387917651339e-06,
+      "loss": 0.9335,
+      "step": 4240
+    },
+    {
+      "epoch": 0.5836774015964767,
+      "grad_norm": 1.5446418939604443,
+      "learning_rate": 7.794040081786289e-06,
+      "loss": 0.9735,
+      "step": 4241
+    },
+    {
+      "epoch": 0.5838150289017341,
+      "grad_norm": 1.6567090020848032,
+      "learning_rate": 7.789692684215284e-06,
+      "loss": 1.0328,
+      "step": 4242
+    },
+    {
+      "epoch": 0.5839526562069914,
+      "grad_norm": 1.6741849837057594,
+      "learning_rate": 7.785345725802081e-06,
+      "loss": 0.9779,
+      "step": 4243
+    },
+    {
+      "epoch": 0.5840902835122488,
+      "grad_norm": 1.7116160809032748,
+      "learning_rate": 7.780999207410372e-06,
+      "loss": 0.9251,
+      "step": 4244
+    },
+    {
+      "epoch": 0.5842279108175062,
+      "grad_norm": 1.9401787891800675,
+      "learning_rate": 7.776653129903744e-06,
+      "loss": 1.0389,
+      "step": 4245
+    },
+    {
+      "epoch": 0.5843655381227636,
+      "grad_norm": 1.7148381799436883,
+      "learning_rate": 7.772307494145704e-06,
+      "loss": 1.0162,
+      "step": 4246
+    },
+    {
+      "epoch": 0.5845031654280209,
+      "grad_norm": 1.6309918658806077,
+      "learning_rate": 7.767962300999668e-06,
+      "loss": 0.9246,
+      "step": 4247
+    },
+    {
+      "epoch": 0.5846407927332783,
+      "grad_norm": 1.7996394202373762,
+      "learning_rate": 7.763617551328976e-06,
+      "loss": 0.9874,
+      "step": 4248
+    },
+    {
+      "epoch": 0.5847784200385356,
+      "grad_norm": 1.7006538590892448,
+      "learning_rate": 7.759273245996858e-06,
+      "loss": 1.0514,
+      "step": 4249
+    },
+    {
+      "epoch": 0.584916047343793,
+      "grad_norm": 1.7406472972814693,
+      "learning_rate": 7.754929385866481e-06,
+      "loss": 0.9165,
+      "step": 4250
+    },
+    {
+      "epoch": 0.5850536746490503,
+      "grad_norm": 1.6930528010112595,
+      "learning_rate": 7.750585971800898e-06,
+      "loss": 0.9681,
+      "step": 4251
+    },
+    {
+      "epoch": 0.5851913019543077,
+      "grad_norm": 1.6927088210177024,
+      "learning_rate": 7.746243004663094e-06,
+      "loss": 0.9921,
+      "step": 4252
+    },
+    {
+      "epoch": 0.5853289292595651,
+      "grad_norm": 1.8772656769757505,
+      "learning_rate": 7.74190048531595e-06,
+      "loss": 1.003,
+      "step": 4253
+    },
+    {
+      "epoch": 0.5854665565648225,
+      "grad_norm": 1.7615518290231331,
+      "learning_rate": 7.737558414622272e-06,
+      "loss": 0.8959,
+      "step": 4254
+    },
+    {
+      "epoch": 0.5856041838700798,
+      "grad_norm": 1.8312780138517997,
+      "learning_rate": 7.73321679344477e-06,
+      "loss": 1.0272,
+      "step": 4255
+    },
+    {
+      "epoch": 0.5857418111753372,
+      "grad_norm": 1.7553243670160097,
+      "learning_rate": 7.728875622646059e-06,
+      "loss": 1.0001,
+      "step": 4256
+    },
+    {
+      "epoch": 0.5858794384805945,
+      "grad_norm": 1.7364185368831442,
+      "learning_rate": 7.724534903088678e-06,
+      "loss": 0.935,
+      "step": 4257
+    },
+    {
+      "epoch": 0.5860170657858519,
+      "grad_norm": 1.6934138870216737,
+      "learning_rate": 7.720194635635058e-06,
+      "loss": 1.0027,
+      "step": 4258
+    },
+    {
+      "epoch": 0.5861546930911092,
+      "grad_norm": 1.7045791871198879,
+      "learning_rate": 7.715854821147562e-06,
+      "loss": 0.9646,
+      "step": 4259
+    },
+    {
+      "epoch": 0.5862923203963666,
+      "grad_norm": 1.87977994917619,
+      "learning_rate": 7.711515460488442e-06,
+      "loss": 0.9629,
+      "step": 4260
+    },
+    {
+      "epoch": 0.586429947701624,
+      "grad_norm": 1.7605783507012922,
+      "learning_rate": 7.707176554519876e-06,
+      "loss": 0.9016,
+      "step": 4261
+    },
+    {
+      "epoch": 0.5865675750068814,
+      "grad_norm": 1.7361617546998254,
+      "learning_rate": 7.702838104103941e-06,
+      "loss": 0.9905,
+      "step": 4262
+    },
+    {
+      "epoch": 0.5867052023121387,
+      "grad_norm": 1.752338629092878,
+      "learning_rate": 7.698500110102634e-06,
+      "loss": 0.9444,
+      "step": 4263
+    },
+    {
+      "epoch": 0.5868428296173961,
+      "grad_norm": 1.9392847696184061,
+      "learning_rate": 7.694162573377848e-06,
+      "loss": 1.077,
+      "step": 4264
+    },
+    {
+      "epoch": 0.5869804569226534,
+      "grad_norm": 1.8650372606218177,
+      "learning_rate": 7.689825494791398e-06,
+      "loss": 1.0288,
+      "step": 4265
+    },
+    {
+      "epoch": 0.5871180842279108,
+      "grad_norm": 1.583125478349895,
+      "learning_rate": 7.685488875204995e-06,
+      "loss": 0.939,
+      "step": 4266
+    },
+    {
+      "epoch": 0.5872557115331681,
+      "grad_norm": 1.8703814047459895,
+      "learning_rate": 7.681152715480276e-06,
+      "loss": 0.929,
+      "step": 4267
+    },
+    {
+      "epoch": 0.5873933388384256,
+      "grad_norm": 1.6012132804403048,
+      "learning_rate": 7.676817016478767e-06,
+      "loss": 1.0117,
+      "step": 4268
+    },
+    {
+      "epoch": 0.5875309661436829,
+      "grad_norm": 1.785121090363919,
+      "learning_rate": 7.672481779061917e-06,
+      "loss": 1.0689,
+      "step": 4269
+    },
+    {
+      "epoch": 0.5876685934489403,
+      "grad_norm": 1.8195958691366931,
+      "learning_rate": 7.668147004091075e-06,
+      "loss": 0.9496,
+      "step": 4270
+    },
+    {
+      "epoch": 0.5878062207541976,
+      "grad_norm": 1.9262855377804398,
+      "learning_rate": 7.663812692427509e-06,
+      "loss": 1.0324,
+      "step": 4271
+    },
+    {
+      "epoch": 0.587943848059455,
+      "grad_norm": 1.8464947465232997,
+      "learning_rate": 7.659478844932382e-06,
+      "loss": 0.9605,
+      "step": 4272
+    },
+    {
+      "epoch": 0.5880814753647123,
+      "grad_norm": 1.9281895897313235,
+      "learning_rate": 7.655145462466764e-06,
+      "loss": 0.9295,
+      "step": 4273
+    },
+    {
+      "epoch": 0.5882191026699697,
+      "grad_norm": 1.6188570430219926,
+      "learning_rate": 7.650812545891651e-06,
+      "loss": 0.9431,
+      "step": 4274
+    },
+    {
+      "epoch": 0.588356729975227,
+      "grad_norm": 1.7908935362669791,
+      "learning_rate": 7.646480096067923e-06,
+      "loss": 0.907,
+      "step": 4275
+    },
+    {
+      "epoch": 0.5884943572804845,
+      "grad_norm": 1.76610690775602,
+      "learning_rate": 7.642148113856387e-06,
+      "loss": 0.9282,
+      "step": 4276
+    },
+    {
+      "epoch": 0.5886319845857418,
+      "grad_norm": 1.9919960725753476,
+      "learning_rate": 7.637816600117743e-06,
+      "loss": 0.9463,
+      "step": 4277
+    },
+    {
+      "epoch": 0.5887696118909992,
+      "grad_norm": 1.77592318698332,
+      "learning_rate": 7.633485555712609e-06,
+      "loss": 0.9517,
+      "step": 4278
+    },
+    {
+      "epoch": 0.5889072391962565,
+      "grad_norm": 1.6463673180196006,
+      "learning_rate": 7.6291549815014975e-06,
+      "loss": 0.9296,
+      "step": 4279
+    },
+    {
+      "epoch": 0.5890448665015139,
+      "grad_norm": 1.7845877395106373,
+      "learning_rate": 7.624824878344842e-06,
+      "loss": 1.0362,
+      "step": 4280
+    },
+    {
+      "epoch": 0.5891824938067712,
+      "grad_norm": 1.6804502036020807,
+      "learning_rate": 7.620495247102966e-06,
+      "loss": 0.9934,
+      "step": 4281
+    },
+    {
+      "epoch": 0.5893201211120286,
+      "grad_norm": 2.0105003746974073,
+      "learning_rate": 7.616166088636117e-06,
+      "loss": 0.9691,
+      "step": 4282
+    },
+    {
+      "epoch": 0.589457748417286,
+      "grad_norm": 1.94255482906479,
+      "learning_rate": 7.611837403804432e-06,
+      "loss": 1.0411,
+      "step": 4283
+    },
+    {
+      "epoch": 0.5895953757225434,
+      "grad_norm": 1.9526622203390758,
+      "learning_rate": 7.607509193467965e-06,
+      "loss": 0.9645,
+      "step": 4284
+    },
+    {
+      "epoch": 0.5897330030278007,
+      "grad_norm": 1.7510249169065462,
+      "learning_rate": 7.603181458486673e-06,
+      "loss": 0.9754,
+      "step": 4285
+    },
+    {
+      "epoch": 0.5898706303330581,
+      "grad_norm": 1.5534972720759255,
+      "learning_rate": 7.598854199720415e-06,
+      "loss": 0.9458,
+      "step": 4286
+    },
+    {
+      "epoch": 0.5900082576383154,
+      "grad_norm": 1.6916293805454217,
+      "learning_rate": 7.594527418028962e-06,
+      "loss": 0.9258,
+      "step": 4287
+    },
+    {
+      "epoch": 0.5901458849435728,
+      "grad_norm": 1.7484705691551685,
+      "learning_rate": 7.590201114271979e-06,
+      "loss": 1.0162,
+      "step": 4288
+    },
+    {
+      "epoch": 0.5902835122488301,
+      "grad_norm": 1.936463477628001,
+      "learning_rate": 7.585875289309053e-06,
+      "loss": 1.0701,
+      "step": 4289
+    },
+    {
+      "epoch": 0.5904211395540875,
+      "grad_norm": 1.7480762669811787,
+      "learning_rate": 7.581549943999656e-06,
+      "loss": 0.9536,
+      "step": 4290
+    },
+    {
+      "epoch": 0.5905587668593449,
+      "grad_norm": 1.7414952981528666,
+      "learning_rate": 7.577225079203183e-06,
+      "loss": 0.9749,
+      "step": 4291
+    },
+    {
+      "epoch": 0.5906963941646023,
+      "grad_norm": 1.8002406990987645,
+      "learning_rate": 7.572900695778923e-06,
+      "loss": 0.9568,
+      "step": 4292
+    },
+    {
+      "epoch": 0.5908340214698596,
+      "grad_norm": 1.8158583573650602,
+      "learning_rate": 7.568576794586071e-06,
+      "loss": 0.9824,
+      "step": 4293
+    },
+    {
+      "epoch": 0.590971648775117,
+      "grad_norm": 1.9571750142903197,
+      "learning_rate": 7.564253376483725e-06,
+      "loss": 0.9835,
+      "step": 4294
+    },
+    {
+      "epoch": 0.5911092760803743,
+      "grad_norm": 2.124615875128621,
+      "learning_rate": 7.559930442330896e-06,
+      "loss": 1.0909,
+      "step": 4295
+    },
+    {
+      "epoch": 0.5912469033856317,
+      "grad_norm": 2.065636774527557,
+      "learning_rate": 7.555607992986484e-06,
+      "loss": 0.9987,
+      "step": 4296
+    },
+    {
+      "epoch": 0.591384530690889,
+      "grad_norm": 1.8207563325056715,
+      "learning_rate": 7.551286029309306e-06,
+      "loss": 1.0214,
+      "step": 4297
+    },
+    {
+      "epoch": 0.5915221579961464,
+      "grad_norm": 1.9015359550159594,
+      "learning_rate": 7.546964552158072e-06,
+      "loss": 0.9814,
+      "step": 4298
+    },
+    {
+      "epoch": 0.5916597853014038,
+      "grad_norm": 1.748315759143318,
+      "learning_rate": 7.542643562391406e-06,
+      "loss": 0.9164,
+      "step": 4299
+    },
+    {
+      "epoch": 0.5917974126066612,
+      "grad_norm": 1.6151843132686623,
+      "learning_rate": 7.538323060867822e-06,
+      "loss": 0.9285,
+      "step": 4300
+    },
+    {
+      "epoch": 0.5919350399119185,
+      "grad_norm": 1.9536274504785336,
+      "learning_rate": 7.534003048445753e-06,
+      "loss": 0.9954,
+      "step": 4301
+    },
+    {
+      "epoch": 0.5920726672171759,
+      "grad_norm": 1.6478925321154179,
+      "learning_rate": 7.52968352598352e-06,
+      "loss": 0.9485,
+      "step": 4302
+    },
+    {
+      "epoch": 0.5922102945224332,
+      "grad_norm": 1.7689648480227445,
+      "learning_rate": 7.525364494339359e-06,
+      "loss": 0.9763,
+      "step": 4303
+    },
+    {
+      "epoch": 0.5923479218276906,
+      "grad_norm": 1.6825615598365653,
+      "learning_rate": 7.521045954371399e-06,
+      "loss": 0.9946,
+      "step": 4304
+    },
+    {
+      "epoch": 0.5924855491329479,
+      "grad_norm": 1.8664072718276692,
+      "learning_rate": 7.516727906937671e-06,
+      "loss": 0.9539,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5926231764382054,
+      "grad_norm": 1.7551972823446285,
+      "learning_rate": 7.5124103528961165e-06,
+      "loss": 0.863,
+      "step": 4306
+    },
+    {
+      "epoch": 0.5927608037434627,
+      "grad_norm": 1.76477509268471,
+      "learning_rate": 7.508093293104575e-06,
+      "loss": 0.9961,
+      "step": 4307
+    },
+    {
+      "epoch": 0.5928984310487201,
+      "grad_norm": 1.9779645762381661,
+      "learning_rate": 7.503776728420784e-06,
+      "loss": 1.0584,
+      "step": 4308
+    },
+    {
+      "epoch": 0.5930360583539774,
+      "grad_norm": 1.6307980395562403,
+      "learning_rate": 7.4994606597023865e-06,
+      "loss": 0.9022,
+      "step": 4309
+    },
+    {
+      "epoch": 0.5931736856592348,
+      "grad_norm": 1.8205406900020848,
+      "learning_rate": 7.495145087806932e-06,
+      "loss": 1.0011,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5933113129644921,
+      "grad_norm": 1.818710751117435,
+      "learning_rate": 7.490830013591856e-06,
+      "loss": 0.9528,
+      "step": 4311
+    },
+    {
+      "epoch": 0.5934489402697495,
+      "grad_norm": 1.6928624422864835,
+      "learning_rate": 7.486515437914513e-06,
+      "loss": 0.9287,
+      "step": 4312
+    },
+    {
+      "epoch": 0.5935865675750068,
+      "grad_norm": 2.1283489058648577,
+      "learning_rate": 7.4822013616321435e-06,
+      "loss": 0.9616,
+      "step": 4313
+    },
+    {
+      "epoch": 0.5937241948802643,
+      "grad_norm": 1.6318718667112082,
+      "learning_rate": 7.4778877856019025e-06,
+      "loss": 0.949,
+      "step": 4314
+    },
+    {
+      "epoch": 0.5938618221855216,
+      "grad_norm": 2.145799510878873,
+      "learning_rate": 7.47357471068083e-06,
+      "loss": 0.9532,
+      "step": 4315
+    },
+    {
+      "epoch": 0.593999449490779,
+      "grad_norm": 1.738205940763011,
+      "learning_rate": 7.469262137725881e-06,
+      "loss": 0.9303,
+      "step": 4316
+    },
+    {
+      "epoch": 0.5941370767960363,
+      "grad_norm": 1.7285295609264977,
+      "learning_rate": 7.464950067593902e-06,
+      "loss": 0.9771,
+      "step": 4317
+    },
+    {
+      "epoch": 0.5942747041012937,
+      "grad_norm": 1.772813683712298,
+      "learning_rate": 7.4606385011416485e-06,
+      "loss": 1.0406,
+      "step": 4318
+    },
+    {
+      "epoch": 0.594412331406551,
+      "grad_norm": 1.8228535288113377,
+      "learning_rate": 7.456327439225763e-06,
+      "loss": 0.889,
+      "step": 4319
+    },
+    {
+      "epoch": 0.5945499587118084,
+      "grad_norm": 1.5869825715352297,
+      "learning_rate": 7.4520168827027926e-06,
+      "loss": 0.9672,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5946875860170658,
+      "grad_norm": 1.8583496870657272,
+      "learning_rate": 7.447706832429194e-06,
+      "loss": 1.0053,
+      "step": 4321
+    },
+    {
+      "epoch": 0.5948252133223232,
+      "grad_norm": 1.7952782508807401,
+      "learning_rate": 7.4433972892613075e-06,
+      "loss": 0.9502,
+      "step": 4322
+    },
+    {
+      "epoch": 0.5949628406275805,
+      "grad_norm": 1.8544550101033679,
+      "learning_rate": 7.4390882540553845e-06,
+      "loss": 0.9822,
+      "step": 4323
+    },
+    {
+      "epoch": 0.5951004679328379,
+      "grad_norm": 1.8104325405066757,
+      "learning_rate": 7.434779727667568e-06,
+      "loss": 1.0139,
+      "step": 4324
+    },
+    {
+      "epoch": 0.5952380952380952,
+      "grad_norm": 1.6269655726096586,
+      "learning_rate": 7.43047171095391e-06,
+      "loss": 0.9645,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5953757225433526,
+      "grad_norm": 1.7851718786574644,
+      "learning_rate": 7.426164204770346e-06,
+      "loss": 0.9876,
+      "step": 4326
+    },
+    {
+      "epoch": 0.5955133498486099,
+      "grad_norm": 1.8209927123482244,
+      "learning_rate": 7.421857209972727e-06,
+      "loss": 0.9199,
+      "step": 4327
+    },
+    {
+      "epoch": 0.5956509771538673,
+      "grad_norm": 1.838872212816762,
+      "learning_rate": 7.417550727416785e-06,
+      "loss": 1.0306,
+      "step": 4328
+    },
+    {
+      "epoch": 0.5957886044591247,
+      "grad_norm": 1.6264105095807433,
+      "learning_rate": 7.413244757958168e-06,
+      "loss": 0.9611,
+      "step": 4329
+    },
+    {
+      "epoch": 0.5959262317643821,
+      "grad_norm": 1.8726173544750835,
+      "learning_rate": 7.4089393024524045e-06,
+      "loss": 1.0441,
+      "step": 4330
+    },
+    {
+      "epoch": 0.5960638590696394,
+      "grad_norm": 1.6498568096359165,
+      "learning_rate": 7.404634361754936e-06,
+      "loss": 0.9539,
+      "step": 4331
+    },
+    {
+      "epoch": 0.5962014863748968,
+      "grad_norm": 1.7129491213979007,
+      "learning_rate": 7.400329936721092e-06,
+      "loss": 0.9375,
+      "step": 4332
+    },
+    {
+      "epoch": 0.5963391136801541,
+      "grad_norm": 1.8402535169501426,
+      "learning_rate": 7.396026028206105e-06,
+      "loss": 1.0239,
+      "step": 4333
+    },
+    {
+      "epoch": 0.5964767409854115,
+      "grad_norm": 1.8244820685360077,
+      "learning_rate": 7.391722637065099e-06,
+      "loss": 0.9607,
+      "step": 4334
+    },
+    {
+      "epoch": 0.5966143682906688,
+      "grad_norm": 1.8909826010998754,
+      "learning_rate": 7.387419764153108e-06,
+      "loss": 0.9027,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5967519955959262,
+      "grad_norm": 1.832300508056641,
+      "learning_rate": 7.383117410325045e-06,
+      "loss": 0.9552,
+      "step": 4336
+    },
+    {
+      "epoch": 0.5968896229011836,
+      "grad_norm": 1.763287267146626,
+      "learning_rate": 7.3788155764357275e-06,
+      "loss": 0.9944,
+      "step": 4337
+    },
+    {
+      "epoch": 0.597027250206441,
+      "grad_norm": 1.7605664383262927,
+      "learning_rate": 7.374514263339878e-06,
+      "loss": 1.0226,
+      "step": 4338
+    },
+    {
+      "epoch": 0.5971648775116983,
+      "grad_norm": 1.7132576845740801,
+      "learning_rate": 7.370213471892102e-06,
+      "loss": 0.9699,
+      "step": 4339
+    },
+    {
+      "epoch": 0.5973025048169557,
+      "grad_norm": 1.8194253714874216,
+      "learning_rate": 7.365913202946917e-06,
+      "loss": 0.9055,
+      "step": 4340
+    },
+    {
+      "epoch": 0.597440132122213,
+      "grad_norm": 1.7900792349213197,
+      "learning_rate": 7.361613457358716e-06,
+      "loss": 0.9567,
+      "step": 4341
+    },
+    {
+      "epoch": 0.5975777594274704,
+      "grad_norm": 1.6912020893008777,
+      "learning_rate": 7.35731423598181e-06,
+      "loss": 0.9962,
+      "step": 4342
+    },
+    {
+      "epoch": 0.5977153867327277,
+      "grad_norm": 1.6675106707434213,
+      "learning_rate": 7.353015539670387e-06,
+      "loss": 0.9514,
+      "step": 4343
+    },
+    {
+      "epoch": 0.5978530140379852,
+      "grad_norm": 1.6020805739304795,
+      "learning_rate": 7.348717369278549e-06,
+      "loss": 0.9323,
+      "step": 4344
+    },
+    {
+      "epoch": 0.5979906413432425,
+      "grad_norm": 1.6162506496601305,
+      "learning_rate": 7.344419725660271e-06,
+      "loss": 0.9382,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5981282686484999,
+      "grad_norm": 1.5864611427240152,
+      "learning_rate": 7.340122609669445e-06,
+      "loss": 0.9801,
+      "step": 4346
+    },
+    {
+      "epoch": 0.5982658959537572,
+      "grad_norm": 2.016888861173644,
+      "learning_rate": 7.335826022159848e-06,
+      "loss": 0.9876,
+      "step": 4347
+    },
+    {
+      "epoch": 0.5984035232590146,
+      "grad_norm": 1.8947269984006183,
+      "learning_rate": 7.331529963985151e-06,
+      "loss": 1.0123,
+      "step": 4348
+    },
+    {
+      "epoch": 0.5985411505642719,
+      "grad_norm": 1.8931015543537988,
+      "learning_rate": 7.32723443599892e-06,
+      "loss": 0.9218,
+      "step": 4349
+    },
+    {
+      "epoch": 0.5986787778695293,
+      "grad_norm": 1.7812862669412417,
+      "learning_rate": 7.322939439054626e-06,
+      "loss": 0.9584,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5988164051747866,
+      "grad_norm": 1.6772236031703374,
+      "learning_rate": 7.31864497400562e-06,
+      "loss": 0.9587,
+      "step": 4351
+    },
+    {
+      "epoch": 0.5989540324800441,
+      "grad_norm": 1.852507241107247,
+      "learning_rate": 7.3143510417051485e-06,
+      "loss": 0.9767,
+      "step": 4352
+    },
+    {
+      "epoch": 0.5990916597853014,
+      "grad_norm": 1.7192984263418716,
+      "learning_rate": 7.310057643006365e-06,
+      "loss": 1.0229,
+      "step": 4353
+    },
+    {
+      "epoch": 0.5992292870905588,
+      "grad_norm": 1.7365212215787655,
+      "learning_rate": 7.305764778762308e-06,
+      "loss": 0.955,
+      "step": 4354
+    },
+    {
+      "epoch": 0.5993669143958161,
+      "grad_norm": 1.7604515272942076,
+      "learning_rate": 7.301472449825909e-06,
+      "loss": 0.8157,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5995045417010735,
+      "grad_norm": 2.0147004853341417,
+      "learning_rate": 7.297180657049992e-06,
+      "loss": 1.0091,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5996421690063308,
+      "grad_norm": 1.7015248211305911,
+      "learning_rate": 7.2928894012872865e-06,
+      "loss": 0.9926,
+      "step": 4357
+    },
+    {
+      "epoch": 0.5997797963115882,
+      "grad_norm": 1.5925282014668032,
+      "learning_rate": 7.288598683390395e-06,
+      "loss": 0.971,
+      "step": 4358
+    },
+    {
+      "epoch": 0.5999174236168456,
+      "grad_norm": 1.8033144291052599,
+      "learning_rate": 7.2843085042118355e-06,
+      "loss": 0.9811,
+      "step": 4359
+    },
+    {
+      "epoch": 0.600055050922103,
+      "grad_norm": 1.9802816832883887,
+      "learning_rate": 7.280018864603997e-06,
+      "loss": 0.9697,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6001926782273603,
+      "grad_norm": 1.7960098141977567,
+      "learning_rate": 7.275729765419184e-06,
+      "loss": 1.0457,
+      "step": 4361
+    },
+    {
+      "epoch": 0.6003303055326177,
+      "grad_norm": 1.974676796621578,
+      "learning_rate": 7.2714412075095696e-06,
+      "loss": 1.0156,
+      "step": 4362
+    },
+    {
+      "epoch": 0.600467932837875,
+      "grad_norm": 1.7742499980059303,
+      "learning_rate": 7.267153191727239e-06,
+      "loss": 0.9013,
+      "step": 4363
+    },
+    {
+      "epoch": 0.6006055601431324,
+      "grad_norm": 1.8747885029689622,
+      "learning_rate": 7.2628657189241594e-06,
+      "loss": 0.9863,
+      "step": 4364
+    },
+    {
+      "epoch": 0.6007431874483897,
+      "grad_norm": 1.8903990814655263,
+      "learning_rate": 7.258578789952198e-06,
+      "loss": 0.9869,
+      "step": 4365
+    },
+    {
+      "epoch": 0.6008808147536471,
+      "grad_norm": 1.5162077621732435,
+      "learning_rate": 7.254292405663101e-06,
+      "loss": 0.9737,
+      "step": 4366
+    },
+    {
+      "epoch": 0.6010184420589045,
+      "grad_norm": 2.2763789014702063,
+      "learning_rate": 7.250006566908523e-06,
+      "loss": 0.9729,
+      "step": 4367
+    },
+    {
+      "epoch": 0.6011560693641619,
+      "grad_norm": 1.6096229140701266,
+      "learning_rate": 7.245721274539996e-06,
+      "loss": 0.9436,
+      "step": 4368
+    },
+    {
+      "epoch": 0.6012936966694192,
+      "grad_norm": 1.6776632166602625,
+      "learning_rate": 7.241436529408947e-06,
+      "loss": 1.041,
+      "step": 4369
+    },
+    {
+      "epoch": 0.6014313239746766,
+      "grad_norm": 1.825915858017262,
+      "learning_rate": 7.2371523323667e-06,
+      "loss": 1.0194,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6015689512799339,
+      "grad_norm": 1.7594090565730396,
+      "learning_rate": 7.232868684264463e-06,
+      "loss": 0.9519,
+      "step": 4371
+    },
+    {
+      "epoch": 0.6017065785851913,
+      "grad_norm": 1.8547440392176489,
+      "learning_rate": 7.228585585953345e-06,
+      "loss": 1.0095,
+      "step": 4372
+    },
+    {
+      "epoch": 0.6018442058904486,
+      "grad_norm": 1.6182771667105258,
+      "learning_rate": 7.22430303828433e-06,
+      "loss": 0.9941,
+      "step": 4373
+    },
+    {
+      "epoch": 0.601981833195706,
+      "grad_norm": 2.185582252121093,
+      "learning_rate": 7.220021042108312e-06,
+      "loss": 0.936,
+      "step": 4374
+    },
+    {
+      "epoch": 0.6021194605009634,
+      "grad_norm": 1.7552957629711297,
+      "learning_rate": 7.215739598276054e-06,
+      "loss": 0.9299,
+      "step": 4375
+    },
+    {
+      "epoch": 0.6022570878062208,
+      "grad_norm": 1.663866373429022,
+      "learning_rate": 7.211458707638229e-06,
+      "loss": 0.918,
+      "step": 4376
+    },
+    {
+      "epoch": 0.6023947151114781,
+      "grad_norm": 1.8833735116909562,
+      "learning_rate": 7.207178371045385e-06,
+      "loss": 0.9441,
+      "step": 4377
+    },
+    {
+      "epoch": 0.6025323424167355,
+      "grad_norm": 1.9364103803296826,
+      "learning_rate": 7.202898589347969e-06,
+      "loss": 1.0047,
+      "step": 4378
+    },
+    {
+      "epoch": 0.6026699697219928,
+      "grad_norm": 2.0507533279218255,
+      "learning_rate": 7.198619363396315e-06,
+      "loss": 0.9313,
+      "step": 4379
+    },
+    {
+      "epoch": 0.6028075970272502,
+      "grad_norm": 1.825470156568926,
+      "learning_rate": 7.1943406940406494e-06,
+      "loss": 0.9446,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6029452243325075,
+      "grad_norm": 1.685280082325007,
+      "learning_rate": 7.19006258213108e-06,
+      "loss": 0.9564,
+      "step": 4381
+    },
+    {
+      "epoch": 0.603082851637765,
+      "grad_norm": 1.735757128505911,
+      "learning_rate": 7.185785028517615e-06,
+      "loss": 1.0317,
+      "step": 4382
+    },
+    {
+      "epoch": 0.6032204789430223,
+      "grad_norm": 2.22157381683179,
+      "learning_rate": 7.181508034050142e-06,
+      "loss": 0.8981,
+      "step": 4383
+    },
+    {
+      "epoch": 0.6033581062482797,
+      "grad_norm": 2.104574167825755,
+      "learning_rate": 7.177231599578438e-06,
+      "loss": 0.9444,
+      "step": 4384
+    },
+    {
+      "epoch": 0.603495733553537,
+      "grad_norm": 1.6069556052540663,
+      "learning_rate": 7.172955725952178e-06,
+      "loss": 0.9892,
+      "step": 4385
+    },
+    {
+      "epoch": 0.6036333608587944,
+      "grad_norm": 1.7541124574746447,
+      "learning_rate": 7.168680414020916e-06,
+      "loss": 1.0048,
+      "step": 4386
+    },
+    {
+      "epoch": 0.6037709881640517,
+      "grad_norm": 1.7680190544540175,
+      "learning_rate": 7.164405664634102e-06,
+      "loss": 0.9538,
+      "step": 4387
+    },
+    {
+      "epoch": 0.6039086154693091,
+      "grad_norm": 1.9532630312841865,
+      "learning_rate": 7.160131478641064e-06,
+      "loss": 0.9807,
+      "step": 4388
+    },
+    {
+      "epoch": 0.6040462427745664,
+      "grad_norm": 1.880946859245638,
+      "learning_rate": 7.155857856891031e-06,
+      "loss": 0.8654,
+      "step": 4389
+    },
+    {
+      "epoch": 0.6041838700798239,
+      "grad_norm": 1.7073062917141437,
+      "learning_rate": 7.1515848002331045e-06,
+      "loss": 0.926,
+      "step": 4390
+    },
+    {
+      "epoch": 0.6043214973850812,
+      "grad_norm": 1.5883682846273928,
+      "learning_rate": 7.147312309516294e-06,
+      "loss": 0.9029,
+      "step": 4391
+    },
+    {
+      "epoch": 0.6044591246903386,
+      "grad_norm": 1.9180296965259906,
+      "learning_rate": 7.1430403855894734e-06,
+      "loss": 0.9242,
+      "step": 4392
+    },
+    {
+      "epoch": 0.6045967519955959,
+      "grad_norm": 1.8985200947725587,
+      "learning_rate": 7.138769029301423e-06,
+      "loss": 1.018,
+      "step": 4393
+    },
+    {
+      "epoch": 0.6047343793008533,
+      "grad_norm": 1.7806839157955971,
+      "learning_rate": 7.1344982415008005e-06,
+      "loss": 0.9889,
+      "step": 4394
+    },
+    {
+      "epoch": 0.6048720066061106,
+      "grad_norm": 2.001200197233345,
+      "learning_rate": 7.1302280230361535e-06,
+      "loss": 0.9604,
+      "step": 4395
+    },
+    {
+      "epoch": 0.605009633911368,
+      "grad_norm": 1.6312555611515147,
+      "learning_rate": 7.1259583747559126e-06,
+      "loss": 0.9958,
+      "step": 4396
+    },
+    {
+      "epoch": 0.6051472612166254,
+      "grad_norm": 1.8313068912147679,
+      "learning_rate": 7.121689297508405e-06,
+      "loss": 0.9894,
+      "step": 4397
+    },
+    {
+      "epoch": 0.6052848885218828,
+      "grad_norm": 1.8854203258406599,
+      "learning_rate": 7.117420792141831e-06,
+      "loss": 0.9894,
+      "step": 4398
+    },
+    {
+      "epoch": 0.6054225158271401,
+      "grad_norm": 1.7206228962587846,
+      "learning_rate": 7.1131528595042905e-06,
+      "loss": 0.9641,
+      "step": 4399
+    },
+    {
+      "epoch": 0.6055601431323975,
+      "grad_norm": 1.8719189460207613,
+      "learning_rate": 7.108885500443761e-06,
+      "loss": 0.911,
+      "step": 4400
+    },
+    {
+      "epoch": 0.6056977704376548,
+      "grad_norm": 1.7999444785641199,
+      "learning_rate": 7.104618715808105e-06,
+      "loss": 0.9951,
+      "step": 4401
+    },
+    {
+      "epoch": 0.6058353977429122,
+      "grad_norm": 1.8340658289209342,
+      "learning_rate": 7.100352506445076e-06,
+      "loss": 0.9934,
+      "step": 4402
+    },
+    {
+      "epoch": 0.6059730250481695,
+      "grad_norm": 1.9565492393285713,
+      "learning_rate": 7.096086873202311e-06,
+      "loss": 1.0034,
+      "step": 4403
+    },
+    {
+      "epoch": 0.6061106523534269,
+      "grad_norm": 1.7186967210968194,
+      "learning_rate": 7.091821816927339e-06,
+      "loss": 0.9682,
+      "step": 4404
+    },
+    {
+      "epoch": 0.6062482796586843,
+      "grad_norm": 1.806181552404213,
+      "learning_rate": 7.0875573384675585e-06,
+      "loss": 0.9595,
+      "step": 4405
+    },
+    {
+      "epoch": 0.6063859069639417,
+      "grad_norm": 1.7817547170038963,
+      "learning_rate": 7.083293438670272e-06,
+      "loss": 0.9962,
+      "step": 4406
+    },
+    {
+      "epoch": 0.606523534269199,
+      "grad_norm": 1.6326658728526442,
+      "learning_rate": 7.07903011838265e-06,
+      "loss": 0.9133,
+      "step": 4407
+    },
+    {
+      "epoch": 0.6066611615744564,
+      "grad_norm": 2.07312511727519,
+      "learning_rate": 7.074767378451761e-06,
+      "loss": 0.9907,
+      "step": 4408
+    },
+    {
+      "epoch": 0.6067987888797137,
+      "grad_norm": 1.7664910893696837,
+      "learning_rate": 7.070505219724553e-06,
+      "loss": 0.9772,
+      "step": 4409
+    },
+    {
+      "epoch": 0.6069364161849711,
+      "grad_norm": 1.6898218280238155,
+      "learning_rate": 7.066243643047857e-06,
+      "loss": 0.9921,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6070740434902284,
+      "grad_norm": 1.6555706143603486,
+      "learning_rate": 7.061982649268389e-06,
+      "loss": 0.9304,
+      "step": 4411
+    },
+    {
+      "epoch": 0.6072116707954858,
+      "grad_norm": 2.1150285239476267,
+      "learning_rate": 7.057722239232754e-06,
+      "loss": 0.9662,
+      "step": 4412
+    },
+    {
+      "epoch": 0.6073492981007432,
+      "grad_norm": 1.5626970202872368,
+      "learning_rate": 7.053462413787431e-06,
+      "loss": 0.9969,
+      "step": 4413
+    },
+    {
+      "epoch": 0.6074869254060006,
+      "grad_norm": 1.6246223800556885,
+      "learning_rate": 7.049203173778798e-06,
+      "loss": 0.9481,
+      "step": 4414
+    },
+    {
+      "epoch": 0.6076245527112579,
+      "grad_norm": 2.092719836286789,
+      "learning_rate": 7.044944520053099e-06,
+      "loss": 1.0512,
+      "step": 4415
+    },
+    {
+      "epoch": 0.6077621800165153,
+      "grad_norm": 1.7756217499405682,
+      "learning_rate": 7.040686453456474e-06,
+      "loss": 1.0425,
+      "step": 4416
+    },
+    {
+      "epoch": 0.6078998073217726,
+      "grad_norm": 1.6013349947595639,
+      "learning_rate": 7.036428974834941e-06,
+      "loss": 0.9337,
+      "step": 4417
+    },
+    {
+      "epoch": 0.60803743462703,
+      "grad_norm": 1.8076674601389076,
+      "learning_rate": 7.0321720850344036e-06,
+      "loss": 0.9515,
+      "step": 4418
+    },
+    {
+      "epoch": 0.6081750619322873,
+      "grad_norm": 1.9579803938295546,
+      "learning_rate": 7.0279157849006516e-06,
+      "loss": 1.0061,
+      "step": 4419
+    },
+    {
+      "epoch": 0.6083126892375448,
+      "grad_norm": 1.6689484957128382,
+      "learning_rate": 7.023660075279345e-06,
+      "loss": 0.9718,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6084503165428021,
+      "grad_norm": 1.8904325452021673,
+      "learning_rate": 7.019404957016045e-06,
+      "loss": 0.913,
+      "step": 4421
+    },
+    {
+      "epoch": 0.6085879438480595,
+      "grad_norm": 1.7361300613475077,
+      "learning_rate": 7.015150430956175e-06,
+      "loss": 0.9936,
+      "step": 4422
+    },
+    {
+      "epoch": 0.6087255711533168,
+      "grad_norm": 1.9477970003946525,
+      "learning_rate": 7.01089649794506e-06,
+      "loss": 0.9446,
+      "step": 4423
+    },
+    {
+      "epoch": 0.6088631984585742,
+      "grad_norm": 1.952911230965377,
+      "learning_rate": 7.006643158827891e-06,
+      "loss": 0.9884,
+      "step": 4424
+    },
+    {
+      "epoch": 0.6090008257638315,
+      "grad_norm": 1.746085588578271,
+      "learning_rate": 7.0023904144497536e-06,
+      "loss": 0.9611,
+      "step": 4425
+    },
+    {
+      "epoch": 0.6091384530690889,
+      "grad_norm": 1.975769632736285,
+      "learning_rate": 6.998138265655605e-06,
+      "loss": 1.0011,
+      "step": 4426
+    },
+    {
+      "epoch": 0.6092760803743462,
+      "grad_norm": 1.8857546055821717,
+      "learning_rate": 6.993886713290299e-06,
+      "loss": 0.9735,
+      "step": 4427
+    },
+    {
+      "epoch": 0.6094137076796037,
+      "grad_norm": 1.6482638552499609,
+      "learning_rate": 6.989635758198547e-06,
+      "loss": 0.9907,
+      "step": 4428
+    },
+    {
+      "epoch": 0.609551334984861,
+      "grad_norm": 1.6880083447059313,
+      "learning_rate": 6.985385401224968e-06,
+      "loss": 1.0295,
+      "step": 4429
+    },
+    {
+      "epoch": 0.6096889622901184,
+      "grad_norm": 1.8446399210200966,
+      "learning_rate": 6.981135643214039e-06,
+      "loss": 0.9844,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6098265895953757,
+      "grad_norm": 1.8342555441409474,
+      "learning_rate": 6.976886485010139e-06,
+      "loss": 0.942,
+      "step": 4431
+    },
+    {
+      "epoch": 0.6099642169006331,
+      "grad_norm": 1.7909740404788146,
+      "learning_rate": 6.9726379274575104e-06,
+      "loss": 0.9842,
+      "step": 4432
+    },
+    {
+      "epoch": 0.6101018442058904,
+      "grad_norm": 2.0052596270347687,
+      "learning_rate": 6.9683899714002826e-06,
+      "loss": 0.9397,
+      "step": 4433
+    },
+    {
+      "epoch": 0.6102394715111478,
+      "grad_norm": 1.7081525349106654,
+      "learning_rate": 6.964142617682475e-06,
+      "loss": 0.9751,
+      "step": 4434
+    },
+    {
+      "epoch": 0.6103770988164052,
+      "grad_norm": 1.8273247300410294,
+      "learning_rate": 6.959895867147967e-06,
+      "loss": 0.96,
+      "step": 4435
+    },
+    {
+      "epoch": 0.6105147261216626,
+      "grad_norm": 1.797382200856444,
+      "learning_rate": 6.955649720640541e-06,
+      "loss": 0.9753,
+      "step": 4436
+    },
+    {
+      "epoch": 0.6106523534269199,
+      "grad_norm": 1.6907715972499473,
+      "learning_rate": 6.951404179003838e-06,
+      "loss": 0.9645,
+      "step": 4437
+    },
+    {
+      "epoch": 0.6107899807321773,
+      "grad_norm": 1.7803995997392006,
+      "learning_rate": 6.947159243081398e-06,
+      "loss": 0.9678,
+      "step": 4438
+    },
+    {
+      "epoch": 0.6109276080374346,
+      "grad_norm": 2.2509109966297958,
+      "learning_rate": 6.942914913716624e-06,
+      "loss": 0.9934,
+      "step": 4439
+    },
+    {
+      "epoch": 0.611065235342692,
+      "grad_norm": 1.6895251721199356,
+      "learning_rate": 6.938671191752812e-06,
+      "loss": 1.0122,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6112028626479493,
+      "grad_norm": 1.9484730731077469,
+      "learning_rate": 6.934428078033126e-06,
+      "loss": 0.8772,
+      "step": 4441
+    },
+    {
+      "epoch": 0.6113404899532067,
+      "grad_norm": 1.6524308979224689,
+      "learning_rate": 6.930185573400623e-06,
+      "loss": 0.9599,
+      "step": 4442
+    },
+    {
+      "epoch": 0.6114781172584641,
+      "grad_norm": 1.6925061691546168,
+      "learning_rate": 6.925943678698221e-06,
+      "loss": 0.9266,
+      "step": 4443
+    },
+    {
+      "epoch": 0.6116157445637215,
+      "grad_norm": 1.768200928568819,
+      "learning_rate": 6.921702394768734e-06,
+      "loss": 0.9887,
+      "step": 4444
+    },
+    {
+      "epoch": 0.6117533718689788,
+      "grad_norm": 1.6590213237060751,
+      "learning_rate": 6.9174617224548416e-06,
+      "loss": 0.9093,
+      "step": 4445
+    },
+    {
+      "epoch": 0.6118909991742362,
+      "grad_norm": 1.7090619134342846,
+      "learning_rate": 6.9132216625991125e-06,
+      "loss": 0.9562,
+      "step": 4446
+    },
+    {
+      "epoch": 0.6120286264794935,
+      "grad_norm": 1.97552275977963,
+      "learning_rate": 6.908982216043983e-06,
+      "loss": 0.9476,
+      "step": 4447
+    },
+    {
+      "epoch": 0.6121662537847509,
+      "grad_norm": 1.7198883601249262,
+      "learning_rate": 6.904743383631774e-06,
+      "loss": 1.0351,
+      "step": 4448
+    },
+    {
+      "epoch": 0.6123038810900082,
+      "grad_norm": 1.7869008364227306,
+      "learning_rate": 6.90050516620469e-06,
+      "loss": 0.9178,
+      "step": 4449
+    },
+    {
+      "epoch": 0.6124415083952656,
+      "grad_norm": 1.705614021275434,
+      "learning_rate": 6.8962675646047985e-06,
+      "loss": 0.9502,
+      "step": 4450
+    },
+    {
+      "epoch": 0.612579135700523,
+      "grad_norm": 1.5587707606082601,
+      "learning_rate": 6.89203057967406e-06,
+      "loss": 0.9387,
+      "step": 4451
+    },
+    {
+      "epoch": 0.6127167630057804,
+      "grad_norm": 1.8932076345798052,
+      "learning_rate": 6.887794212254297e-06,
+      "loss": 1.0069,
+      "step": 4452
+    },
+    {
+      "epoch": 0.6128543903110377,
+      "grad_norm": 1.580039796501389,
+      "learning_rate": 6.8835584631872276e-06,
+      "loss": 0.959,
+      "step": 4453
+    },
+    {
+      "epoch": 0.6129920176162951,
+      "grad_norm": 1.8924956166714384,
+      "learning_rate": 6.879323333314427e-06,
+      "loss": 1.0559,
+      "step": 4454
+    },
+    {
+      "epoch": 0.6131296449215524,
+      "grad_norm": 1.9875788518435897,
+      "learning_rate": 6.875088823477365e-06,
+      "loss": 1.0566,
+      "step": 4455
+    },
+    {
+      "epoch": 0.6132672722268098,
+      "grad_norm": 1.5732188298905556,
+      "learning_rate": 6.8708549345173795e-06,
+      "loss": 0.8837,
+      "step": 4456
+    },
+    {
+      "epoch": 0.6134048995320671,
+      "grad_norm": 1.6918133696364572,
+      "learning_rate": 6.866621667275683e-06,
+      "loss": 0.9745,
+      "step": 4457
+    },
+    {
+      "epoch": 0.6135425268373246,
+      "grad_norm": 1.8224494788772838,
+      "learning_rate": 6.86238902259337e-06,
+      "loss": 0.9576,
+      "step": 4458
+    },
+    {
+      "epoch": 0.6136801541425819,
+      "grad_norm": 1.4540768667079262,
+      "learning_rate": 6.8581570013114125e-06,
+      "loss": 0.9997,
+      "step": 4459
+    },
+    {
+      "epoch": 0.6138177814478393,
+      "grad_norm": 2.1547969127174826,
+      "learning_rate": 6.853925604270648e-06,
+      "loss": 1.0207,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6139554087530966,
+      "grad_norm": 1.744539613788188,
+      "learning_rate": 6.849694832311804e-06,
+      "loss": 0.9955,
+      "step": 4461
+    },
+    {
+      "epoch": 0.614093036058354,
+      "grad_norm": 1.7096847131376491,
+      "learning_rate": 6.84546468627547e-06,
+      "loss": 0.8209,
+      "step": 4462
+    },
+    {
+      "epoch": 0.6142306633636113,
+      "grad_norm": 1.8980210482945494,
+      "learning_rate": 6.841235167002127e-06,
+      "loss": 0.9711,
+      "step": 4463
+    },
+    {
+      "epoch": 0.6143682906688687,
+      "grad_norm": 2.124712588386739,
+      "learning_rate": 6.837006275332114e-06,
+      "loss": 1.0104,
+      "step": 4464
+    },
+    {
+      "epoch": 0.614505917974126,
+      "grad_norm": 1.7323357989751254,
+      "learning_rate": 6.832778012105657e-06,
+      "loss": 1.0109,
+      "step": 4465
+    },
+    {
+      "epoch": 0.6146435452793835,
+      "grad_norm": 1.5890815648818166,
+      "learning_rate": 6.828550378162857e-06,
+      "loss": 0.8971,
+      "step": 4466
+    },
+    {
+      "epoch": 0.6147811725846408,
+      "grad_norm": 2.1736918727993526,
+      "learning_rate": 6.8243233743436806e-06,
+      "loss": 0.9837,
+      "step": 4467
+    },
+    {
+      "epoch": 0.6149187998898982,
+      "grad_norm": 1.6121732670162712,
+      "learning_rate": 6.820097001487984e-06,
+      "loss": 0.9142,
+      "step": 4468
+    },
+    {
+      "epoch": 0.6150564271951555,
+      "grad_norm": 1.6670486922126246,
+      "learning_rate": 6.8158712604354806e-06,
+      "loss": 0.9048,
+      "step": 4469
+    },
+    {
+      "epoch": 0.6151940545004129,
+      "grad_norm": 1.6911687757933196,
+      "learning_rate": 6.811646152025777e-06,
+      "loss": 0.9842,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6153316818056702,
+      "grad_norm": 1.8159766276600469,
+      "learning_rate": 6.807421677098334e-06,
+      "loss": 0.9047,
+      "step": 4471
+    },
+    {
+      "epoch": 0.6154693091109276,
+      "grad_norm": 1.8077067506063895,
+      "learning_rate": 6.8031978364925045e-06,
+      "loss": 1.0123,
+      "step": 4472
+    },
+    {
+      "epoch": 0.615606936416185,
+      "grad_norm": 1.893495562227542,
+      "learning_rate": 6.798974631047504e-06,
+      "loss": 0.9302,
+      "step": 4473
+    },
+    {
+      "epoch": 0.6157445637214424,
+      "grad_norm": 1.6024456499052941,
+      "learning_rate": 6.7947520616024295e-06,
+      "loss": 0.9906,
+      "step": 4474
+    },
+    {
+      "epoch": 0.6158821910266997,
+      "grad_norm": 1.7644480193381793,
+      "learning_rate": 6.7905301289962425e-06,
+      "loss": 0.8543,
+      "step": 4475
+    },
+    {
+      "epoch": 0.6160198183319571,
+      "grad_norm": 1.7900186539467675,
+      "learning_rate": 6.78630883406779e-06,
+      "loss": 0.9619,
+      "step": 4476
+    },
+    {
+      "epoch": 0.6161574456372144,
+      "grad_norm": 1.99877642023352,
+      "learning_rate": 6.782088177655777e-06,
+      "loss": 0.9704,
+      "step": 4477
+    },
+    {
+      "epoch": 0.6162950729424718,
+      "grad_norm": 1.7145831055222354,
+      "learning_rate": 6.777868160598799e-06,
+      "loss": 0.9768,
+      "step": 4478
+    },
+    {
+      "epoch": 0.6164327002477291,
+      "grad_norm": 1.7617800142825604,
+      "learning_rate": 6.77364878373531e-06,
+      "loss": 0.8817,
+      "step": 4479
+    },
+    {
+      "epoch": 0.6165703275529865,
+      "grad_norm": 1.8139685877842213,
+      "learning_rate": 6.76943004790364e-06,
+      "loss": 0.9247,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6167079548582439,
+      "grad_norm": 1.7115076760300016,
+      "learning_rate": 6.765211953942004e-06,
+      "loss": 1.0141,
+      "step": 4481
+    },
+    {
+      "epoch": 0.6168455821635013,
+      "grad_norm": 1.7982185982642773,
+      "learning_rate": 6.7609945026884674e-06,
+      "loss": 0.9318,
+      "step": 4482
+    },
+    {
+      "epoch": 0.6169832094687586,
+      "grad_norm": 1.7299220290593724,
+      "learning_rate": 6.75677769498099e-06,
+      "loss": 0.8956,
+      "step": 4483
+    },
+    {
+      "epoch": 0.617120836774016,
+      "grad_norm": 1.850374189195348,
+      "learning_rate": 6.752561531657386e-06,
+      "loss": 1.0139,
+      "step": 4484
+    },
+    {
+      "epoch": 0.6172584640792733,
+      "grad_norm": 1.821946795404957,
+      "learning_rate": 6.748346013555358e-06,
+      "loss": 0.9814,
+      "step": 4485
+    },
+    {
+      "epoch": 0.6173960913845307,
+      "grad_norm": 2.08896702829991,
+      "learning_rate": 6.744131141512462e-06,
+      "loss": 1.0086,
+      "step": 4486
+    },
+    {
+      "epoch": 0.617533718689788,
+      "grad_norm": 1.7118137650396255,
+      "learning_rate": 6.739916916366142e-06,
+      "loss": 0.8922,
+      "step": 4487
+    },
+    {
+      "epoch": 0.6176713459950454,
+      "grad_norm": 1.7211392664959346,
+      "learning_rate": 6.735703338953704e-06,
+      "loss": 0.9187,
+      "step": 4488
+    },
+    {
+      "epoch": 0.6178089733003028,
+      "grad_norm": 1.9276258108013018,
+      "learning_rate": 6.731490410112334e-06,
+      "loss": 0.8237,
+      "step": 4489
+    },
+    {
+      "epoch": 0.6179466006055602,
+      "grad_norm": 1.8139520730128054,
+      "learning_rate": 6.727278130679072e-06,
+      "loss": 0.9943,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6180842279108175,
+      "grad_norm": 1.8441529772662606,
+      "learning_rate": 6.723066501490853e-06,
+      "loss": 0.9508,
+      "step": 4491
+    },
+    {
+      "epoch": 0.6182218552160749,
+      "grad_norm": 1.6653433667927398,
+      "learning_rate": 6.718855523384458e-06,
+      "loss": 0.9674,
+      "step": 4492
+    },
+    {
+      "epoch": 0.6183594825213322,
+      "grad_norm": 1.7288863719573833,
+      "learning_rate": 6.714645197196561e-06,
+      "loss": 0.9297,
+      "step": 4493
+    },
+    {
+      "epoch": 0.6184971098265896,
+      "grad_norm": 1.6829657495423669,
+      "learning_rate": 6.710435523763689e-06,
+      "loss": 0.9254,
+      "step": 4494
+    },
+    {
+      "epoch": 0.6186347371318469,
+      "grad_norm": 1.7299596814673934,
+      "learning_rate": 6.70622650392225e-06,
+      "loss": 0.8965,
+      "step": 4495
+    },
+    {
+      "epoch": 0.6187723644371044,
+      "grad_norm": 1.7981768030661358,
+      "learning_rate": 6.7020181385085195e-06,
+      "loss": 0.9646,
+      "step": 4496
+    },
+    {
+      "epoch": 0.6189099917423617,
+      "grad_norm": 2.3041212474505937,
+      "learning_rate": 6.697810428358638e-06,
+      "loss": 0.9812,
+      "step": 4497
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 1.9122901584221164,
+      "learning_rate": 6.693603374308624e-06,
+      "loss": 0.9456,
+      "step": 4498
+    },
+    {
+      "epoch": 0.6191852463528764,
+      "grad_norm": 1.7736810331251673,
+      "learning_rate": 6.689396977194358e-06,
+      "loss": 0.953,
+      "step": 4499
+    },
+    {
+      "epoch": 0.6193228736581338,
+      "grad_norm": 1.7676134033034956,
+      "learning_rate": 6.685191237851598e-06,
+      "loss": 0.981,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6194605009633911,
+      "grad_norm": 1.7979574732727979,
+      "learning_rate": 6.680986157115961e-06,
+      "loss": 0.9373,
+      "step": 4501
+    },
+    {
+      "epoch": 0.6195981282686485,
+      "grad_norm": 1.6432602785409371,
+      "learning_rate": 6.676781735822944e-06,
+      "loss": 0.9636,
+      "step": 4502
+    },
+    {
+      "epoch": 0.6197357555739058,
+      "grad_norm": 1.6935718056399358,
+      "learning_rate": 6.672577974807907e-06,
+      "loss": 0.9365,
+      "step": 4503
+    },
+    {
+      "epoch": 0.6198733828791633,
+      "grad_norm": 1.6598125217366497,
+      "learning_rate": 6.6683748749060784e-06,
+      "loss": 0.9962,
+      "step": 4504
+    },
+    {
+      "epoch": 0.6200110101844206,
+      "grad_norm": 1.6046752343593662,
+      "learning_rate": 6.664172436952554e-06,
+      "loss": 0.9597,
+      "step": 4505
+    },
+    {
+      "epoch": 0.620148637489678,
+      "grad_norm": 1.953949662707906,
+      "learning_rate": 6.659970661782309e-06,
+      "loss": 0.8744,
+      "step": 4506
+    },
+    {
+      "epoch": 0.6202862647949353,
+      "grad_norm": 1.6218952101345383,
+      "learning_rate": 6.65576955023017e-06,
+      "loss": 0.9892,
+      "step": 4507
+    },
+    {
+      "epoch": 0.6204238921001927,
+      "grad_norm": 1.9563776290406325,
+      "learning_rate": 6.651569103130847e-06,
+      "loss": 1.0087,
+      "step": 4508
+    },
+    {
+      "epoch": 0.62056151940545,
+      "grad_norm": 1.6391478543875306,
+      "learning_rate": 6.647369321318904e-06,
+      "loss": 0.9507,
+      "step": 4509
+    },
+    {
+      "epoch": 0.6206991467107074,
+      "grad_norm": 1.6207910585644774,
+      "learning_rate": 6.643170205628787e-06,
+      "loss": 0.8633,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6208367740159648,
+      "grad_norm": 1.6319776164770134,
+      "learning_rate": 6.638971756894803e-06,
+      "loss": 0.9128,
+      "step": 4511
+    },
+    {
+      "epoch": 0.6209744013212222,
+      "grad_norm": 1.7280362428740252,
+      "learning_rate": 6.634773975951118e-06,
+      "loss": 0.9664,
+      "step": 4512
+    },
+    {
+      "epoch": 0.6211120286264795,
+      "grad_norm": 1.8667595451984118,
+      "learning_rate": 6.630576863631783e-06,
+      "loss": 0.9421,
+      "step": 4513
+    },
+    {
+      "epoch": 0.6212496559317369,
+      "grad_norm": 1.5948902789487673,
+      "learning_rate": 6.6263804207706995e-06,
+      "loss": 0.8514,
+      "step": 4514
+    },
+    {
+      "epoch": 0.6213872832369942,
+      "grad_norm": 1.5296976537817324,
+      "learning_rate": 6.622184648201651e-06,
+      "loss": 0.9862,
+      "step": 4515
+    },
+    {
+      "epoch": 0.6215249105422516,
+      "grad_norm": 1.7670016729894655,
+      "learning_rate": 6.61798954675827e-06,
+      "loss": 0.9553,
+      "step": 4516
+    },
+    {
+      "epoch": 0.6216625378475089,
+      "grad_norm": 1.7798781127399919,
+      "learning_rate": 6.6137951172740745e-06,
+      "loss": 0.9537,
+      "step": 4517
+    },
+    {
+      "epoch": 0.6218001651527663,
+      "grad_norm": 1.3624423703696258,
+      "learning_rate": 6.609601360582436e-06,
+      "loss": 0.9837,
+      "step": 4518
+    },
+    {
+      "epoch": 0.6219377924580237,
+      "grad_norm": 1.8268211969022745,
+      "learning_rate": 6.605408277516599e-06,
+      "loss": 0.9496,
+      "step": 4519
+    },
+    {
+      "epoch": 0.6220754197632811,
+      "grad_norm": 1.5860370335182972,
+      "learning_rate": 6.6012158689096685e-06,
+      "loss": 0.9463,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6222130470685384,
+      "grad_norm": 1.685284875608895,
+      "learning_rate": 6.5970241355946244e-06,
+      "loss": 0.9448,
+      "step": 4521
+    },
+    {
+      "epoch": 0.6223506743737958,
+      "grad_norm": 1.733551905001973,
+      "learning_rate": 6.592833078404298e-06,
+      "loss": 0.9065,
+      "step": 4522
+    },
+    {
+      "epoch": 0.6224883016790531,
+      "grad_norm": 1.6223166778702045,
+      "learning_rate": 6.588642698171404e-06,
+      "loss": 0.937,
+      "step": 4523
+    },
+    {
+      "epoch": 0.6226259289843105,
+      "grad_norm": 2.073413283770982,
+      "learning_rate": 6.584452995728506e-06,
+      "loss": 1.01,
+      "step": 4524
+    },
+    {
+      "epoch": 0.6227635562895678,
+      "grad_norm": 1.9076989896953125,
+      "learning_rate": 6.5802639719080475e-06,
+      "loss": 0.9775,
+      "step": 4525
+    },
+    {
+      "epoch": 0.6229011835948252,
+      "grad_norm": 2.061698074796893,
+      "learning_rate": 6.576075627542324e-06,
+      "loss": 0.9633,
+      "step": 4526
+    },
+    {
+      "epoch": 0.6230388109000826,
+      "grad_norm": 2.150175309088167,
+      "learning_rate": 6.571887963463506e-06,
+      "loss": 0.9074,
+      "step": 4527
+    },
+    {
+      "epoch": 0.62317643820534,
+      "grad_norm": 1.9783326260263014,
+      "learning_rate": 6.567700980503626e-06,
+      "loss": 0.9102,
+      "step": 4528
+    },
+    {
+      "epoch": 0.6233140655105973,
+      "grad_norm": 1.9963996691508377,
+      "learning_rate": 6.5635146794945736e-06,
+      "loss": 0.9373,
+      "step": 4529
+    },
+    {
+      "epoch": 0.6234516928158547,
+      "grad_norm": 1.9874746409003496,
+      "learning_rate": 6.5593290612681184e-06,
+      "loss": 1.0362,
+      "step": 4530
+    },
+    {
+      "epoch": 0.623589320121112,
+      "grad_norm": 1.829447604064256,
+      "learning_rate": 6.5551441266558766e-06,
+      "loss": 0.8754,
+      "step": 4531
+    },
+    {
+      "epoch": 0.6237269474263694,
+      "grad_norm": 1.8293180758493182,
+      "learning_rate": 6.550959876489346e-06,
+      "loss": 0.9472,
+      "step": 4532
+    },
+    {
+      "epoch": 0.6238645747316267,
+      "grad_norm": 1.840591837441197,
+      "learning_rate": 6.546776311599874e-06,
+      "loss": 1.0114,
+      "step": 4533
+    },
+    {
+      "epoch": 0.6240022020368842,
+      "grad_norm": 2.0662437512294747,
+      "learning_rate": 6.542593432818678e-06,
+      "loss": 0.9999,
+      "step": 4534
+    },
+    {
+      "epoch": 0.6241398293421415,
+      "grad_norm": 1.9158264094918873,
+      "learning_rate": 6.538411240976841e-06,
+      "loss": 0.8872,
+      "step": 4535
+    },
+    {
+      "epoch": 0.6242774566473989,
+      "grad_norm": 1.6829820373568665,
+      "learning_rate": 6.534229736905308e-06,
+      "loss": 0.905,
+      "step": 4536
+    },
+    {
+      "epoch": 0.6244150839526562,
+      "grad_norm": 1.8190642229261782,
+      "learning_rate": 6.530048921434883e-06,
+      "loss": 0.9393,
+      "step": 4537
+    },
+    {
+      "epoch": 0.6245527112579136,
+      "grad_norm": 1.9568265821547344,
+      "learning_rate": 6.525868795396242e-06,
+      "loss": 0.9962,
+      "step": 4538
+    },
+    {
+      "epoch": 0.6246903385631709,
+      "grad_norm": 1.8598101357555294,
+      "learning_rate": 6.521689359619912e-06,
+      "loss": 0.8569,
+      "step": 4539
+    },
+    {
+      "epoch": 0.6248279658684283,
+      "grad_norm": 1.9915750387644138,
+      "learning_rate": 6.517510614936296e-06,
+      "loss": 0.9665,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6249655931736856,
+      "grad_norm": 1.6363578414706583,
+      "learning_rate": 6.5133325621756485e-06,
+      "loss": 0.9135,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6251032204789431,
+      "grad_norm": 1.7323935384462967,
+      "learning_rate": 6.509155202168095e-06,
+      "loss": 0.8315,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6252408477842004,
+      "grad_norm": 2.2013494508324647,
+      "learning_rate": 6.5049785357436205e-06,
+      "loss": 0.8225,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6253784750894578,
+      "grad_norm": 1.7597843871901941,
+      "learning_rate": 6.500802563732065e-06,
+      "loss": 0.8615,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6255161023947151,
+      "grad_norm": 1.7705422104814144,
+      "learning_rate": 6.496627286963145e-06,
+      "loss": 0.9228,
+      "step": 4545
+    },
+    {
+      "epoch": 0.6256537296999725,
+      "grad_norm": 1.5983986625591127,
+      "learning_rate": 6.4924527062664235e-06,
+      "loss": 0.948,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6257913570052298,
+      "grad_norm": 1.8358494483854382,
+      "learning_rate": 6.4882788224713415e-06,
+      "loss": 0.9824,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6259289843104872,
+      "grad_norm": 1.7224441243078317,
+      "learning_rate": 6.484105636407181e-06,
+      "loss": 0.9076,
+      "step": 4548
+    },
+    {
+      "epoch": 0.6260666116157446,
+      "grad_norm": 1.8110841628987042,
+      "learning_rate": 6.479933148903106e-06,
+      "loss": 0.9458,
+      "step": 4549
+    },
+    {
+      "epoch": 0.626204238921002,
+      "grad_norm": 1.9328605269516776,
+      "learning_rate": 6.475761360788127e-06,
+      "loss": 0.9925,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6263418662262593,
+      "grad_norm": 1.7210512511024718,
+      "learning_rate": 6.471590272891131e-06,
+      "loss": 1.013,
+      "step": 4551
+    },
+    {
+      "epoch": 0.6264794935315167,
+      "grad_norm": 1.5195711747140455,
+      "learning_rate": 6.467419886040842e-06,
+      "loss": 0.9257,
+      "step": 4552
+    },
+    {
+      "epoch": 0.626617120836774,
+      "grad_norm": 1.8144414991938833,
+      "learning_rate": 6.463250201065872e-06,
+      "loss": 0.9851,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6267547481420314,
+      "grad_norm": 1.7400891548170292,
+      "learning_rate": 6.4590812187946685e-06,
+      "loss": 0.8825,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6268923754472887,
+      "grad_norm": 1.685406882312589,
+      "learning_rate": 6.454912940055562e-06,
+      "loss": 0.9014,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6270300027525461,
+      "grad_norm": 2.083533024765111,
+      "learning_rate": 6.450745365676723e-06,
+      "loss": 0.9408,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6271676300578035,
+      "grad_norm": 1.85691320224333,
+      "learning_rate": 6.446578496486198e-06,
+      "loss": 1.0137,
+      "step": 4557
+    },
+    {
+      "epoch": 0.6273052573630609,
+      "grad_norm": 1.6404055103543915,
+      "learning_rate": 6.442412333311885e-06,
+      "loss": 0.9602,
+      "step": 4558
+    },
+    {
+      "epoch": 0.6274428846683182,
+      "grad_norm": 1.692965959448579,
+      "learning_rate": 6.4382468769815425e-06,
+      "loss": 0.9705,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6275805119735756,
+      "grad_norm": 1.920463055552961,
+      "learning_rate": 6.434082128322793e-06,
+      "loss": 0.9434,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6277181392788329,
+      "grad_norm": 1.609780908307586,
+      "learning_rate": 6.429918088163108e-06,
+      "loss": 0.822,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6278557665840903,
+      "grad_norm": 1.729881477276873,
+      "learning_rate": 6.425754757329834e-06,
+      "loss": 1.0114,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6279933938893476,
+      "grad_norm": 1.6764058134535513,
+      "learning_rate": 6.4215921366501595e-06,
+      "loss": 0.9317,
+      "step": 4563
+    },
+    {
+      "epoch": 0.628131021194605,
+      "grad_norm": 1.8227690862787205,
+      "learning_rate": 6.417430226951147e-06,
+      "loss": 0.8814,
+      "step": 4564
+    },
+    {
+      "epoch": 0.6282686484998624,
+      "grad_norm": 2.370103510048762,
+      "learning_rate": 6.413269029059708e-06,
+      "loss": 0.9715,
+      "step": 4565
+    },
+    {
+      "epoch": 0.6284062758051198,
+      "grad_norm": 1.9714898548898367,
+      "learning_rate": 6.409108543802616e-06,
+      "loss": 0.9869,
+      "step": 4566
+    },
+    {
+      "epoch": 0.6285439031103771,
+      "grad_norm": 1.6301782557324682,
+      "learning_rate": 6.404948772006501e-06,
+      "loss": 0.9634,
+      "step": 4567
+    },
+    {
+      "epoch": 0.6286815304156345,
+      "grad_norm": 1.7232377612483616,
+      "learning_rate": 6.400789714497857e-06,
+      "loss": 1.0178,
+      "step": 4568
+    },
+    {
+      "epoch": 0.6288191577208918,
+      "grad_norm": 1.6399466304210857,
+      "learning_rate": 6.396631372103025e-06,
+      "loss": 0.8581,
+      "step": 4569
+    },
+    {
+      "epoch": 0.6289567850261492,
+      "grad_norm": 1.7227874479247651,
+      "learning_rate": 6.39247374564822e-06,
+      "loss": 0.906,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6290944123314065,
+      "grad_norm": 1.7129212534598153,
+      "learning_rate": 6.388316835959494e-06,
+      "loss": 0.941,
+      "step": 4571
+    },
+    {
+      "epoch": 0.629232039636664,
+      "grad_norm": 1.8264263292231637,
+      "learning_rate": 6.38416064386278e-06,
+      "loss": 0.992,
+      "step": 4572
+    },
+    {
+      "epoch": 0.6293696669419213,
+      "grad_norm": 1.7966949252964057,
+      "learning_rate": 6.380005170183845e-06,
+      "loss": 0.8975,
+      "step": 4573
+    },
+    {
+      "epoch": 0.6295072942471787,
+      "grad_norm": 1.6673329806483295,
+      "learning_rate": 6.375850415748334e-06,
+      "loss": 0.9535,
+      "step": 4574
+    },
+    {
+      "epoch": 0.629644921552436,
+      "grad_norm": 1.6944808000957843,
+      "learning_rate": 6.371696381381738e-06,
+      "loss": 0.9815,
+      "step": 4575
+    },
+    {
+      "epoch": 0.6297825488576934,
+      "grad_norm": 2.15224034144866,
+      "learning_rate": 6.367543067909398e-06,
+      "loss": 1.0617,
+      "step": 4576
+    },
+    {
+      "epoch": 0.6299201761629507,
+      "grad_norm": 1.6628423210212386,
+      "learning_rate": 6.363390476156535e-06,
+      "loss": 0.9383,
+      "step": 4577
+    },
+    {
+      "epoch": 0.630057803468208,
+      "grad_norm": 1.9157529346847277,
+      "learning_rate": 6.359238606948199e-06,
+      "loss": 0.9862,
+      "step": 4578
+    },
+    {
+      "epoch": 0.6301954307734654,
+      "grad_norm": 1.7379222830485246,
+      "learning_rate": 6.355087461109319e-06,
+      "loss": 0.9597,
+      "step": 4579
+    },
+    {
+      "epoch": 0.6303330580787229,
+      "grad_norm": 1.9179897370554069,
+      "learning_rate": 6.350937039464665e-06,
+      "loss": 0.8966,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6304706853839802,
+      "grad_norm": 1.840516061012518,
+      "learning_rate": 6.346787342838872e-06,
+      "loss": 0.9306,
+      "step": 4581
+    },
+    {
+      "epoch": 0.6306083126892376,
+      "grad_norm": 1.7019370051973535,
+      "learning_rate": 6.342638372056424e-06,
+      "loss": 0.9521,
+      "step": 4582
+    },
+    {
+      "epoch": 0.6307459399944949,
+      "grad_norm": 1.9441645010860646,
+      "learning_rate": 6.338490127941672e-06,
+      "loss": 0.9325,
+      "step": 4583
+    },
+    {
+      "epoch": 0.6308835672997523,
+      "grad_norm": 1.9475061161166787,
+      "learning_rate": 6.334342611318809e-06,
+      "loss": 0.9587,
+      "step": 4584
+    },
+    {
+      "epoch": 0.6310211946050096,
+      "grad_norm": 1.7841902203061712,
+      "learning_rate": 6.330195823011893e-06,
+      "loss": 0.9737,
+      "step": 4585
+    },
+    {
+      "epoch": 0.631158821910267,
+      "grad_norm": 1.5401862918848828,
+      "learning_rate": 6.32604976384483e-06,
+      "loss": 0.852,
+      "step": 4586
+    },
+    {
+      "epoch": 0.6312964492155244,
+      "grad_norm": 1.7047538176401293,
+      "learning_rate": 6.321904434641393e-06,
+      "loss": 0.9467,
+      "step": 4587
+    },
+    {
+      "epoch": 0.6314340765207818,
+      "grad_norm": 1.8788755055549675,
+      "learning_rate": 6.317759836225192e-06,
+      "loss": 0.9741,
+      "step": 4588
+    },
+    {
+      "epoch": 0.6315717038260391,
+      "grad_norm": 1.8784692919010602,
+      "learning_rate": 6.313615969419709e-06,
+      "loss": 0.946,
+      "step": 4589
+    },
+    {
+      "epoch": 0.6317093311312965,
+      "grad_norm": 1.8456604305112205,
+      "learning_rate": 6.309472835048269e-06,
+      "loss": 0.9952,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6318469584365538,
+      "grad_norm": 1.6877454775084553,
+      "learning_rate": 6.305330433934065e-06,
+      "loss": 0.9624,
+      "step": 4591
+    },
+    {
+      "epoch": 0.6319845857418112,
+      "grad_norm": 1.6633000884945794,
+      "learning_rate": 6.301188766900128e-06,
+      "loss": 0.9883,
+      "step": 4592
+    },
+    {
+      "epoch": 0.6321222130470685,
+      "grad_norm": 1.803549913995877,
+      "learning_rate": 6.297047834769348e-06,
+      "loss": 0.9171,
+      "step": 4593
+    },
+    {
+      "epoch": 0.6322598403523259,
+      "grad_norm": 1.8479320922201288,
+      "learning_rate": 6.2929076383644794e-06,
+      "loss": 0.9336,
+      "step": 4594
+    },
+    {
+      "epoch": 0.6323974676575833,
+      "grad_norm": 2.1215087666031884,
+      "learning_rate": 6.288768178508113e-06,
+      "loss": 0.8441,
+      "step": 4595
+    },
+    {
+      "epoch": 0.6325350949628407,
+      "grad_norm": 2.1747931518428607,
+      "learning_rate": 6.284629456022712e-06,
+      "loss": 0.9496,
+      "step": 4596
+    },
+    {
+      "epoch": 0.632672722268098,
+      "grad_norm": 1.9032727138305383,
+      "learning_rate": 6.280491471730579e-06,
+      "loss": 0.9382,
+      "step": 4597
+    },
+    {
+      "epoch": 0.6328103495733554,
+      "grad_norm": 1.7584107220230383,
+      "learning_rate": 6.276354226453879e-06,
+      "loss": 0.8954,
+      "step": 4598
+    },
+    {
+      "epoch": 0.6329479768786127,
+      "grad_norm": 1.741230785725227,
+      "learning_rate": 6.27221772101462e-06,
+      "loss": 0.9235,
+      "step": 4599
+    },
+    {
+      "epoch": 0.63308560418387,
+      "grad_norm": 1.7397058796892202,
+      "learning_rate": 6.268081956234674e-06,
+      "loss": 0.9996,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6332232314891274,
+      "grad_norm": 2.0605057841510077,
+      "learning_rate": 6.263946932935756e-06,
+      "loss": 0.9307,
+      "step": 4601
+    },
+    {
+      "epoch": 0.6333608587943848,
+      "grad_norm": 1.9098459535955392,
+      "learning_rate": 6.259812651939445e-06,
+      "loss": 0.9737,
+      "step": 4602
+    },
+    {
+      "epoch": 0.6334984860996422,
+      "grad_norm": 1.9034677852534794,
+      "learning_rate": 6.255679114067159e-06,
+      "loss": 0.9931,
+      "step": 4603
+    },
+    {
+      "epoch": 0.6336361134048996,
+      "grad_norm": 2.0639150194046647,
+      "learning_rate": 6.25154632014018e-06,
+      "loss": 1.0295,
+      "step": 4604
+    },
+    {
+      "epoch": 0.6337737407101569,
+      "grad_norm": 2.3136478769144464,
+      "learning_rate": 6.247414270979637e-06,
+      "loss": 0.9085,
+      "step": 4605
+    },
+    {
+      "epoch": 0.6339113680154143,
+      "grad_norm": 1.8152414942433093,
+      "learning_rate": 6.243282967406512e-06,
+      "loss": 0.9932,
+      "step": 4606
+    },
+    {
+      "epoch": 0.6340489953206716,
+      "grad_norm": 1.6842077952622179,
+      "learning_rate": 6.23915241024164e-06,
+      "loss": 0.9545,
+      "step": 4607
+    },
+    {
+      "epoch": 0.634186622625929,
+      "grad_norm": 1.8967381480379688,
+      "learning_rate": 6.235022600305698e-06,
+      "loss": 0.9918,
+      "step": 4608
+    },
+    {
+      "epoch": 0.6343242499311863,
+      "grad_norm": 1.664866459014443,
+      "learning_rate": 6.230893538419233e-06,
+      "loss": 0.9336,
+      "step": 4609
+    },
+    {
+      "epoch": 0.6344618772364438,
+      "grad_norm": 2.2302153639348594,
+      "learning_rate": 6.226765225402624e-06,
+      "loss": 0.9025,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6345995045417011,
+      "grad_norm": 1.9569268871282208,
+      "learning_rate": 6.222637662076116e-06,
+      "loss": 0.9543,
+      "step": 4611
+    },
+    {
+      "epoch": 0.6347371318469585,
+      "grad_norm": 1.7715802281594577,
+      "learning_rate": 6.218510849259798e-06,
+      "loss": 0.9082,
+      "step": 4612
+    },
+    {
+      "epoch": 0.6348747591522158,
+      "grad_norm": 2.0242036357095765,
+      "learning_rate": 6.214384787773615e-06,
+      "loss": 0.9581,
+      "step": 4613
+    },
+    {
+      "epoch": 0.6350123864574732,
+      "grad_norm": 1.7431041495277726,
+      "learning_rate": 6.2102594784373505e-06,
+      "loss": 0.9796,
+      "step": 4614
+    },
+    {
+      "epoch": 0.6351500137627305,
+      "grad_norm": 1.6086591294961665,
+      "learning_rate": 6.206134922070657e-06,
+      "loss": 0.9637,
+      "step": 4615
+    },
+    {
+      "epoch": 0.6352876410679879,
+      "grad_norm": 1.6835302459481012,
+      "learning_rate": 6.202011119493017e-06,
+      "loss": 0.9507,
+      "step": 4616
+    },
+    {
+      "epoch": 0.6354252683732452,
+      "grad_norm": 1.7933742370024826,
+      "learning_rate": 6.197888071523782e-06,
+      "loss": 0.9404,
+      "step": 4617
+    },
+    {
+      "epoch": 0.6355628956785027,
+      "grad_norm": 1.6417429052534318,
+      "learning_rate": 6.193765778982139e-06,
+      "loss": 0.8917,
+      "step": 4618
+    },
+    {
+      "epoch": 0.63570052298376,
+      "grad_norm": 1.6273818894142318,
+      "learning_rate": 6.189644242687136e-06,
+      "loss": 0.9408,
+      "step": 4619
+    },
+    {
+      "epoch": 0.6358381502890174,
+      "grad_norm": 1.8426928385151873,
+      "learning_rate": 6.185523463457663e-06,
+      "loss": 0.9577,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6359757775942747,
+      "grad_norm": 1.657653882119092,
+      "learning_rate": 6.181403442112465e-06,
+      "loss": 0.9274,
+      "step": 4621
+    },
+    {
+      "epoch": 0.636113404899532,
+      "grad_norm": 1.810994741495603,
+      "learning_rate": 6.177284179470129e-06,
+      "loss": 0.92,
+      "step": 4622
+    },
+    {
+      "epoch": 0.6362510322047894,
+      "grad_norm": 1.800592057941373,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.8795,
+      "step": 4623
+    },
+    {
+      "epoch": 0.6363886595100468,
+      "grad_norm": 1.8843958809791,
+      "learning_rate": 6.1690479335676746e-06,
+      "loss": 0.9191,
+      "step": 4624
+    },
+    {
+      "epoch": 0.6365262868153042,
+      "grad_norm": 1.8889468027793805,
+      "learning_rate": 6.164930951943977e-06,
+      "loss": 0.9807,
+      "step": 4625
+    },
+    {
+      "epoch": 0.6366639141205616,
+      "grad_norm": 1.736093471525083,
+      "learning_rate": 6.160814732296006e-06,
+      "loss": 0.8687,
+      "step": 4626
+    },
+    {
+      "epoch": 0.6368015414258189,
+      "grad_norm": 1.9294585106080533,
+      "learning_rate": 6.156699275441593e-06,
+      "loss": 0.9567,
+      "step": 4627
+    },
+    {
+      "epoch": 0.6369391687310763,
+      "grad_norm": 1.7588852504478198,
+      "learning_rate": 6.152584582198426e-06,
+      "loss": 0.9945,
+      "step": 4628
+    },
+    {
+      "epoch": 0.6370767960363336,
+      "grad_norm": 1.8681902407435589,
+      "learning_rate": 6.148470653384035e-06,
+      "loss": 0.9822,
+      "step": 4629
+    },
+    {
+      "epoch": 0.637214423341591,
+      "grad_norm": 1.9439027840303185,
+      "learning_rate": 6.144357489815807e-06,
+      "loss": 0.9364,
+      "step": 4630
+    },
+    {
+      "epoch": 0.6373520506468483,
+      "grad_norm": 1.7824840622220914,
+      "learning_rate": 6.140245092310965e-06,
+      "loss": 1.0202,
+      "step": 4631
+    },
+    {
+      "epoch": 0.6374896779521056,
+      "grad_norm": 1.9785826832150997,
+      "learning_rate": 6.136133461686592e-06,
+      "loss": 1.0364,
+      "step": 4632
+    },
+    {
+      "epoch": 0.6376273052573631,
+      "grad_norm": 1.6291993178184687,
+      "learning_rate": 6.132022598759606e-06,
+      "loss": 0.9153,
+      "step": 4633
+    },
+    {
+      "epoch": 0.6377649325626205,
+      "grad_norm": 1.6061192849148591,
+      "learning_rate": 6.127912504346787e-06,
+      "loss": 0.9707,
+      "step": 4634
+    },
+    {
+      "epoch": 0.6379025598678778,
+      "grad_norm": 1.8211819100195648,
+      "learning_rate": 6.123803179264745e-06,
+      "loss": 0.9209,
+      "step": 4635
+    },
+    {
+      "epoch": 0.6380401871731352,
+      "grad_norm": 1.974506030104459,
+      "learning_rate": 6.119694624329955e-06,
+      "loss": 0.9405,
+      "step": 4636
+    },
+    {
+      "epoch": 0.6381778144783925,
+      "grad_norm": 1.6682661537780281,
+      "learning_rate": 6.1155868403587244e-06,
+      "loss": 0.8608,
+      "step": 4637
+    },
+    {
+      "epoch": 0.6383154417836499,
+      "grad_norm": 1.8300911745380444,
+      "learning_rate": 6.111479828167223e-06,
+      "loss": 0.9059,
+      "step": 4638
+    },
+    {
+      "epoch": 0.6384530690889072,
+      "grad_norm": 2.067664248250286,
+      "learning_rate": 6.107373588571446e-06,
+      "loss": 0.8844,
+      "step": 4639
+    },
+    {
+      "epoch": 0.6385906963941645,
+      "grad_norm": 1.6141235217150596,
+      "learning_rate": 6.103268122387257e-06,
+      "loss": 0.949,
+      "step": 4640
+    },
+    {
+      "epoch": 0.638728323699422,
+      "grad_norm": 1.619387738715072,
+      "learning_rate": 6.0991634304303525e-06,
+      "loss": 0.8603,
+      "step": 4641
+    },
+    {
+      "epoch": 0.6388659510046794,
+      "grad_norm": 1.7858257145957521,
+      "learning_rate": 6.095059513516273e-06,
+      "loss": 0.9699,
+      "step": 4642
+    },
+    {
+      "epoch": 0.6390035783099367,
+      "grad_norm": 2.0177160073397618,
+      "learning_rate": 6.090956372460417e-06,
+      "loss": 0.9108,
+      "step": 4643
+    },
+    {
+      "epoch": 0.639141205615194,
+      "grad_norm": 2.3850421478173214,
+      "learning_rate": 6.086854008078018e-06,
+      "loss": 0.931,
+      "step": 4644
+    },
+    {
+      "epoch": 0.6392788329204514,
+      "grad_norm": 1.8381084393735032,
+      "learning_rate": 6.082752421184168e-06,
+      "loss": 0.8905,
+      "step": 4645
+    },
+    {
+      "epoch": 0.6394164602257087,
+      "grad_norm": 1.6825705245239608,
+      "learning_rate": 6.078651612593786e-06,
+      "loss": 0.9231,
+      "step": 4646
+    },
+    {
+      "epoch": 0.6395540875309661,
+      "grad_norm": 1.8278742209284728,
+      "learning_rate": 6.0745515831216564e-06,
+      "loss": 0.8227,
+      "step": 4647
+    },
+    {
+      "epoch": 0.6396917148362236,
+      "grad_norm": 1.7003716983282071,
+      "learning_rate": 6.070452333582388e-06,
+      "loss": 0.9219,
+      "step": 4648
+    },
+    {
+      "epoch": 0.6398293421414809,
+      "grad_norm": 1.7822249434901423,
+      "learning_rate": 6.0663538647904575e-06,
+      "loss": 0.9658,
+      "step": 4649
+    },
+    {
+      "epoch": 0.6399669694467383,
+      "grad_norm": 1.7221117461989217,
+      "learning_rate": 6.062256177560162e-06,
+      "loss": 0.9094,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6401045967519956,
+      "grad_norm": 1.709032045874571,
+      "learning_rate": 6.058159272705665e-06,
+      "loss": 1.0093,
+      "step": 4651
+    },
+    {
+      "epoch": 0.640242224057253,
+      "grad_norm": 1.6033615323747255,
+      "learning_rate": 6.054063151040961e-06,
+      "loss": 0.9761,
+      "step": 4652
+    },
+    {
+      "epoch": 0.6403798513625103,
+      "grad_norm": 1.9429281380778696,
+      "learning_rate": 6.049967813379898e-06,
+      "loss": 1.0152,
+      "step": 4653
+    },
+    {
+      "epoch": 0.6405174786677676,
+      "grad_norm": 1.94251012865841,
+      "learning_rate": 6.045873260536157e-06,
+      "loss": 0.9515,
+      "step": 4654
+    },
+    {
+      "epoch": 0.640655105973025,
+      "grad_norm": 2.0562120657176126,
+      "learning_rate": 6.0417794933232764e-06,
+      "loss": 1.0149,
+      "step": 4655
+    },
+    {
+      "epoch": 0.6407927332782825,
+      "grad_norm": 1.8418920801983083,
+      "learning_rate": 6.03768651255463e-06,
+      "loss": 0.9306,
+      "step": 4656
+    },
+    {
+      "epoch": 0.6409303605835398,
+      "grad_norm": 1.9130854029445379,
+      "learning_rate": 6.0335943190434295e-06,
+      "loss": 0.8917,
+      "step": 4657
+    },
+    {
+      "epoch": 0.6410679878887972,
+      "grad_norm": 1.923950068728734,
+      "learning_rate": 6.029502913602747e-06,
+      "loss": 0.9722,
+      "step": 4658
+    },
+    {
+      "epoch": 0.6412056151940545,
+      "grad_norm": 1.6943391679054847,
+      "learning_rate": 6.025412297045483e-06,
+      "loss": 0.9813,
+      "step": 4659
+    },
+    {
+      "epoch": 0.6413432424993118,
+      "grad_norm": 1.92277797136287,
+      "learning_rate": 6.021322470184392e-06,
+      "loss": 0.9817,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6414808698045692,
+      "grad_norm": 1.6847458692955486,
+      "learning_rate": 6.017233433832062e-06,
+      "loss": 0.9577,
+      "step": 4661
+    },
+    {
+      "epoch": 0.6416184971098265,
+      "grad_norm": 1.7684009714172524,
+      "learning_rate": 6.013145188800933e-06,
+      "loss": 0.9794,
+      "step": 4662
+    },
+    {
+      "epoch": 0.641756124415084,
+      "grad_norm": 1.6344184981483505,
+      "learning_rate": 6.009057735903279e-06,
+      "loss": 0.9203,
+      "step": 4663
+    },
+    {
+      "epoch": 0.6418937517203414,
+      "grad_norm": 1.6437118974050084,
+      "learning_rate": 6.004971075951225e-06,
+      "loss": 0.9245,
+      "step": 4664
+    },
+    {
+      "epoch": 0.6420313790255987,
+      "grad_norm": 1.4991400381238833,
+      "learning_rate": 6.000885209756729e-06,
+      "loss": 0.9327,
+      "step": 4665
+    },
+    {
+      "epoch": 0.642169006330856,
+      "grad_norm": 1.6376648247514098,
+      "learning_rate": 5.996800138131601e-06,
+      "loss": 0.8793,
+      "step": 4666
+    },
+    {
+      "epoch": 0.6423066336361134,
+      "grad_norm": 1.7087378081005948,
+      "learning_rate": 5.992715861887488e-06,
+      "loss": 0.9331,
+      "step": 4667
+    },
+    {
+      "epoch": 0.6424442609413707,
+      "grad_norm": 1.8381711735532147,
+      "learning_rate": 5.98863238183588e-06,
+      "loss": 0.9613,
+      "step": 4668
+    },
+    {
+      "epoch": 0.6425818882466281,
+      "grad_norm": 1.5953842704243024,
+      "learning_rate": 5.984549698788107e-06,
+      "loss": 0.9075,
+      "step": 4669
+    },
+    {
+      "epoch": 0.6427195155518854,
+      "grad_norm": 1.5735097453533011,
+      "learning_rate": 5.980467813555346e-06,
+      "loss": 0.9128,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 1.6565054989385288,
+      "learning_rate": 5.9763867269486045e-06,
+      "loss": 0.9598,
+      "step": 4671
+    },
+    {
+      "epoch": 0.6429947701624003,
+      "grad_norm": 1.868156698608928,
+      "learning_rate": 5.972306439778747e-06,
+      "loss": 0.9345,
+      "step": 4672
+    },
+    {
+      "epoch": 0.6431323974676576,
+      "grad_norm": 1.8653699995880506,
+      "learning_rate": 5.9682269528564674e-06,
+      "loss": 1.0512,
+      "step": 4673
+    },
+    {
+      "epoch": 0.643270024772915,
+      "grad_norm": 1.6922773472644,
+      "learning_rate": 5.9641482669923e-06,
+      "loss": 0.9788,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6434076520781723,
+      "grad_norm": 1.6564431915129343,
+      "learning_rate": 5.96007038299663e-06,
+      "loss": 0.93,
+      "step": 4675
+    },
+    {
+      "epoch": 0.6435452793834296,
+      "grad_norm": 1.8075376492203272,
+      "learning_rate": 5.9559933016796715e-06,
+      "loss": 0.8591,
+      "step": 4676
+    },
+    {
+      "epoch": 0.643682906688687,
+      "grad_norm": 1.715258834675489,
+      "learning_rate": 5.95191702385149e-06,
+      "loss": 0.9834,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6438205339939443,
+      "grad_norm": 1.622766964087439,
+      "learning_rate": 5.947841550321982e-06,
+      "loss": 0.962,
+      "step": 4678
+    },
+    {
+      "epoch": 0.6439581612992018,
+      "grad_norm": 1.7459788265544067,
+      "learning_rate": 5.943766881900892e-06,
+      "loss": 0.9196,
+      "step": 4679
+    },
+    {
+      "epoch": 0.6440957886044592,
+      "grad_norm": 1.7014460448796573,
+      "learning_rate": 5.939693019397797e-06,
+      "loss": 0.9242,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6442334159097165,
+      "grad_norm": 1.8972065704131083,
+      "learning_rate": 5.935619963622119e-06,
+      "loss": 1.013,
+      "step": 4681
+    },
+    {
+      "epoch": 0.6443710432149738,
+      "grad_norm": 2.023762240829979,
+      "learning_rate": 5.931547715383119e-06,
+      "loss": 0.9532,
+      "step": 4682
+    },
+    {
+      "epoch": 0.6445086705202312,
+      "grad_norm": 1.7810926212959912,
+      "learning_rate": 5.9274762754898985e-06,
+      "loss": 0.9196,
+      "step": 4683
+    },
+    {
+      "epoch": 0.6446462978254885,
+      "grad_norm": 1.719199410346551,
+      "learning_rate": 5.923405644751392e-06,
+      "loss": 0.9617,
+      "step": 4684
+    },
+    {
+      "epoch": 0.6447839251307459,
+      "grad_norm": 1.672050553692242,
+      "learning_rate": 5.919335823976386e-06,
+      "loss": 0.9804,
+      "step": 4685
+    },
+    {
+      "epoch": 0.6449215524360034,
+      "grad_norm": 3.1500105986941715,
+      "learning_rate": 5.9152668139734904e-06,
+      "loss": 0.9469,
+      "step": 4686
+    },
+    {
+      "epoch": 0.6450591797412607,
+      "grad_norm": 1.7401597977646142,
+      "learning_rate": 5.911198615551168e-06,
+      "loss": 0.9504,
+      "step": 4687
+    },
+    {
+      "epoch": 0.645196807046518,
+      "grad_norm": 1.508635890595875,
+      "learning_rate": 5.9071312295177085e-06,
+      "loss": 0.8874,
+      "step": 4688
+    },
+    {
+      "epoch": 0.6453344343517754,
+      "grad_norm": 1.6663587876081523,
+      "learning_rate": 5.90306465668125e-06,
+      "loss": 0.9077,
+      "step": 4689
+    },
+    {
+      "epoch": 0.6454720616570327,
+      "grad_norm": 1.8801021530484494,
+      "learning_rate": 5.8989988978497616e-06,
+      "loss": 0.92,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6456096889622901,
+      "grad_norm": 1.6241074841735041,
+      "learning_rate": 5.8949339538310544e-06,
+      "loss": 0.9418,
+      "step": 4691
+    },
+    {
+      "epoch": 0.6457473162675474,
+      "grad_norm": 1.7110146643569286,
+      "learning_rate": 5.890869825432784e-06,
+      "loss": 0.8564,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6458849435728048,
+      "grad_norm": 1.6619081632275132,
+      "learning_rate": 5.8868065134624244e-06,
+      "loss": 0.9687,
+      "step": 4693
+    },
+    {
+      "epoch": 0.6460225708780623,
+      "grad_norm": 1.773332522569468,
+      "learning_rate": 5.882744018727313e-06,
+      "loss": 0.9919,
+      "step": 4694
+    },
+    {
+      "epoch": 0.6461601981833196,
+      "grad_norm": 1.8127500232206128,
+      "learning_rate": 5.8786823420346e-06,
+      "loss": 0.9672,
+      "step": 4695
+    },
+    {
+      "epoch": 0.646297825488577,
+      "grad_norm": 1.7038749585149044,
+      "learning_rate": 5.874621484191298e-06,
+      "loss": 0.9024,
+      "step": 4696
+    },
+    {
+      "epoch": 0.6464354527938343,
+      "grad_norm": 1.8796920781012176,
+      "learning_rate": 5.87056144600423e-06,
+      "loss": 0.9546,
+      "step": 4697
+    },
+    {
+      "epoch": 0.6465730800990916,
+      "grad_norm": 1.7182549181074176,
+      "learning_rate": 5.866502228280081e-06,
+      "loss": 0.9269,
+      "step": 4698
+    },
+    {
+      "epoch": 0.646710707404349,
+      "grad_norm": 1.9853619760095602,
+      "learning_rate": 5.862443831825355e-06,
+      "loss": 0.9305,
+      "step": 4699
+    },
+    {
+      "epoch": 0.6468483347096063,
+      "grad_norm": 1.827084484117126,
+      "learning_rate": 5.858386257446407e-06,
+      "loss": 0.9402,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6469859620148638,
+      "grad_norm": 1.7445711613971626,
+      "learning_rate": 5.854329505949412e-06,
+      "loss": 0.9596,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6471235893201212,
+      "grad_norm": 1.9044111836036044,
+      "learning_rate": 5.850273578140402e-06,
+      "loss": 0.8499,
+      "step": 4702
+    },
+    {
+      "epoch": 0.6472612166253785,
+      "grad_norm": 1.9652352776464437,
+      "learning_rate": 5.846218474825223e-06,
+      "loss": 0.9789,
+      "step": 4703
+    },
+    {
+      "epoch": 0.6473988439306358,
+      "grad_norm": 1.8453069534738111,
+      "learning_rate": 5.842164196809578e-06,
+      "loss": 0.9696,
+      "step": 4704
+    },
+    {
+      "epoch": 0.6475364712358932,
+      "grad_norm": 1.8822558013626276,
+      "learning_rate": 5.8381107448989926e-06,
+      "loss": 0.9415,
+      "step": 4705
+    },
+    {
+      "epoch": 0.6476740985411505,
+      "grad_norm": 2.084590747919869,
+      "learning_rate": 5.8340581198988265e-06,
+      "loss": 1.0655,
+      "step": 4706
+    },
+    {
+      "epoch": 0.6478117258464079,
+      "grad_norm": 1.9309659309017047,
+      "learning_rate": 5.8300063226142945e-06,
+      "loss": 0.9664,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6479493531516652,
+      "grad_norm": 2.2183599264132163,
+      "learning_rate": 5.825955353850417e-06,
+      "loss": 1.013,
+      "step": 4708
+    },
+    {
+      "epoch": 0.6480869804569227,
+      "grad_norm": 1.710352144999003,
+      "learning_rate": 5.821905214412078e-06,
+      "loss": 0.9546,
+      "step": 4709
+    },
+    {
+      "epoch": 0.64822460776218,
+      "grad_norm": 1.8604922652304907,
+      "learning_rate": 5.81785590510398e-06,
+      "loss": 1.0412,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6483622350674374,
+      "grad_norm": 1.9653711384531218,
+      "learning_rate": 5.813807426730669e-06,
+      "loss": 0.8976,
+      "step": 4711
+    },
+    {
+      "epoch": 0.6484998623726947,
+      "grad_norm": 1.5291957626161208,
+      "learning_rate": 5.809759780096518e-06,
+      "loss": 0.894,
+      "step": 4712
+    },
+    {
+      "epoch": 0.6486374896779521,
+      "grad_norm": 1.7653736633462207,
+      "learning_rate": 5.8057129660057395e-06,
+      "loss": 0.9613,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6487751169832094,
+      "grad_norm": 1.7100631108672826,
+      "learning_rate": 5.801666985262382e-06,
+      "loss": 0.9161,
+      "step": 4714
+    },
+    {
+      "epoch": 0.6489127442884668,
+      "grad_norm": 1.743464950414026,
+      "learning_rate": 5.797621838670327e-06,
+      "loss": 0.9422,
+      "step": 4715
+    },
+    {
+      "epoch": 0.6490503715937241,
+      "grad_norm": 1.6825908282723192,
+      "learning_rate": 5.793577527033287e-06,
+      "loss": 0.8859,
+      "step": 4716
+    },
+    {
+      "epoch": 0.6491879988989816,
+      "grad_norm": 1.5861789880968253,
+      "learning_rate": 5.789534051154812e-06,
+      "loss": 0.9208,
+      "step": 4717
+    },
+    {
+      "epoch": 0.649325626204239,
+      "grad_norm": 1.7545750361860968,
+      "learning_rate": 5.7854914118382875e-06,
+      "loss": 0.877,
+      "step": 4718
+    },
+    {
+      "epoch": 0.6494632535094963,
+      "grad_norm": 1.8820652439866667,
+      "learning_rate": 5.781449609886929e-06,
+      "loss": 0.8799,
+      "step": 4719
+    },
+    {
+      "epoch": 0.6496008808147536,
+      "grad_norm": 1.643839326782363,
+      "learning_rate": 5.777408646103787e-06,
+      "loss": 0.942,
+      "step": 4720
+    },
+    {
+      "epoch": 0.649738508120011,
+      "grad_norm": 1.8147122938479316,
+      "learning_rate": 5.773368521291747e-06,
+      "loss": 0.9166,
+      "step": 4721
+    },
+    {
+      "epoch": 0.6498761354252683,
+      "grad_norm": 1.8078875124530347,
+      "learning_rate": 5.7693292362535235e-06,
+      "loss": 0.9616,
+      "step": 4722
+    },
+    {
+      "epoch": 0.6500137627305257,
+      "grad_norm": 2.0379238169140153,
+      "learning_rate": 5.7652907917916716e-06,
+      "loss": 0.9568,
+      "step": 4723
+    },
+    {
+      "epoch": 0.6501513900357831,
+      "grad_norm": 1.7467298298810137,
+      "learning_rate": 5.7612531887085696e-06,
+      "loss": 0.9781,
+      "step": 4724
+    },
+    {
+      "epoch": 0.6502890173410405,
+      "grad_norm": 1.7668215924698745,
+      "learning_rate": 5.757216427806436e-06,
+      "loss": 0.9078,
+      "step": 4725
+    },
+    {
+      "epoch": 0.6504266446462978,
+      "grad_norm": 1.7423680113086488,
+      "learning_rate": 5.753180509887326e-06,
+      "loss": 0.9372,
+      "step": 4726
+    },
+    {
+      "epoch": 0.6505642719515552,
+      "grad_norm": 1.7393768333476862,
+      "learning_rate": 5.749145435753109e-06,
+      "loss": 0.9453,
+      "step": 4727
+    },
+    {
+      "epoch": 0.6507018992568125,
+      "grad_norm": 1.8641347058759012,
+      "learning_rate": 5.745111206205516e-06,
+      "loss": 0.8891,
+      "step": 4728
+    },
+    {
+      "epoch": 0.6508395265620699,
+      "grad_norm": 1.7571265197369372,
+      "learning_rate": 5.741077822046073e-06,
+      "loss": 0.9899,
+      "step": 4729
+    },
+    {
+      "epoch": 0.6509771538673272,
+      "grad_norm": 1.8297773399279351,
+      "learning_rate": 5.737045284076177e-06,
+      "loss": 0.8556,
+      "step": 4730
+    },
+    {
+      "epoch": 0.6511147811725846,
+      "grad_norm": 1.7359387469685557,
+      "learning_rate": 5.733013593097021e-06,
+      "loss": 0.9097,
+      "step": 4731
+    },
+    {
+      "epoch": 0.651252408477842,
+      "grad_norm": 1.9387900736890944,
+      "learning_rate": 5.7289827499096606e-06,
+      "loss": 0.8859,
+      "step": 4732
+    },
+    {
+      "epoch": 0.6513900357830994,
+      "grad_norm": 1.7203412998219079,
+      "learning_rate": 5.724952755314963e-06,
+      "loss": 0.9505,
+      "step": 4733
+    },
+    {
+      "epoch": 0.6515276630883567,
+      "grad_norm": 1.9342835919485395,
+      "learning_rate": 5.720923610113636e-06,
+      "loss": 0.9664,
+      "step": 4734
+    },
+    {
+      "epoch": 0.6516652903936141,
+      "grad_norm": 2.1792824439155836,
+      "learning_rate": 5.71689531510621e-06,
+      "loss": 0.9726,
+      "step": 4735
+    },
+    {
+      "epoch": 0.6518029176988714,
+      "grad_norm": 1.5905732080850266,
+      "learning_rate": 5.7128678710930576e-06,
+      "loss": 0.945,
+      "step": 4736
+    },
+    {
+      "epoch": 0.6519405450041288,
+      "grad_norm": 1.7088038238450163,
+      "learning_rate": 5.70884127887438e-06,
+      "loss": 0.9233,
+      "step": 4737
+    },
+    {
+      "epoch": 0.6520781723093861,
+      "grad_norm": 1.5713960702488403,
+      "learning_rate": 5.70481553925019e-06,
+      "loss": 0.9563,
+      "step": 4738
+    },
+    {
+      "epoch": 0.6522157996146436,
+      "grad_norm": 1.6719472569196379,
+      "learning_rate": 5.700790653020363e-06,
+      "loss": 0.9359,
+      "step": 4739
+    },
+    {
+      "epoch": 0.652353426919901,
+      "grad_norm": 1.9278555371991215,
+      "learning_rate": 5.696766620984584e-06,
+      "loss": 0.8846,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6524910542251583,
+      "grad_norm": 1.7381745370144501,
+      "learning_rate": 5.692743443942372e-06,
+      "loss": 0.9174,
+      "step": 4741
+    },
+    {
+      "epoch": 0.6526286815304156,
+      "grad_norm": 1.9247989791258253,
+      "learning_rate": 5.6887211226930775e-06,
+      "loss": 0.9176,
+      "step": 4742
+    },
+    {
+      "epoch": 0.652766308835673,
+      "grad_norm": 1.7044798008380282,
+      "learning_rate": 5.684699658035881e-06,
+      "loss": 0.917,
+      "step": 4743
+    },
+    {
+      "epoch": 0.6529039361409303,
+      "grad_norm": 1.6269841549903865,
+      "learning_rate": 5.6806790507697926e-06,
+      "loss": 0.8471,
+      "step": 4744
+    },
+    {
+      "epoch": 0.6530415634461877,
+      "grad_norm": 2.348984834001434,
+      "learning_rate": 5.676659301693652e-06,
+      "loss": 1.0082,
+      "step": 4745
+    },
+    {
+      "epoch": 0.653179190751445,
+      "grad_norm": 1.9103381392471932,
+      "learning_rate": 5.672640411606127e-06,
+      "loss": 1.0059,
+      "step": 4746
+    },
+    {
+      "epoch": 0.6533168180567025,
+      "grad_norm": 2.021722357731002,
+      "learning_rate": 5.66862238130572e-06,
+      "loss": 0.8592,
+      "step": 4747
+    },
+    {
+      "epoch": 0.6534544453619598,
+      "grad_norm": 1.7710243885406995,
+      "learning_rate": 5.6646052115907524e-06,
+      "loss": 0.9133,
+      "step": 4748
+    },
+    {
+      "epoch": 0.6535920726672172,
+      "grad_norm": 1.5313561846396435,
+      "learning_rate": 5.660588903259385e-06,
+      "loss": 0.9385,
+      "step": 4749
+    },
+    {
+      "epoch": 0.6537296999724745,
+      "grad_norm": 1.6744837350781705,
+      "learning_rate": 5.6565734571096e-06,
+      "loss": 0.9253,
+      "step": 4750
+    },
+    {
+      "epoch": 0.6538673272777319,
+      "grad_norm": 1.7481336595352612,
+      "learning_rate": 5.65255887393922e-06,
+      "loss": 0.9638,
+      "step": 4751
+    },
+    {
+      "epoch": 0.6540049545829892,
+      "grad_norm": 1.6889358517692095,
+      "learning_rate": 5.648545154545879e-06,
+      "loss": 0.8283,
+      "step": 4752
+    },
+    {
+      "epoch": 0.6541425818882466,
+      "grad_norm": 1.6245962274552208,
+      "learning_rate": 5.644532299727051e-06,
+      "loss": 1.0306,
+      "step": 4753
+    },
+    {
+      "epoch": 0.6542802091935039,
+      "grad_norm": 1.8931121557903652,
+      "learning_rate": 5.640520310280036e-06,
+      "loss": 0.9319,
+      "step": 4754
+    },
+    {
+      "epoch": 0.6544178364987614,
+      "grad_norm": 1.9735605247289432,
+      "learning_rate": 5.636509187001958e-06,
+      "loss": 0.9889,
+      "step": 4755
+    },
+    {
+      "epoch": 0.6545554638040187,
+      "grad_norm": 1.9029307130516868,
+      "learning_rate": 5.632498930689776e-06,
+      "loss": 0.9192,
+      "step": 4756
+    },
+    {
+      "epoch": 0.6546930911092761,
+      "grad_norm": 1.6870658867971808,
+      "learning_rate": 5.628489542140269e-06,
+      "loss": 0.9097,
+      "step": 4757
+    },
+    {
+      "epoch": 0.6548307184145334,
+      "grad_norm": 1.9398313396404252,
+      "learning_rate": 5.624481022150057e-06,
+      "loss": 0.982,
+      "step": 4758
+    },
+    {
+      "epoch": 0.6549683457197908,
+      "grad_norm": 1.6461949603165773,
+      "learning_rate": 5.620473371515564e-06,
+      "loss": 0.9559,
+      "step": 4759
+    },
+    {
+      "epoch": 0.6551059730250481,
+      "grad_norm": 3.046982130521264,
+      "learning_rate": 5.6164665910330715e-06,
+      "loss": 0.9246,
+      "step": 4760
+    },
+    {
+      "epoch": 0.6552436003303055,
+      "grad_norm": 1.921365374427787,
+      "learning_rate": 5.612460681498655e-06,
+      "loss": 0.8761,
+      "step": 4761
+    },
+    {
+      "epoch": 0.655381227635563,
+      "grad_norm": 1.7877611803003675,
+      "learning_rate": 5.608455643708251e-06,
+      "loss": 0.9784,
+      "step": 4762
+    },
+    {
+      "epoch": 0.6555188549408203,
+      "grad_norm": 1.8622379279682406,
+      "learning_rate": 5.6044514784575865e-06,
+      "loss": 0.9699,
+      "step": 4763
+    },
+    {
+      "epoch": 0.6556564822460776,
+      "grad_norm": 1.798972256952548,
+      "learning_rate": 5.600448186542251e-06,
+      "loss": 0.9484,
+      "step": 4764
+    },
+    {
+      "epoch": 0.655794109551335,
+      "grad_norm": 1.7873423292286488,
+      "learning_rate": 5.5964457687576366e-06,
+      "loss": 0.9781,
+      "step": 4765
+    },
+    {
+      "epoch": 0.6559317368565923,
+      "grad_norm": 1.6487300758632697,
+      "learning_rate": 5.5924442258989694e-06,
+      "loss": 0.9011,
+      "step": 4766
+    },
+    {
+      "epoch": 0.6560693641618497,
+      "grad_norm": 2.016685073026264,
+      "learning_rate": 5.588443558761303e-06,
+      "loss": 0.9134,
+      "step": 4767
+    },
+    {
+      "epoch": 0.656206991467107,
+      "grad_norm": 1.6595368082724218,
+      "learning_rate": 5.584443768139514e-06,
+      "loss": 0.9201,
+      "step": 4768
+    },
+    {
+      "epoch": 0.6563446187723644,
+      "grad_norm": 1.9181399705764264,
+      "learning_rate": 5.580444854828309e-06,
+      "loss": 0.935,
+      "step": 4769
+    },
+    {
+      "epoch": 0.6564822460776218,
+      "grad_norm": 1.6575017485135672,
+      "learning_rate": 5.576446819622205e-06,
+      "loss": 0.8864,
+      "step": 4770
+    },
+    {
+      "epoch": 0.6566198733828792,
+      "grad_norm": 1.5596366349219368,
+      "learning_rate": 5.572449663315572e-06,
+      "loss": 0.9618,
+      "step": 4771
+    },
+    {
+      "epoch": 0.6567575006881365,
+      "grad_norm": 1.5915940632181482,
+      "learning_rate": 5.5684533867025836e-06,
+      "loss": 0.9346,
+      "step": 4772
+    },
+    {
+      "epoch": 0.6568951279933939,
+      "grad_norm": 1.6786747500734258,
+      "learning_rate": 5.564457990577246e-06,
+      "loss": 0.9246,
+      "step": 4773
+    },
+    {
+      "epoch": 0.6570327552986512,
+      "grad_norm": 2.0542785043894902,
+      "learning_rate": 5.560463475733391e-06,
+      "loss": 0.8635,
+      "step": 4774
+    },
+    {
+      "epoch": 0.6571703826039086,
+      "grad_norm": 1.8286086758025915,
+      "learning_rate": 5.556469842964671e-06,
+      "loss": 0.9226,
+      "step": 4775
+    },
+    {
+      "epoch": 0.6573080099091659,
+      "grad_norm": 1.9070787855225195,
+      "learning_rate": 5.552477093064569e-06,
+      "loss": 0.926,
+      "step": 4776
+    },
+    {
+      "epoch": 0.6574456372144234,
+      "grad_norm": 1.8197117485145144,
+      "learning_rate": 5.548485226826388e-06,
+      "loss": 0.9327,
+      "step": 4777
+    },
+    {
+      "epoch": 0.6575832645196807,
+      "grad_norm": 1.817641231659796,
+      "learning_rate": 5.544494245043257e-06,
+      "loss": 0.9512,
+      "step": 4778
+    },
+    {
+      "epoch": 0.6577208918249381,
+      "grad_norm": 1.6158656313706663,
+      "learning_rate": 5.5405041485081325e-06,
+      "loss": 0.9718,
+      "step": 4779
+    },
+    {
+      "epoch": 0.6578585191301954,
+      "grad_norm": 1.7958837150032625,
+      "learning_rate": 5.536514938013788e-06,
+      "loss": 1.0335,
+      "step": 4780
+    },
+    {
+      "epoch": 0.6579961464354528,
+      "grad_norm": 1.8111482609644687,
+      "learning_rate": 5.5325266143528275e-06,
+      "loss": 0.993,
+      "step": 4781
+    },
+    {
+      "epoch": 0.6581337737407101,
+      "grad_norm": 1.7682189690482637,
+      "learning_rate": 5.528539178317671e-06,
+      "loss": 1.0005,
+      "step": 4782
+    },
+    {
+      "epoch": 0.6582714010459675,
+      "grad_norm": 2.3681347559499333,
+      "learning_rate": 5.524552630700582e-06,
+      "loss": 0.9315,
+      "step": 4783
+    },
+    {
+      "epoch": 0.6584090283512248,
+      "grad_norm": 2.273217520412821,
+      "learning_rate": 5.520566972293617e-06,
+      "loss": 0.9656,
+      "step": 4784
+    },
+    {
+      "epoch": 0.6585466556564823,
+      "grad_norm": 1.6436123867671102,
+      "learning_rate": 5.516582203888679e-06,
+      "loss": 0.9339,
+      "step": 4785
+    },
+    {
+      "epoch": 0.6586842829617396,
+      "grad_norm": 1.7825602949204846,
+      "learning_rate": 5.512598326277485e-06,
+      "loss": 0.8829,
+      "step": 4786
+    },
+    {
+      "epoch": 0.658821910266997,
+      "grad_norm": 1.586233092149229,
+      "learning_rate": 5.50861534025158e-06,
+      "loss": 0.8386,
+      "step": 4787
+    },
+    {
+      "epoch": 0.6589595375722543,
+      "grad_norm": 1.5519737846349984,
+      "learning_rate": 5.504633246602325e-06,
+      "loss": 0.8585,
+      "step": 4788
+    },
+    {
+      "epoch": 0.6590971648775117,
+      "grad_norm": 2.8139886685347775,
+      "learning_rate": 5.500652046120904e-06,
+      "loss": 0.9331,
+      "step": 4789
+    },
+    {
+      "epoch": 0.659234792182769,
+      "grad_norm": 2.0849848563577753,
+      "learning_rate": 5.49667173959834e-06,
+      "loss": 0.9474,
+      "step": 4790
+    },
+    {
+      "epoch": 0.6593724194880264,
+      "grad_norm": 1.9472216192364793,
+      "learning_rate": 5.49269232782545e-06,
+      "loss": 0.9759,
+      "step": 4791
+    },
+    {
+      "epoch": 0.6595100467932837,
+      "grad_norm": 1.5223370743680098,
+      "learning_rate": 5.488713811592905e-06,
+      "loss": 0.95,
+      "step": 4792
+    },
+    {
+      "epoch": 0.6596476740985412,
+      "grad_norm": 1.7795115276815248,
+      "learning_rate": 5.484736191691164e-06,
+      "loss": 0.9644,
+      "step": 4793
+    },
+    {
+      "epoch": 0.6597853014037985,
+      "grad_norm": 1.8301772376442782,
+      "learning_rate": 5.480759468910539e-06,
+      "loss": 0.9963,
+      "step": 4794
+    },
+    {
+      "epoch": 0.6599229287090559,
+      "grad_norm": 2.0207218472046042,
+      "learning_rate": 5.476783644041146e-06,
+      "loss": 1.046,
+      "step": 4795
+    },
+    {
+      "epoch": 0.6600605560143132,
+      "grad_norm": 1.6322878989117735,
+      "learning_rate": 5.472808717872927e-06,
+      "loss": 0.9139,
+      "step": 4796
+    },
+    {
+      "epoch": 0.6601981833195706,
+      "grad_norm": 1.8705520577060168,
+      "learning_rate": 5.468834691195646e-06,
+      "loss": 0.9237,
+      "step": 4797
+    },
+    {
+      "epoch": 0.6603358106248279,
+      "grad_norm": 1.55841029130605,
+      "learning_rate": 5.464861564798885e-06,
+      "loss": 0.9807,
+      "step": 4798
+    },
+    {
+      "epoch": 0.6604734379300853,
+      "grad_norm": 2.231216787937271,
+      "learning_rate": 5.460889339472053e-06,
+      "loss": 0.8948,
+      "step": 4799
+    },
+    {
+      "epoch": 0.6606110652353427,
+      "grad_norm": 1.7642161742655231,
+      "learning_rate": 5.4569180160043744e-06,
+      "loss": 0.9431,
+      "step": 4800
+    },
+    {
+      "epoch": 0.6607486925406001,
+      "grad_norm": 1.8862589333636366,
+      "learning_rate": 5.452947595184899e-06,
+      "loss": 0.9465,
+      "step": 4801
+    },
+    {
+      "epoch": 0.6608863198458574,
+      "grad_norm": 1.684327523303295,
+      "learning_rate": 5.448978077802493e-06,
+      "loss": 0.922,
+      "step": 4802
+    },
+    {
+      "epoch": 0.6610239471511148,
+      "grad_norm": 1.586718451754625,
+      "learning_rate": 5.445009464645847e-06,
+      "loss": 0.9331,
+      "step": 4803
+    },
+    {
+      "epoch": 0.6611615744563721,
+      "grad_norm": 1.5648252421752016,
+      "learning_rate": 5.441041756503471e-06,
+      "loss": 0.985,
+      "step": 4804
+    },
+    {
+      "epoch": 0.6612992017616295,
+      "grad_norm": 1.9870894219262807,
+      "learning_rate": 5.43707495416369e-06,
+      "loss": 0.8787,
+      "step": 4805
+    },
+    {
+      "epoch": 0.6614368290668868,
+      "grad_norm": 1.8217231945777181,
+      "learning_rate": 5.433109058414658e-06,
+      "loss": 0.9183,
+      "step": 4806
+    },
+    {
+      "epoch": 0.6615744563721442,
+      "grad_norm": 1.733092520996274,
+      "learning_rate": 5.429144070044341e-06,
+      "loss": 0.9478,
+      "step": 4807
+    },
+    {
+      "epoch": 0.6617120836774016,
+      "grad_norm": 1.6807428712357377,
+      "learning_rate": 5.42517998984053e-06,
+      "loss": 0.9346,
+      "step": 4808
+    },
+    {
+      "epoch": 0.661849710982659,
+      "grad_norm": 1.9261952148208565,
+      "learning_rate": 5.4212168185908306e-06,
+      "loss": 0.9717,
+      "step": 4809
+    },
+    {
+      "epoch": 0.6619873382879163,
+      "grad_norm": 1.7070469195944993,
+      "learning_rate": 5.417254557082674e-06,
+      "loss": 0.9289,
+      "step": 4810
+    },
+    {
+      "epoch": 0.6621249655931737,
+      "grad_norm": 1.6527808900697794,
+      "learning_rate": 5.413293206103307e-06,
+      "loss": 0.858,
+      "step": 4811
+    },
+    {
+      "epoch": 0.662262592898431,
+      "grad_norm": 1.6352672618739776,
+      "learning_rate": 5.40933276643979e-06,
+      "loss": 0.9326,
+      "step": 4812
+    },
+    {
+      "epoch": 0.6624002202036884,
+      "grad_norm": 1.6388311857160311,
+      "learning_rate": 5.405373238879019e-06,
+      "loss": 0.8872,
+      "step": 4813
+    },
+    {
+      "epoch": 0.6625378475089457,
+      "grad_norm": 1.7798790121996448,
+      "learning_rate": 5.401414624207686e-06,
+      "loss": 0.9936,
+      "step": 4814
+    },
+    {
+      "epoch": 0.6626754748142032,
+      "grad_norm": 1.7964817196908192,
+      "learning_rate": 5.3974569232123276e-06,
+      "loss": 0.9169,
+      "step": 4815
+    },
+    {
+      "epoch": 0.6628131021194605,
+      "grad_norm": 1.7539945311181884,
+      "learning_rate": 5.39350013667927e-06,
+      "loss": 0.8869,
+      "step": 4816
+    },
+    {
+      "epoch": 0.6629507294247179,
+      "grad_norm": 1.837124684683883,
+      "learning_rate": 5.3895442653946815e-06,
+      "loss": 0.9352,
+      "step": 4817
+    },
+    {
+      "epoch": 0.6630883567299752,
+      "grad_norm": 1.884975813977823,
+      "learning_rate": 5.385589310144537e-06,
+      "loss": 0.93,
+      "step": 4818
+    },
+    {
+      "epoch": 0.6632259840352326,
+      "grad_norm": 2.0147527899623223,
+      "learning_rate": 5.381635271714628e-06,
+      "loss": 0.9091,
+      "step": 4819
+    },
+    {
+      "epoch": 0.6633636113404899,
+      "grad_norm": 1.9933332905998353,
+      "learning_rate": 5.37768215089058e-06,
+      "loss": 0.9507,
+      "step": 4820
+    },
+    {
+      "epoch": 0.6635012386457473,
+      "grad_norm": 1.6162683540936273,
+      "learning_rate": 5.373729948457808e-06,
+      "loss": 0.9423,
+      "step": 4821
+    },
+    {
+      "epoch": 0.6636388659510046,
+      "grad_norm": 1.8077022990854756,
+      "learning_rate": 5.369778665201576e-06,
+      "loss": 0.8766,
+      "step": 4822
+    },
+    {
+      "epoch": 0.6637764932562621,
+      "grad_norm": 1.8565527119308636,
+      "learning_rate": 5.365828301906935e-06,
+      "loss": 0.9864,
+      "step": 4823
+    },
+    {
+      "epoch": 0.6639141205615194,
+      "grad_norm": 1.9682076812298217,
+      "learning_rate": 5.361878859358782e-06,
+      "loss": 1.0084,
+      "step": 4824
+    },
+    {
+      "epoch": 0.6640517478667768,
+      "grad_norm": 1.8094138519308847,
+      "learning_rate": 5.3579303383418055e-06,
+      "loss": 0.9184,
+      "step": 4825
+    },
+    {
+      "epoch": 0.6641893751720341,
+      "grad_norm": 1.70571074518677,
+      "learning_rate": 5.35398273964053e-06,
+      "loss": 0.9156,
+      "step": 4826
+    },
+    {
+      "epoch": 0.6643270024772915,
+      "grad_norm": 1.8078830540605337,
+      "learning_rate": 5.350036064039287e-06,
+      "loss": 0.9213,
+      "step": 4827
+    },
+    {
+      "epoch": 0.6644646297825488,
+      "grad_norm": 1.7880119449978071,
+      "learning_rate": 5.346090312322227e-06,
+      "loss": 0.8809,
+      "step": 4828
+    },
+    {
+      "epoch": 0.6646022570878062,
+      "grad_norm": 1.7348864464791165,
+      "learning_rate": 5.342145485273318e-06,
+      "loss": 0.9622,
+      "step": 4829
+    },
+    {
+      "epoch": 0.6647398843930635,
+      "grad_norm": 1.8730222989355196,
+      "learning_rate": 5.338201583676341e-06,
+      "loss": 0.9168,
+      "step": 4830
+    },
+    {
+      "epoch": 0.664877511698321,
+      "grad_norm": 1.8779717580668454,
+      "learning_rate": 5.334258608314896e-06,
+      "loss": 0.9868,
+      "step": 4831
+    },
+    {
+      "epoch": 0.6650151390035783,
+      "grad_norm": 1.8596894935275075,
+      "learning_rate": 5.330316559972399e-06,
+      "loss": 0.9811,
+      "step": 4832
+    },
+    {
+      "epoch": 0.6651527663088357,
+      "grad_norm": 1.5671035282518169,
+      "learning_rate": 5.326375439432081e-06,
+      "loss": 0.8558,
+      "step": 4833
+    },
+    {
+      "epoch": 0.665290393614093,
+      "grad_norm": 1.7469413819214,
+      "learning_rate": 5.322435247476987e-06,
+      "loss": 0.9223,
+      "step": 4834
+    },
+    {
+      "epoch": 0.6654280209193504,
+      "grad_norm": 1.7965477684775375,
+      "learning_rate": 5.318495984889981e-06,
+      "loss": 0.9751,
+      "step": 4835
+    },
+    {
+      "epoch": 0.6655656482246077,
+      "grad_norm": 1.7216704915826047,
+      "learning_rate": 5.31455765245374e-06,
+      "loss": 0.9299,
+      "step": 4836
+    },
+    {
+      "epoch": 0.6657032755298651,
+      "grad_norm": 1.872358180203065,
+      "learning_rate": 5.310620250950757e-06,
+      "loss": 0.9166,
+      "step": 4837
+    },
+    {
+      "epoch": 0.6658409028351225,
+      "grad_norm": 1.7665845013952979,
+      "learning_rate": 5.30668378116334e-06,
+      "loss": 0.9693,
+      "step": 4838
+    },
+    {
+      "epoch": 0.6659785301403799,
+      "grad_norm": 1.6258215132595784,
+      "learning_rate": 5.30274824387361e-06,
+      "loss": 0.9233,
+      "step": 4839
+    },
+    {
+      "epoch": 0.6661161574456372,
+      "grad_norm": 2.2642381979066646,
+      "learning_rate": 5.298813639863506e-06,
+      "loss": 0.9443,
+      "step": 4840
+    },
+    {
+      "epoch": 0.6662537847508946,
+      "grad_norm": 1.6263913166049058,
+      "learning_rate": 5.294879969914779e-06,
+      "loss": 0.9225,
+      "step": 4841
+    },
+    {
+      "epoch": 0.6663914120561519,
+      "grad_norm": 1.7302759276664266,
+      "learning_rate": 5.2909472348089966e-06,
+      "loss": 0.8868,
+      "step": 4842
+    },
+    {
+      "epoch": 0.6665290393614093,
+      "grad_norm": 1.681557386524565,
+      "learning_rate": 5.287015435327538e-06,
+      "loss": 0.9594,
+      "step": 4843
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.5932125643263908,
+      "learning_rate": 5.2830845722515965e-06,
+      "loss": 0.9608,
+      "step": 4844
+    },
+    {
+      "epoch": 0.666804293971924,
+      "grad_norm": 1.966749014603279,
+      "learning_rate": 5.27915464636219e-06,
+      "loss": 0.9553,
+      "step": 4845
+    },
+    {
+      "epoch": 0.6669419212771814,
+      "grad_norm": 1.856777001811676,
+      "learning_rate": 5.275225658440126e-06,
+      "loss": 0.9877,
+      "step": 4846
+    },
+    {
+      "epoch": 0.6670795485824388,
+      "grad_norm": 1.7926240840334098,
+      "learning_rate": 5.271297609266055e-06,
+      "loss": 0.9351,
+      "step": 4847
+    },
+    {
+      "epoch": 0.6672171758876961,
+      "grad_norm": 2.0506245590916583,
+      "learning_rate": 5.267370499620419e-06,
+      "loss": 0.9382,
+      "step": 4848
+    },
+    {
+      "epoch": 0.6673548031929535,
+      "grad_norm": 1.521391167830223,
+      "learning_rate": 5.263444330283481e-06,
+      "loss": 0.9135,
+      "step": 4849
+    },
+    {
+      "epoch": 0.6674924304982108,
+      "grad_norm": 1.7534530916343891,
+      "learning_rate": 5.2595191020353165e-06,
+      "loss": 0.9862,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6676300578034682,
+      "grad_norm": 1.5138644780689012,
+      "learning_rate": 5.255594815655813e-06,
+      "loss": 0.8674,
+      "step": 4851
+    },
+    {
+      "epoch": 0.6677676851087255,
+      "grad_norm": 2.0913163258831897,
+      "learning_rate": 5.251671471924684e-06,
+      "loss": 0.9312,
+      "step": 4852
+    },
+    {
+      "epoch": 0.667905312413983,
+      "grad_norm": 1.6595909635099093,
+      "learning_rate": 5.247749071621429e-06,
+      "loss": 0.9631,
+      "step": 4853
+    },
+    {
+      "epoch": 0.6680429397192403,
+      "grad_norm": 1.8304869811787914,
+      "learning_rate": 5.2438276155253895e-06,
+      "loss": 0.9758,
+      "step": 4854
+    },
+    {
+      "epoch": 0.6681805670244977,
+      "grad_norm": 1.6917234365057052,
+      "learning_rate": 5.23990710441569e-06,
+      "loss": 0.975,
+      "step": 4855
+    },
+    {
+      "epoch": 0.668318194329755,
+      "grad_norm": 1.9442752869120192,
+      "learning_rate": 5.235987539071299e-06,
+      "loss": 0.9268,
+      "step": 4856
+    },
+    {
+      "epoch": 0.6684558216350124,
+      "grad_norm": 1.82269803077331,
+      "learning_rate": 5.232068920270963e-06,
+      "loss": 1.0476,
+      "step": 4857
+    },
+    {
+      "epoch": 0.6685934489402697,
+      "grad_norm": 1.7286872373511077,
+      "learning_rate": 5.22815124879327e-06,
+      "loss": 0.991,
+      "step": 4858
+    },
+    {
+      "epoch": 0.6687310762455271,
+      "grad_norm": 1.9100540474384782,
+      "learning_rate": 5.224234525416605e-06,
+      "loss": 0.9912,
+      "step": 4859
+    },
+    {
+      "epoch": 0.6688687035507844,
+      "grad_norm": 1.7559940599718344,
+      "learning_rate": 5.2203187509191665e-06,
+      "loss": 0.9198,
+      "step": 4860
+    },
+    {
+      "epoch": 0.6690063308560419,
+      "grad_norm": 2.0016348354342153,
+      "learning_rate": 5.216403926078965e-06,
+      "loss": 1.0055,
+      "step": 4861
+    },
+    {
+      "epoch": 0.6691439581612992,
+      "grad_norm": 2.0452095138914785,
+      "learning_rate": 5.212490051673825e-06,
+      "loss": 0.968,
+      "step": 4862
+    },
+    {
+      "epoch": 0.6692815854665566,
+      "grad_norm": 1.567994824741222,
+      "learning_rate": 5.208577128481377e-06,
+      "loss": 0.8547,
+      "step": 4863
+    },
+    {
+      "epoch": 0.6694192127718139,
+      "grad_norm": 1.5134339630187161,
+      "learning_rate": 5.204665157279065e-06,
+      "loss": 0.933,
+      "step": 4864
+    },
+    {
+      "epoch": 0.6695568400770713,
+      "grad_norm": 1.666285144543641,
+      "learning_rate": 5.200754138844146e-06,
+      "loss": 0.9787,
+      "step": 4865
+    },
+    {
+      "epoch": 0.6696944673823286,
+      "grad_norm": 1.9828382305226402,
+      "learning_rate": 5.196844073953686e-06,
+      "loss": 0.9891,
+      "step": 4866
+    },
+    {
+      "epoch": 0.669832094687586,
+      "grad_norm": 2.0634522205797183,
+      "learning_rate": 5.1929349633845595e-06,
+      "loss": 0.9344,
+      "step": 4867
+    },
+    {
+      "epoch": 0.6699697219928433,
+      "grad_norm": 1.840569318680048,
+      "learning_rate": 5.189026807913454e-06,
+      "loss": 0.8883,
+      "step": 4868
+    },
+    {
+      "epoch": 0.6701073492981008,
+      "grad_norm": 1.891805822897526,
+      "learning_rate": 5.185119608316868e-06,
+      "loss": 0.9621,
+      "step": 4869
+    },
+    {
+      "epoch": 0.6702449766033581,
+      "grad_norm": 1.7694740575556107,
+      "learning_rate": 5.181213365371106e-06,
+      "loss": 0.9102,
+      "step": 4870
+    },
+    {
+      "epoch": 0.6703826039086155,
+      "grad_norm": 1.7168621508469908,
+      "learning_rate": 5.177308079852287e-06,
+      "loss": 0.8503,
+      "step": 4871
+    },
+    {
+      "epoch": 0.6705202312138728,
+      "grad_norm": 1.8110756654453317,
+      "learning_rate": 5.1734037525363365e-06,
+      "loss": 1.0106,
+      "step": 4872
+    },
+    {
+      "epoch": 0.6706578585191302,
+      "grad_norm": 1.7150932683300406,
+      "learning_rate": 5.169500384198991e-06,
+      "loss": 0.9444,
+      "step": 4873
+    },
+    {
+      "epoch": 0.6707954858243875,
+      "grad_norm": 1.6850579383751274,
+      "learning_rate": 5.165597975615795e-06,
+      "loss": 0.9829,
+      "step": 4874
+    },
+    {
+      "epoch": 0.6709331131296449,
+      "grad_norm": 1.741830437601638,
+      "learning_rate": 5.161696527562111e-06,
+      "loss": 0.8876,
+      "step": 4875
+    },
+    {
+      "epoch": 0.6710707404349023,
+      "grad_norm": 1.7626312821494987,
+      "learning_rate": 5.1577960408130915e-06,
+      "loss": 0.9304,
+      "step": 4876
+    },
+    {
+      "epoch": 0.6712083677401597,
+      "grad_norm": 1.8535563987135282,
+      "learning_rate": 5.153896516143724e-06,
+      "loss": 0.95,
+      "step": 4877
+    },
+    {
+      "epoch": 0.671345995045417,
+      "grad_norm": 1.6673524819972623,
+      "learning_rate": 5.149997954328774e-06,
+      "loss": 0.9201,
+      "step": 4878
+    },
+    {
+      "epoch": 0.6714836223506744,
+      "grad_norm": 1.7347341637172635,
+      "learning_rate": 5.146100356142849e-06,
+      "loss": 0.9301,
+      "step": 4879
+    },
+    {
+      "epoch": 0.6716212496559317,
+      "grad_norm": 1.814521382583184,
+      "learning_rate": 5.142203722360337e-06,
+      "loss": 0.963,
+      "step": 4880
+    },
+    {
+      "epoch": 0.6717588769611891,
+      "grad_norm": 1.744258831797634,
+      "learning_rate": 5.138308053755446e-06,
+      "loss": 0.9504,
+      "step": 4881
+    },
+    {
+      "epoch": 0.6718965042664464,
+      "grad_norm": 1.8947185062577048,
+      "learning_rate": 5.134413351102203e-06,
+      "loss": 0.8865,
+      "step": 4882
+    },
+    {
+      "epoch": 0.6720341315717038,
+      "grad_norm": 1.642824926409088,
+      "learning_rate": 5.130519615174417e-06,
+      "loss": 0.9286,
+      "step": 4883
+    },
+    {
+      "epoch": 0.6721717588769612,
+      "grad_norm": 1.6325174431026037,
+      "learning_rate": 5.126626846745735e-06,
+      "loss": 0.9526,
+      "step": 4884
+    },
+    {
+      "epoch": 0.6723093861822186,
+      "grad_norm": 1.6860398501437657,
+      "learning_rate": 5.122735046589579e-06,
+      "loss": 0.9136,
+      "step": 4885
+    },
+    {
+      "epoch": 0.6724470134874759,
+      "grad_norm": 1.5908606501820293,
+      "learning_rate": 5.118844215479218e-06,
+      "loss": 0.9358,
+      "step": 4886
+    },
+    {
+      "epoch": 0.6725846407927333,
+      "grad_norm": 1.7091069150230902,
+      "learning_rate": 5.114954354187687e-06,
+      "loss": 0.9479,
+      "step": 4887
+    },
+    {
+      "epoch": 0.6727222680979906,
+      "grad_norm": 1.8393342977331808,
+      "learning_rate": 5.111065463487861e-06,
+      "loss": 1.0103,
+      "step": 4888
+    },
+    {
+      "epoch": 0.672859895403248,
+      "grad_norm": 1.6980355620182184,
+      "learning_rate": 5.107177544152404e-06,
+      "loss": 0.9685,
+      "step": 4889
+    },
+    {
+      "epoch": 0.6729975227085053,
+      "grad_norm": 1.5326246337283154,
+      "learning_rate": 5.103290596953795e-06,
+      "loss": 0.926,
+      "step": 4890
+    },
+    {
+      "epoch": 0.6731351500137628,
+      "grad_norm": 1.7614009824483032,
+      "learning_rate": 5.099404622664314e-06,
+      "loss": 0.9024,
+      "step": 4891
+    },
+    {
+      "epoch": 0.6732727773190201,
+      "grad_norm": 1.8199055096495367,
+      "learning_rate": 5.095519622056054e-06,
+      "loss": 0.8748,
+      "step": 4892
+    },
+    {
+      "epoch": 0.6734104046242775,
+      "grad_norm": 1.797990214991825,
+      "learning_rate": 5.09163559590091e-06,
+      "loss": 0.9356,
+      "step": 4893
+    },
+    {
+      "epoch": 0.6735480319295348,
+      "grad_norm": 1.922115466134818,
+      "learning_rate": 5.087752544970584e-06,
+      "loss": 0.9569,
+      "step": 4894
+    },
+    {
+      "epoch": 0.6736856592347922,
+      "grad_norm": 1.7594958535625165,
+      "learning_rate": 5.083870470036587e-06,
+      "loss": 0.9034,
+      "step": 4895
+    },
+    {
+      "epoch": 0.6738232865400495,
+      "grad_norm": 1.8920957941838494,
+      "learning_rate": 5.079989371870233e-06,
+      "loss": 0.9709,
+      "step": 4896
+    },
+    {
+      "epoch": 0.6739609138453069,
+      "grad_norm": 2.0314050550495257,
+      "learning_rate": 5.076109251242642e-06,
+      "loss": 0.9525,
+      "step": 4897
+    },
+    {
+      "epoch": 0.6740985411505642,
+      "grad_norm": 1.6096477438334647,
+      "learning_rate": 5.072230108924744e-06,
+      "loss": 0.9215,
+      "step": 4898
+    },
+    {
+      "epoch": 0.6742361684558217,
+      "grad_norm": 1.9153485492985267,
+      "learning_rate": 5.068351945687269e-06,
+      "loss": 0.9813,
+      "step": 4899
+    },
+    {
+      "epoch": 0.674373795761079,
+      "grad_norm": 1.6668128465768288,
+      "learning_rate": 5.064474762300755e-06,
+      "loss": 0.9972,
+      "step": 4900
+    },
+    {
+      "epoch": 0.6745114230663364,
+      "grad_norm": 3.741843183742772,
+      "learning_rate": 5.060598559535546e-06,
+      "loss": 0.918,
+      "step": 4901
+    },
+    {
+      "epoch": 0.6746490503715937,
+      "grad_norm": 1.9024416342728618,
+      "learning_rate": 5.056723338161788e-06,
+      "loss": 0.9131,
+      "step": 4902
+    },
+    {
+      "epoch": 0.6747866776768511,
+      "grad_norm": 1.7603050121508654,
+      "learning_rate": 5.052849098949439e-06,
+      "loss": 0.9179,
+      "step": 4903
+    },
+    {
+      "epoch": 0.6749243049821084,
+      "grad_norm": 1.6410275524886213,
+      "learning_rate": 5.048975842668255e-06,
+      "loss": 0.9476,
+      "step": 4904
+    },
+    {
+      "epoch": 0.6750619322873658,
+      "grad_norm": 2.299042220984978,
+      "learning_rate": 5.0451035700877994e-06,
+      "loss": 0.9264,
+      "step": 4905
+    },
+    {
+      "epoch": 0.6751995595926231,
+      "grad_norm": 1.9758463227506775,
+      "learning_rate": 5.041232281977433e-06,
+      "loss": 0.9327,
+      "step": 4906
+    },
+    {
+      "epoch": 0.6753371868978806,
+      "grad_norm": 1.732042578657454,
+      "learning_rate": 5.037361979106343e-06,
+      "loss": 0.9925,
+      "step": 4907
+    },
+    {
+      "epoch": 0.6754748142031379,
+      "grad_norm": 1.6958795087144407,
+      "learning_rate": 5.03349266224349e-06,
+      "loss": 0.9383,
+      "step": 4908
+    },
+    {
+      "epoch": 0.6756124415083953,
+      "grad_norm": 1.88049279989569,
+      "learning_rate": 5.029624332157667e-06,
+      "loss": 1.0068,
+      "step": 4909
+    },
+    {
+      "epoch": 0.6757500688136526,
+      "grad_norm": 1.962142678853269,
+      "learning_rate": 5.0257569896174455e-06,
+      "loss": 0.9781,
+      "step": 4910
+    },
+    {
+      "epoch": 0.67588769611891,
+      "grad_norm": 1.7102186948348521,
+      "learning_rate": 5.021890635391228e-06,
+      "loss": 1.0075,
+      "step": 4911
+    },
+    {
+      "epoch": 0.6760253234241673,
+      "grad_norm": 1.89182648444078,
+      "learning_rate": 5.0180252702471935e-06,
+      "loss": 0.9129,
+      "step": 4912
+    },
+    {
+      "epoch": 0.6761629507294247,
+      "grad_norm": 1.895372922650316,
+      "learning_rate": 5.014160894953339e-06,
+      "loss": 0.9253,
+      "step": 4913
+    },
+    {
+      "epoch": 0.6763005780346821,
+      "grad_norm": 1.8217178997996266,
+      "learning_rate": 5.010297510277473e-06,
+      "loss": 1.0025,
+      "step": 4914
+    },
+    {
+      "epoch": 0.6764382053399395,
+      "grad_norm": 1.9861275749238183,
+      "learning_rate": 5.006435116987183e-06,
+      "loss": 0.8884,
+      "step": 4915
+    },
+    {
+      "epoch": 0.6765758326451968,
+      "grad_norm": 1.8083563593812109,
+      "learning_rate": 5.002573715849889e-06,
+      "loss": 0.9211,
+      "step": 4916
+    },
+    {
+      "epoch": 0.6767134599504542,
+      "grad_norm": 1.8625348875238106,
+      "learning_rate": 4.998713307632782e-06,
+      "loss": 0.92,
+      "step": 4917
+    },
+    {
+      "epoch": 0.6768510872557115,
+      "grad_norm": 2.0793553117396484,
+      "learning_rate": 4.994853893102888e-06,
+      "loss": 0.9728,
+      "step": 4918
+    },
+    {
+      "epoch": 0.6769887145609689,
+      "grad_norm": 2.145997614393466,
+      "learning_rate": 4.990995473027005e-06,
+      "loss": 1.0395,
+      "step": 4919
+    },
+    {
+      "epoch": 0.6771263418662262,
+      "grad_norm": 1.5999667976502856,
+      "learning_rate": 4.9871380481717605e-06,
+      "loss": 0.9383,
+      "step": 4920
+    },
+    {
+      "epoch": 0.6772639691714836,
+      "grad_norm": 1.798350427548774,
+      "learning_rate": 4.983281619303566e-06,
+      "loss": 0.9576,
+      "step": 4921
+    },
+    {
+      "epoch": 0.677401596476741,
+      "grad_norm": 1.719450648615924,
+      "learning_rate": 4.979426187188643e-06,
+      "loss": 0.9151,
+      "step": 4922
+    },
+    {
+      "epoch": 0.6775392237819984,
+      "grad_norm": 1.8590760704400113,
+      "learning_rate": 4.9755717525930114e-06,
+      "loss": 0.9367,
+      "step": 4923
+    },
+    {
+      "epoch": 0.6776768510872557,
+      "grad_norm": 1.7103590697132074,
+      "learning_rate": 4.971718316282497e-06,
+      "loss": 0.9423,
+      "step": 4924
+    },
+    {
+      "epoch": 0.6778144783925131,
+      "grad_norm": 1.8103921355254051,
+      "learning_rate": 4.967865879022723e-06,
+      "loss": 0.8955,
+      "step": 4925
+    },
+    {
+      "epoch": 0.6779521056977704,
+      "grad_norm": 1.7951062928202395,
+      "learning_rate": 4.964014441579116e-06,
+      "loss": 0.8939,
+      "step": 4926
+    },
+    {
+      "epoch": 0.6780897330030278,
+      "grad_norm": 1.6226869437066,
+      "learning_rate": 4.960164004716904e-06,
+      "loss": 0.9275,
+      "step": 4927
+    },
+    {
+      "epoch": 0.6782273603082851,
+      "grad_norm": 1.5371496661751622,
+      "learning_rate": 4.956314569201117e-06,
+      "loss": 1.0389,
+      "step": 4928
+    },
+    {
+      "epoch": 0.6783649876135426,
+      "grad_norm": 1.6719937490941819,
+      "learning_rate": 4.9524661357965855e-06,
+      "loss": 1.0125,
+      "step": 4929
+    },
+    {
+      "epoch": 0.6785026149187999,
+      "grad_norm": 1.7697414861521257,
+      "learning_rate": 4.948618705267939e-06,
+      "loss": 0.9539,
+      "step": 4930
+    },
+    {
+      "epoch": 0.6786402422240573,
+      "grad_norm": 1.7599469289461205,
+      "learning_rate": 4.944772278379609e-06,
+      "loss": 0.9656,
+      "step": 4931
+    },
+    {
+      "epoch": 0.6787778695293146,
+      "grad_norm": 1.7408539388808886,
+      "learning_rate": 4.94092685589583e-06,
+      "loss": 0.9336,
+      "step": 4932
+    },
+    {
+      "epoch": 0.678915496834572,
+      "grad_norm": 1.7822578431839784,
+      "learning_rate": 4.937082438580632e-06,
+      "loss": 0.887,
+      "step": 4933
+    },
+    {
+      "epoch": 0.6790531241398293,
+      "grad_norm": 1.5934620256104077,
+      "learning_rate": 4.9332390271978515e-06,
+      "loss": 0.9021,
+      "step": 4934
+    },
+    {
+      "epoch": 0.6791907514450867,
+      "grad_norm": 1.6045986034569957,
+      "learning_rate": 4.929396622511119e-06,
+      "loss": 0.9195,
+      "step": 4935
+    },
+    {
+      "epoch": 0.679328378750344,
+      "grad_norm": 1.8531309235422835,
+      "learning_rate": 4.925555225283865e-06,
+      "loss": 0.972,
+      "step": 4936
+    },
+    {
+      "epoch": 0.6794660060556015,
+      "grad_norm": 1.915395223416936,
+      "learning_rate": 4.9217148362793335e-06,
+      "loss": 0.9471,
+      "step": 4937
+    },
+    {
+      "epoch": 0.6796036333608588,
+      "grad_norm": 1.922853491302613,
+      "learning_rate": 4.917875456260542e-06,
+      "loss": 0.9584,
+      "step": 4938
+    },
+    {
+      "epoch": 0.6797412606661162,
+      "grad_norm": 1.6358694617553937,
+      "learning_rate": 4.914037085990339e-06,
+      "loss": 0.8697,
+      "step": 4939
+    },
+    {
+      "epoch": 0.6798788879713735,
+      "grad_norm": 1.7488284082949492,
+      "learning_rate": 4.910199726231339e-06,
+      "loss": 0.9605,
+      "step": 4940
+    },
+    {
+      "epoch": 0.6800165152766309,
+      "grad_norm": 1.7174729768910724,
+      "learning_rate": 4.906363377745991e-06,
+      "loss": 0.9204,
+      "step": 4941
+    },
+    {
+      "epoch": 0.6801541425818882,
+      "grad_norm": 1.869540283414227,
+      "learning_rate": 4.902528041296506e-06,
+      "loss": 0.9735,
+      "step": 4942
+    },
+    {
+      "epoch": 0.6802917698871456,
+      "grad_norm": 1.7786879702782017,
+      "learning_rate": 4.898693717644928e-06,
+      "loss": 1.033,
+      "step": 4943
+    },
+    {
+      "epoch": 0.6804293971924029,
+      "grad_norm": 1.7560619812285323,
+      "learning_rate": 4.894860407553082e-06,
+      "loss": 0.9593,
+      "step": 4944
+    },
+    {
+      "epoch": 0.6805670244976604,
+      "grad_norm": 1.852669501691022,
+      "learning_rate": 4.891028111782583e-06,
+      "loss": 0.9441,
+      "step": 4945
+    },
+    {
+      "epoch": 0.6807046518029177,
+      "grad_norm": 1.7898855348291138,
+      "learning_rate": 4.887196831094872e-06,
+      "loss": 0.9291,
+      "step": 4946
+    },
+    {
+      "epoch": 0.6808422791081751,
+      "grad_norm": 1.7644383690418188,
+      "learning_rate": 4.883366566251158e-06,
+      "loss": 0.9396,
+      "step": 4947
+    },
+    {
+      "epoch": 0.6809799064134324,
+      "grad_norm": 1.7808462557909008,
+      "learning_rate": 4.879537318012474e-06,
+      "loss": 0.9524,
+      "step": 4948
+    },
+    {
+      "epoch": 0.6811175337186898,
+      "grad_norm": 1.7009945488856582,
+      "learning_rate": 4.875709087139626e-06,
+      "loss": 0.9261,
+      "step": 4949
+    },
+    {
+      "epoch": 0.6812551610239471,
+      "grad_norm": 1.7448583202706687,
+      "learning_rate": 4.871881874393243e-06,
+      "loss": 0.9503,
+      "step": 4950
+    },
+    {
+      "epoch": 0.6813927883292045,
+      "grad_norm": 1.6337230405515122,
+      "learning_rate": 4.868055680533734e-06,
+      "loss": 0.8795,
+      "step": 4951
+    },
+    {
+      "epoch": 0.6815304156344619,
+      "grad_norm": 1.8367386551234477,
+      "learning_rate": 4.864230506321314e-06,
+      "loss": 0.8959,
+      "step": 4952
+    },
+    {
+      "epoch": 0.6816680429397193,
+      "grad_norm": 1.725004620380599,
+      "learning_rate": 4.860406352515991e-06,
+      "loss": 0.9269,
+      "step": 4953
+    },
+    {
+      "epoch": 0.6818056702449766,
+      "grad_norm": 1.8678371795404125,
+      "learning_rate": 4.856583219877572e-06,
+      "loss": 0.9849,
+      "step": 4954
+    },
+    {
+      "epoch": 0.681943297550234,
+      "grad_norm": 1.623119185277133,
+      "learning_rate": 4.852761109165662e-06,
+      "loss": 0.9306,
+      "step": 4955
+    },
+    {
+      "epoch": 0.6820809248554913,
+      "grad_norm": 1.8623989488170152,
+      "learning_rate": 4.848940021139661e-06,
+      "loss": 0.9511,
+      "step": 4956
+    },
+    {
+      "epoch": 0.6822185521607487,
+      "grad_norm": 1.611358420717568,
+      "learning_rate": 4.845119956558767e-06,
+      "loss": 0.8748,
+      "step": 4957
+    },
+    {
+      "epoch": 0.682356179466006,
+      "grad_norm": 1.595792769997048,
+      "learning_rate": 4.841300916181977e-06,
+      "loss": 0.8806,
+      "step": 4958
+    },
+    {
+      "epoch": 0.6824938067712634,
+      "grad_norm": 1.770378644564124,
+      "learning_rate": 4.837482900768078e-06,
+      "loss": 0.8901,
+      "step": 4959
+    },
+    {
+      "epoch": 0.6826314340765208,
+      "grad_norm": 2.05373699386087,
+      "learning_rate": 4.83366591107566e-06,
+      "loss": 0.9358,
+      "step": 4960
+    },
+    {
+      "epoch": 0.6827690613817782,
+      "grad_norm": 2.0458388177753055,
+      "learning_rate": 4.829849947863107e-06,
+      "loss": 0.9092,
+      "step": 4961
+    },
+    {
+      "epoch": 0.6829066886870355,
+      "grad_norm": 1.6861740382248513,
+      "learning_rate": 4.826035011888597e-06,
+      "loss": 0.9347,
+      "step": 4962
+    },
+    {
+      "epoch": 0.6830443159922929,
+      "grad_norm": 1.7151779799031506,
+      "learning_rate": 4.822221103910107e-06,
+      "loss": 0.8523,
+      "step": 4963
+    },
+    {
+      "epoch": 0.6831819432975502,
+      "grad_norm": 1.64012526341682,
+      "learning_rate": 4.818408224685406e-06,
+      "loss": 0.8926,
+      "step": 4964
+    },
+    {
+      "epoch": 0.6833195706028076,
+      "grad_norm": 1.7226300292390875,
+      "learning_rate": 4.814596374972064e-06,
+      "loss": 0.9122,
+      "step": 4965
+    },
+    {
+      "epoch": 0.6834571979080649,
+      "grad_norm": 1.9547004259546499,
+      "learning_rate": 4.810785555527441e-06,
+      "loss": 0.9161,
+      "step": 4966
+    },
+    {
+      "epoch": 0.6835948252133224,
+      "grad_norm": 1.9039369685779217,
+      "learning_rate": 4.806975767108696e-06,
+      "loss": 1.005,
+      "step": 4967
+    },
+    {
+      "epoch": 0.6837324525185797,
+      "grad_norm": 2.6622065196069027,
+      "learning_rate": 4.803167010472777e-06,
+      "loss": 0.924,
+      "step": 4968
+    },
+    {
+      "epoch": 0.6838700798238371,
+      "grad_norm": 1.566835386579845,
+      "learning_rate": 4.7993592863764435e-06,
+      "loss": 0.9378,
+      "step": 4969
+    },
+    {
+      "epoch": 0.6840077071290944,
+      "grad_norm": 1.8946839165839233,
+      "learning_rate": 4.795552595576224e-06,
+      "loss": 0.9778,
+      "step": 4970
+    },
+    {
+      "epoch": 0.6841453344343518,
+      "grad_norm": 1.9449562192240912,
+      "learning_rate": 4.79174693882847e-06,
+      "loss": 0.9486,
+      "step": 4971
+    },
+    {
+      "epoch": 0.6842829617396091,
+      "grad_norm": 2.249459941137504,
+      "learning_rate": 4.787942316889299e-06,
+      "loss": 0.9884,
+      "step": 4972
+    },
+    {
+      "epoch": 0.6844205890448665,
+      "grad_norm": 1.8506237219140045,
+      "learning_rate": 4.7841387305146515e-06,
+      "loss": 0.9586,
+      "step": 4973
+    },
+    {
+      "epoch": 0.6845582163501238,
+      "grad_norm": 1.8264772842282788,
+      "learning_rate": 4.780336180460232e-06,
+      "loss": 0.8529,
+      "step": 4974
+    },
+    {
+      "epoch": 0.6846958436553813,
+      "grad_norm": 1.741485205526366,
+      "learning_rate": 4.776534667481569e-06,
+      "loss": 0.8661,
+      "step": 4975
+    },
+    {
+      "epoch": 0.6848334709606386,
+      "grad_norm": 2.141915861588948,
+      "learning_rate": 4.772734192333971e-06,
+      "loss": 0.9758,
+      "step": 4976
+    },
+    {
+      "epoch": 0.684971098265896,
+      "grad_norm": 1.899312132190996,
+      "learning_rate": 4.768934755772526e-06,
+      "loss": 0.8742,
+      "step": 4977
+    },
+    {
+      "epoch": 0.6851087255711533,
+      "grad_norm": 1.683148728759335,
+      "learning_rate": 4.765136358552148e-06,
+      "loss": 0.8962,
+      "step": 4978
+    },
+    {
+      "epoch": 0.6852463528764107,
+      "grad_norm": 2.29957107817339,
+      "learning_rate": 4.761339001427511e-06,
+      "loss": 0.9208,
+      "step": 4979
+    },
+    {
+      "epoch": 0.685383980181668,
+      "grad_norm": 1.6102643760526887,
+      "learning_rate": 4.757542685153112e-06,
+      "loss": 0.9178,
+      "step": 4980
+    },
+    {
+      "epoch": 0.6855216074869254,
+      "grad_norm": 1.8786862723461684,
+      "learning_rate": 4.7537474104832125e-06,
+      "loss": 0.8999,
+      "step": 4981
+    },
+    {
+      "epoch": 0.6856592347921827,
+      "grad_norm": 1.7206997719140997,
+      "learning_rate": 4.749953178171892e-06,
+      "loss": 0.8343,
+      "step": 4982
+    },
+    {
+      "epoch": 0.6857968620974402,
+      "grad_norm": 1.8596850690952889,
+      "learning_rate": 4.746159988973012e-06,
+      "loss": 0.9411,
+      "step": 4983
+    },
+    {
+      "epoch": 0.6859344894026975,
+      "grad_norm": 2.1515285955309347,
+      "learning_rate": 4.742367843640223e-06,
+      "loss": 0.9138,
+      "step": 4984
+    },
+    {
+      "epoch": 0.6860721167079549,
+      "grad_norm": 2.028593179935735,
+      "learning_rate": 4.738576742926975e-06,
+      "loss": 1.0088,
+      "step": 4985
+    },
+    {
+      "epoch": 0.6862097440132122,
+      "grad_norm": 1.76748842799167,
+      "learning_rate": 4.7347866875865064e-06,
+      "loss": 0.9316,
+      "step": 4986
+    },
+    {
+      "epoch": 0.6863473713184696,
+      "grad_norm": 2.0105651170064993,
+      "learning_rate": 4.73099767837185e-06,
+      "loss": 0.9381,
+      "step": 4987
+    },
+    {
+      "epoch": 0.6864849986237269,
+      "grad_norm": 1.6086823834985942,
+      "learning_rate": 4.727209716035832e-06,
+      "loss": 0.8929,
+      "step": 4988
+    },
+    {
+      "epoch": 0.6866226259289843,
+      "grad_norm": 1.9198759077251246,
+      "learning_rate": 4.723422801331066e-06,
+      "loss": 0.9064,
+      "step": 4989
+    },
+    {
+      "epoch": 0.6867602532342417,
+      "grad_norm": 1.6752439003143222,
+      "learning_rate": 4.719636935009961e-06,
+      "loss": 0.8506,
+      "step": 4990
+    },
+    {
+      "epoch": 0.6868978805394991,
+      "grad_norm": 1.689782379137984,
+      "learning_rate": 4.715852117824715e-06,
+      "loss": 0.9439,
+      "step": 4991
+    },
+    {
+      "epoch": 0.6870355078447564,
+      "grad_norm": 3.4026186438204697,
+      "learning_rate": 4.712068350527322e-06,
+      "loss": 0.945,
+      "step": 4992
+    },
+    {
+      "epoch": 0.6871731351500138,
+      "grad_norm": 2.0516754735560943,
+      "learning_rate": 4.708285633869565e-06,
+      "loss": 0.967,
+      "step": 4993
+    },
+    {
+      "epoch": 0.6873107624552711,
+      "grad_norm": 1.7024857442643633,
+      "learning_rate": 4.704503968603016e-06,
+      "loss": 0.8804,
+      "step": 4994
+    },
+    {
+      "epoch": 0.6874483897605285,
+      "grad_norm": 1.8275348483722258,
+      "learning_rate": 4.70072335547904e-06,
+      "loss": 0.9127,
+      "step": 4995
+    },
+    {
+      "epoch": 0.6875860170657858,
+      "grad_norm": 1.8199309236566261,
+      "learning_rate": 4.696943795248795e-06,
+      "loss": 0.9535,
+      "step": 4996
+    },
+    {
+      "epoch": 0.6877236443710432,
+      "grad_norm": 1.7490506924266702,
+      "learning_rate": 4.693165288663227e-06,
+      "loss": 0.9471,
+      "step": 4997
+    },
+    {
+      "epoch": 0.6878612716763006,
+      "grad_norm": 1.5475710490741603,
+      "learning_rate": 4.68938783647307e-06,
+      "loss": 0.9254,
+      "step": 4998
+    },
+    {
+      "epoch": 0.687998898981558,
+      "grad_norm": 1.927495751280919,
+      "learning_rate": 4.685611439428863e-06,
+      "loss": 0.9643,
+      "step": 4999
+    },
+    {
+      "epoch": 0.6881365262868153,
+      "grad_norm": 1.996139310903076,
+      "learning_rate": 4.68183609828091e-06,
+      "loss": 0.9565,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6882741535920727,
+      "grad_norm": 1.759831811215478,
+      "learning_rate": 4.6780618137793334e-06,
+      "loss": 0.935,
+      "step": 5001
+    },
+    {
+      "epoch": 0.68841178089733,
+      "grad_norm": 1.9466430789924138,
+      "learning_rate": 4.6742885866740165e-06,
+      "loss": 0.8576,
+      "step": 5002
+    },
+    {
+      "epoch": 0.6885494082025874,
+      "grad_norm": 1.6898647834180371,
+      "learning_rate": 4.6705164177146656e-06,
+      "loss": 0.968,
+      "step": 5003
+    },
+    {
+      "epoch": 0.6886870355078447,
+      "grad_norm": 1.8551985436705014,
+      "learning_rate": 4.6667453076507405e-06,
+      "loss": 0.964,
+      "step": 5004
+    },
+    {
+      "epoch": 0.6888246628131022,
+      "grad_norm": 1.5802135984593486,
+      "learning_rate": 4.662975257231523e-06,
+      "loss": 0.9254,
+      "step": 5005
+    },
+    {
+      "epoch": 0.6889622901183595,
+      "grad_norm": 1.845109994576382,
+      "learning_rate": 4.659206267206066e-06,
+      "loss": 0.973,
+      "step": 5006
+    },
+    {
+      "epoch": 0.6890999174236169,
+      "grad_norm": 1.9123954312094356,
+      "learning_rate": 4.655438338323217e-06,
+      "loss": 0.9074,
+      "step": 5007
+    },
+    {
+      "epoch": 0.6892375447288742,
+      "grad_norm": 1.688824307352302,
+      "learning_rate": 4.6516714713316145e-06,
+      "loss": 0.9427,
+      "step": 5008
+    },
+    {
+      "epoch": 0.6893751720341316,
+      "grad_norm": 1.7489037958701934,
+      "learning_rate": 4.6479056669796715e-06,
+      "loss": 0.9485,
+      "step": 5009
+    },
+    {
+      "epoch": 0.6895127993393889,
+      "grad_norm": 1.6149190711416013,
+      "learning_rate": 4.644140926015619e-06,
+      "loss": 0.8967,
+      "step": 5010
+    },
+    {
+      "epoch": 0.6896504266446463,
+      "grad_norm": 1.6458682322280005,
+      "learning_rate": 4.640377249187443e-06,
+      "loss": 0.9403,
+      "step": 5011
+    },
+    {
+      "epoch": 0.6897880539499036,
+      "grad_norm": 1.557411567682636,
+      "learning_rate": 4.636614637242946e-06,
+      "loss": 0.9067,
+      "step": 5012
+    },
+    {
+      "epoch": 0.6899256812551611,
+      "grad_norm": 1.7229437615335952,
+      "learning_rate": 4.632853090929704e-06,
+      "loss": 0.9069,
+      "step": 5013
+    },
+    {
+      "epoch": 0.6900633085604184,
+      "grad_norm": 1.6347499584680492,
+      "learning_rate": 4.629092610995084e-06,
+      "loss": 0.99,
+      "step": 5014
+    },
+    {
+      "epoch": 0.6902009358656758,
+      "grad_norm": 1.7735717243426252,
+      "learning_rate": 4.625333198186243e-06,
+      "loss": 0.9079,
+      "step": 5015
+    },
+    {
+      "epoch": 0.6903385631709331,
+      "grad_norm": 1.856301730144119,
+      "learning_rate": 4.621574853250125e-06,
+      "loss": 0.9557,
+      "step": 5016
+    },
+    {
+      "epoch": 0.6904761904761905,
+      "grad_norm": 1.749418650221802,
+      "learning_rate": 4.61781757693346e-06,
+      "loss": 0.9423,
+      "step": 5017
+    },
+    {
+      "epoch": 0.6906138177814478,
+      "grad_norm": 1.666314006955715,
+      "learning_rate": 4.614061369982769e-06,
+      "loss": 0.9135,
+      "step": 5018
+    },
+    {
+      "epoch": 0.6907514450867052,
+      "grad_norm": 1.9457679624688526,
+      "learning_rate": 4.610306233144357e-06,
+      "loss": 0.9279,
+      "step": 5019
+    },
+    {
+      "epoch": 0.6908890723919625,
+      "grad_norm": 2.1120479310881137,
+      "learning_rate": 4.606552167164321e-06,
+      "loss": 0.9483,
+      "step": 5020
+    },
+    {
+      "epoch": 0.69102669969722,
+      "grad_norm": 1.8082121175251815,
+      "learning_rate": 4.60279917278854e-06,
+      "loss": 0.8935,
+      "step": 5021
+    },
+    {
+      "epoch": 0.6911643270024773,
+      "grad_norm": 1.627884443410556,
+      "learning_rate": 4.599047250762685e-06,
+      "loss": 0.885,
+      "step": 5022
+    },
+    {
+      "epoch": 0.6913019543077347,
+      "grad_norm": 1.7698937400070622,
+      "learning_rate": 4.5952964018322055e-06,
+      "loss": 0.9503,
+      "step": 5023
+    },
+    {
+      "epoch": 0.691439581612992,
+      "grad_norm": 1.7996573663514066,
+      "learning_rate": 4.591546626742357e-06,
+      "loss": 0.9461,
+      "step": 5024
+    },
+    {
+      "epoch": 0.6915772089182494,
+      "grad_norm": 2.011102894061317,
+      "learning_rate": 4.587797926238156e-06,
+      "loss": 0.926,
+      "step": 5025
+    },
+    {
+      "epoch": 0.6917148362235067,
+      "grad_norm": 1.7717564800851904,
+      "learning_rate": 4.584050301064422e-06,
+      "loss": 0.9995,
+      "step": 5026
+    },
+    {
+      "epoch": 0.6918524635287641,
+      "grad_norm": 1.71233771187298,
+      "learning_rate": 4.580303751965757e-06,
+      "loss": 0.8939,
+      "step": 5027
+    },
+    {
+      "epoch": 0.6919900908340215,
+      "grad_norm": 1.6034244274634826,
+      "learning_rate": 4.576558279686549e-06,
+      "loss": 0.9052,
+      "step": 5028
+    },
+    {
+      "epoch": 0.6921277181392789,
+      "grad_norm": 1.5799361891399502,
+      "learning_rate": 4.572813884970971e-06,
+      "loss": 0.8869,
+      "step": 5029
+    },
+    {
+      "epoch": 0.6922653454445362,
+      "grad_norm": 1.7128902298303348,
+      "learning_rate": 4.569070568562982e-06,
+      "loss": 0.9527,
+      "step": 5030
+    },
+    {
+      "epoch": 0.6924029727497936,
+      "grad_norm": 1.828669881167549,
+      "learning_rate": 4.565328331206336e-06,
+      "loss": 0.8584,
+      "step": 5031
+    },
+    {
+      "epoch": 0.6925406000550509,
+      "grad_norm": 2.002968598115233,
+      "learning_rate": 4.56158717364455e-06,
+      "loss": 0.8489,
+      "step": 5032
+    },
+    {
+      "epoch": 0.6926782273603083,
+      "grad_norm": 1.7867045368787395,
+      "learning_rate": 4.5578470966209585e-06,
+      "loss": 0.9316,
+      "step": 5033
+    },
+    {
+      "epoch": 0.6928158546655656,
+      "grad_norm": 1.9342150930189863,
+      "learning_rate": 4.5541081008786445e-06,
+      "loss": 0.9264,
+      "step": 5034
+    },
+    {
+      "epoch": 0.692953481970823,
+      "grad_norm": 1.6393870728396391,
+      "learning_rate": 4.550370187160514e-06,
+      "loss": 0.9125,
+      "step": 5035
+    },
+    {
+      "epoch": 0.6930911092760804,
+      "grad_norm": 1.6411814818934647,
+      "learning_rate": 4.54663335620922e-06,
+      "loss": 0.9215,
+      "step": 5036
+    },
+    {
+      "epoch": 0.6932287365813378,
+      "grad_norm": 1.8609114076736357,
+      "learning_rate": 4.542897608767235e-06,
+      "loss": 0.954,
+      "step": 5037
+    },
+    {
+      "epoch": 0.6933663638865951,
+      "grad_norm": 1.9402138704474456,
+      "learning_rate": 4.539162945576796e-06,
+      "loss": 0.8912,
+      "step": 5038
+    },
+    {
+      "epoch": 0.6935039911918525,
+      "grad_norm": 1.7212138435155977,
+      "learning_rate": 4.5354293673799275e-06,
+      "loss": 0.9657,
+      "step": 5039
+    },
+    {
+      "epoch": 0.6936416184971098,
+      "grad_norm": 1.9488583388004803,
+      "learning_rate": 4.531696874918447e-06,
+      "loss": 0.9979,
+      "step": 5040
+    },
+    {
+      "epoch": 0.6937792458023672,
+      "grad_norm": 1.9733601395233151,
+      "learning_rate": 4.527965468933936e-06,
+      "loss": 0.9101,
+      "step": 5041
+    },
+    {
+      "epoch": 0.6939168731076245,
+      "grad_norm": 1.6537495669445832,
+      "learning_rate": 4.524235150167791e-06,
+      "loss": 0.9451,
+      "step": 5042
+    },
+    {
+      "epoch": 0.694054500412882,
+      "grad_norm": 1.9992087087807289,
+      "learning_rate": 4.520505919361158e-06,
+      "loss": 0.8664,
+      "step": 5043
+    },
+    {
+      "epoch": 0.6941921277181393,
+      "grad_norm": 1.707762027041643,
+      "learning_rate": 4.516777777254997e-06,
+      "loss": 0.9105,
+      "step": 5044
+    },
+    {
+      "epoch": 0.6943297550233967,
+      "grad_norm": 1.6470742170830577,
+      "learning_rate": 4.513050724590034e-06,
+      "loss": 0.9396,
+      "step": 5045
+    },
+    {
+      "epoch": 0.694467382328654,
+      "grad_norm": 1.5505674665324045,
+      "learning_rate": 4.509324762106785e-06,
+      "loss": 0.9758,
+      "step": 5046
+    },
+    {
+      "epoch": 0.6946050096339114,
+      "grad_norm": 1.689757768716327,
+      "learning_rate": 4.505599890545546e-06,
+      "loss": 0.8687,
+      "step": 5047
+    },
+    {
+      "epoch": 0.6947426369391687,
+      "grad_norm": 1.750502030657462,
+      "learning_rate": 4.501876110646398e-06,
+      "loss": 0.9558,
+      "step": 5048
+    },
+    {
+      "epoch": 0.6948802642444261,
+      "grad_norm": 1.6881035053459295,
+      "learning_rate": 4.498153423149206e-06,
+      "loss": 0.8956,
+      "step": 5049
+    },
+    {
+      "epoch": 0.6950178915496834,
+      "grad_norm": 1.600464187255,
+      "learning_rate": 4.494431828793617e-06,
+      "loss": 0.9392,
+      "step": 5050
+    },
+    {
+      "epoch": 0.6951555188549409,
+      "grad_norm": 1.8454503549916428,
+      "learning_rate": 4.490711328319061e-06,
+      "loss": 0.8438,
+      "step": 5051
+    },
+    {
+      "epoch": 0.6952931461601982,
+      "grad_norm": 1.7924179949631098,
+      "learning_rate": 4.48699192246475e-06,
+      "loss": 0.8354,
+      "step": 5052
+    },
+    {
+      "epoch": 0.6954307734654556,
+      "grad_norm": 2.7429979845297585,
+      "learning_rate": 4.483273611969678e-06,
+      "loss": 0.9427,
+      "step": 5053
+    },
+    {
+      "epoch": 0.6955684007707129,
+      "grad_norm": 1.80973788396394,
+      "learning_rate": 4.479556397572625e-06,
+      "loss": 0.9501,
+      "step": 5054
+    },
+    {
+      "epoch": 0.6957060280759703,
+      "grad_norm": 1.7180215191928057,
+      "learning_rate": 4.475840280012145e-06,
+      "loss": 0.8818,
+      "step": 5055
+    },
+    {
+      "epoch": 0.6958436553812276,
+      "grad_norm": 1.757184681171232,
+      "learning_rate": 4.4721252600265906e-06,
+      "loss": 0.8963,
+      "step": 5056
+    },
+    {
+      "epoch": 0.695981282686485,
+      "grad_norm": 2.0366286461346457,
+      "learning_rate": 4.468411338354076e-06,
+      "loss": 0.9522,
+      "step": 5057
+    },
+    {
+      "epoch": 0.6961189099917423,
+      "grad_norm": 1.6238520625772739,
+      "learning_rate": 4.464698515732509e-06,
+      "loss": 0.9386,
+      "step": 5058
+    },
+    {
+      "epoch": 0.6962565372969998,
+      "grad_norm": 1.6463525888851822,
+      "learning_rate": 4.460986792899577e-06,
+      "loss": 0.8367,
+      "step": 5059
+    },
+    {
+      "epoch": 0.6963941646022571,
+      "grad_norm": 1.8406637374589985,
+      "learning_rate": 4.457276170592747e-06,
+      "loss": 0.8345,
+      "step": 5060
+    },
+    {
+      "epoch": 0.6965317919075145,
+      "grad_norm": 1.8972674833865015,
+      "learning_rate": 4.453566649549272e-06,
+      "loss": 0.9294,
+      "step": 5061
+    },
+    {
+      "epoch": 0.6966694192127718,
+      "grad_norm": 2.0991956045004225,
+      "learning_rate": 4.4498582305061766e-06,
+      "loss": 0.864,
+      "step": 5062
+    },
+    {
+      "epoch": 0.6968070465180292,
+      "grad_norm": 1.770954857217113,
+      "learning_rate": 4.4461509142002855e-06,
+      "loss": 0.9226,
+      "step": 5063
+    },
+    {
+      "epoch": 0.6969446738232865,
+      "grad_norm": 1.6099766852750965,
+      "learning_rate": 4.442444701368176e-06,
+      "loss": 0.8943,
+      "step": 5064
+    },
+    {
+      "epoch": 0.6970823011285439,
+      "grad_norm": 1.8417392691570047,
+      "learning_rate": 4.4387395927462386e-06,
+      "loss": 0.9849,
+      "step": 5065
+    },
+    {
+      "epoch": 0.6972199284338013,
+      "grad_norm": 1.735280765924145,
+      "learning_rate": 4.43503558907061e-06,
+      "loss": 1.0523,
+      "step": 5066
+    },
+    {
+      "epoch": 0.6973575557390587,
+      "grad_norm": 1.7382444912259905,
+      "learning_rate": 4.431332691077237e-06,
+      "loss": 0.8983,
+      "step": 5067
+    },
+    {
+      "epoch": 0.697495183044316,
+      "grad_norm": 1.981834560500647,
+      "learning_rate": 4.4276308995018324e-06,
+      "loss": 0.8794,
+      "step": 5068
+    },
+    {
+      "epoch": 0.6976328103495734,
+      "grad_norm": 1.7853412053778503,
+      "learning_rate": 4.423930215079891e-06,
+      "loss": 0.8786,
+      "step": 5069
+    },
+    {
+      "epoch": 0.6977704376548307,
+      "grad_norm": 1.8210074861693657,
+      "learning_rate": 4.420230638546686e-06,
+      "loss": 0.9105,
+      "step": 5070
+    },
+    {
+      "epoch": 0.6979080649600881,
+      "grad_norm": 1.6171687846001375,
+      "learning_rate": 4.416532170637277e-06,
+      "loss": 0.8906,
+      "step": 5071
+    },
+    {
+      "epoch": 0.6980456922653454,
+      "grad_norm": 1.5679187883466783,
+      "learning_rate": 4.412834812086498e-06,
+      "loss": 0.8864,
+      "step": 5072
+    },
+    {
+      "epoch": 0.6981833195706028,
+      "grad_norm": 1.7711300429214918,
+      "learning_rate": 4.409138563628955e-06,
+      "loss": 0.9302,
+      "step": 5073
+    },
+    {
+      "epoch": 0.6983209468758602,
+      "grad_norm": 1.6338100080453581,
+      "learning_rate": 4.405443425999053e-06,
+      "loss": 0.9209,
+      "step": 5074
+    },
+    {
+      "epoch": 0.6984585741811176,
+      "grad_norm": 1.8470469128309877,
+      "learning_rate": 4.401749399930961e-06,
+      "loss": 0.9292,
+      "step": 5075
+    },
+    {
+      "epoch": 0.6985962014863749,
+      "grad_norm": 1.7601740652576645,
+      "learning_rate": 4.398056486158632e-06,
+      "loss": 0.9023,
+      "step": 5076
+    },
+    {
+      "epoch": 0.6987338287916323,
+      "grad_norm": 1.9813217669251713,
+      "learning_rate": 4.394364685415796e-06,
+      "loss": 0.9543,
+      "step": 5077
+    },
+    {
+      "epoch": 0.6988714560968896,
+      "grad_norm": 1.7073496724290795,
+      "learning_rate": 4.3906739984359645e-06,
+      "loss": 0.8983,
+      "step": 5078
+    },
+    {
+      "epoch": 0.699009083402147,
+      "grad_norm": 3.2418186032007266,
+      "learning_rate": 4.386984425952426e-06,
+      "loss": 0.9684,
+      "step": 5079
+    },
+    {
+      "epoch": 0.6991467107074043,
+      "grad_norm": 1.5655246795437379,
+      "learning_rate": 4.383295968698249e-06,
+      "loss": 0.9298,
+      "step": 5080
+    },
+    {
+      "epoch": 0.6992843380126618,
+      "grad_norm": 1.6625161097264447,
+      "learning_rate": 4.379608627406279e-06,
+      "loss": 0.9324,
+      "step": 5081
+    },
+    {
+      "epoch": 0.6994219653179191,
+      "grad_norm": 1.6955130966414405,
+      "learning_rate": 4.375922402809139e-06,
+      "loss": 0.9216,
+      "step": 5082
+    },
+    {
+      "epoch": 0.6995595926231765,
+      "grad_norm": 2.004391818743147,
+      "learning_rate": 4.372237295639233e-06,
+      "loss": 1.0068,
+      "step": 5083
+    },
+    {
+      "epoch": 0.6996972199284338,
+      "grad_norm": 1.635449374095777,
+      "learning_rate": 4.368553306628739e-06,
+      "loss": 0.8396,
+      "step": 5084
+    },
+    {
+      "epoch": 0.6998348472336912,
+      "grad_norm": 2.6487348985895145,
+      "learning_rate": 4.364870436509615e-06,
+      "loss": 0.8152,
+      "step": 5085
+    },
+    {
+      "epoch": 0.6999724745389485,
+      "grad_norm": 1.5964588132176627,
+      "learning_rate": 4.3611886860136056e-06,
+      "loss": 0.9415,
+      "step": 5086
+    },
+    {
+      "epoch": 0.7001101018442059,
+      "grad_norm": 1.578979348422104,
+      "learning_rate": 4.35750805587221e-06,
+      "loss": 0.992,
+      "step": 5087
+    },
+    {
+      "epoch": 0.7002477291494632,
+      "grad_norm": 1.982151703279505,
+      "learning_rate": 4.353828546816733e-06,
+      "loss": 0.846,
+      "step": 5088
+    },
+    {
+      "epoch": 0.7003853564547207,
+      "grad_norm": 1.9971555101594234,
+      "learning_rate": 4.350150159578233e-06,
+      "loss": 0.9517,
+      "step": 5089
+    },
+    {
+      "epoch": 0.700522983759978,
+      "grad_norm": 1.8198494209290095,
+      "learning_rate": 4.346472894887559e-06,
+      "loss": 0.9937,
+      "step": 5090
+    },
+    {
+      "epoch": 0.7006606110652354,
+      "grad_norm": 1.776615992516107,
+      "learning_rate": 4.342796753475333e-06,
+      "loss": 0.9639,
+      "step": 5091
+    },
+    {
+      "epoch": 0.7007982383704927,
+      "grad_norm": 1.9055024267091587,
+      "learning_rate": 4.3391217360719476e-06,
+      "loss": 0.9868,
+      "step": 5092
+    },
+    {
+      "epoch": 0.7009358656757501,
+      "grad_norm": 1.8910497045944668,
+      "learning_rate": 4.335447843407592e-06,
+      "loss": 1.039,
+      "step": 5093
+    },
+    {
+      "epoch": 0.7010734929810074,
+      "grad_norm": 2.380984242710699,
+      "learning_rate": 4.331775076212205e-06,
+      "loss": 0.9229,
+      "step": 5094
+    },
+    {
+      "epoch": 0.7012111202862648,
+      "grad_norm": 1.689714486241484,
+      "learning_rate": 4.328103435215526e-06,
+      "loss": 0.9378,
+      "step": 5095
+    },
+    {
+      "epoch": 0.7013487475915221,
+      "grad_norm": 1.7371153469063594,
+      "learning_rate": 4.324432921147048e-06,
+      "loss": 0.9182,
+      "step": 5096
+    },
+    {
+      "epoch": 0.7014863748967796,
+      "grad_norm": 1.421112508405305,
+      "learning_rate": 4.320763534736067e-06,
+      "loss": 0.8385,
+      "step": 5097
+    },
+    {
+      "epoch": 0.7016240022020369,
+      "grad_norm": 1.7944259310051087,
+      "learning_rate": 4.317095276711623e-06,
+      "loss": 1.0481,
+      "step": 5098
+    },
+    {
+      "epoch": 0.7017616295072943,
+      "grad_norm": 1.594604658905283,
+      "learning_rate": 4.313428147802561e-06,
+      "loss": 0.8667,
+      "step": 5099
+    },
+    {
+      "epoch": 0.7018992568125516,
+      "grad_norm": 1.9431659822182392,
+      "learning_rate": 4.3097621487374855e-06,
+      "loss": 0.9655,
+      "step": 5100
+    },
+    {
+      "epoch": 0.702036884117809,
+      "grad_norm": 1.8322117271678973,
+      "learning_rate": 4.3060972802447794e-06,
+      "loss": 0.8521,
+      "step": 5101
+    },
+    {
+      "epoch": 0.7021745114230663,
+      "grad_norm": 1.6653951450820863,
+      "learning_rate": 4.302433543052603e-06,
+      "loss": 0.9038,
+      "step": 5102
+    },
+    {
+      "epoch": 0.7023121387283237,
+      "grad_norm": 1.6769778864380802,
+      "learning_rate": 4.298770937888891e-06,
+      "loss": 0.9502,
+      "step": 5103
+    },
+    {
+      "epoch": 0.7024497660335811,
+      "grad_norm": 1.7461942484728894,
+      "learning_rate": 4.295109465481355e-06,
+      "loss": 0.9513,
+      "step": 5104
+    },
+    {
+      "epoch": 0.7025873933388385,
+      "grad_norm": 1.7661378265176921,
+      "learning_rate": 4.291449126557471e-06,
+      "loss": 0.9389,
+      "step": 5105
+    },
+    {
+      "epoch": 0.7027250206440958,
+      "grad_norm": 2.195692478834936,
+      "learning_rate": 4.287789921844506e-06,
+      "loss": 0.884,
+      "step": 5106
+    },
+    {
+      "epoch": 0.7028626479493532,
+      "grad_norm": 1.7561965044384864,
+      "learning_rate": 4.284131852069493e-06,
+      "loss": 0.9098,
+      "step": 5107
+    },
+    {
+      "epoch": 0.7030002752546105,
+      "grad_norm": 1.9272378152556409,
+      "learning_rate": 4.280474917959239e-06,
+      "loss": 0.9392,
+      "step": 5108
+    },
+    {
+      "epoch": 0.7031379025598679,
+      "grad_norm": 1.8936805287275464,
+      "learning_rate": 4.276819120240326e-06,
+      "loss": 0.8729,
+      "step": 5109
+    },
+    {
+      "epoch": 0.7032755298651252,
+      "grad_norm": 1.6259954911558936,
+      "learning_rate": 4.273164459639114e-06,
+      "loss": 0.875,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7034131571703826,
+      "grad_norm": 1.616431255779718,
+      "learning_rate": 4.269510936881731e-06,
+      "loss": 0.8197,
+      "step": 5111
+    },
+    {
+      "epoch": 0.70355078447564,
+      "grad_norm": 2.013099888258806,
+      "learning_rate": 4.265858552694082e-06,
+      "loss": 0.8914,
+      "step": 5112
+    },
+    {
+      "epoch": 0.7036884117808974,
+      "grad_norm": 2.0170750350652105,
+      "learning_rate": 4.262207307801849e-06,
+      "loss": 0.9402,
+      "step": 5113
+    },
+    {
+      "epoch": 0.7038260390861547,
+      "grad_norm": 1.7612946796947122,
+      "learning_rate": 4.25855720293048e-06,
+      "loss": 1.0239,
+      "step": 5114
+    },
+    {
+      "epoch": 0.7039636663914121,
+      "grad_norm": 1.7104142569718108,
+      "learning_rate": 4.254908238805204e-06,
+      "loss": 0.9417,
+      "step": 5115
+    },
+    {
+      "epoch": 0.7041012936966694,
+      "grad_norm": 2.0731302610217788,
+      "learning_rate": 4.251260416151018e-06,
+      "loss": 0.9292,
+      "step": 5116
+    },
+    {
+      "epoch": 0.7042389210019268,
+      "grad_norm": 2.029551884520438,
+      "learning_rate": 4.2476137356926935e-06,
+      "loss": 0.9174,
+      "step": 5117
+    },
+    {
+      "epoch": 0.7043765483071841,
+      "grad_norm": 1.6968039093407905,
+      "learning_rate": 4.2439681981547855e-06,
+      "loss": 0.9774,
+      "step": 5118
+    },
+    {
+      "epoch": 0.7045141756124416,
+      "grad_norm": 1.8179240033331696,
+      "learning_rate": 4.240323804261597e-06,
+      "loss": 0.8942,
+      "step": 5119
+    },
+    {
+      "epoch": 0.7046518029176989,
+      "grad_norm": 1.821617109513332,
+      "learning_rate": 4.236680554737235e-06,
+      "loss": 0.8086,
+      "step": 5120
+    },
+    {
+      "epoch": 0.7047894302229563,
+      "grad_norm": 1.7937915729278946,
+      "learning_rate": 4.233038450305551e-06,
+      "loss": 0.9329,
+      "step": 5121
+    },
+    {
+      "epoch": 0.7049270575282136,
+      "grad_norm": 1.9325743767502186,
+      "learning_rate": 4.229397491690187e-06,
+      "loss": 0.9061,
+      "step": 5122
+    },
+    {
+      "epoch": 0.705064684833471,
+      "grad_norm": 1.8430262209993065,
+      "learning_rate": 4.225757679614549e-06,
+      "loss": 0.9044,
+      "step": 5123
+    },
+    {
+      "epoch": 0.7052023121387283,
+      "grad_norm": 1.7578583148029037,
+      "learning_rate": 4.222119014801816e-06,
+      "loss": 0.9916,
+      "step": 5124
+    },
+    {
+      "epoch": 0.7053399394439857,
+      "grad_norm": 1.9375013779738766,
+      "learning_rate": 4.218481497974953e-06,
+      "loss": 0.8992,
+      "step": 5125
+    },
+    {
+      "epoch": 0.705477566749243,
+      "grad_norm": 1.579422166887886,
+      "learning_rate": 4.2148451298566675e-06,
+      "loss": 0.9285,
+      "step": 5126
+    },
+    {
+      "epoch": 0.7056151940545005,
+      "grad_norm": 1.709043378152004,
+      "learning_rate": 4.211209911169472e-06,
+      "loss": 0.9081,
+      "step": 5127
+    },
+    {
+      "epoch": 0.7057528213597578,
+      "grad_norm": 1.75980777070148,
+      "learning_rate": 4.207575842635621e-06,
+      "loss": 0.9395,
+      "step": 5128
+    },
+    {
+      "epoch": 0.7058904486650152,
+      "grad_norm": 1.867885763014054,
+      "learning_rate": 4.203942924977166e-06,
+      "loss": 0.934,
+      "step": 5129
+    },
+    {
+      "epoch": 0.7060280759702725,
+      "grad_norm": 1.6065633190259931,
+      "learning_rate": 4.200311158915906e-06,
+      "loss": 0.8635,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7061657032755299,
+      "grad_norm": 1.9430742321776113,
+      "learning_rate": 4.1966805451734326e-06,
+      "loss": 0.9337,
+      "step": 5131
+    },
+    {
+      "epoch": 0.7063033305807872,
+      "grad_norm": 1.899291306153583,
+      "learning_rate": 4.193051084471097e-06,
+      "loss": 0.8886,
+      "step": 5132
+    },
+    {
+      "epoch": 0.7064409578860446,
+      "grad_norm": 1.74447418673423,
+      "learning_rate": 4.189422777530022e-06,
+      "loss": 0.9348,
+      "step": 5133
+    },
+    {
+      "epoch": 0.7065785851913019,
+      "grad_norm": 1.5858299933739237,
+      "learning_rate": 4.185795625071103e-06,
+      "loss": 0.9333,
+      "step": 5134
+    },
+    {
+      "epoch": 0.7067162124965594,
+      "grad_norm": 4.57845572930728,
+      "learning_rate": 4.182169627815006e-06,
+      "loss": 0.9476,
+      "step": 5135
+    },
+    {
+      "epoch": 0.7068538398018167,
+      "grad_norm": 1.7944037692801242,
+      "learning_rate": 4.178544786482166e-06,
+      "loss": 0.965,
+      "step": 5136
+    },
+    {
+      "epoch": 0.7069914671070741,
+      "grad_norm": 2.1741245882885614,
+      "learning_rate": 4.17492110179279e-06,
+      "loss": 0.8316,
+      "step": 5137
+    },
+    {
+      "epoch": 0.7071290944123314,
+      "grad_norm": 1.740339624162474,
+      "learning_rate": 4.171298574466854e-06,
+      "loss": 0.839,
+      "step": 5138
+    },
+    {
+      "epoch": 0.7072667217175888,
+      "grad_norm": 2.118429588149702,
+      "learning_rate": 4.1676772052241056e-06,
+      "loss": 0.9695,
+      "step": 5139
+    },
+    {
+      "epoch": 0.7074043490228461,
+      "grad_norm": 1.9406191468978404,
+      "learning_rate": 4.16405699478406e-06,
+      "loss": 0.9821,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7075419763281035,
+      "grad_norm": 1.7216561690217567,
+      "learning_rate": 4.160437943866005e-06,
+      "loss": 0.8829,
+      "step": 5141
+    },
+    {
+      "epoch": 0.7076796036333609,
+      "grad_norm": 1.7945984824970234,
+      "learning_rate": 4.1568200531889955e-06,
+      "loss": 0.9487,
+      "step": 5142
+    },
+    {
+      "epoch": 0.7078172309386183,
+      "grad_norm": 1.7062678903133426,
+      "learning_rate": 4.153203323471857e-06,
+      "loss": 0.8869,
+      "step": 5143
+    },
+    {
+      "epoch": 0.7079548582438756,
+      "grad_norm": 1.5643552707532848,
+      "learning_rate": 4.149587755433184e-06,
+      "loss": 0.9421,
+      "step": 5144
+    },
+    {
+      "epoch": 0.708092485549133,
+      "grad_norm": 1.6653434254427788,
+      "learning_rate": 4.14597334979134e-06,
+      "loss": 0.8991,
+      "step": 5145
+    },
+    {
+      "epoch": 0.7082301128543903,
+      "grad_norm": 1.8347264388190192,
+      "learning_rate": 4.1423601072644595e-06,
+      "loss": 0.9462,
+      "step": 5146
+    },
+    {
+      "epoch": 0.7083677401596477,
+      "grad_norm": 1.6634482687510321,
+      "learning_rate": 4.1387480285704394e-06,
+      "loss": 0.9437,
+      "step": 5147
+    },
+    {
+      "epoch": 0.708505367464905,
+      "grad_norm": 1.9346315543403325,
+      "learning_rate": 4.1351371144269615e-06,
+      "loss": 0.8925,
+      "step": 5148
+    },
+    {
+      "epoch": 0.7086429947701623,
+      "grad_norm": 1.5991762264317202,
+      "learning_rate": 4.131527365551451e-06,
+      "loss": 1.0165,
+      "step": 5149
+    },
+    {
+      "epoch": 0.7087806220754198,
+      "grad_norm": 1.8558370226776015,
+      "learning_rate": 4.127918782661129e-06,
+      "loss": 0.9249,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7089182493806772,
+      "grad_norm": 1.780712934907872,
+      "learning_rate": 4.124311366472957e-06,
+      "loss": 0.9777,
+      "step": 5151
+    },
+    {
+      "epoch": 0.7090558766859345,
+      "grad_norm": 1.763501091453496,
+      "learning_rate": 4.120705117703695e-06,
+      "loss": 0.8792,
+      "step": 5152
+    },
+    {
+      "epoch": 0.7091935039911919,
+      "grad_norm": 1.7723132838551234,
+      "learning_rate": 4.117100037069843e-06,
+      "loss": 0.9014,
+      "step": 5153
+    },
+    {
+      "epoch": 0.7093311312964492,
+      "grad_norm": 1.7648997383276515,
+      "learning_rate": 4.113496125287684e-06,
+      "loss": 0.8744,
+      "step": 5154
+    },
+    {
+      "epoch": 0.7094687586017066,
+      "grad_norm": 1.8450302584433758,
+      "learning_rate": 4.109893383073272e-06,
+      "loss": 0.9234,
+      "step": 5155
+    },
+    {
+      "epoch": 0.7096063859069639,
+      "grad_norm": 1.7314708998318946,
+      "learning_rate": 4.1062918111424125e-06,
+      "loss": 0.8855,
+      "step": 5156
+    },
+    {
+      "epoch": 0.7097440132122214,
+      "grad_norm": 2.0483126995367673,
+      "learning_rate": 4.102691410210701e-06,
+      "loss": 0.9015,
+      "step": 5157
+    },
+    {
+      "epoch": 0.7098816405174787,
+      "grad_norm": 2.0716274036129074,
+      "learning_rate": 4.099092180993473e-06,
+      "loss": 0.9164,
+      "step": 5158
+    },
+    {
+      "epoch": 0.710019267822736,
+      "grad_norm": 1.8253615518829598,
+      "learning_rate": 4.095494124205862e-06,
+      "loss": 0.9709,
+      "step": 5159
+    },
+    {
+      "epoch": 0.7101568951279934,
+      "grad_norm": 4.3472928954049115,
+      "learning_rate": 4.091897240562738e-06,
+      "loss": 0.8435,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7102945224332508,
+      "grad_norm": 1.9431924866373664,
+      "learning_rate": 4.088301530778761e-06,
+      "loss": 0.9033,
+      "step": 5161
+    },
+    {
+      "epoch": 0.7104321497385081,
+      "grad_norm": 1.6864649958934186,
+      "learning_rate": 4.084706995568348e-06,
+      "loss": 0.83,
+      "step": 5162
+    },
+    {
+      "epoch": 0.7105697770437654,
+      "grad_norm": 1.5950939216258953,
+      "learning_rate": 4.081113635645685e-06,
+      "loss": 0.896,
+      "step": 5163
+    },
+    {
+      "epoch": 0.7107074043490228,
+      "grad_norm": 1.7827142663105238,
+      "learning_rate": 4.077521451724721e-06,
+      "loss": 0.8448,
+      "step": 5164
+    },
+    {
+      "epoch": 0.7108450316542803,
+      "grad_norm": 1.6387713114811087,
+      "learning_rate": 4.073930444519174e-06,
+      "loss": 0.9589,
+      "step": 5165
+    },
+    {
+      "epoch": 0.7109826589595376,
+      "grad_norm": 1.8345285254965196,
+      "learning_rate": 4.0703406147425285e-06,
+      "loss": 0.864,
+      "step": 5166
+    },
+    {
+      "epoch": 0.711120286264795,
+      "grad_norm": 1.9354392742878037,
+      "learning_rate": 4.066751963108036e-06,
+      "loss": 0.9052,
+      "step": 5167
+    },
+    {
+      "epoch": 0.7112579135700523,
+      "grad_norm": 1.8679777348058537,
+      "learning_rate": 4.063164490328709e-06,
+      "loss": 0.9281,
+      "step": 5168
+    },
+    {
+      "epoch": 0.7113955408753097,
+      "grad_norm": 1.7562475436384717,
+      "learning_rate": 4.059578197117333e-06,
+      "loss": 0.8879,
+      "step": 5169
+    },
+    {
+      "epoch": 0.711533168180567,
+      "grad_norm": 2.171488725153095,
+      "learning_rate": 4.055993084186454e-06,
+      "loss": 0.896,
+      "step": 5170
+    },
+    {
+      "epoch": 0.7116707954858243,
+      "grad_norm": 1.7364546113594967,
+      "learning_rate": 4.052409152248384e-06,
+      "loss": 0.9404,
+      "step": 5171
+    },
+    {
+      "epoch": 0.7118084227910817,
+      "grad_norm": 1.7820678815272881,
+      "learning_rate": 4.048826402015202e-06,
+      "loss": 0.8297,
+      "step": 5172
+    },
+    {
+      "epoch": 0.7119460500963392,
+      "grad_norm": 1.891700810845978,
+      "learning_rate": 4.045244834198751e-06,
+      "loss": 0.8718,
+      "step": 5173
+    },
+    {
+      "epoch": 0.7120836774015965,
+      "grad_norm": 2.100898732909389,
+      "learning_rate": 4.04166444951064e-06,
+      "loss": 0.8939,
+      "step": 5174
+    },
+    {
+      "epoch": 0.7122213047068539,
+      "grad_norm": 1.8969501973323752,
+      "learning_rate": 4.038085248662241e-06,
+      "loss": 0.9474,
+      "step": 5175
+    },
+    {
+      "epoch": 0.7123589320121112,
+      "grad_norm": 1.6402772845274616,
+      "learning_rate": 4.0345072323646935e-06,
+      "loss": 0.8497,
+      "step": 5176
+    },
+    {
+      "epoch": 0.7124965593173685,
+      "grad_norm": 1.8318917346709507,
+      "learning_rate": 4.0309304013289015e-06,
+      "loss": 0.8924,
+      "step": 5177
+    },
+    {
+      "epoch": 0.7126341866226259,
+      "grad_norm": 1.6992771362082488,
+      "learning_rate": 4.027354756265532e-06,
+      "loss": 0.9864,
+      "step": 5178
+    },
+    {
+      "epoch": 0.7127718139278832,
+      "grad_norm": 1.7491826815977427,
+      "learning_rate": 4.023780297885009e-06,
+      "loss": 0.8895,
+      "step": 5179
+    },
+    {
+      "epoch": 0.7129094412331407,
+      "grad_norm": 1.69115979570301,
+      "learning_rate": 4.020207026897545e-06,
+      "loss": 0.9106,
+      "step": 5180
+    },
+    {
+      "epoch": 0.713047068538398,
+      "grad_norm": 1.6464588206448079,
+      "learning_rate": 4.016634944013083e-06,
+      "loss": 0.9605,
+      "step": 5181
+    },
+    {
+      "epoch": 0.7131846958436554,
+      "grad_norm": 1.7087088345827548,
+      "learning_rate": 4.0130640499413595e-06,
+      "loss": 0.9327,
+      "step": 5182
+    },
+    {
+      "epoch": 0.7133223231489128,
+      "grad_norm": 1.8971866147173888,
+      "learning_rate": 4.009494345391852e-06,
+      "loss": 0.9411,
+      "step": 5183
+    },
+    {
+      "epoch": 0.7134599504541701,
+      "grad_norm": 1.684924141482463,
+      "learning_rate": 4.005925831073823e-06,
+      "loss": 0.9301,
+      "step": 5184
+    },
+    {
+      "epoch": 0.7135975777594274,
+      "grad_norm": 1.7695544878711318,
+      "learning_rate": 4.002358507696279e-06,
+      "loss": 0.8487,
+      "step": 5185
+    },
+    {
+      "epoch": 0.7137352050646848,
+      "grad_norm": 2.0193482303143755,
+      "learning_rate": 3.998792375967996e-06,
+      "loss": 0.9124,
+      "step": 5186
+    },
+    {
+      "epoch": 0.7138728323699421,
+      "grad_norm": 1.79156740294054,
+      "learning_rate": 3.995227436597529e-06,
+      "loss": 0.9038,
+      "step": 5187
+    },
+    {
+      "epoch": 0.7140104596751996,
+      "grad_norm": 1.8287373740371755,
+      "learning_rate": 3.9916636902931645e-06,
+      "loss": 0.9675,
+      "step": 5188
+    },
+    {
+      "epoch": 0.714148086980457,
+      "grad_norm": 1.6469279548988085,
+      "learning_rate": 3.988101137762989e-06,
+      "loss": 0.9761,
+      "step": 5189
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 1.6399448636270328,
+      "learning_rate": 3.984539779714815e-06,
+      "loss": 0.9704,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7144233415909717,
+      "grad_norm": 1.9456805609355556,
+      "learning_rate": 3.980979616856251e-06,
+      "loss": 0.8817,
+      "step": 5191
+    },
+    {
+      "epoch": 0.714560968896229,
+      "grad_norm": 1.67026520352583,
+      "learning_rate": 3.977420649894639e-06,
+      "loss": 0.9282,
+      "step": 5192
+    },
+    {
+      "epoch": 0.7146985962014863,
+      "grad_norm": 1.759640764178995,
+      "learning_rate": 3.973862879537107e-06,
+      "loss": 0.9224,
+      "step": 5193
+    },
+    {
+      "epoch": 0.7148362235067437,
+      "grad_norm": 2.1250237513389494,
+      "learning_rate": 3.9703063064905334e-06,
+      "loss": 0.9761,
+      "step": 5194
+    },
+    {
+      "epoch": 0.7149738508120012,
+      "grad_norm": 1.754815361304875,
+      "learning_rate": 3.966750931461558e-06,
+      "loss": 0.9232,
+      "step": 5195
+    },
+    {
+      "epoch": 0.7151114781172585,
+      "grad_norm": 1.7770231839810173,
+      "learning_rate": 3.9631967551565865e-06,
+      "loss": 0.9118,
+      "step": 5196
+    },
+    {
+      "epoch": 0.7152491054225159,
+      "grad_norm": 1.669203149221461,
+      "learning_rate": 3.9596437782817845e-06,
+      "loss": 0.8883,
+      "step": 5197
+    },
+    {
+      "epoch": 0.7153867327277732,
+      "grad_norm": 1.9620900659568794,
+      "learning_rate": 3.956092001543081e-06,
+      "loss": 0.9039,
+      "step": 5198
+    },
+    {
+      "epoch": 0.7155243600330305,
+      "grad_norm": 1.716769340671222,
+      "learning_rate": 3.952541425646165e-06,
+      "loss": 0.8577,
+      "step": 5199
+    },
+    {
+      "epoch": 0.7156619873382879,
+      "grad_norm": 1.865992111411231,
+      "learning_rate": 3.948992051296486e-06,
+      "loss": 0.9694,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7157996146435452,
+      "grad_norm": 2.204792558830319,
+      "learning_rate": 3.945443879199258e-06,
+      "loss": 0.9082,
+      "step": 5201
+    },
+    {
+      "epoch": 0.7159372419488026,
+      "grad_norm": 1.5725319299339242,
+      "learning_rate": 3.941896910059453e-06,
+      "loss": 0.8553,
+      "step": 5202
+    },
+    {
+      "epoch": 0.71607486925406,
+      "grad_norm": 2.362840551724885,
+      "learning_rate": 3.938351144581805e-06,
+      "loss": 0.906,
+      "step": 5203
+    },
+    {
+      "epoch": 0.7162124965593174,
+      "grad_norm": 1.66163411165644,
+      "learning_rate": 3.93480658347081e-06,
+      "loss": 0.8495,
+      "step": 5204
+    },
+    {
+      "epoch": 0.7163501238645748,
+      "grad_norm": 2.6247269543727496,
+      "learning_rate": 3.9312632274307235e-06,
+      "loss": 0.9239,
+      "step": 5205
+    },
+    {
+      "epoch": 0.7164877511698321,
+      "grad_norm": 1.7039754157142608,
+      "learning_rate": 3.9277210771655626e-06,
+      "loss": 0.9438,
+      "step": 5206
+    },
+    {
+      "epoch": 0.7166253784750894,
+      "grad_norm": 2.1254439404088887,
+      "learning_rate": 3.924180133379102e-06,
+      "loss": 0.9236,
+      "step": 5207
+    },
+    {
+      "epoch": 0.7167630057803468,
+      "grad_norm": 1.65718300312885,
+      "learning_rate": 3.920640396774881e-06,
+      "loss": 0.8456,
+      "step": 5208
+    },
+    {
+      "epoch": 0.7169006330856041,
+      "grad_norm": 2.017130873129969,
+      "learning_rate": 3.917101868056191e-06,
+      "loss": 0.8775,
+      "step": 5209
+    },
+    {
+      "epoch": 0.7170382603908615,
+      "grad_norm": 1.8569807887756613,
+      "learning_rate": 3.913564547926102e-06,
+      "loss": 0.97,
+      "step": 5210
+    },
+    {
+      "epoch": 0.717175887696119,
+      "grad_norm": 1.7345055043261586,
+      "learning_rate": 3.910028437087417e-06,
+      "loss": 0.8458,
+      "step": 5211
+    },
+    {
+      "epoch": 0.7173135150013763,
+      "grad_norm": 1.6035081867123977,
+      "learning_rate": 3.906493536242723e-06,
+      "loss": 0.9072,
+      "step": 5212
+    },
+    {
+      "epoch": 0.7174511423066336,
+      "grad_norm": 1.757623442915531,
+      "learning_rate": 3.9029598460943484e-06,
+      "loss": 0.9535,
+      "step": 5213
+    },
+    {
+      "epoch": 0.717588769611891,
+      "grad_norm": 1.7661737896475227,
+      "learning_rate": 3.899427367344399e-06,
+      "loss": 0.9248,
+      "step": 5214
+    },
+    {
+      "epoch": 0.7177263969171483,
+      "grad_norm": 1.7459796318533067,
+      "learning_rate": 3.895896100694716e-06,
+      "loss": 0.9615,
+      "step": 5215
+    },
+    {
+      "epoch": 0.7178640242224057,
+      "grad_norm": 2.136993783408053,
+      "learning_rate": 3.892366046846926e-06,
+      "loss": 0.9191,
+      "step": 5216
+    },
+    {
+      "epoch": 0.718001651527663,
+      "grad_norm": 2.320579839527249,
+      "learning_rate": 3.888837206502402e-06,
+      "loss": 0.8957,
+      "step": 5217
+    },
+    {
+      "epoch": 0.7181392788329205,
+      "grad_norm": 1.6961020435138954,
+      "learning_rate": 3.8853095803622645e-06,
+      "loss": 0.8478,
+      "step": 5218
+    },
+    {
+      "epoch": 0.7182769061381779,
+      "grad_norm": 1.7318511221437074,
+      "learning_rate": 3.88178316912742e-06,
+      "loss": 0.9662,
+      "step": 5219
+    },
+    {
+      "epoch": 0.7184145334434352,
+      "grad_norm": 1.8465130333206523,
+      "learning_rate": 3.878257973498504e-06,
+      "loss": 0.9646,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7185521607486925,
+      "grad_norm": 1.7602364506001495,
+      "learning_rate": 3.874733994175936e-06,
+      "loss": 0.9402,
+      "step": 5221
+    },
+    {
+      "epoch": 0.7186897880539499,
+      "grad_norm": 1.6146063521295564,
+      "learning_rate": 3.871211231859872e-06,
+      "loss": 0.9452,
+      "step": 5222
+    },
+    {
+      "epoch": 0.7188274153592072,
+      "grad_norm": 2.0963344165175104,
+      "learning_rate": 3.867689687250246e-06,
+      "loss": 0.9228,
+      "step": 5223
+    },
+    {
+      "epoch": 0.7189650426644646,
+      "grad_norm": 1.7265714229237468,
+      "learning_rate": 3.864169361046738e-06,
+      "loss": 0.9128,
+      "step": 5224
+    },
+    {
+      "epoch": 0.7191026699697219,
+      "grad_norm": 1.7187370554733907,
+      "learning_rate": 3.860650253948785e-06,
+      "loss": 0.9007,
+      "step": 5225
+    },
+    {
+      "epoch": 0.7192402972749794,
+      "grad_norm": 1.853105589795603,
+      "learning_rate": 3.8571323666555906e-06,
+      "loss": 0.9387,
+      "step": 5226
+    },
+    {
+      "epoch": 0.7193779245802367,
+      "grad_norm": 1.5260829526036763,
+      "learning_rate": 3.8536156998661066e-06,
+      "loss": 0.9076,
+      "step": 5227
+    },
+    {
+      "epoch": 0.7195155518854941,
+      "grad_norm": 1.6612488991374663,
+      "learning_rate": 3.850100254279049e-06,
+      "loss": 0.9173,
+      "step": 5228
+    },
+    {
+      "epoch": 0.7196531791907514,
+      "grad_norm": 1.658363019234065,
+      "learning_rate": 3.846586030592887e-06,
+      "loss": 0.8663,
+      "step": 5229
+    },
+    {
+      "epoch": 0.7197908064960088,
+      "grad_norm": 1.7004386431691416,
+      "learning_rate": 3.843073029505851e-06,
+      "loss": 0.8976,
+      "step": 5230
+    },
+    {
+      "epoch": 0.7199284338012661,
+      "grad_norm": 1.6639051727664846,
+      "learning_rate": 3.839561251715924e-06,
+      "loss": 0.8598,
+      "step": 5231
+    },
+    {
+      "epoch": 0.7200660611065235,
+      "grad_norm": 6.208591695681176,
+      "learning_rate": 3.836050697920849e-06,
+      "loss": 0.9584,
+      "step": 5232
+    },
+    {
+      "epoch": 0.720203688411781,
+      "grad_norm": 1.8747018687196448,
+      "learning_rate": 3.832541368818126e-06,
+      "loss": 0.8802,
+      "step": 5233
+    },
+    {
+      "epoch": 0.7203413157170383,
+      "grad_norm": 1.6778380549433567,
+      "learning_rate": 3.8290332651050075e-06,
+      "loss": 0.9663,
+      "step": 5234
+    },
+    {
+      "epoch": 0.7204789430222956,
+      "grad_norm": 1.9800282132513551,
+      "learning_rate": 3.825526387478509e-06,
+      "loss": 0.934,
+      "step": 5235
+    },
+    {
+      "epoch": 0.720616570327553,
+      "grad_norm": 1.6718434352999254,
+      "learning_rate": 3.822020736635399e-06,
+      "loss": 0.8407,
+      "step": 5236
+    },
+    {
+      "epoch": 0.7207541976328103,
+      "grad_norm": 1.9797734589933966,
+      "learning_rate": 3.818516313272202e-06,
+      "loss": 0.9502,
+      "step": 5237
+    },
+    {
+      "epoch": 0.7208918249380677,
+      "grad_norm": 1.6937231758547295,
+      "learning_rate": 3.815013118085196e-06,
+      "loss": 0.9156,
+      "step": 5238
+    },
+    {
+      "epoch": 0.721029452243325,
+      "grad_norm": 1.8042615207641088,
+      "learning_rate": 3.8115111517704207e-06,
+      "loss": 0.9367,
+      "step": 5239
+    },
+    {
+      "epoch": 0.7211670795485824,
+      "grad_norm": 1.579846358890038,
+      "learning_rate": 3.8080104150236686e-06,
+      "loss": 0.9142,
+      "step": 5240
+    },
+    {
+      "epoch": 0.7213047068538398,
+      "grad_norm": 1.9842851711564975,
+      "learning_rate": 3.8045109085404853e-06,
+      "loss": 0.8966,
+      "step": 5241
+    },
+    {
+      "epoch": 0.7214423341590972,
+      "grad_norm": 1.7983023009526422,
+      "learning_rate": 3.8010126330161835e-06,
+      "loss": 0.878,
+      "step": 5242
+    },
+    {
+      "epoch": 0.7215799614643545,
+      "grad_norm": 1.8152979543153747,
+      "learning_rate": 3.7975155891458093e-06,
+      "loss": 0.8809,
+      "step": 5243
+    },
+    {
+      "epoch": 0.7217175887696119,
+      "grad_norm": 2.0685041369383566,
+      "learning_rate": 3.7940197776241926e-06,
+      "loss": 0.9268,
+      "step": 5244
+    },
+    {
+      "epoch": 0.7218552160748692,
+      "grad_norm": 1.879852493942682,
+      "learning_rate": 3.790525199145888e-06,
+      "loss": 0.8955,
+      "step": 5245
+    },
+    {
+      "epoch": 0.7219928433801266,
+      "grad_norm": 2.0435675705074616,
+      "learning_rate": 3.7870318544052354e-06,
+      "loss": 0.9387,
+      "step": 5246
+    },
+    {
+      "epoch": 0.7221304706853839,
+      "grad_norm": 1.8461888059817466,
+      "learning_rate": 3.7835397440962996e-06,
+      "loss": 0.9438,
+      "step": 5247
+    },
+    {
+      "epoch": 0.7222680979906413,
+      "grad_norm": 1.7121294634244952,
+      "learning_rate": 3.7800488689129267e-06,
+      "loss": 0.8898,
+      "step": 5248
+    },
+    {
+      "epoch": 0.7224057252958987,
+      "grad_norm": 1.8573986553593866,
+      "learning_rate": 3.7765592295487043e-06,
+      "loss": 0.9717,
+      "step": 5249
+    },
+    {
+      "epoch": 0.7225433526011561,
+      "grad_norm": 1.848516250989898,
+      "learning_rate": 3.7730708266969673e-06,
+      "loss": 0.9513,
+      "step": 5250
+    },
+    {
+      "epoch": 0.7226809799064134,
+      "grad_norm": 1.6329369215329173,
+      "learning_rate": 3.7695836610508264e-06,
+      "loss": 0.933,
+      "step": 5251
+    },
+    {
+      "epoch": 0.7228186072116708,
+      "grad_norm": 1.8335791719487562,
+      "learning_rate": 3.766097733303119e-06,
+      "loss": 0.9349,
+      "step": 5252
+    },
+    {
+      "epoch": 0.7229562345169281,
+      "grad_norm": 1.7547533046067223,
+      "learning_rate": 3.762613044146467e-06,
+      "loss": 0.8945,
+      "step": 5253
+    },
+    {
+      "epoch": 0.7230938618221855,
+      "grad_norm": 2.013973372815274,
+      "learning_rate": 3.7591295942732153e-06,
+      "loss": 0.9604,
+      "step": 5254
+    },
+    {
+      "epoch": 0.7232314891274428,
+      "grad_norm": 1.7212157840811677,
+      "learning_rate": 3.7556473843754893e-06,
+      "loss": 0.999,
+      "step": 5255
+    },
+    {
+      "epoch": 0.7233691164327003,
+      "grad_norm": 1.693696731476249,
+      "learning_rate": 3.752166415145151e-06,
+      "loss": 0.9312,
+      "step": 5256
+    },
+    {
+      "epoch": 0.7235067437379576,
+      "grad_norm": 1.8853649645866906,
+      "learning_rate": 3.7486866872738236e-06,
+      "loss": 0.8978,
+      "step": 5257
+    },
+    {
+      "epoch": 0.723644371043215,
+      "grad_norm": 1.8665831157373645,
+      "learning_rate": 3.745208201452879e-06,
+      "loss": 0.9871,
+      "step": 5258
+    },
+    {
+      "epoch": 0.7237819983484723,
+      "grad_norm": 1.6916700185985052,
+      "learning_rate": 3.7417309583734475e-06,
+      "loss": 0.8862,
+      "step": 5259
+    },
+    {
+      "epoch": 0.7239196256537297,
+      "grad_norm": 1.5910627863982651,
+      "learning_rate": 3.738254958726407e-06,
+      "loss": 0.9534,
+      "step": 5260
+    },
+    {
+      "epoch": 0.724057252958987,
+      "grad_norm": 1.6892806461852596,
+      "learning_rate": 3.7347802032023927e-06,
+      "loss": 0.9363,
+      "step": 5261
+    },
+    {
+      "epoch": 0.7241948802642444,
+      "grad_norm": 1.6945488960301496,
+      "learning_rate": 3.7313066924917906e-06,
+      "loss": 0.9018,
+      "step": 5262
+    },
+    {
+      "epoch": 0.7243325075695017,
+      "grad_norm": 1.7948871348035533,
+      "learning_rate": 3.72783442728474e-06,
+      "loss": 0.9148,
+      "step": 5263
+    },
+    {
+      "epoch": 0.7244701348747592,
+      "grad_norm": 1.7637580360175407,
+      "learning_rate": 3.724363408271131e-06,
+      "loss": 0.9596,
+      "step": 5264
+    },
+    {
+      "epoch": 0.7246077621800165,
+      "grad_norm": 1.5531617042455883,
+      "learning_rate": 3.7208936361406088e-06,
+      "loss": 0.9206,
+      "step": 5265
+    },
+    {
+      "epoch": 0.7247453894852739,
+      "grad_norm": 1.9635847378187228,
+      "learning_rate": 3.717425111582571e-06,
+      "loss": 0.9251,
+      "step": 5266
+    },
+    {
+      "epoch": 0.7248830167905312,
+      "grad_norm": 1.6387599373641766,
+      "learning_rate": 3.713957835286163e-06,
+      "loss": 0.9414,
+      "step": 5267
+    },
+    {
+      "epoch": 0.7250206440957886,
+      "grad_norm": 1.8786223482022815,
+      "learning_rate": 3.7104918079402874e-06,
+      "loss": 0.9167,
+      "step": 5268
+    },
+    {
+      "epoch": 0.7251582714010459,
+      "grad_norm": 1.9298772009633662,
+      "learning_rate": 3.7070270302335954e-06,
+      "loss": 0.9694,
+      "step": 5269
+    },
+    {
+      "epoch": 0.7252958987063033,
+      "grad_norm": 2.0868091426204853,
+      "learning_rate": 3.70356350285449e-06,
+      "loss": 0.8991,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7254335260115607,
+      "grad_norm": 1.6350046170287333,
+      "learning_rate": 3.700101226491126e-06,
+      "loss": 0.8414,
+      "step": 5271
+    },
+    {
+      "epoch": 0.7255711533168181,
+      "grad_norm": 1.75295336139188,
+      "learning_rate": 3.6966402018314187e-06,
+      "loss": 0.9623,
+      "step": 5272
+    },
+    {
+      "epoch": 0.7257087806220754,
+      "grad_norm": 1.8242044394605565,
+      "learning_rate": 3.6931804295630115e-06,
+      "loss": 0.8499,
+      "step": 5273
+    },
+    {
+      "epoch": 0.7258464079273328,
+      "grad_norm": 1.715079109892309,
+      "learning_rate": 3.6897219103733294e-06,
+      "loss": 0.8921,
+      "step": 5274
+    },
+    {
+      "epoch": 0.7259840352325901,
+      "grad_norm": 1.6603639259547738,
+      "learning_rate": 3.6862646449495187e-06,
+      "loss": 0.9137,
+      "step": 5275
+    },
+    {
+      "epoch": 0.7261216625378475,
+      "grad_norm": 1.8731215829523309,
+      "learning_rate": 3.682808633978505e-06,
+      "loss": 0.9039,
+      "step": 5276
+    },
+    {
+      "epoch": 0.7262592898431048,
+      "grad_norm": 1.6633463610622765,
+      "learning_rate": 3.6793538781469353e-06,
+      "loss": 0.8874,
+      "step": 5277
+    },
+    {
+      "epoch": 0.7263969171483622,
+      "grad_norm": 1.8778534024817988,
+      "learning_rate": 3.6759003781412327e-06,
+      "loss": 0.9715,
+      "step": 5278
+    },
+    {
+      "epoch": 0.7265345444536196,
+      "grad_norm": 1.9260690641432139,
+      "learning_rate": 3.672448134647558e-06,
+      "loss": 0.8807,
+      "step": 5279
+    },
+    {
+      "epoch": 0.726672171758877,
+      "grad_norm": 2.0817387566057257,
+      "learning_rate": 3.668997148351825e-06,
+      "loss": 0.859,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7268097990641343,
+      "grad_norm": 2.1115394378825676,
+      "learning_rate": 3.665547419939699e-06,
+      "loss": 0.9809,
+      "step": 5281
+    },
+    {
+      "epoch": 0.7269474263693917,
+      "grad_norm": 1.6655224131489796,
+      "learning_rate": 3.6620989500965853e-06,
+      "loss": 0.9162,
+      "step": 5282
+    },
+    {
+      "epoch": 0.727085053674649,
+      "grad_norm": 1.71990596088117,
+      "learning_rate": 3.6586517395076605e-06,
+      "loss": 0.8823,
+      "step": 5283
+    },
+    {
+      "epoch": 0.7272226809799064,
+      "grad_norm": 1.5776383793454134,
+      "learning_rate": 3.655205788857824e-06,
+      "loss": 0.8971,
+      "step": 5284
+    },
+    {
+      "epoch": 0.7273603082851637,
+      "grad_norm": 1.8201200880919435,
+      "learning_rate": 3.65176109883175e-06,
+      "loss": 0.9235,
+      "step": 5285
+    },
+    {
+      "epoch": 0.7274979355904211,
+      "grad_norm": 1.7898470375281073,
+      "learning_rate": 3.648317670113848e-06,
+      "loss": 0.8691,
+      "step": 5286
+    },
+    {
+      "epoch": 0.7276355628956785,
+      "grad_norm": 2.1412492397768728,
+      "learning_rate": 3.6448755033882797e-06,
+      "loss": 0.9068,
+      "step": 5287
+    },
+    {
+      "epoch": 0.7277731902009359,
+      "grad_norm": 1.6840438736877315,
+      "learning_rate": 3.641434599338958e-06,
+      "loss": 0.9903,
+      "step": 5288
+    },
+    {
+      "epoch": 0.7279108175061932,
+      "grad_norm": 1.460442993483951,
+      "learning_rate": 3.637994958649541e-06,
+      "loss": 0.9597,
+      "step": 5289
+    },
+    {
+      "epoch": 0.7280484448114506,
+      "grad_norm": 2.087061697669228,
+      "learning_rate": 3.634556582003439e-06,
+      "loss": 0.9406,
+      "step": 5290
+    },
+    {
+      "epoch": 0.7281860721167079,
+      "grad_norm": 2.0188184569801573,
+      "learning_rate": 3.6311194700838113e-06,
+      "loss": 0.9161,
+      "step": 5291
+    },
+    {
+      "epoch": 0.7283236994219653,
+      "grad_norm": 1.8125083634099686,
+      "learning_rate": 3.6276836235735646e-06,
+      "loss": 0.8558,
+      "step": 5292
+    },
+    {
+      "epoch": 0.7284613267272226,
+      "grad_norm": 1.6142611852611792,
+      "learning_rate": 3.6242490431553536e-06,
+      "loss": 0.9602,
+      "step": 5293
+    },
+    {
+      "epoch": 0.7285989540324801,
+      "grad_norm": 1.561917549846892,
+      "learning_rate": 3.620815729511584e-06,
+      "loss": 0.9175,
+      "step": 5294
+    },
+    {
+      "epoch": 0.7287365813377374,
+      "grad_norm": 1.9969988879096154,
+      "learning_rate": 3.617383683324407e-06,
+      "loss": 0.8478,
+      "step": 5295
+    },
+    {
+      "epoch": 0.7288742086429948,
+      "grad_norm": 2.2055649871786773,
+      "learning_rate": 3.613952905275724e-06,
+      "loss": 0.8514,
+      "step": 5296
+    },
+    {
+      "epoch": 0.7290118359482521,
+      "grad_norm": 1.712659926677815,
+      "learning_rate": 3.610523396047183e-06,
+      "loss": 0.9044,
+      "step": 5297
+    },
+    {
+      "epoch": 0.7291494632535095,
+      "grad_norm": 1.8692593447289574,
+      "learning_rate": 3.6070951563201805e-06,
+      "loss": 0.9659,
+      "step": 5298
+    },
+    {
+      "epoch": 0.7292870905587668,
+      "grad_norm": 1.82367916255235,
+      "learning_rate": 3.6036681867758618e-06,
+      "loss": 0.9261,
+      "step": 5299
+    },
+    {
+      "epoch": 0.7294247178640242,
+      "grad_norm": 1.7360811856324996,
+      "learning_rate": 3.600242488095117e-06,
+      "loss": 0.8825,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7295623451692815,
+      "grad_norm": 1.7300883478966302,
+      "learning_rate": 3.596818060958588e-06,
+      "loss": 0.896,
+      "step": 5301
+    },
+    {
+      "epoch": 0.729699972474539,
+      "grad_norm": 1.8700646782066308,
+      "learning_rate": 3.593394906046659e-06,
+      "loss": 0.7985,
+      "step": 5302
+    },
+    {
+      "epoch": 0.7298375997797963,
+      "grad_norm": 1.6467434806308108,
+      "learning_rate": 3.5899730240394628e-06,
+      "loss": 0.9176,
+      "step": 5303
+    },
+    {
+      "epoch": 0.7299752270850537,
+      "grad_norm": 1.6630454812127828,
+      "learning_rate": 3.5865524156168894e-06,
+      "loss": 0.9313,
+      "step": 5304
+    },
+    {
+      "epoch": 0.730112854390311,
+      "grad_norm": 1.755643099263133,
+      "learning_rate": 3.583133081458554e-06,
+      "loss": 0.8778,
+      "step": 5305
+    },
+    {
+      "epoch": 0.7302504816955684,
+      "grad_norm": 1.6602541788271965,
+      "learning_rate": 3.579715022243845e-06,
+      "loss": 0.9221,
+      "step": 5306
+    },
+    {
+      "epoch": 0.7303881090008257,
+      "grad_norm": 1.5903015203608266,
+      "learning_rate": 3.5762982386518695e-06,
+      "loss": 0.8873,
+      "step": 5307
+    },
+    {
+      "epoch": 0.7305257363060831,
+      "grad_norm": 1.6925348902802826,
+      "learning_rate": 3.57288273136151e-06,
+      "loss": 0.9088,
+      "step": 5308
+    },
+    {
+      "epoch": 0.7306633636113405,
+      "grad_norm": 1.855645528680667,
+      "learning_rate": 3.5694685010513686e-06,
+      "loss": 0.9022,
+      "step": 5309
+    },
+    {
+      "epoch": 0.7308009909165979,
+      "grad_norm": 1.726876241284321,
+      "learning_rate": 3.5660555483998137e-06,
+      "loss": 0.9479,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7309386182218552,
+      "grad_norm": 1.7173795202580493,
+      "learning_rate": 3.56264387408495e-06,
+      "loss": 0.9005,
+      "step": 5311
+    },
+    {
+      "epoch": 0.7310762455271126,
+      "grad_norm": 1.722255610882931,
+      "learning_rate": 3.5592334787846314e-06,
+      "loss": 0.9018,
+      "step": 5312
+    },
+    {
+      "epoch": 0.7312138728323699,
+      "grad_norm": 1.6791395519326309,
+      "learning_rate": 3.55582436317646e-06,
+      "loss": 0.9206,
+      "step": 5313
+    },
+    {
+      "epoch": 0.7313515001376273,
+      "grad_norm": 1.7869267242749116,
+      "learning_rate": 3.552416527937769e-06,
+      "loss": 0.9804,
+      "step": 5314
+    },
+    {
+      "epoch": 0.7314891274428846,
+      "grad_norm": 1.649235723816545,
+      "learning_rate": 3.549009973745663e-06,
+      "loss": 0.9486,
+      "step": 5315
+    },
+    {
+      "epoch": 0.731626754748142,
+      "grad_norm": 1.9403884309205994,
+      "learning_rate": 3.5456047012769632e-06,
+      "loss": 0.9408,
+      "step": 5316
+    },
+    {
+      "epoch": 0.7317643820533994,
+      "grad_norm": 1.7170603858133688,
+      "learning_rate": 3.5422007112082624e-06,
+      "loss": 0.899,
+      "step": 5317
+    },
+    {
+      "epoch": 0.7319020093586568,
+      "grad_norm": 1.8249777785884589,
+      "learning_rate": 3.538798004215883e-06,
+      "loss": 0.9159,
+      "step": 5318
+    },
+    {
+      "epoch": 0.7320396366639141,
+      "grad_norm": 2.420433625816806,
+      "learning_rate": 3.535396580975895e-06,
+      "loss": 0.9549,
+      "step": 5319
+    },
+    {
+      "epoch": 0.7321772639691715,
+      "grad_norm": 1.5553825700930723,
+      "learning_rate": 3.5319964421641163e-06,
+      "loss": 0.8696,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7323148912744288,
+      "grad_norm": 1.9392250757763636,
+      "learning_rate": 3.5285975884561075e-06,
+      "loss": 0.9615,
+      "step": 5321
+    },
+    {
+      "epoch": 0.7324525185796862,
+      "grad_norm": 1.8613023387684744,
+      "learning_rate": 3.525200020527173e-06,
+      "loss": 0.9537,
+      "step": 5322
+    },
+    {
+      "epoch": 0.7325901458849435,
+      "grad_norm": 1.7425791223313267,
+      "learning_rate": 3.5218037390523653e-06,
+      "loss": 0.8905,
+      "step": 5323
+    },
+    {
+      "epoch": 0.7327277731902009,
+      "grad_norm": 1.5807277348544768,
+      "learning_rate": 3.5184087447064763e-06,
+      "loss": 0.8795,
+      "step": 5324
+    },
+    {
+      "epoch": 0.7328654004954583,
+      "grad_norm": 1.892101501467385,
+      "learning_rate": 3.515015038164048e-06,
+      "loss": 0.9652,
+      "step": 5325
+    },
+    {
+      "epoch": 0.7330030278007157,
+      "grad_norm": 1.6750855341760533,
+      "learning_rate": 3.5116226200993606e-06,
+      "loss": 0.8535,
+      "step": 5326
+    },
+    {
+      "epoch": 0.733140655105973,
+      "grad_norm": 1.6148431943533017,
+      "learning_rate": 3.5082314911864422e-06,
+      "loss": 0.9851,
+      "step": 5327
+    },
+    {
+      "epoch": 0.7332782824112304,
+      "grad_norm": 1.8507295280267404,
+      "learning_rate": 3.5048416520990645e-06,
+      "loss": 0.8968,
+      "step": 5328
+    },
+    {
+      "epoch": 0.7334159097164877,
+      "grad_norm": 1.7676870106796847,
+      "learning_rate": 3.5014531035107423e-06,
+      "loss": 0.9001,
+      "step": 5329
+    },
+    {
+      "epoch": 0.7335535370217451,
+      "grad_norm": 1.7335978897367443,
+      "learning_rate": 3.4980658460947326e-06,
+      "loss": 0.9168,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7336911643270024,
+      "grad_norm": 1.8490527682830393,
+      "learning_rate": 3.494679880524037e-06,
+      "loss": 0.9198,
+      "step": 5331
+    },
+    {
+      "epoch": 0.7338287916322599,
+      "grad_norm": 1.523958992740363,
+      "learning_rate": 3.4912952074714014e-06,
+      "loss": 0.87,
+      "step": 5332
+    },
+    {
+      "epoch": 0.7339664189375172,
+      "grad_norm": 1.6361506164946327,
+      "learning_rate": 3.4879118276093126e-06,
+      "loss": 0.933,
+      "step": 5333
+    },
+    {
+      "epoch": 0.7341040462427746,
+      "grad_norm": 1.764890623843533,
+      "learning_rate": 3.4845297416100044e-06,
+      "loss": 0.8634,
+      "step": 5334
+    },
+    {
+      "epoch": 0.7342416735480319,
+      "grad_norm": 1.6735032683532114,
+      "learning_rate": 3.4811489501454444e-06,
+      "loss": 0.9088,
+      "step": 5335
+    },
+    {
+      "epoch": 0.7343793008532893,
+      "grad_norm": 1.9644629983346653,
+      "learning_rate": 3.477769453887363e-06,
+      "loss": 0.9836,
+      "step": 5336
+    },
+    {
+      "epoch": 0.7345169281585466,
+      "grad_norm": 1.8158415007835387,
+      "learning_rate": 3.4743912535072042e-06,
+      "loss": 0.8275,
+      "step": 5337
+    },
+    {
+      "epoch": 0.734654555463804,
+      "grad_norm": 1.8794224657216105,
+      "learning_rate": 3.471014349676184e-06,
+      "loss": 0.9107,
+      "step": 5338
+    },
+    {
+      "epoch": 0.7347921827690613,
+      "grad_norm": 1.843173646543842,
+      "learning_rate": 3.4676387430652335e-06,
+      "loss": 0.9727,
+      "step": 5339
+    },
+    {
+      "epoch": 0.7349298100743188,
+      "grad_norm": 1.6410817311317818,
+      "learning_rate": 3.4642644343450503e-06,
+      "loss": 0.9845,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7350674373795761,
+      "grad_norm": 1.757701676761118,
+      "learning_rate": 3.4608914241860593e-06,
+      "loss": 0.8836,
+      "step": 5341
+    },
+    {
+      "epoch": 0.7352050646848335,
+      "grad_norm": 1.760059814249015,
+      "learning_rate": 3.4575197132584316e-06,
+      "loss": 0.9001,
+      "step": 5342
+    },
+    {
+      "epoch": 0.7353426919900908,
+      "grad_norm": 1.7351843979098676,
+      "learning_rate": 3.4541493022320803e-06,
+      "loss": 0.8596,
+      "step": 5343
+    },
+    {
+      "epoch": 0.7354803192953482,
+      "grad_norm": 1.9743375169592805,
+      "learning_rate": 3.4507801917766605e-06,
+      "loss": 0.938,
+      "step": 5344
+    },
+    {
+      "epoch": 0.7356179466006055,
+      "grad_norm": 1.8370298988349438,
+      "learning_rate": 3.44741238256157e-06,
+      "loss": 0.9139,
+      "step": 5345
+    },
+    {
+      "epoch": 0.7357555739058629,
+      "grad_norm": 2.065942388534253,
+      "learning_rate": 3.444045875255937e-06,
+      "loss": 0.9235,
+      "step": 5346
+    },
+    {
+      "epoch": 0.7358932012111203,
+      "grad_norm": 1.8479359650419447,
+      "learning_rate": 3.440680670528651e-06,
+      "loss": 0.8183,
+      "step": 5347
+    },
+    {
+      "epoch": 0.7360308285163777,
+      "grad_norm": 1.8722118740478073,
+      "learning_rate": 3.437316769048329e-06,
+      "loss": 0.9297,
+      "step": 5348
+    },
+    {
+      "epoch": 0.736168455821635,
+      "grad_norm": 1.9216855865907785,
+      "learning_rate": 3.43395417148333e-06,
+      "loss": 0.938,
+      "step": 5349
+    },
+    {
+      "epoch": 0.7363060831268924,
+      "grad_norm": 1.6383505618484056,
+      "learning_rate": 3.4305928785017595e-06,
+      "loss": 0.9385,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7364437104321497,
+      "grad_norm": 1.8605079699724014,
+      "learning_rate": 3.427232890771458e-06,
+      "loss": 0.9587,
+      "step": 5351
+    },
+    {
+      "epoch": 0.7365813377374071,
+      "grad_norm": 1.9585010423187021,
+      "learning_rate": 3.4238742089600098e-06,
+      "loss": 0.8927,
+      "step": 5352
+    },
+    {
+      "epoch": 0.7367189650426644,
+      "grad_norm": 1.9797679417380907,
+      "learning_rate": 3.4205168337347393e-06,
+      "loss": 0.9388,
+      "step": 5353
+    },
+    {
+      "epoch": 0.7368565923479218,
+      "grad_norm": 1.7583741814972624,
+      "learning_rate": 3.417160765762709e-06,
+      "loss": 0.9243,
+      "step": 5354
+    },
+    {
+      "epoch": 0.7369942196531792,
+      "grad_norm": 1.7405355366294266,
+      "learning_rate": 3.413806005710727e-06,
+      "loss": 0.9315,
+      "step": 5355
+    },
+    {
+      "epoch": 0.7371318469584366,
+      "grad_norm": 1.597786272578749,
+      "learning_rate": 3.410452554245336e-06,
+      "loss": 0.9375,
+      "step": 5356
+    },
+    {
+      "epoch": 0.7372694742636939,
+      "grad_norm": 1.6616529422224957,
+      "learning_rate": 3.4071004120328223e-06,
+      "loss": 0.9527,
+      "step": 5357
+    },
+    {
+      "epoch": 0.7374071015689513,
+      "grad_norm": 1.8896497108057202,
+      "learning_rate": 3.403749579739205e-06,
+      "loss": 0.842,
+      "step": 5358
+    },
+    {
+      "epoch": 0.7375447288742086,
+      "grad_norm": 1.7703938245966926,
+      "learning_rate": 3.4004000580302608e-06,
+      "loss": 0.9347,
+      "step": 5359
+    },
+    {
+      "epoch": 0.737682356179466,
+      "grad_norm": 1.7263967035827457,
+      "learning_rate": 3.397051847571483e-06,
+      "loss": 0.8795,
+      "step": 5360
+    },
+    {
+      "epoch": 0.7378199834847233,
+      "grad_norm": 1.9059037793469038,
+      "learning_rate": 3.3937049490281184e-06,
+      "loss": 0.9569,
+      "step": 5361
+    },
+    {
+      "epoch": 0.7379576107899807,
+      "grad_norm": 1.9953453096096039,
+      "learning_rate": 3.39035936306515e-06,
+      "loss": 0.8431,
+      "step": 5362
+    },
+    {
+      "epoch": 0.7380952380952381,
+      "grad_norm": 1.6740389680721992,
+      "learning_rate": 3.3870150903473e-06,
+      "loss": 0.8983,
+      "step": 5363
+    },
+    {
+      "epoch": 0.7382328654004955,
+      "grad_norm": 1.8230905771735781,
+      "learning_rate": 3.38367213153903e-06,
+      "loss": 0.8949,
+      "step": 5364
+    },
+    {
+      "epoch": 0.7383704927057528,
+      "grad_norm": 1.7597530190851718,
+      "learning_rate": 3.380330487304536e-06,
+      "loss": 0.8838,
+      "step": 5365
+    },
+    {
+      "epoch": 0.7385081200110102,
+      "grad_norm": 1.8585719908041678,
+      "learning_rate": 3.376990158307768e-06,
+      "loss": 0.887,
+      "step": 5366
+    },
+    {
+      "epoch": 0.7386457473162675,
+      "grad_norm": 1.7260618173731745,
+      "learning_rate": 3.373651145212389e-06,
+      "loss": 0.9083,
+      "step": 5367
+    },
+    {
+      "epoch": 0.7387833746215249,
+      "grad_norm": 1.6602032472903947,
+      "learning_rate": 3.37031344868183e-06,
+      "loss": 0.9326,
+      "step": 5368
+    },
+    {
+      "epoch": 0.7389210019267822,
+      "grad_norm": 1.6617601151026693,
+      "learning_rate": 3.3669770693792313e-06,
+      "loss": 0.9389,
+      "step": 5369
+    },
+    {
+      "epoch": 0.7390586292320397,
+      "grad_norm": 1.7039987133921115,
+      "learning_rate": 3.363642007967499e-06,
+      "loss": 0.9255,
+      "step": 5370
+    },
+    {
+      "epoch": 0.739196256537297,
+      "grad_norm": 1.8051054749125792,
+      "learning_rate": 3.360308265109252e-06,
+      "loss": 0.8974,
+      "step": 5371
+    },
+    {
+      "epoch": 0.7393338838425544,
+      "grad_norm": 1.9555580519007822,
+      "learning_rate": 3.3569758414668686e-06,
+      "loss": 0.8786,
+      "step": 5372
+    },
+    {
+      "epoch": 0.7394715111478117,
+      "grad_norm": 1.7647423184782687,
+      "learning_rate": 3.3536447377024516e-06,
+      "loss": 0.9408,
+      "step": 5373
+    },
+    {
+      "epoch": 0.7396091384530691,
+      "grad_norm": 1.8573980690437624,
+      "learning_rate": 3.350314954477847e-06,
+      "loss": 0.8428,
+      "step": 5374
+    },
+    {
+      "epoch": 0.7397467657583264,
+      "grad_norm": 1.6184287065613328,
+      "learning_rate": 3.346986492454637e-06,
+      "loss": 0.9617,
+      "step": 5375
+    },
+    {
+      "epoch": 0.7398843930635838,
+      "grad_norm": 1.773773314211498,
+      "learning_rate": 3.343659352294141e-06,
+      "loss": 0.915,
+      "step": 5376
+    },
+    {
+      "epoch": 0.7400220203688411,
+      "grad_norm": 1.6272666452804558,
+      "learning_rate": 3.3403335346574186e-06,
+      "loss": 0.9132,
+      "step": 5377
+    },
+    {
+      "epoch": 0.7401596476740986,
+      "grad_norm": 2.037348616917277,
+      "learning_rate": 3.3370090402052557e-06,
+      "loss": 0.9709,
+      "step": 5378
+    },
+    {
+      "epoch": 0.7402972749793559,
+      "grad_norm": 1.8313727575610053,
+      "learning_rate": 3.3336858695981944e-06,
+      "loss": 0.949,
+      "step": 5379
+    },
+    {
+      "epoch": 0.7404349022846133,
+      "grad_norm": 1.8379224312058455,
+      "learning_rate": 3.3303640234964985e-06,
+      "loss": 0.8437,
+      "step": 5380
+    },
+    {
+      "epoch": 0.7405725295898706,
+      "grad_norm": 1.6492160972115861,
+      "learning_rate": 3.327043502560173e-06,
+      "loss": 0.9047,
+      "step": 5381
+    },
+    {
+      "epoch": 0.740710156895128,
+      "grad_norm": 1.6887732087357337,
+      "learning_rate": 3.323724307448961e-06,
+      "loss": 0.9511,
+      "step": 5382
+    },
+    {
+      "epoch": 0.7408477842003853,
+      "grad_norm": 1.814184828178591,
+      "learning_rate": 3.320406438822341e-06,
+      "loss": 0.9093,
+      "step": 5383
+    },
+    {
+      "epoch": 0.7409854115056427,
+      "grad_norm": 1.6067990511749097,
+      "learning_rate": 3.317089897339528e-06,
+      "loss": 0.8933,
+      "step": 5384
+    },
+    {
+      "epoch": 0.7411230388109001,
+      "grad_norm": 1.8594203825447861,
+      "learning_rate": 3.313774683659474e-06,
+      "loss": 0.9051,
+      "step": 5385
+    },
+    {
+      "epoch": 0.7412606661161575,
+      "grad_norm": 1.6909211227845824,
+      "learning_rate": 3.3104607984408656e-06,
+      "loss": 0.868,
+      "step": 5386
+    },
+    {
+      "epoch": 0.7413982934214148,
+      "grad_norm": 1.8876855767360883,
+      "learning_rate": 3.3071482423421263e-06,
+      "loss": 0.9713,
+      "step": 5387
+    },
+    {
+      "epoch": 0.7415359207266722,
+      "grad_norm": 1.5619637083852707,
+      "learning_rate": 3.303837016021416e-06,
+      "loss": 0.9227,
+      "step": 5388
+    },
+    {
+      "epoch": 0.7416735480319295,
+      "grad_norm": 1.8841475413765907,
+      "learning_rate": 3.3005271201366308e-06,
+      "loss": 0.818,
+      "step": 5389
+    },
+    {
+      "epoch": 0.7418111753371869,
+      "grad_norm": 1.8620515112203149,
+      "learning_rate": 3.297218555345397e-06,
+      "loss": 0.8327,
+      "step": 5390
+    },
+    {
+      "epoch": 0.7419488026424442,
+      "grad_norm": 1.6206115198537818,
+      "learning_rate": 3.293911322305092e-06,
+      "loss": 0.9443,
+      "step": 5391
+    },
+    {
+      "epoch": 0.7420864299477016,
+      "grad_norm": 1.8720287500960495,
+      "learning_rate": 3.290605421672808e-06,
+      "loss": 0.9364,
+      "step": 5392
+    },
+    {
+      "epoch": 0.742224057252959,
+      "grad_norm": 1.8095489134616263,
+      "learning_rate": 3.287300854105385e-06,
+      "loss": 0.9045,
+      "step": 5393
+    },
+    {
+      "epoch": 0.7423616845582164,
+      "grad_norm": 1.5721489510668316,
+      "learning_rate": 3.283997620259395e-06,
+      "loss": 0.903,
+      "step": 5394
+    },
+    {
+      "epoch": 0.7424993118634737,
+      "grad_norm": 1.5783819784498287,
+      "learning_rate": 3.280695720791146e-06,
+      "loss": 0.879,
+      "step": 5395
+    },
+    {
+      "epoch": 0.7426369391687311,
+      "grad_norm": 1.6398086761186235,
+      "learning_rate": 3.277395156356681e-06,
+      "loss": 0.9216,
+      "step": 5396
+    },
+    {
+      "epoch": 0.7427745664739884,
+      "grad_norm": 1.8808392007775487,
+      "learning_rate": 3.274095927611771e-06,
+      "loss": 0.9967,
+      "step": 5397
+    },
+    {
+      "epoch": 0.7429121937792458,
+      "grad_norm": 2.019374211119897,
+      "learning_rate": 3.2707980352119406e-06,
+      "loss": 0.8459,
+      "step": 5398
+    },
+    {
+      "epoch": 0.7430498210845031,
+      "grad_norm": 1.7630190873828178,
+      "learning_rate": 3.2675014798124207e-06,
+      "loss": 0.9641,
+      "step": 5399
+    },
+    {
+      "epoch": 0.7431874483897605,
+      "grad_norm": 1.6530262546829535,
+      "learning_rate": 3.264206262068205e-06,
+      "loss": 0.8669,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7433250756950179,
+      "grad_norm": 1.9922344982339548,
+      "learning_rate": 3.2609123826339963e-06,
+      "loss": 0.8798,
+      "step": 5401
+    },
+    {
+      "epoch": 0.7434627030002753,
+      "grad_norm": 1.8617786728056946,
+      "learning_rate": 3.2576198421642546e-06,
+      "loss": 0.9115,
+      "step": 5402
+    },
+    {
+      "epoch": 0.7436003303055326,
+      "grad_norm": 1.8934102227703686,
+      "learning_rate": 3.2543286413131514e-06,
+      "loss": 0.9282,
+      "step": 5403
+    },
+    {
+      "epoch": 0.74373795761079,
+      "grad_norm": 2.122019153131676,
+      "learning_rate": 3.251038780734612e-06,
+      "loss": 0.9306,
+      "step": 5404
+    },
+    {
+      "epoch": 0.7438755849160473,
+      "grad_norm": 1.905915958054961,
+      "learning_rate": 3.247750261082284e-06,
+      "loss": 0.8845,
+      "step": 5405
+    },
+    {
+      "epoch": 0.7440132122213047,
+      "grad_norm": 1.7314551044514395,
+      "learning_rate": 3.244463083009551e-06,
+      "loss": 0.8916,
+      "step": 5406
+    },
+    {
+      "epoch": 0.744150839526562,
+      "grad_norm": 2.073314583404827,
+      "learning_rate": 3.2411772471695304e-06,
+      "loss": 0.8556,
+      "step": 5407
+    },
+    {
+      "epoch": 0.7442884668318195,
+      "grad_norm": 2.126073146155783,
+      "learning_rate": 3.237892754215072e-06,
+      "loss": 0.9129,
+      "step": 5408
+    },
+    {
+      "epoch": 0.7444260941370768,
+      "grad_norm": 2.1715746779703746,
+      "learning_rate": 3.2346096047987597e-06,
+      "loss": 0.9109,
+      "step": 5409
+    },
+    {
+      "epoch": 0.7445637214423342,
+      "grad_norm": 1.9741336782657246,
+      "learning_rate": 3.231327799572912e-06,
+      "loss": 1.0158,
+      "step": 5410
+    },
+    {
+      "epoch": 0.7447013487475915,
+      "grad_norm": 1.8554458166182735,
+      "learning_rate": 3.2280473391895763e-06,
+      "loss": 0.9327,
+      "step": 5411
+    },
+    {
+      "epoch": 0.7448389760528489,
+      "grad_norm": 1.7460568193649428,
+      "learning_rate": 3.2247682243005374e-06,
+      "loss": 1.014,
+      "step": 5412
+    },
+    {
+      "epoch": 0.7449766033581062,
+      "grad_norm": 1.9326093311009171,
+      "learning_rate": 3.221490455557309e-06,
+      "loss": 0.92,
+      "step": 5413
+    },
+    {
+      "epoch": 0.7451142306633636,
+      "grad_norm": 2.2598726718785245,
+      "learning_rate": 3.218214033611139e-06,
+      "loss": 0.9249,
+      "step": 5414
+    },
+    {
+      "epoch": 0.7452518579686209,
+      "grad_norm": 1.6448314284586654,
+      "learning_rate": 3.2149389591130085e-06,
+      "loss": 0.9373,
+      "step": 5415
+    },
+    {
+      "epoch": 0.7453894852738784,
+      "grad_norm": 1.6574154005581918,
+      "learning_rate": 3.21166523271363e-06,
+      "loss": 0.8915,
+      "step": 5416
+    },
+    {
+      "epoch": 0.7455271125791357,
+      "grad_norm": 1.9045173034951033,
+      "learning_rate": 3.208392855063447e-06,
+      "loss": 0.9118,
+      "step": 5417
+    },
+    {
+      "epoch": 0.7456647398843931,
+      "grad_norm": 1.9639611833758877,
+      "learning_rate": 3.205121826812636e-06,
+      "loss": 0.8621,
+      "step": 5418
+    },
+    {
+      "epoch": 0.7458023671896504,
+      "grad_norm": 1.7553652473411705,
+      "learning_rate": 3.2018521486111074e-06,
+      "loss": 0.8826,
+      "step": 5419
+    },
+    {
+      "epoch": 0.7459399944949078,
+      "grad_norm": 1.8359439343973158,
+      "learning_rate": 3.1985838211084963e-06,
+      "loss": 0.9303,
+      "step": 5420
+    },
+    {
+      "epoch": 0.7460776218001651,
+      "grad_norm": 2.0571710726519603,
+      "learning_rate": 3.195316844954186e-06,
+      "loss": 0.8348,
+      "step": 5421
+    },
+    {
+      "epoch": 0.7462152491054225,
+      "grad_norm": 1.9130877046369414,
+      "learning_rate": 3.1920512207972655e-06,
+      "loss": 0.9857,
+      "step": 5422
+    },
+    {
+      "epoch": 0.7463528764106799,
+      "grad_norm": 1.89377256644532,
+      "learning_rate": 3.188786949286583e-06,
+      "loss": 0.9464,
+      "step": 5423
+    },
+    {
+      "epoch": 0.7464905037159373,
+      "grad_norm": 1.9778426903734947,
+      "learning_rate": 3.1855240310706958e-06,
+      "loss": 0.8644,
+      "step": 5424
+    },
+    {
+      "epoch": 0.7466281310211946,
+      "grad_norm": 1.8355999065289066,
+      "learning_rate": 3.182262466797904e-06,
+      "loss": 0.8305,
+      "step": 5425
+    },
+    {
+      "epoch": 0.746765758326452,
+      "grad_norm": 1.698401098014406,
+      "learning_rate": 3.179002257116235e-06,
+      "loss": 0.9272,
+      "step": 5426
+    },
+    {
+      "epoch": 0.7469033856317093,
+      "grad_norm": 1.7610010664913602,
+      "learning_rate": 3.1757434026734446e-06,
+      "loss": 0.9721,
+      "step": 5427
+    },
+    {
+      "epoch": 0.7470410129369667,
+      "grad_norm": 1.9070882551386175,
+      "learning_rate": 3.172485904117033e-06,
+      "loss": 0.9008,
+      "step": 5428
+    },
+    {
+      "epoch": 0.747178640242224,
+      "grad_norm": 1.966270733438746,
+      "learning_rate": 3.169229762094207e-06,
+      "loss": 0.9327,
+      "step": 5429
+    },
+    {
+      "epoch": 0.7473162675474814,
+      "grad_norm": 1.8998540306605942,
+      "learning_rate": 3.165974977251931e-06,
+      "loss": 0.9322,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7474538948527388,
+      "grad_norm": 1.6680477255976567,
+      "learning_rate": 3.1627215502368726e-06,
+      "loss": 0.8748,
+      "step": 5431
+    },
+    {
+      "epoch": 0.7475915221579962,
+      "grad_norm": 1.7629237799269026,
+      "learning_rate": 3.159469481695456e-06,
+      "loss": 0.9392,
+      "step": 5432
+    },
+    {
+      "epoch": 0.7477291494632535,
+      "grad_norm": 1.782760631063199,
+      "learning_rate": 3.1562187722738113e-06,
+      "loss": 0.9774,
+      "step": 5433
+    },
+    {
+      "epoch": 0.7478667767685109,
+      "grad_norm": 1.9147972160786704,
+      "learning_rate": 3.1529694226178186e-06,
+      "loss": 0.9148,
+      "step": 5434
+    },
+    {
+      "epoch": 0.7480044040737682,
+      "grad_norm": 2.038299124603252,
+      "learning_rate": 3.149721433373075e-06,
+      "loss": 0.892,
+      "step": 5435
+    },
+    {
+      "epoch": 0.7481420313790256,
+      "grad_norm": 1.850721578082473,
+      "learning_rate": 3.1464748051849135e-06,
+      "loss": 0.8826,
+      "step": 5436
+    },
+    {
+      "epoch": 0.7482796586842829,
+      "grad_norm": 2.2890755680903503,
+      "learning_rate": 3.1432295386983935e-06,
+      "loss": 0.8822,
+      "step": 5437
+    },
+    {
+      "epoch": 0.7484172859895403,
+      "grad_norm": 1.8356698759222332,
+      "learning_rate": 3.1399856345583046e-06,
+      "loss": 0.9434,
+      "step": 5438
+    },
+    {
+      "epoch": 0.7485549132947977,
+      "grad_norm": 1.6864293688590377,
+      "learning_rate": 3.136743093409168e-06,
+      "loss": 0.8998,
+      "step": 5439
+    },
+    {
+      "epoch": 0.7486925406000551,
+      "grad_norm": 1.8881792360705087,
+      "learning_rate": 3.1335019158952296e-06,
+      "loss": 0.8823,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7488301679053124,
+      "grad_norm": 1.7727556113488412,
+      "learning_rate": 3.1302621026604696e-06,
+      "loss": 0.9498,
+      "step": 5441
+    },
+    {
+      "epoch": 0.7489677952105698,
+      "grad_norm": 1.5928286656453894,
+      "learning_rate": 3.127023654348592e-06,
+      "loss": 0.8553,
+      "step": 5442
+    },
+    {
+      "epoch": 0.7491054225158271,
+      "grad_norm": 1.7012005203422216,
+      "learning_rate": 3.1237865716030347e-06,
+      "loss": 0.957,
+      "step": 5443
+    },
+    {
+      "epoch": 0.7492430498210845,
+      "grad_norm": 1.7533626663946758,
+      "learning_rate": 3.1205508550669596e-06,
+      "loss": 0.915,
+      "step": 5444
+    },
+    {
+      "epoch": 0.7493806771263418,
+      "grad_norm": 1.8168089720860214,
+      "learning_rate": 3.117316505383261e-06,
+      "loss": 0.8651,
+      "step": 5445
+    },
+    {
+      "epoch": 0.7495183044315993,
+      "grad_norm": 2.1551005442819045,
+      "learning_rate": 3.114083523194559e-06,
+      "loss": 0.9812,
+      "step": 5446
+    },
+    {
+      "epoch": 0.7496559317368566,
+      "grad_norm": 1.7441247420169146,
+      "learning_rate": 3.110851909143202e-06,
+      "loss": 0.8795,
+      "step": 5447
+    },
+    {
+      "epoch": 0.749793559042114,
+      "grad_norm": 1.5585055988851828,
+      "learning_rate": 3.1076216638712696e-06,
+      "loss": 0.8621,
+      "step": 5448
+    },
+    {
+      "epoch": 0.7499311863473713,
+      "grad_norm": 1.692233338774258,
+      "learning_rate": 3.104392788020566e-06,
+      "loss": 0.9136,
+      "step": 5449
+    },
+    {
+      "epoch": 0.7500688136526287,
+      "grad_norm": 1.7578892913906397,
+      "learning_rate": 3.101165282232624e-06,
+      "loss": 0.8977,
+      "step": 5450
+    },
+    {
+      "epoch": 0.750206440957886,
+      "grad_norm": 1.899213855859315,
+      "learning_rate": 3.097939147148706e-06,
+      "loss": 0.96,
+      "step": 5451
+    },
+    {
+      "epoch": 0.7503440682631434,
+      "grad_norm": 1.6731229211911267,
+      "learning_rate": 3.0947143834097983e-06,
+      "loss": 0.9947,
+      "step": 5452
+    },
+    {
+      "epoch": 0.7504816955684007,
+      "grad_norm": 2.065107727898624,
+      "learning_rate": 3.091490991656625e-06,
+      "loss": 0.9305,
+      "step": 5453
+    },
+    {
+      "epoch": 0.7506193228736582,
+      "grad_norm": 1.79581842507103,
+      "learning_rate": 3.088268972529619e-06,
+      "loss": 0.9765,
+      "step": 5454
+    },
+    {
+      "epoch": 0.7507569501789155,
+      "grad_norm": 1.8675997290600785,
+      "learning_rate": 3.0850483266689623e-06,
+      "loss": 0.907,
+      "step": 5455
+    },
+    {
+      "epoch": 0.7508945774841729,
+      "grad_norm": 1.8161503562128376,
+      "learning_rate": 3.0818290547145447e-06,
+      "loss": 0.9117,
+      "step": 5456
+    },
+    {
+      "epoch": 0.7510322047894302,
+      "grad_norm": 1.7219725302622113,
+      "learning_rate": 3.078611157305995e-06,
+      "loss": 0.7934,
+      "step": 5457
+    },
+    {
+      "epoch": 0.7511698320946876,
+      "grad_norm": 1.7410702982127422,
+      "learning_rate": 3.075394635082665e-06,
+      "loss": 0.8775,
+      "step": 5458
+    },
+    {
+      "epoch": 0.7513074593999449,
+      "grad_norm": 1.7522922381402224,
+      "learning_rate": 3.07217948868363e-06,
+      "loss": 0.8636,
+      "step": 5459
+    },
+    {
+      "epoch": 0.7514450867052023,
+      "grad_norm": 1.9211295230424688,
+      "learning_rate": 3.068965718747705e-06,
+      "loss": 0.8494,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7515827140104597,
+      "grad_norm": 1.9091904274798164,
+      "learning_rate": 3.0657533259134098e-06,
+      "loss": 0.9094,
+      "step": 5461
+    },
+    {
+      "epoch": 0.7517203413157171,
+      "grad_norm": 1.8109550401946803,
+      "learning_rate": 3.0625423108190155e-06,
+      "loss": 0.9752,
+      "step": 5462
+    },
+    {
+      "epoch": 0.7518579686209744,
+      "grad_norm": 1.944562176496586,
+      "learning_rate": 3.0593326741024955e-06,
+      "loss": 0.9199,
+      "step": 5463
+    },
+    {
+      "epoch": 0.7519955959262318,
+      "grad_norm": 2.0765669766029458,
+      "learning_rate": 3.0561244164015714e-06,
+      "loss": 0.8715,
+      "step": 5464
+    },
+    {
+      "epoch": 0.7521332232314891,
+      "grad_norm": 1.7798393655381646,
+      "learning_rate": 3.052917538353667e-06,
+      "loss": 0.9554,
+      "step": 5465
+    },
+    {
+      "epoch": 0.7522708505367465,
+      "grad_norm": 1.779900375297692,
+      "learning_rate": 3.049712040595957e-06,
+      "loss": 0.9382,
+      "step": 5466
+    },
+    {
+      "epoch": 0.7524084778420038,
+      "grad_norm": 1.6912992199611379,
+      "learning_rate": 3.046507923765325e-06,
+      "loss": 0.9452,
+      "step": 5467
+    },
+    {
+      "epoch": 0.7525461051472612,
+      "grad_norm": 1.6794501772391903,
+      "learning_rate": 3.0433051884983856e-06,
+      "loss": 0.9039,
+      "step": 5468
+    },
+    {
+      "epoch": 0.7526837324525186,
+      "grad_norm": 1.9748418667695296,
+      "learning_rate": 3.0401038354314794e-06,
+      "loss": 0.8883,
+      "step": 5469
+    },
+    {
+      "epoch": 0.752821359757776,
+      "grad_norm": 1.5999923957778552,
+      "learning_rate": 3.036903865200669e-06,
+      "loss": 0.952,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7529589870630333,
+      "grad_norm": 1.7495976309053185,
+      "learning_rate": 3.0337052784417477e-06,
+      "loss": 0.8686,
+      "step": 5471
+    },
+    {
+      "epoch": 0.7530966143682907,
+      "grad_norm": 1.8197347386555367,
+      "learning_rate": 3.030508075790228e-06,
+      "loss": 0.874,
+      "step": 5472
+    },
+    {
+      "epoch": 0.753234241673548,
+      "grad_norm": 1.5545893909023443,
+      "learning_rate": 3.027312257881353e-06,
+      "loss": 0.9594,
+      "step": 5473
+    },
+    {
+      "epoch": 0.7533718689788054,
+      "grad_norm": 1.6522512550449608,
+      "learning_rate": 3.0241178253500857e-06,
+      "loss": 0.8892,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7535094962840627,
+      "grad_norm": 1.7011374341843077,
+      "learning_rate": 3.020924778831117e-06,
+      "loss": 0.8872,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7536471235893201,
+      "grad_norm": 1.7006397961277306,
+      "learning_rate": 3.017733118958862e-06,
+      "loss": 0.8801,
+      "step": 5476
+    },
+    {
+      "epoch": 0.7537847508945775,
+      "grad_norm": 1.6647266524058915,
+      "learning_rate": 3.0145428463674597e-06,
+      "loss": 0.9,
+      "step": 5477
+    },
+    {
+      "epoch": 0.7539223781998349,
+      "grad_norm": 1.7982351484736916,
+      "learning_rate": 3.0113539616907726e-06,
+      "loss": 0.9467,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7540600055050922,
+      "grad_norm": 1.8227822116820565,
+      "learning_rate": 3.0081664655623897e-06,
+      "loss": 0.9322,
+      "step": 5479
+    },
+    {
+      "epoch": 0.7541976328103496,
+      "grad_norm": 1.6843341248558537,
+      "learning_rate": 3.0049803586156224e-06,
+      "loss": 0.9248,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7543352601156069,
+      "grad_norm": 1.6785061662391683,
+      "learning_rate": 3.001795641483506e-06,
+      "loss": 0.9072,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7544728874208643,
+      "grad_norm": 1.6926379270842984,
+      "learning_rate": 2.9986123147987978e-06,
+      "loss": 0.9043,
+      "step": 5482
+    },
+    {
+      "epoch": 0.7546105147261216,
+      "grad_norm": 1.8821567394872063,
+      "learning_rate": 2.995430379193991e-06,
+      "loss": 0.8038,
+      "step": 5483
+    },
+    {
+      "epoch": 0.7547481420313791,
+      "grad_norm": 1.8331406712016125,
+      "learning_rate": 2.9922498353012786e-06,
+      "loss": 0.9074,
+      "step": 5484
+    },
+    {
+      "epoch": 0.7548857693366364,
+      "grad_norm": 2.015963975382853,
+      "learning_rate": 2.9890706837526052e-06,
+      "loss": 0.8145,
+      "step": 5485
+    },
+    {
+      "epoch": 0.7550233966418938,
+      "grad_norm": 1.792575701786076,
+      "learning_rate": 2.985892925179611e-06,
+      "loss": 0.9103,
+      "step": 5486
+    },
+    {
+      "epoch": 0.7551610239471511,
+      "grad_norm": 1.7821774333884204,
+      "learning_rate": 2.9827165602136854e-06,
+      "loss": 0.9331,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7552986512524085,
+      "grad_norm": 1.6431258839444023,
+      "learning_rate": 2.9795415894859216e-06,
+      "loss": 0.8412,
+      "step": 5488
+    },
+    {
+      "epoch": 0.7554362785576658,
+      "grad_norm": 1.5925118471929725,
+      "learning_rate": 2.97636801362714e-06,
+      "loss": 0.8684,
+      "step": 5489
+    },
+    {
+      "epoch": 0.7555739058629232,
+      "grad_norm": 1.7743398979774374,
+      "learning_rate": 2.9731958332678977e-06,
+      "loss": 0.9485,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7557115331681805,
+      "grad_norm": 1.886578798633562,
+      "learning_rate": 2.9700250490384507e-06,
+      "loss": 0.8733,
+      "step": 5491
+    },
+    {
+      "epoch": 0.755849160473438,
+      "grad_norm": 1.6380299942802359,
+      "learning_rate": 2.9668556615688037e-06,
+      "loss": 0.9024,
+      "step": 5492
+    },
+    {
+      "epoch": 0.7559867877786953,
+      "grad_norm": 1.6773900852635475,
+      "learning_rate": 2.9636876714886562e-06,
+      "loss": 0.8915,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7561244150839527,
+      "grad_norm": 1.6217614547621944,
+      "learning_rate": 2.9605210794274588e-06,
+      "loss": 0.9215,
+      "step": 5494
+    },
+    {
+      "epoch": 0.75626204238921,
+      "grad_norm": 1.7554982483425272,
+      "learning_rate": 2.957355886014356e-06,
+      "loss": 0.8722,
+      "step": 5495
+    },
+    {
+      "epoch": 0.7563996696944674,
+      "grad_norm": 1.7928245987801514,
+      "learning_rate": 2.9541920918782375e-06,
+      "loss": 0.9093,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7565372969997247,
+      "grad_norm": 1.7615719020511682,
+      "learning_rate": 2.9510296976477047e-06,
+      "loss": 1.0024,
+      "step": 5497
+    },
+    {
+      "epoch": 0.7566749243049821,
+      "grad_norm": 1.5613812349936373,
+      "learning_rate": 2.9478687039510823e-06,
+      "loss": 0.9034,
+      "step": 5498
+    },
+    {
+      "epoch": 0.7568125516102395,
+      "grad_norm": 1.8099573500175263,
+      "learning_rate": 2.9447091114164137e-06,
+      "loss": 0.9357,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7569501789154969,
+      "grad_norm": 1.691815261150095,
+      "learning_rate": 2.9415509206714686e-06,
+      "loss": 0.8604,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7570878062207542,
+      "grad_norm": 1.6306825811415881,
+      "learning_rate": 2.938394132343736e-06,
+      "loss": 0.9175,
+      "step": 5501
+    },
+    {
+      "epoch": 0.7572254335260116,
+      "grad_norm": 1.5867109916951256,
+      "learning_rate": 2.9352387470604272e-06,
+      "loss": 0.8968,
+      "step": 5502
+    },
+    {
+      "epoch": 0.7573630608312689,
+      "grad_norm": 1.615124336960378,
+      "learning_rate": 2.9320847654484717e-06,
+      "loss": 0.9449,
+      "step": 5503
+    },
+    {
+      "epoch": 0.7575006881365263,
+      "grad_norm": 1.6235135069323288,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.9055,
+      "step": 5504
+    },
+    {
+      "epoch": 0.7576383154417836,
+      "grad_norm": 1.744416354215708,
+      "learning_rate": 2.9257810157449607e-06,
+      "loss": 0.9511,
+      "step": 5505
+    },
+    {
+      "epoch": 0.757775942747041,
+      "grad_norm": 1.7654497908380193,
+      "learning_rate": 2.922631248905873e-06,
+      "loss": 0.9349,
+      "step": 5506
+    },
+    {
+      "epoch": 0.7579135700522984,
+      "grad_norm": 1.8887775225031076,
+      "learning_rate": 2.919482888243077e-06,
+      "loss": 0.8787,
+      "step": 5507
+    },
+    {
+      "epoch": 0.7580511973575558,
+      "grad_norm": 2.0751444106240404,
+      "learning_rate": 2.9163359343821097e-06,
+      "loss": 0.9322,
+      "step": 5508
+    },
+    {
+      "epoch": 0.7581888246628131,
+      "grad_norm": 2.0425270185351634,
+      "learning_rate": 2.913190387948227e-06,
+      "loss": 0.906,
+      "step": 5509
+    },
+    {
+      "epoch": 0.7583264519680705,
+      "grad_norm": 1.8045879966559795,
+      "learning_rate": 2.9100462495664063e-06,
+      "loss": 0.8724,
+      "step": 5510
+    },
+    {
+      "epoch": 0.7584640792733278,
+      "grad_norm": 1.5413003269395515,
+      "learning_rate": 2.906903519861345e-06,
+      "loss": 0.9475,
+      "step": 5511
+    },
+    {
+      "epoch": 0.7586017065785852,
+      "grad_norm": 1.9239341678929622,
+      "learning_rate": 2.9037621994574607e-06,
+      "loss": 1.0228,
+      "step": 5512
+    },
+    {
+      "epoch": 0.7587393338838425,
+      "grad_norm": 1.6944041848823597,
+      "learning_rate": 2.9006222889788903e-06,
+      "loss": 0.9037,
+      "step": 5513
+    },
+    {
+      "epoch": 0.7588769611890999,
+      "grad_norm": 1.9024791369340794,
+      "learning_rate": 2.8974837890494876e-06,
+      "loss": 0.9412,
+      "step": 5514
+    },
+    {
+      "epoch": 0.7590145884943573,
+      "grad_norm": 1.9203269161579444,
+      "learning_rate": 2.8943467002928404e-06,
+      "loss": 0.9163,
+      "step": 5515
+    },
+    {
+      "epoch": 0.7591522157996147,
+      "grad_norm": 1.8710500457059325,
+      "learning_rate": 2.891211023332231e-06,
+      "loss": 0.8501,
+      "step": 5516
+    },
+    {
+      "epoch": 0.759289843104872,
+      "grad_norm": 1.7651689388637386,
+      "learning_rate": 2.88807675879069e-06,
+      "loss": 0.9424,
+      "step": 5517
+    },
+    {
+      "epoch": 0.7594274704101294,
+      "grad_norm": 1.6679558549842892,
+      "learning_rate": 2.884943907290938e-06,
+      "loss": 0.8733,
+      "step": 5518
+    },
+    {
+      "epoch": 0.7595650977153867,
+      "grad_norm": 1.8705669217832368,
+      "learning_rate": 2.881812469455444e-06,
+      "loss": 0.8998,
+      "step": 5519
+    },
+    {
+      "epoch": 0.7597027250206441,
+      "grad_norm": 1.892494155693648,
+      "learning_rate": 2.878682445906368e-06,
+      "loss": 0.8899,
+      "step": 5520
+    },
+    {
+      "epoch": 0.7598403523259014,
+      "grad_norm": 1.5493013241351175,
+      "learning_rate": 2.8755538372656132e-06,
+      "loss": 0.8785,
+      "step": 5521
+    },
+    {
+      "epoch": 0.7599779796311589,
+      "grad_norm": 1.8177794558789961,
+      "learning_rate": 2.872426644154791e-06,
+      "loss": 0.9468,
+      "step": 5522
+    },
+    {
+      "epoch": 0.7601156069364162,
+      "grad_norm": 1.5859483503571836,
+      "learning_rate": 2.8693008671952225e-06,
+      "loss": 0.9438,
+      "step": 5523
+    },
+    {
+      "epoch": 0.7602532342416736,
+      "grad_norm": 1.5978646805733026,
+      "learning_rate": 2.8661765070079695e-06,
+      "loss": 0.8968,
+      "step": 5524
+    },
+    {
+      "epoch": 0.7603908615469309,
+      "grad_norm": 1.744257849747236,
+      "learning_rate": 2.8630535642137867e-06,
+      "loss": 0.9645,
+      "step": 5525
+    },
+    {
+      "epoch": 0.7605284888521883,
+      "grad_norm": 1.6489470983330505,
+      "learning_rate": 2.8599320394331733e-06,
+      "loss": 0.8865,
+      "step": 5526
+    },
+    {
+      "epoch": 0.7606661161574456,
+      "grad_norm": 1.6862950081011878,
+      "learning_rate": 2.85681193328632e-06,
+      "loss": 0.9298,
+      "step": 5527
+    },
+    {
+      "epoch": 0.760803743462703,
+      "grad_norm": 1.8536580323135852,
+      "learning_rate": 2.853693246393159e-06,
+      "loss": 0.9197,
+      "step": 5528
+    },
+    {
+      "epoch": 0.7609413707679603,
+      "grad_norm": 1.802102131842383,
+      "learning_rate": 2.8505759793733277e-06,
+      "loss": 0.9562,
+      "step": 5529
+    },
+    {
+      "epoch": 0.7610789980732178,
+      "grad_norm": 7.4980599994822015,
+      "learning_rate": 2.8474601328461847e-06,
+      "loss": 0.9761,
+      "step": 5530
+    },
+    {
+      "epoch": 0.7612166253784751,
+      "grad_norm": 1.7240928427906432,
+      "learning_rate": 2.8443457074308055e-06,
+      "loss": 0.9187,
+      "step": 5531
+    },
+    {
+      "epoch": 0.7613542526837325,
+      "grad_norm": 1.6609117529759112,
+      "learning_rate": 2.841232703745984e-06,
+      "loss": 0.9496,
+      "step": 5532
+    },
+    {
+      "epoch": 0.7614918799889898,
+      "grad_norm": 1.7544038310263668,
+      "learning_rate": 2.8381211224102313e-06,
+      "loss": 0.9298,
+      "step": 5533
+    },
+    {
+      "epoch": 0.7616295072942472,
+      "grad_norm": 1.5965033029381335,
+      "learning_rate": 2.835010964041777e-06,
+      "loss": 0.9409,
+      "step": 5534
+    },
+    {
+      "epoch": 0.7617671345995045,
+      "grad_norm": 1.5423369374412463,
+      "learning_rate": 2.831902229258565e-06,
+      "loss": 0.8482,
+      "step": 5535
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 2.0701862558609245,
+      "learning_rate": 2.8287949186782594e-06,
+      "loss": 0.8841,
+      "step": 5536
+    },
+    {
+      "epoch": 0.7620423892100193,
+      "grad_norm": 1.9792824414926642,
+      "learning_rate": 2.8256890329182417e-06,
+      "loss": 0.8336,
+      "step": 5537
+    },
+    {
+      "epoch": 0.7621800165152767,
+      "grad_norm": 1.6662758550941312,
+      "learning_rate": 2.822584572595606e-06,
+      "loss": 0.9393,
+      "step": 5538
+    },
+    {
+      "epoch": 0.762317643820534,
+      "grad_norm": 1.7330159306255433,
+      "learning_rate": 2.819481538327169e-06,
+      "loss": 0.8795,
+      "step": 5539
+    },
+    {
+      "epoch": 0.7624552711257914,
+      "grad_norm": 1.6559720243117015,
+      "learning_rate": 2.816379930729458e-06,
+      "loss": 0.8982,
+      "step": 5540
+    },
+    {
+      "epoch": 0.7625928984310487,
+      "grad_norm": 1.7976406610939581,
+      "learning_rate": 2.813279750418724e-06,
+      "loss": 0.991,
+      "step": 5541
+    },
+    {
+      "epoch": 0.7627305257363061,
+      "grad_norm": 1.6581970044420524,
+      "learning_rate": 2.8101809980109264e-06,
+      "loss": 0.9047,
+      "step": 5542
+    },
+    {
+      "epoch": 0.7628681530415634,
+      "grad_norm": 1.9072540086696796,
+      "learning_rate": 2.8070836741217466e-06,
+      "loss": 0.8754,
+      "step": 5543
+    },
+    {
+      "epoch": 0.7630057803468208,
+      "grad_norm": 1.8397962915543937,
+      "learning_rate": 2.803987779366578e-06,
+      "loss": 0.8954,
+      "step": 5544
+    },
+    {
+      "epoch": 0.7631434076520782,
+      "grad_norm": 1.7882342981135908,
+      "learning_rate": 2.800893314360541e-06,
+      "loss": 0.846,
+      "step": 5545
+    },
+    {
+      "epoch": 0.7632810349573356,
+      "grad_norm": 1.6480205598859798,
+      "learning_rate": 2.797800279718451e-06,
+      "loss": 0.8966,
+      "step": 5546
+    },
+    {
+      "epoch": 0.7634186622625929,
+      "grad_norm": 1.8147891241535186,
+      "learning_rate": 2.7947086760548647e-06,
+      "loss": 0.9242,
+      "step": 5547
+    },
+    {
+      "epoch": 0.7635562895678503,
+      "grad_norm": 1.4550626352429987,
+      "learning_rate": 2.7916185039840282e-06,
+      "loss": 0.9313,
+      "step": 5548
+    },
+    {
+      "epoch": 0.7636939168731076,
+      "grad_norm": 1.741547953639759,
+      "learning_rate": 2.7885297641199293e-06,
+      "loss": 0.9002,
+      "step": 5549
+    },
+    {
+      "epoch": 0.763831544178365,
+      "grad_norm": 2.057413370539198,
+      "learning_rate": 2.7854424570762463e-06,
+      "loss": 0.8646,
+      "step": 5550
+    },
+    {
+      "epoch": 0.7639691714836223,
+      "grad_norm": 1.8737370235337334,
+      "learning_rate": 2.7823565834663937e-06,
+      "loss": 0.894,
+      "step": 5551
+    },
+    {
+      "epoch": 0.7641067987888797,
+      "grad_norm": 1.633193244391118,
+      "learning_rate": 2.7792721439034873e-06,
+      "loss": 0.8694,
+      "step": 5552
+    },
+    {
+      "epoch": 0.7642444260941371,
+      "grad_norm": 1.7475690868797231,
+      "learning_rate": 2.776189139000367e-06,
+      "loss": 0.9291,
+      "step": 5553
+    },
+    {
+      "epoch": 0.7643820533993945,
+      "grad_norm": 2.215703890484703,
+      "learning_rate": 2.773107569369584e-06,
+      "loss": 0.9099,
+      "step": 5554
+    },
+    {
+      "epoch": 0.7645196807046518,
+      "grad_norm": 1.5391569738851028,
+      "learning_rate": 2.7700274356233958e-06,
+      "loss": 0.945,
+      "step": 5555
+    },
+    {
+      "epoch": 0.7646573080099092,
+      "grad_norm": 1.5810782493414817,
+      "learning_rate": 2.7669487383737936e-06,
+      "loss": 0.8103,
+      "step": 5556
+    },
+    {
+      "epoch": 0.7647949353151665,
+      "grad_norm": 1.8017350757286585,
+      "learning_rate": 2.7638714782324627e-06,
+      "loss": 0.8304,
+      "step": 5557
+    },
+    {
+      "epoch": 0.7649325626204239,
+      "grad_norm": 1.646079182649469,
+      "learning_rate": 2.7607956558108197e-06,
+      "loss": 0.935,
+      "step": 5558
+    },
+    {
+      "epoch": 0.7650701899256812,
+      "grad_norm": 1.8350396285951502,
+      "learning_rate": 2.7577212717199853e-06,
+      "loss": 0.8771,
+      "step": 5559
+    },
+    {
+      "epoch": 0.7652078172309387,
+      "grad_norm": 1.8036222782929265,
+      "learning_rate": 2.754648326570799e-06,
+      "loss": 0.8798,
+      "step": 5560
+    },
+    {
+      "epoch": 0.765345444536196,
+      "grad_norm": 1.798053074340803,
+      "learning_rate": 2.7515768209738116e-06,
+      "loss": 0.8635,
+      "step": 5561
+    },
+    {
+      "epoch": 0.7654830718414534,
+      "grad_norm": 1.8278432340871467,
+      "learning_rate": 2.7485067555392907e-06,
+      "loss": 0.9069,
+      "step": 5562
+    },
+    {
+      "epoch": 0.7656206991467107,
+      "grad_norm": 1.6124224736592274,
+      "learning_rate": 2.745438130877215e-06,
+      "loss": 0.9207,
+      "step": 5563
+    },
+    {
+      "epoch": 0.7657583264519681,
+      "grad_norm": 1.659774624739692,
+      "learning_rate": 2.742370947597278e-06,
+      "loss": 0.9167,
+      "step": 5564
+    },
+    {
+      "epoch": 0.7658959537572254,
+      "grad_norm": 1.741125298830129,
+      "learning_rate": 2.7393052063088877e-06,
+      "loss": 0.9615,
+      "step": 5565
+    },
+    {
+      "epoch": 0.7660335810624828,
+      "grad_norm": 2.021968986307654,
+      "learning_rate": 2.736240907621165e-06,
+      "loss": 0.9185,
+      "step": 5566
+    },
+    {
+      "epoch": 0.7661712083677401,
+      "grad_norm": 1.7137102649838716,
+      "learning_rate": 2.733178052142944e-06,
+      "loss": 0.9211,
+      "step": 5567
+    },
+    {
+      "epoch": 0.7663088356729976,
+      "grad_norm": 1.6547548816984932,
+      "learning_rate": 2.7301166404827716e-06,
+      "loss": 0.851,
+      "step": 5568
+    },
+    {
+      "epoch": 0.7664464629782549,
+      "grad_norm": 1.8962841242426112,
+      "learning_rate": 2.72705667324891e-06,
+      "loss": 0.8603,
+      "step": 5569
+    },
+    {
+      "epoch": 0.7665840902835123,
+      "grad_norm": 1.6615353939241628,
+      "learning_rate": 2.7239981510493306e-06,
+      "loss": 0.8274,
+      "step": 5570
+    },
+    {
+      "epoch": 0.7667217175887696,
+      "grad_norm": 2.0019135520893783,
+      "learning_rate": 2.720941074491722e-06,
+      "loss": 0.9123,
+      "step": 5571
+    },
+    {
+      "epoch": 0.766859344894027,
+      "grad_norm": 1.7858963227401488,
+      "learning_rate": 2.7178854441834813e-06,
+      "loss": 0.9391,
+      "step": 5572
+    },
+    {
+      "epoch": 0.7669969721992843,
+      "grad_norm": 1.8425243572647436,
+      "learning_rate": 2.7148312607317227e-06,
+      "loss": 0.9804,
+      "step": 5573
+    },
+    {
+      "epoch": 0.7671345995045417,
+      "grad_norm": 1.6712628732838954,
+      "learning_rate": 2.711778524743268e-06,
+      "loss": 0.8998,
+      "step": 5574
+    },
+    {
+      "epoch": 0.7672722268097991,
+      "grad_norm": 1.9097480668124736,
+      "learning_rate": 2.7087272368246574e-06,
+      "loss": 0.9437,
+      "step": 5575
+    },
+    {
+      "epoch": 0.7674098541150565,
+      "grad_norm": 1.8137934509041271,
+      "learning_rate": 2.705677397582134e-06,
+      "loss": 0.909,
+      "step": 5576
+    },
+    {
+      "epoch": 0.7675474814203138,
+      "grad_norm": 1.6433406686753138,
+      "learning_rate": 2.7026290076216697e-06,
+      "loss": 0.9014,
+      "step": 5577
+    },
+    {
+      "epoch": 0.7676851087255712,
+      "grad_norm": 1.695455110697961,
+      "learning_rate": 2.699582067548925e-06,
+      "loss": 0.9432,
+      "step": 5578
+    },
+    {
+      "epoch": 0.7678227360308285,
+      "grad_norm": 1.9173339554630964,
+      "learning_rate": 2.6965365779692987e-06,
+      "loss": 0.8568,
+      "step": 5579
+    },
+    {
+      "epoch": 0.7679603633360859,
+      "grad_norm": 1.7213110531528557,
+      "learning_rate": 2.6934925394878743e-06,
+      "loss": 0.9077,
+      "step": 5580
+    },
+    {
+      "epoch": 0.7680979906413432,
+      "grad_norm": 1.6161078213263744,
+      "learning_rate": 2.690449952709472e-06,
+      "loss": 0.9079,
+      "step": 5581
+    },
+    {
+      "epoch": 0.7682356179466006,
+      "grad_norm": 1.7350256739527512,
+      "learning_rate": 2.6874088182386027e-06,
+      "loss": 0.8484,
+      "step": 5582
+    },
+    {
+      "epoch": 0.768373245251858,
+      "grad_norm": 1.7351729364416828,
+      "learning_rate": 2.6843691366795054e-06,
+      "loss": 0.8732,
+      "step": 5583
+    },
+    {
+      "epoch": 0.7685108725571154,
+      "grad_norm": 1.829546330306816,
+      "learning_rate": 2.6813309086361196e-06,
+      "loss": 0.9107,
+      "step": 5584
+    },
+    {
+      "epoch": 0.7686484998623727,
+      "grad_norm": 1.8368923817370513,
+      "learning_rate": 2.678294134712102e-06,
+      "loss": 0.8851,
+      "step": 5585
+    },
+    {
+      "epoch": 0.7687861271676301,
+      "grad_norm": 1.820002017620623,
+      "learning_rate": 2.6752588155108183e-06,
+      "loss": 0.9457,
+      "step": 5586
+    },
+    {
+      "epoch": 0.7689237544728874,
+      "grad_norm": 1.8044278849154478,
+      "learning_rate": 2.6722249516353382e-06,
+      "loss": 0.859,
+      "step": 5587
+    },
+    {
+      "epoch": 0.7690613817781448,
+      "grad_norm": 1.7890049334154676,
+      "learning_rate": 2.6691925436884603e-06,
+      "loss": 0.9252,
+      "step": 5588
+    },
+    {
+      "epoch": 0.7691990090834021,
+      "grad_norm": 1.724298974855598,
+      "learning_rate": 2.6661615922726692e-06,
+      "loss": 0.8093,
+      "step": 5589
+    },
+    {
+      "epoch": 0.7693366363886595,
+      "grad_norm": 1.8361263640653,
+      "learning_rate": 2.663132097990183e-06,
+      "loss": 0.8658,
+      "step": 5590
+    },
+    {
+      "epoch": 0.7694742636939169,
+      "grad_norm": 1.7118805276650495,
+      "learning_rate": 2.6601040614429186e-06,
+      "loss": 0.8975,
+      "step": 5591
+    },
+    {
+      "epoch": 0.7696118909991743,
+      "grad_norm": 2.1599420951068535,
+      "learning_rate": 2.657077483232505e-06,
+      "loss": 0.9081,
+      "step": 5592
+    },
+    {
+      "epoch": 0.7697495183044316,
+      "grad_norm": 1.72506789788939,
+      "learning_rate": 2.6540523639602823e-06,
+      "loss": 0.8892,
+      "step": 5593
+    },
+    {
+      "epoch": 0.769887145609689,
+      "grad_norm": 1.6583440723975187,
+      "learning_rate": 2.6510287042272996e-06,
+      "loss": 0.9345,
+      "step": 5594
+    },
+    {
+      "epoch": 0.7700247729149463,
+      "grad_norm": 1.9851081427458022,
+      "learning_rate": 2.6480065046343173e-06,
+      "loss": 0.8677,
+      "step": 5595
+    },
+    {
+      "epoch": 0.7701624002202037,
+      "grad_norm": 1.8524054530379794,
+      "learning_rate": 2.6449857657818034e-06,
+      "loss": 0.9382,
+      "step": 5596
+    },
+    {
+      "epoch": 0.770300027525461,
+      "grad_norm": 1.7945422263820077,
+      "learning_rate": 2.6419664882699404e-06,
+      "loss": 0.9493,
+      "step": 5597
+    },
+    {
+      "epoch": 0.7704376548307185,
+      "grad_norm": 1.7363904107031078,
+      "learning_rate": 2.6389486726986148e-06,
+      "loss": 0.8726,
+      "step": 5598
+    },
+    {
+      "epoch": 0.7705752821359758,
+      "grad_norm": 1.7928677871559875,
+      "learning_rate": 2.6359323196674256e-06,
+      "loss": 0.8477,
+      "step": 5599
+    },
+    {
+      "epoch": 0.7707129094412332,
+      "grad_norm": 1.7928181227030888,
+      "learning_rate": 2.632917429775682e-06,
+      "loss": 0.9066,
+      "step": 5600
+    },
+    {
+      "epoch": 0.7708505367464905,
+      "grad_norm": 1.7588482953144533,
+      "learning_rate": 2.629904003622401e-06,
+      "loss": 0.9483,
+      "step": 5601
+    },
+    {
+      "epoch": 0.7709881640517479,
+      "grad_norm": 1.6116037731665678,
+      "learning_rate": 2.626892041806308e-06,
+      "loss": 0.805,
+      "step": 5602
+    },
+    {
+      "epoch": 0.7711257913570052,
+      "grad_norm": 1.8703430124872893,
+      "learning_rate": 2.6238815449258404e-06,
+      "loss": 0.8838,
+      "step": 5603
+    },
+    {
+      "epoch": 0.7712634186622626,
+      "grad_norm": 1.7650914118866157,
+      "learning_rate": 2.62087251357914e-06,
+      "loss": 0.8391,
+      "step": 5604
+    },
+    {
+      "epoch": 0.7714010459675199,
+      "grad_norm": 1.8101223184832043,
+      "learning_rate": 2.617864948364064e-06,
+      "loss": 0.9124,
+      "step": 5605
+    },
+    {
+      "epoch": 0.7715386732727774,
+      "grad_norm": 1.8190085255924182,
+      "learning_rate": 2.614858849878168e-06,
+      "loss": 0.9157,
+      "step": 5606
+    },
+    {
+      "epoch": 0.7716763005780347,
+      "grad_norm": 1.7057193708720535,
+      "learning_rate": 2.6118542187187334e-06,
+      "loss": 0.9527,
+      "step": 5607
+    },
+    {
+      "epoch": 0.7718139278832921,
+      "grad_norm": 1.7819112297806747,
+      "learning_rate": 2.608851055482727e-06,
+      "loss": 0.8622,
+      "step": 5608
+    },
+    {
+      "epoch": 0.7719515551885494,
+      "grad_norm": 1.7284774725171599,
+      "learning_rate": 2.605849360766849e-06,
+      "loss": 0.9234,
+      "step": 5609
+    },
+    {
+      "epoch": 0.7720891824938068,
+      "grad_norm": 1.6394413171149156,
+      "learning_rate": 2.60284913516748e-06,
+      "loss": 0.9267,
+      "step": 5610
+    },
+    {
+      "epoch": 0.7722268097990641,
+      "grad_norm": 2.025148633733877,
+      "learning_rate": 2.5998503792807393e-06,
+      "loss": 0.8786,
+      "step": 5611
+    },
+    {
+      "epoch": 0.7723644371043215,
+      "grad_norm": 1.7945584103420753,
+      "learning_rate": 2.596853093702425e-06,
+      "loss": 0.8855,
+      "step": 5612
+    },
+    {
+      "epoch": 0.7725020644095789,
+      "grad_norm": 1.8653908296322677,
+      "learning_rate": 2.593857279028066e-06,
+      "loss": 0.9461,
+      "step": 5613
+    },
+    {
+      "epoch": 0.7726396917148363,
+      "grad_norm": 1.9124334605461848,
+      "learning_rate": 2.5908629358528867e-06,
+      "loss": 0.902,
+      "step": 5614
+    },
+    {
+      "epoch": 0.7727773190200936,
+      "grad_norm": 1.7076409722048267,
+      "learning_rate": 2.587870064771821e-06,
+      "loss": 0.9507,
+      "step": 5615
+    },
+    {
+      "epoch": 0.772914946325351,
+      "grad_norm": 1.6930001798786676,
+      "learning_rate": 2.5848786663795135e-06,
+      "loss": 0.8912,
+      "step": 5616
+    },
+    {
+      "epoch": 0.7730525736306083,
+      "grad_norm": 1.7914430724307597,
+      "learning_rate": 2.581888741270313e-06,
+      "loss": 0.977,
+      "step": 5617
+    },
+    {
+      "epoch": 0.7731902009358657,
+      "grad_norm": 1.5723486108659253,
+      "learning_rate": 2.57890029003828e-06,
+      "loss": 0.8891,
+      "step": 5618
+    },
+    {
+      "epoch": 0.773327828241123,
+      "grad_norm": 1.6821338842556035,
+      "learning_rate": 2.5759133132771683e-06,
+      "loss": 0.8841,
+      "step": 5619
+    },
+    {
+      "epoch": 0.7734654555463804,
+      "grad_norm": 1.6942857003889098,
+      "learning_rate": 2.57292781158046e-06,
+      "loss": 0.9376,
+      "step": 5620
+    },
+    {
+      "epoch": 0.7736030828516378,
+      "grad_norm": 1.810474348173228,
+      "learning_rate": 2.569943785541331e-06,
+      "loss": 0.8849,
+      "step": 5621
+    },
+    {
+      "epoch": 0.7737407101568952,
+      "grad_norm": 1.7717786010031422,
+      "learning_rate": 2.5669612357526653e-06,
+      "loss": 0.8849,
+      "step": 5622
+    },
+    {
+      "epoch": 0.7738783374621525,
+      "grad_norm": 1.8893991687967917,
+      "learning_rate": 2.5639801628070536e-06,
+      "loss": 0.89,
+      "step": 5623
+    },
+    {
+      "epoch": 0.7740159647674099,
+      "grad_norm": 1.863082107768605,
+      "learning_rate": 2.5610005672967963e-06,
+      "loss": 0.9423,
+      "step": 5624
+    },
+    {
+      "epoch": 0.7741535920726672,
+      "grad_norm": 1.8509632780334762,
+      "learning_rate": 2.5580224498138974e-06,
+      "loss": 0.9954,
+      "step": 5625
+    },
+    {
+      "epoch": 0.7742912193779246,
+      "grad_norm": 1.9855437890213528,
+      "learning_rate": 2.5550458109500686e-06,
+      "loss": 0.9401,
+      "step": 5626
+    },
+    {
+      "epoch": 0.7744288466831819,
+      "grad_norm": 1.9998381035141575,
+      "learning_rate": 2.5520706512967263e-06,
+      "loss": 0.8521,
+      "step": 5627
+    },
+    {
+      "epoch": 0.7745664739884393,
+      "grad_norm": 1.5360365863602614,
+      "learning_rate": 2.5490969714449943e-06,
+      "loss": 0.8472,
+      "step": 5628
+    },
+    {
+      "epoch": 0.7747041012936967,
+      "grad_norm": 1.7083724134298932,
+      "learning_rate": 2.546124771985703e-06,
+      "loss": 0.8776,
+      "step": 5629
+    },
+    {
+      "epoch": 0.7748417285989541,
+      "grad_norm": 1.6137406059567772,
+      "learning_rate": 2.5431540535093867e-06,
+      "loss": 0.9242,
+      "step": 5630
+    },
+    {
+      "epoch": 0.7749793559042114,
+      "grad_norm": 1.6145806971785712,
+      "learning_rate": 2.5401848166062836e-06,
+      "loss": 0.8878,
+      "step": 5631
+    },
+    {
+      "epoch": 0.7751169832094688,
+      "grad_norm": 1.9040232827042867,
+      "learning_rate": 2.5372170618663504e-06,
+      "loss": 0.8714,
+      "step": 5632
+    },
+    {
+      "epoch": 0.7752546105147261,
+      "grad_norm": 2.0084052315653587,
+      "learning_rate": 2.534250789879229e-06,
+      "loss": 0.8892,
+      "step": 5633
+    },
+    {
+      "epoch": 0.7753922378199835,
+      "grad_norm": 1.624318677920536,
+      "learning_rate": 2.531286001234281e-06,
+      "loss": 0.9416,
+      "step": 5634
+    },
+    {
+      "epoch": 0.7755298651252408,
+      "grad_norm": 1.6700854469476591,
+      "learning_rate": 2.528322696520569e-06,
+      "loss": 0.9023,
+      "step": 5635
+    },
+    {
+      "epoch": 0.7756674924304983,
+      "grad_norm": 1.7348497042133397,
+      "learning_rate": 2.525360876326861e-06,
+      "loss": 0.8352,
+      "step": 5636
+    },
+    {
+      "epoch": 0.7758051197357556,
+      "grad_norm": 1.5838852729537705,
+      "learning_rate": 2.522400541241631e-06,
+      "loss": 0.9079,
+      "step": 5637
+    },
+    {
+      "epoch": 0.775942747041013,
+      "grad_norm": 1.8364188128816892,
+      "learning_rate": 2.5194416918530517e-06,
+      "loss": 0.9438,
+      "step": 5638
+    },
+    {
+      "epoch": 0.7760803743462703,
+      "grad_norm": 1.7938104679276492,
+      "learning_rate": 2.516484328749017e-06,
+      "loss": 0.8966,
+      "step": 5639
+    },
+    {
+      "epoch": 0.7762180016515277,
+      "grad_norm": 1.9577092266120515,
+      "learning_rate": 2.5135284525171035e-06,
+      "loss": 0.9792,
+      "step": 5640
+    },
+    {
+      "epoch": 0.776355628956785,
+      "grad_norm": 1.812112909079803,
+      "learning_rate": 2.510574063744612e-06,
+      "loss": 0.9432,
+      "step": 5641
+    },
+    {
+      "epoch": 0.7764932562620424,
+      "grad_norm": 1.9984890838769855,
+      "learning_rate": 2.50762116301853e-06,
+      "loss": 0.854,
+      "step": 5642
+    },
+    {
+      "epoch": 0.7766308835672997,
+      "grad_norm": 1.5635363944828542,
+      "learning_rate": 2.504669750925569e-06,
+      "loss": 0.904,
+      "step": 5643
+    },
+    {
+      "epoch": 0.7767685108725572,
+      "grad_norm": 1.8614043696546663,
+      "learning_rate": 2.5017198280521226e-06,
+      "loss": 0.8146,
+      "step": 5644
+    },
+    {
+      "epoch": 0.7769061381778145,
+      "grad_norm": 1.7536775280405275,
+      "learning_rate": 2.4987713949843094e-06,
+      "loss": 0.9122,
+      "step": 5645
+    },
+    {
+      "epoch": 0.7770437654830719,
+      "grad_norm": 1.7410977633888929,
+      "learning_rate": 2.495824452307939e-06,
+      "loss": 0.9236,
+      "step": 5646
+    },
+    {
+      "epoch": 0.7771813927883292,
+      "grad_norm": 1.6980131592890475,
+      "learning_rate": 2.492879000608529e-06,
+      "loss": 0.8699,
+      "step": 5647
+    },
+    {
+      "epoch": 0.7773190200935866,
+      "grad_norm": 1.664811907289621,
+      "learning_rate": 2.4899350404713007e-06,
+      "loss": 0.8821,
+      "step": 5648
+    },
+    {
+      "epoch": 0.7774566473988439,
+      "grad_norm": 1.8417078142949366,
+      "learning_rate": 2.4869925724811785e-06,
+      "loss": 0.8031,
+      "step": 5649
+    },
+    {
+      "epoch": 0.7775942747041013,
+      "grad_norm": 1.705066495538552,
+      "learning_rate": 2.484051597222792e-06,
+      "loss": 0.8165,
+      "step": 5650
+    },
+    {
+      "epoch": 0.7777319020093587,
+      "grad_norm": 1.5617473498109369,
+      "learning_rate": 2.4811121152804664e-06,
+      "loss": 0.9289,
+      "step": 5651
+    },
+    {
+      "epoch": 0.7778695293146161,
+      "grad_norm": 1.816195503709991,
+      "learning_rate": 2.478174127238243e-06,
+      "loss": 0.8948,
+      "step": 5652
+    },
+    {
+      "epoch": 0.7780071566198734,
+      "grad_norm": 1.7917560323364952,
+      "learning_rate": 2.4752376336798588e-06,
+      "loss": 0.8446,
+      "step": 5653
+    },
+    {
+      "epoch": 0.7781447839251308,
+      "grad_norm": 2.1766044274448446,
+      "learning_rate": 2.4723026351887536e-06,
+      "loss": 0.9575,
+      "step": 5654
+    },
+    {
+      "epoch": 0.7782824112303881,
+      "grad_norm": 1.6319135129739313,
+      "learning_rate": 2.4693691323480715e-06,
+      "loss": 0.8698,
+      "step": 5655
+    },
+    {
+      "epoch": 0.7784200385356455,
+      "grad_norm": 2.076508732634166,
+      "learning_rate": 2.4664371257406605e-06,
+      "loss": 0.8961,
+      "step": 5656
+    },
+    {
+      "epoch": 0.7785576658409028,
+      "grad_norm": 1.8780679334400128,
+      "learning_rate": 2.4635066159490697e-06,
+      "loss": 0.8931,
+      "step": 5657
+    },
+    {
+      "epoch": 0.7786952931461602,
+      "grad_norm": 1.7022418941069022,
+      "learning_rate": 2.4605776035555494e-06,
+      "loss": 0.89,
+      "step": 5658
+    },
+    {
+      "epoch": 0.7788329204514176,
+      "grad_norm": 3.167284517220072,
+      "learning_rate": 2.4576500891420575e-06,
+      "loss": 0.8092,
+      "step": 5659
+    },
+    {
+      "epoch": 0.778970547756675,
+      "grad_norm": 1.7055413307815244,
+      "learning_rate": 2.4547240732902476e-06,
+      "loss": 0.8636,
+      "step": 5660
+    },
+    {
+      "epoch": 0.7791081750619323,
+      "grad_norm": 1.7916968231017,
+      "learning_rate": 2.4517995565814822e-06,
+      "loss": 0.8575,
+      "step": 5661
+    },
+    {
+      "epoch": 0.7792458023671897,
+      "grad_norm": 1.584960798913686,
+      "learning_rate": 2.4488765395968208e-06,
+      "loss": 0.8497,
+      "step": 5662
+    },
+    {
+      "epoch": 0.779383429672447,
+      "grad_norm": 1.8444840039101915,
+      "learning_rate": 2.4459550229170246e-06,
+      "loss": 0.9777,
+      "step": 5663
+    },
+    {
+      "epoch": 0.7795210569777044,
+      "grad_norm": 1.8697744934647558,
+      "learning_rate": 2.443035007122567e-06,
+      "loss": 0.9333,
+      "step": 5664
+    },
+    {
+      "epoch": 0.7796586842829617,
+      "grad_norm": 2.0038320970223555,
+      "learning_rate": 2.4401164927936083e-06,
+      "loss": 0.9492,
+      "step": 5665
+    },
+    {
+      "epoch": 0.779796311588219,
+      "grad_norm": 1.7320728861443688,
+      "learning_rate": 2.4371994805100175e-06,
+      "loss": 0.8968,
+      "step": 5666
+    },
+    {
+      "epoch": 0.7799339388934765,
+      "grad_norm": 1.744886514912419,
+      "learning_rate": 2.434283970851368e-06,
+      "loss": 0.8675,
+      "step": 5667
+    },
+    {
+      "epoch": 0.7800715661987339,
+      "grad_norm": 1.7311785654617562,
+      "learning_rate": 2.431369964396928e-06,
+      "loss": 0.9157,
+      "step": 5668
+    },
+    {
+      "epoch": 0.7802091935039912,
+      "grad_norm": 1.9055684895788865,
+      "learning_rate": 2.428457461725674e-06,
+      "loss": 0.8588,
+      "step": 5669
+    },
+    {
+      "epoch": 0.7803468208092486,
+      "grad_norm": 2.0419914824490495,
+      "learning_rate": 2.4255464634162773e-06,
+      "loss": 0.9294,
+      "step": 5670
+    },
+    {
+      "epoch": 0.7804844481145059,
+      "grad_norm": 1.7147177156197309,
+      "learning_rate": 2.4226369700471196e-06,
+      "loss": 0.9188,
+      "step": 5671
+    },
+    {
+      "epoch": 0.7806220754197633,
+      "grad_norm": 1.6998529697999107,
+      "learning_rate": 2.419728982196268e-06,
+      "loss": 0.9129,
+      "step": 5672
+    },
+    {
+      "epoch": 0.7807597027250206,
+      "grad_norm": 2.0237679585053563,
+      "learning_rate": 2.4168225004415115e-06,
+      "loss": 0.9253,
+      "step": 5673
+    },
+    {
+      "epoch": 0.7808973300302781,
+      "grad_norm": 1.9113046309646293,
+      "learning_rate": 2.413917525360314e-06,
+      "loss": 0.8852,
+      "step": 5674
+    },
+    {
+      "epoch": 0.7810349573355354,
+      "grad_norm": 1.9674009097248133,
+      "learning_rate": 2.411014057529867e-06,
+      "loss": 0.8477,
+      "step": 5675
+    },
+    {
+      "epoch": 0.7811725846407928,
+      "grad_norm": 1.516258302483341,
+      "learning_rate": 2.4081120975270442e-06,
+      "loss": 0.9133,
+      "step": 5676
+    },
+    {
+      "epoch": 0.7813102119460501,
+      "grad_norm": 2.0046964125621414,
+      "learning_rate": 2.4052116459284257e-06,
+      "loss": 0.9429,
+      "step": 5677
+    },
+    {
+      "epoch": 0.7814478392513075,
+      "grad_norm": 1.6817476824072486,
+      "learning_rate": 2.402312703310292e-06,
+      "loss": 0.9645,
+      "step": 5678
+    },
+    {
+      "epoch": 0.7815854665565648,
+      "grad_norm": 1.9041409398463784,
+      "learning_rate": 2.3994152702486228e-06,
+      "loss": 0.9038,
+      "step": 5679
+    },
+    {
+      "epoch": 0.7817230938618221,
+      "grad_norm": 1.6123951924332274,
+      "learning_rate": 2.3965193473190996e-06,
+      "loss": 0.8399,
+      "step": 5680
+    },
+    {
+      "epoch": 0.7818607211670795,
+      "grad_norm": 1.5387291868194872,
+      "learning_rate": 2.3936249350971007e-06,
+      "loss": 0.9179,
+      "step": 5681
+    },
+    {
+      "epoch": 0.781998348472337,
+      "grad_norm": 1.7478429694282993,
+      "learning_rate": 2.390732034157708e-06,
+      "loss": 0.8842,
+      "step": 5682
+    },
+    {
+      "epoch": 0.7821359757775943,
+      "grad_norm": 1.7128381559419894,
+      "learning_rate": 2.387840645075701e-06,
+      "loss": 0.8614,
+      "step": 5683
+    },
+    {
+      "epoch": 0.7822736030828517,
+      "grad_norm": 1.6044023200434703,
+      "learning_rate": 2.384950768425558e-06,
+      "loss": 0.9315,
+      "step": 5684
+    },
+    {
+      "epoch": 0.782411230388109,
+      "grad_norm": 1.785939029748251,
+      "learning_rate": 2.382062404781459e-06,
+      "loss": 0.8939,
+      "step": 5685
+    },
+    {
+      "epoch": 0.7825488576933664,
+      "grad_norm": 1.955873870699468,
+      "learning_rate": 2.379175554717281e-06,
+      "loss": 0.8861,
+      "step": 5686
+    },
+    {
+      "epoch": 0.7826864849986237,
+      "grad_norm": 1.6821006302263313,
+      "learning_rate": 2.376290218806604e-06,
+      "loss": 0.92,
+      "step": 5687
+    },
+    {
+      "epoch": 0.782824112303881,
+      "grad_norm": 1.702860491628254,
+      "learning_rate": 2.373406397622703e-06,
+      "loss": 0.9125,
+      "step": 5688
+    },
+    {
+      "epoch": 0.7829617396091385,
+      "grad_norm": 1.7771988857000252,
+      "learning_rate": 2.370524091738553e-06,
+      "loss": 0.892,
+      "step": 5689
+    },
+    {
+      "epoch": 0.7830993669143959,
+      "grad_norm": 1.7576332417122207,
+      "learning_rate": 2.3676433017268308e-06,
+      "loss": 0.9052,
+      "step": 5690
+    },
+    {
+      "epoch": 0.7832369942196532,
+      "grad_norm": 1.804647223195512,
+      "learning_rate": 2.364764028159908e-06,
+      "loss": 0.9309,
+      "step": 5691
+    },
+    {
+      "epoch": 0.7833746215249106,
+      "grad_norm": 1.674640801124432,
+      "learning_rate": 2.361886271609859e-06,
+      "loss": 0.9089,
+      "step": 5692
+    },
+    {
+      "epoch": 0.7835122488301679,
+      "grad_norm": 1.7165826302480307,
+      "learning_rate": 2.35901003264845e-06,
+      "loss": 0.8386,
+      "step": 5693
+    },
+    {
+      "epoch": 0.7836498761354252,
+      "grad_norm": 1.7513371863292337,
+      "learning_rate": 2.3561353118471596e-06,
+      "loss": 0.8476,
+      "step": 5694
+    },
+    {
+      "epoch": 0.7837875034406826,
+      "grad_norm": 1.8624768326433874,
+      "learning_rate": 2.3532621097771448e-06,
+      "loss": 0.9475,
+      "step": 5695
+    },
+    {
+      "epoch": 0.78392513074594,
+      "grad_norm": 1.6691853346499397,
+      "learning_rate": 2.35039042700928e-06,
+      "loss": 0.9967,
+      "step": 5696
+    },
+    {
+      "epoch": 0.7840627580511974,
+      "grad_norm": 1.8612420685936584,
+      "learning_rate": 2.3475202641141248e-06,
+      "loss": 0.8986,
+      "step": 5697
+    },
+    {
+      "epoch": 0.7842003853564548,
+      "grad_norm": 1.896230289758656,
+      "learning_rate": 2.344651621661941e-06,
+      "loss": 0.9309,
+      "step": 5698
+    },
+    {
+      "epoch": 0.7843380126617121,
+      "grad_norm": 1.6803406896865396,
+      "learning_rate": 2.3417845002226903e-06,
+      "loss": 0.9493,
+      "step": 5699
+    },
+    {
+      "epoch": 0.7844756399669695,
+      "grad_norm": 1.778858753550019,
+      "learning_rate": 2.3389189003660264e-06,
+      "loss": 0.9163,
+      "step": 5700
+    },
+    {
+      "epoch": 0.7846132672722268,
+      "grad_norm": 1.748730321807752,
+      "learning_rate": 2.3360548226613144e-06,
+      "loss": 0.9292,
+      "step": 5701
+    },
+    {
+      "epoch": 0.7847508945774841,
+      "grad_norm": 1.5734463538477559,
+      "learning_rate": 2.333192267677594e-06,
+      "loss": 0.9007,
+      "step": 5702
+    },
+    {
+      "epoch": 0.7848885218827415,
+      "grad_norm": 1.7386047483261822,
+      "learning_rate": 2.3303312359836304e-06,
+      "loss": 0.9398,
+      "step": 5703
+    },
+    {
+      "epoch": 0.7850261491879988,
+      "grad_norm": 1.7657455631158028,
+      "learning_rate": 2.3274717281478577e-06,
+      "loss": 0.9041,
+      "step": 5704
+    },
+    {
+      "epoch": 0.7851637764932563,
+      "grad_norm": 1.6641602036958703,
+      "learning_rate": 2.324613744738432e-06,
+      "loss": 0.853,
+      "step": 5705
+    },
+    {
+      "epoch": 0.7853014037985137,
+      "grad_norm": 1.8917856918876181,
+      "learning_rate": 2.3217572863231865e-06,
+      "loss": 0.8645,
+      "step": 5706
+    },
+    {
+      "epoch": 0.785439031103771,
+      "grad_norm": 2.269500278468905,
+      "learning_rate": 2.3189023534696665e-06,
+      "loss": 0.9252,
+      "step": 5707
+    },
+    {
+      "epoch": 0.7855766584090283,
+      "grad_norm": 1.855431700378933,
+      "learning_rate": 2.3160489467451055e-06,
+      "loss": 0.8976,
+      "step": 5708
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 1.8463924273169978,
+      "learning_rate": 2.313197066716438e-06,
+      "loss": 0.9456,
+      "step": 5709
+    },
+    {
+      "epoch": 0.785851913019543,
+      "grad_norm": 1.7355428628575433,
+      "learning_rate": 2.3103467139502923e-06,
+      "loss": 0.889,
+      "step": 5710
+    },
+    {
+      "epoch": 0.7859895403248004,
+      "grad_norm": 1.7168671542186673,
+      "learning_rate": 2.307497889012994e-06,
+      "loss": 0.9068,
+      "step": 5711
+    },
+    {
+      "epoch": 0.7861271676300579,
+      "grad_norm": 1.6843731596242044,
+      "learning_rate": 2.3046505924705663e-06,
+      "loss": 0.8975,
+      "step": 5712
+    },
+    {
+      "epoch": 0.7862647949353152,
+      "grad_norm": 1.7727656423788418,
+      "learning_rate": 2.301804824888728e-06,
+      "loss": 0.8913,
+      "step": 5713
+    },
+    {
+      "epoch": 0.7864024222405726,
+      "grad_norm": 1.7043768190739303,
+      "learning_rate": 2.2989605868328946e-06,
+      "loss": 0.8717,
+      "step": 5714
+    },
+    {
+      "epoch": 0.7865400495458299,
+      "grad_norm": 1.7368669364710854,
+      "learning_rate": 2.296117878868177e-06,
+      "loss": 0.8674,
+      "step": 5715
+    },
+    {
+      "epoch": 0.7866776768510872,
+      "grad_norm": 1.6279000351259212,
+      "learning_rate": 2.2932767015593825e-06,
+      "loss": 0.8503,
+      "step": 5716
+    },
+    {
+      "epoch": 0.7868153041563446,
+      "grad_norm": 1.7677759697512399,
+      "learning_rate": 2.290437055471013e-06,
+      "loss": 0.8804,
+      "step": 5717
+    },
+    {
+      "epoch": 0.786952931461602,
+      "grad_norm": 1.8136116789332974,
+      "learning_rate": 2.2875989411672685e-06,
+      "loss": 0.9024,
+      "step": 5718
+    },
+    {
+      "epoch": 0.7870905587668593,
+      "grad_norm": 1.8059062505541048,
+      "learning_rate": 2.2847623592120438e-06,
+      "loss": 0.8758,
+      "step": 5719
+    },
+    {
+      "epoch": 0.7872281860721168,
+      "grad_norm": 1.620611231830067,
+      "learning_rate": 2.2819273101689275e-06,
+      "loss": 0.9234,
+      "step": 5720
+    },
+    {
+      "epoch": 0.7873658133773741,
+      "grad_norm": 1.8366504984543195,
+      "learning_rate": 2.279093794601206e-06,
+      "loss": 0.9294,
+      "step": 5721
+    },
+    {
+      "epoch": 0.7875034406826315,
+      "grad_norm": 1.6037415523283522,
+      "learning_rate": 2.2762618130718593e-06,
+      "loss": 0.9001,
+      "step": 5722
+    },
+    {
+      "epoch": 0.7876410679878888,
+      "grad_norm": 1.7321544975857024,
+      "learning_rate": 2.273431366143564e-06,
+      "loss": 0.9166,
+      "step": 5723
+    },
+    {
+      "epoch": 0.7877786952931461,
+      "grad_norm": 1.565691253138757,
+      "learning_rate": 2.270602454378692e-06,
+      "loss": 0.8748,
+      "step": 5724
+    },
+    {
+      "epoch": 0.7879163225984035,
+      "grad_norm": 1.7957358237971197,
+      "learning_rate": 2.267775078339305e-06,
+      "loss": 0.9041,
+      "step": 5725
+    },
+    {
+      "epoch": 0.7880539499036608,
+      "grad_norm": 1.814772134891799,
+      "learning_rate": 2.264949238587172e-06,
+      "loss": 0.9158,
+      "step": 5726
+    },
+    {
+      "epoch": 0.7881915772089183,
+      "grad_norm": 1.7379802261885697,
+      "learning_rate": 2.2621249356837392e-06,
+      "loss": 0.9487,
+      "step": 5727
+    },
+    {
+      "epoch": 0.7883292045141757,
+      "grad_norm": 1.7314051157602728,
+      "learning_rate": 2.2593021701901675e-06,
+      "loss": 0.9234,
+      "step": 5728
+    },
+    {
+      "epoch": 0.788466831819433,
+      "grad_norm": 1.7160552476975768,
+      "learning_rate": 2.256480942667293e-06,
+      "loss": 0.8495,
+      "step": 5729
+    },
+    {
+      "epoch": 0.7886044591246903,
+      "grad_norm": 1.9693532279873003,
+      "learning_rate": 2.2536612536756575e-06,
+      "loss": 0.833,
+      "step": 5730
+    },
+    {
+      "epoch": 0.7887420864299477,
+      "grad_norm": 1.9405339215326936,
+      "learning_rate": 2.2508431037754962e-06,
+      "loss": 0.8606,
+      "step": 5731
+    },
+    {
+      "epoch": 0.788879713735205,
+      "grad_norm": 1.6147178471871133,
+      "learning_rate": 2.2480264935267327e-06,
+      "loss": 0.9222,
+      "step": 5732
+    },
+    {
+      "epoch": 0.7890173410404624,
+      "grad_norm": 1.8512429648104651,
+      "learning_rate": 2.2452114234889965e-06,
+      "loss": 0.9067,
+      "step": 5733
+    },
+    {
+      "epoch": 0.7891549683457197,
+      "grad_norm": 1.730654512654597,
+      "learning_rate": 2.242397894221594e-06,
+      "loss": 0.9085,
+      "step": 5734
+    },
+    {
+      "epoch": 0.7892925956509772,
+      "grad_norm": 1.6025425105305533,
+      "learning_rate": 2.2395859062835456e-06,
+      "loss": 0.939,
+      "step": 5735
+    },
+    {
+      "epoch": 0.7894302229562346,
+      "grad_norm": 1.876534461362236,
+      "learning_rate": 2.2367754602335433e-06,
+      "loss": 0.8725,
+      "step": 5736
+    },
+    {
+      "epoch": 0.7895678502614919,
+      "grad_norm": 2.1448669559243254,
+      "learning_rate": 2.2339665566299964e-06,
+      "loss": 0.9618,
+      "step": 5737
+    },
+    {
+      "epoch": 0.7897054775667492,
+      "grad_norm": 1.6015259154823496,
+      "learning_rate": 2.231159196030983e-06,
+      "loss": 0.9259,
+      "step": 5738
+    },
+    {
+      "epoch": 0.7898431048720066,
+      "grad_norm": 1.9376382373618881,
+      "learning_rate": 2.228353378994297e-06,
+      "loss": 0.969,
+      "step": 5739
+    },
+    {
+      "epoch": 0.789980732177264,
+      "grad_norm": 1.9299867463568943,
+      "learning_rate": 2.2255491060774114e-06,
+      "loss": 0.9422,
+      "step": 5740
+    },
+    {
+      "epoch": 0.7901183594825213,
+      "grad_norm": 1.8277545734372316,
+      "learning_rate": 2.2227463778374982e-06,
+      "loss": 0.8798,
+      "step": 5741
+    },
+    {
+      "epoch": 0.7902559867877786,
+      "grad_norm": 1.6821103502960886,
+      "learning_rate": 2.219945194831421e-06,
+      "loss": 0.8928,
+      "step": 5742
+    },
+    {
+      "epoch": 0.7903936140930361,
+      "grad_norm": 1.7839520223074594,
+      "learning_rate": 2.2171455576157352e-06,
+      "loss": 0.9468,
+      "step": 5743
+    },
+    {
+      "epoch": 0.7905312413982934,
+      "grad_norm": 1.7569475291331316,
+      "learning_rate": 2.2143474667466923e-06,
+      "loss": 0.7968,
+      "step": 5744
+    },
+    {
+      "epoch": 0.7906688687035508,
+      "grad_norm": 1.6136561774721396,
+      "learning_rate": 2.2115509227802344e-06,
+      "loss": 0.9441,
+      "step": 5745
+    },
+    {
+      "epoch": 0.7908064960088081,
+      "grad_norm": 1.6332189271382906,
+      "learning_rate": 2.208755926271995e-06,
+      "loss": 0.8015,
+      "step": 5746
+    },
+    {
+      "epoch": 0.7909441233140655,
+      "grad_norm": 1.7630987948270362,
+      "learning_rate": 2.2059624777773025e-06,
+      "loss": 0.9036,
+      "step": 5747
+    },
+    {
+      "epoch": 0.7910817506193228,
+      "grad_norm": 1.556219579450756,
+      "learning_rate": 2.203170577851177e-06,
+      "loss": 0.8694,
+      "step": 5748
+    },
+    {
+      "epoch": 0.7912193779245802,
+      "grad_norm": 1.643646723029824,
+      "learning_rate": 2.200380227048331e-06,
+      "loss": 0.9136,
+      "step": 5749
+    },
+    {
+      "epoch": 0.7913570052298377,
+      "grad_norm": 1.7429030273120034,
+      "learning_rate": 2.197591425923169e-06,
+      "loss": 0.9515,
+      "step": 5750
+    },
+    {
+      "epoch": 0.791494632535095,
+      "grad_norm": 1.7933806071766738,
+      "learning_rate": 2.1948041750297865e-06,
+      "loss": 0.8431,
+      "step": 5751
+    },
+    {
+      "epoch": 0.7916322598403523,
+      "grad_norm": 2.023977653860513,
+      "learning_rate": 2.192018474921973e-06,
+      "loss": 0.8813,
+      "step": 5752
+    },
+    {
+      "epoch": 0.7917698871456097,
+      "grad_norm": 1.8091641908881209,
+      "learning_rate": 2.18923432615321e-06,
+      "loss": 0.8759,
+      "step": 5753
+    },
+    {
+      "epoch": 0.791907514450867,
+      "grad_norm": 1.6464042893643078,
+      "learning_rate": 2.1864517292766675e-06,
+      "loss": 0.971,
+      "step": 5754
+    },
+    {
+      "epoch": 0.7920451417561244,
+      "grad_norm": 1.8331215150744646,
+      "learning_rate": 2.1836706848452083e-06,
+      "loss": 0.8271,
+      "step": 5755
+    },
+    {
+      "epoch": 0.7921827690613817,
+      "grad_norm": 1.538778870034622,
+      "learning_rate": 2.1808911934113953e-06,
+      "loss": 0.9176,
+      "step": 5756
+    },
+    {
+      "epoch": 0.7923203963666391,
+      "grad_norm": 1.8075699031383436,
+      "learning_rate": 2.1781132555274655e-06,
+      "loss": 0.8947,
+      "step": 5757
+    },
+    {
+      "epoch": 0.7924580236718965,
+      "grad_norm": 1.5949101611696803,
+      "learning_rate": 2.175336871745367e-06,
+      "loss": 0.8874,
+      "step": 5758
+    },
+    {
+      "epoch": 0.7925956509771539,
+      "grad_norm": 1.7462093208357699,
+      "learning_rate": 2.1725620426167182e-06,
+      "loss": 0.9444,
+      "step": 5759
+    },
+    {
+      "epoch": 0.7927332782824112,
+      "grad_norm": 1.943647042430855,
+      "learning_rate": 2.169788768692852e-06,
+      "loss": 0.9316,
+      "step": 5760
+    },
+    {
+      "epoch": 0.7928709055876686,
+      "grad_norm": 1.6185710884205202,
+      "learning_rate": 2.16701705052477e-06,
+      "loss": 0.904,
+      "step": 5761
+    },
+    {
+      "epoch": 0.7930085328929259,
+      "grad_norm": 1.6501847676662038,
+      "learning_rate": 2.164246888663175e-06,
+      "loss": 0.8872,
+      "step": 5762
+    },
+    {
+      "epoch": 0.7931461601981833,
+      "grad_norm": 1.642174776231618,
+      "learning_rate": 2.1614782836584704e-06,
+      "loss": 0.9344,
+      "step": 5763
+    },
+    {
+      "epoch": 0.7932837875034406,
+      "grad_norm": 1.7579063549168723,
+      "learning_rate": 2.1587112360607266e-06,
+      "loss": 0.8939,
+      "step": 5764
+    },
+    {
+      "epoch": 0.7934214148086981,
+      "grad_norm": 1.7195236278333423,
+      "learning_rate": 2.15594574641973e-06,
+      "loss": 0.8999,
+      "step": 5765
+    },
+    {
+      "epoch": 0.7935590421139554,
+      "grad_norm": 1.8298942607841038,
+      "learning_rate": 2.153181815284935e-06,
+      "loss": 0.8715,
+      "step": 5766
+    },
+    {
+      "epoch": 0.7936966694192128,
+      "grad_norm": 1.6106279118175921,
+      "learning_rate": 2.150419443205508e-06,
+      "loss": 0.9171,
+      "step": 5767
+    },
+    {
+      "epoch": 0.7938342967244701,
+      "grad_norm": 1.6477527748350782,
+      "learning_rate": 2.147658630730283e-06,
+      "loss": 0.8812,
+      "step": 5768
+    },
+    {
+      "epoch": 0.7939719240297275,
+      "grad_norm": 1.6849315639765694,
+      "learning_rate": 2.1448993784078032e-06,
+      "loss": 0.9084,
+      "step": 5769
+    },
+    {
+      "epoch": 0.7941095513349848,
+      "grad_norm": 1.8182103203060942,
+      "learning_rate": 2.142141686786293e-06,
+      "loss": 0.9511,
+      "step": 5770
+    },
+    {
+      "epoch": 0.7942471786402422,
+      "grad_norm": 1.6362614548676786,
+      "learning_rate": 2.1393855564136667e-06,
+      "loss": 0.8674,
+      "step": 5771
+    },
+    {
+      "epoch": 0.7943848059454995,
+      "grad_norm": 1.6535075823372474,
+      "learning_rate": 2.1366309878375312e-06,
+      "loss": 0.9471,
+      "step": 5772
+    },
+    {
+      "epoch": 0.794522433250757,
+      "grad_norm": 1.7744839212539176,
+      "learning_rate": 2.13387798160518e-06,
+      "loss": 0.922,
+      "step": 5773
+    },
+    {
+      "epoch": 0.7946600605560143,
+      "grad_norm": 1.812703148405666,
+      "learning_rate": 2.131126538263598e-06,
+      "loss": 0.9517,
+      "step": 5774
+    },
+    {
+      "epoch": 0.7947976878612717,
+      "grad_norm": 1.593813810566864,
+      "learning_rate": 2.1283766583594614e-06,
+      "loss": 0.8723,
+      "step": 5775
+    },
+    {
+      "epoch": 0.794935315166529,
+      "grad_norm": 1.5812688147205307,
+      "learning_rate": 2.12562834243913e-06,
+      "loss": 0.9825,
+      "step": 5776
+    },
+    {
+      "epoch": 0.7950729424717864,
+      "grad_norm": 1.8100084926795617,
+      "learning_rate": 2.1228815910486598e-06,
+      "loss": 0.922,
+      "step": 5777
+    },
+    {
+      "epoch": 0.7952105697770437,
+      "grad_norm": 1.736572879235143,
+      "learning_rate": 2.1201364047337903e-06,
+      "loss": 0.8666,
+      "step": 5778
+    },
+    {
+      "epoch": 0.7953481970823011,
+      "grad_norm": 1.7188091775967367,
+      "learning_rate": 2.1173927840399545e-06,
+      "loss": 0.9618,
+      "step": 5779
+    },
+    {
+      "epoch": 0.7954858243875584,
+      "grad_norm": 1.6622362083282916,
+      "learning_rate": 2.114650729512271e-06,
+      "loss": 0.9194,
+      "step": 5780
+    },
+    {
+      "epoch": 0.7956234516928159,
+      "grad_norm": 1.663692058937918,
+      "learning_rate": 2.1119102416955494e-06,
+      "loss": 0.9283,
+      "step": 5781
+    },
+    {
+      "epoch": 0.7957610789980732,
+      "grad_norm": 2.008943598092258,
+      "learning_rate": 2.1091713211342858e-06,
+      "loss": 0.8421,
+      "step": 5782
+    },
+    {
+      "epoch": 0.7958987063033306,
+      "grad_norm": 1.9618857803223995,
+      "learning_rate": 2.1064339683726676e-06,
+      "loss": 0.8748,
+      "step": 5783
+    },
+    {
+      "epoch": 0.7960363336085879,
+      "grad_norm": 1.692099976881477,
+      "learning_rate": 2.103698183954569e-06,
+      "loss": 0.8417,
+      "step": 5784
+    },
+    {
+      "epoch": 0.7961739609138453,
+      "grad_norm": 2.0726841784474432,
+      "learning_rate": 2.1009639684235517e-06,
+      "loss": 0.9339,
+      "step": 5785
+    },
+    {
+      "epoch": 0.7963115882191026,
+      "grad_norm": 1.7564590888513487,
+      "learning_rate": 2.098231322322869e-06,
+      "loss": 0.9047,
+      "step": 5786
+    },
+    {
+      "epoch": 0.79644921552436,
+      "grad_norm": 1.7459695049479844,
+      "learning_rate": 2.0955002461954553e-06,
+      "loss": 0.9444,
+      "step": 5787
+    },
+    {
+      "epoch": 0.7965868428296174,
+      "grad_norm": 1.9404370933779793,
+      "learning_rate": 2.092770740583948e-06,
+      "loss": 0.9186,
+      "step": 5788
+    },
+    {
+      "epoch": 0.7967244701348748,
+      "grad_norm": 2.1050569627942957,
+      "learning_rate": 2.090042806030651e-06,
+      "loss": 0.9406,
+      "step": 5789
+    },
+    {
+      "epoch": 0.7968620974401321,
+      "grad_norm": 1.9276813082150905,
+      "learning_rate": 2.0873164430775784e-06,
+      "loss": 0.868,
+      "step": 5790
+    },
+    {
+      "epoch": 0.7969997247453895,
+      "grad_norm": 1.8577950037527218,
+      "learning_rate": 2.08459165226641e-06,
+      "loss": 0.8581,
+      "step": 5791
+    },
+    {
+      "epoch": 0.7971373520506468,
+      "grad_norm": 1.931541308844158,
+      "learning_rate": 2.081868434138535e-06,
+      "loss": 0.8859,
+      "step": 5792
+    },
+    {
+      "epoch": 0.7972749793559042,
+      "grad_norm": 1.7838710933451485,
+      "learning_rate": 2.0791467892350126e-06,
+      "loss": 0.8966,
+      "step": 5793
+    },
+    {
+      "epoch": 0.7974126066611615,
+      "grad_norm": 1.8312082785085597,
+      "learning_rate": 2.076426718096596e-06,
+      "loss": 0.8958,
+      "step": 5794
+    },
+    {
+      "epoch": 0.7975502339664189,
+      "grad_norm": 1.7330071478684497,
+      "learning_rate": 2.0737082212637337e-06,
+      "loss": 0.839,
+      "step": 5795
+    },
+    {
+      "epoch": 0.7976878612716763,
+      "grad_norm": 1.7158596860386541,
+      "learning_rate": 2.0709912992765436e-06,
+      "loss": 0.8682,
+      "step": 5796
+    },
+    {
+      "epoch": 0.7978254885769337,
+      "grad_norm": 1.7433704637448035,
+      "learning_rate": 2.068275952674851e-06,
+      "loss": 0.8584,
+      "step": 5797
+    },
+    {
+      "epoch": 0.797963115882191,
+      "grad_norm": 1.6101770404296827,
+      "learning_rate": 2.065562181998149e-06,
+      "loss": 0.8529,
+      "step": 5798
+    },
+    {
+      "epoch": 0.7981007431874484,
+      "grad_norm": 1.5318458639270005,
+      "learning_rate": 2.0628499877856346e-06,
+      "loss": 0.9195,
+      "step": 5799
+    },
+    {
+      "epoch": 0.7982383704927057,
+      "grad_norm": 1.6417307829788035,
+      "learning_rate": 2.0601393705761754e-06,
+      "loss": 0.8414,
+      "step": 5800
+    },
+    {
+      "epoch": 0.7983759977979631,
+      "grad_norm": 1.7609775755954107,
+      "learning_rate": 2.057430330908341e-06,
+      "loss": 0.9198,
+      "step": 5801
+    },
+    {
+      "epoch": 0.7985136251032204,
+      "grad_norm": 1.979612608927621,
+      "learning_rate": 2.054722869320377e-06,
+      "loss": 0.9494,
+      "step": 5802
+    },
+    {
+      "epoch": 0.7986512524084779,
+      "grad_norm": 1.7960735342312453,
+      "learning_rate": 2.0520169863502214e-06,
+      "loss": 0.8674,
+      "step": 5803
+    },
+    {
+      "epoch": 0.7987888797137352,
+      "grad_norm": 1.732292598579743,
+      "learning_rate": 2.0493126825354934e-06,
+      "loss": 0.9355,
+      "step": 5804
+    },
+    {
+      "epoch": 0.7989265070189926,
+      "grad_norm": 1.580893857829072,
+      "learning_rate": 2.0466099584135024e-06,
+      "loss": 0.9332,
+      "step": 5805
+    },
+    {
+      "epoch": 0.7990641343242499,
+      "grad_norm": 1.6447649439953038,
+      "learning_rate": 2.0439088145212438e-06,
+      "loss": 0.8763,
+      "step": 5806
+    },
+    {
+      "epoch": 0.7992017616295073,
+      "grad_norm": 2.0027457305309535,
+      "learning_rate": 2.0412092513953952e-06,
+      "loss": 0.8681,
+      "step": 5807
+    },
+    {
+      "epoch": 0.7993393889347646,
+      "grad_norm": 1.8660975859850022,
+      "learning_rate": 2.038511269572325e-06,
+      "loss": 0.9252,
+      "step": 5808
+    },
+    {
+      "epoch": 0.799477016240022,
+      "grad_norm": 1.8150541563347637,
+      "learning_rate": 2.0358148695880853e-06,
+      "loss": 0.8893,
+      "step": 5809
+    },
+    {
+      "epoch": 0.7996146435452793,
+      "grad_norm": 1.6236366564260942,
+      "learning_rate": 2.033120051978412e-06,
+      "loss": 0.8592,
+      "step": 5810
+    },
+    {
+      "epoch": 0.7997522708505368,
+      "grad_norm": 1.6418653559040328,
+      "learning_rate": 2.0304268172787313e-06,
+      "loss": 0.8279,
+      "step": 5811
+    },
+    {
+      "epoch": 0.7998898981557941,
+      "grad_norm": 1.9315930460308892,
+      "learning_rate": 2.0277351660241484e-06,
+      "loss": 0.9168,
+      "step": 5812
+    },
+    {
+      "epoch": 0.8000275254610515,
+      "grad_norm": 2.0745937668202985,
+      "learning_rate": 2.0250450987494607e-06,
+      "loss": 0.8631,
+      "step": 5813
+    },
+    {
+      "epoch": 0.8001651527663088,
+      "grad_norm": 2.1185366023113925,
+      "learning_rate": 2.022356615989147e-06,
+      "loss": 0.8707,
+      "step": 5814
+    },
+    {
+      "epoch": 0.8003027800715662,
+      "grad_norm": 1.9653393164715884,
+      "learning_rate": 2.0196697182773706e-06,
+      "loss": 0.9311,
+      "step": 5815
+    },
+    {
+      "epoch": 0.8004404073768235,
+      "grad_norm": 1.8463436268007716,
+      "learning_rate": 2.0169844061479837e-06,
+      "loss": 0.8803,
+      "step": 5816
+    },
+    {
+      "epoch": 0.8005780346820809,
+      "grad_norm": 1.778360640051144,
+      "learning_rate": 2.0143006801345167e-06,
+      "loss": 0.9151,
+      "step": 5817
+    },
+    {
+      "epoch": 0.8007156619873382,
+      "grad_norm": 1.7444193636759626,
+      "learning_rate": 2.011618540770197e-06,
+      "loss": 0.9078,
+      "step": 5818
+    },
+    {
+      "epoch": 0.8008532892925957,
+      "grad_norm": 1.7519083598862955,
+      "learning_rate": 2.008937988587919e-06,
+      "loss": 0.9222,
+      "step": 5819
+    },
+    {
+      "epoch": 0.800990916597853,
+      "grad_norm": 1.8951360656669713,
+      "learning_rate": 2.006259024120282e-06,
+      "loss": 0.885,
+      "step": 5820
+    },
+    {
+      "epoch": 0.8011285439031104,
+      "grad_norm": 1.7861673865907457,
+      "learning_rate": 2.0035816478995496e-06,
+      "loss": 0.9535,
+      "step": 5821
+    },
+    {
+      "epoch": 0.8012661712083677,
+      "grad_norm": 1.6623013456331268,
+      "learning_rate": 2.0009058604576903e-06,
+      "loss": 0.9634,
+      "step": 5822
+    },
+    {
+      "epoch": 0.8014037985136251,
+      "grad_norm": 1.7096603913023378,
+      "learning_rate": 1.998231662326335e-06,
+      "loss": 0.9222,
+      "step": 5823
+    },
+    {
+      "epoch": 0.8015414258188824,
+      "grad_norm": 1.7711963173911813,
+      "learning_rate": 1.9955590540368187e-06,
+      "loss": 0.9009,
+      "step": 5824
+    },
+    {
+      "epoch": 0.8016790531241398,
+      "grad_norm": 1.7230338572149915,
+      "learning_rate": 1.992888036120153e-06,
+      "loss": 0.8874,
+      "step": 5825
+    },
+    {
+      "epoch": 0.8018166804293972,
+      "grad_norm": 1.956773094566117,
+      "learning_rate": 1.990218609107023e-06,
+      "loss": 0.9341,
+      "step": 5826
+    },
+    {
+      "epoch": 0.8019543077346546,
+      "grad_norm": 1.7187761383556588,
+      "learning_rate": 1.9875507735278187e-06,
+      "loss": 1.0304,
+      "step": 5827
+    },
+    {
+      "epoch": 0.8020919350399119,
+      "grad_norm": 2.0995296221517843,
+      "learning_rate": 1.984884529912593e-06,
+      "loss": 0.9649,
+      "step": 5828
+    },
+    {
+      "epoch": 0.8022295623451693,
+      "grad_norm": 1.621479112429921,
+      "learning_rate": 1.9822198787911006e-06,
+      "loss": 0.9002,
+      "step": 5829
+    },
+    {
+      "epoch": 0.8023671896504266,
+      "grad_norm": 2.026368435735778,
+      "learning_rate": 1.9795568206927617e-06,
+      "loss": 0.8656,
+      "step": 5830
+    },
+    {
+      "epoch": 0.802504816955684,
+      "grad_norm": 1.6145256606961436,
+      "learning_rate": 1.9768953561466974e-06,
+      "loss": 0.9355,
+      "step": 5831
+    },
+    {
+      "epoch": 0.8026424442609413,
+      "grad_norm": 1.6326892756475844,
+      "learning_rate": 1.9742354856817014e-06,
+      "loss": 0.9161,
+      "step": 5832
+    },
+    {
+      "epoch": 0.8027800715661987,
+      "grad_norm": 1.6287183983640545,
+      "learning_rate": 1.9715772098262532e-06,
+      "loss": 0.8841,
+      "step": 5833
+    },
+    {
+      "epoch": 0.8029176988714561,
+      "grad_norm": 1.6764543506373148,
+      "learning_rate": 1.968920529108517e-06,
+      "loss": 0.9763,
+      "step": 5834
+    },
+    {
+      "epoch": 0.8030553261767135,
+      "grad_norm": 1.9399447585245437,
+      "learning_rate": 1.966265444056338e-06,
+      "loss": 0.9007,
+      "step": 5835
+    },
+    {
+      "epoch": 0.8031929534819708,
+      "grad_norm": 1.7121274564987772,
+      "learning_rate": 1.9636119551972456e-06,
+      "loss": 0.873,
+      "step": 5836
+    },
+    {
+      "epoch": 0.8033305807872282,
+      "grad_norm": 1.6913946390734245,
+      "learning_rate": 1.9609600630584512e-06,
+      "loss": 0.907,
+      "step": 5837
+    },
+    {
+      "epoch": 0.8034682080924855,
+      "grad_norm": 1.5623650560277835,
+      "learning_rate": 1.9583097681668496e-06,
+      "loss": 0.9258,
+      "step": 5838
+    },
+    {
+      "epoch": 0.8036058353977429,
+      "grad_norm": 1.8447460407470182,
+      "learning_rate": 1.9556610710490187e-06,
+      "loss": 0.9115,
+      "step": 5839
+    },
+    {
+      "epoch": 0.8037434627030002,
+      "grad_norm": 1.6222939145439987,
+      "learning_rate": 1.953013972231218e-06,
+      "loss": 0.8439,
+      "step": 5840
+    },
+    {
+      "epoch": 0.8038810900082577,
+      "grad_norm": 1.6872103806423533,
+      "learning_rate": 1.950368472239389e-06,
+      "loss": 0.8592,
+      "step": 5841
+    },
+    {
+      "epoch": 0.804018717313515,
+      "grad_norm": 1.828137062107912,
+      "learning_rate": 1.9477245715991575e-06,
+      "loss": 0.824,
+      "step": 5842
+    },
+    {
+      "epoch": 0.8041563446187724,
+      "grad_norm": 1.9217331968027502,
+      "learning_rate": 1.9450822708358306e-06,
+      "loss": 0.8858,
+      "step": 5843
+    },
+    {
+      "epoch": 0.8042939719240297,
+      "grad_norm": 1.914723633195201,
+      "learning_rate": 1.942441570474396e-06,
+      "loss": 0.9277,
+      "step": 5844
+    },
+    {
+      "epoch": 0.8044315992292871,
+      "grad_norm": 1.8003670722393237,
+      "learning_rate": 1.9398024710395257e-06,
+      "loss": 0.829,
+      "step": 5845
+    },
+    {
+      "epoch": 0.8045692265345444,
+      "grad_norm": 1.7710641832129357,
+      "learning_rate": 1.937164973055573e-06,
+      "loss": 0.9154,
+      "step": 5846
+    },
+    {
+      "epoch": 0.8047068538398018,
+      "grad_norm": 1.8695000149820071,
+      "learning_rate": 1.934529077046571e-06,
+      "loss": 0.9421,
+      "step": 5847
+    },
+    {
+      "epoch": 0.8048444811450591,
+      "grad_norm": 1.8075071508449876,
+      "learning_rate": 1.931894783536238e-06,
+      "loss": 0.8905,
+      "step": 5848
+    },
+    {
+      "epoch": 0.8049821084503166,
+      "grad_norm": 1.7352542004283655,
+      "learning_rate": 1.929262093047968e-06,
+      "loss": 0.9094,
+      "step": 5849
+    },
+    {
+      "epoch": 0.8051197357555739,
+      "grad_norm": 2.190709142167543,
+      "learning_rate": 1.9266310061048486e-06,
+      "loss": 0.9659,
+      "step": 5850
+    },
+    {
+      "epoch": 0.8052573630608313,
+      "grad_norm": 1.8656533491277172,
+      "learning_rate": 1.9240015232296305e-06,
+      "loss": 0.8352,
+      "step": 5851
+    },
+    {
+      "epoch": 0.8053949903660886,
+      "grad_norm": 1.7607209554900634,
+      "learning_rate": 1.9213736449447675e-06,
+      "loss": 0.9439,
+      "step": 5852
+    },
+    {
+      "epoch": 0.805532617671346,
+      "grad_norm": 1.7056651613337617,
+      "learning_rate": 1.9187473717723716e-06,
+      "loss": 0.8869,
+      "step": 5853
+    },
+    {
+      "epoch": 0.8056702449766033,
+      "grad_norm": 1.9177048920950863,
+      "learning_rate": 1.9161227042342566e-06,
+      "loss": 0.8952,
+      "step": 5854
+    },
+    {
+      "epoch": 0.8058078722818607,
+      "grad_norm": 1.831669068943791,
+      "learning_rate": 1.9134996428519006e-06,
+      "loss": 0.8489,
+      "step": 5855
+    },
+    {
+      "epoch": 0.805945499587118,
+      "grad_norm": 1.8837379302674335,
+      "learning_rate": 1.9108781881464756e-06,
+      "loss": 0.88,
+      "step": 5856
+    },
+    {
+      "epoch": 0.8060831268923755,
+      "grad_norm": 1.7456635796187767,
+      "learning_rate": 1.9082583406388277e-06,
+      "loss": 0.8467,
+      "step": 5857
+    },
+    {
+      "epoch": 0.8062207541976328,
+      "grad_norm": 1.7199221601296177,
+      "learning_rate": 1.9056401008494807e-06,
+      "loss": 0.9648,
+      "step": 5858
+    },
+    {
+      "epoch": 0.8063583815028902,
+      "grad_norm": 1.740004351770258,
+      "learning_rate": 1.9030234692986494e-06,
+      "loss": 0.995,
+      "step": 5859
+    },
+    {
+      "epoch": 0.8064960088081475,
+      "grad_norm": 1.6498204229552145,
+      "learning_rate": 1.9004084465062144e-06,
+      "loss": 0.917,
+      "step": 5860
+    },
+    {
+      "epoch": 0.8066336361134049,
+      "grad_norm": 1.9696647073405886,
+      "learning_rate": 1.897795032991756e-06,
+      "loss": 0.9082,
+      "step": 5861
+    },
+    {
+      "epoch": 0.8067712634186622,
+      "grad_norm": 1.7585075112999122,
+      "learning_rate": 1.895183229274511e-06,
+      "loss": 0.9068,
+      "step": 5862
+    },
+    {
+      "epoch": 0.8069088907239196,
+      "grad_norm": 1.7347279731697578,
+      "learning_rate": 1.8925730358734185e-06,
+      "loss": 0.8912,
+      "step": 5863
+    },
+    {
+      "epoch": 0.807046518029177,
+      "grad_norm": 2.208274573462775,
+      "learning_rate": 1.8899644533070848e-06,
+      "loss": 0.9535,
+      "step": 5864
+    },
+    {
+      "epoch": 0.8071841453344344,
+      "grad_norm": 2.3093315498095044,
+      "learning_rate": 1.8873574820938011e-06,
+      "loss": 0.9285,
+      "step": 5865
+    },
+    {
+      "epoch": 0.8073217726396917,
+      "grad_norm": 2.0805032960040535,
+      "learning_rate": 1.8847521227515354e-06,
+      "loss": 0.9191,
+      "step": 5866
+    },
+    {
+      "epoch": 0.8074593999449491,
+      "grad_norm": 1.6375995549097349,
+      "learning_rate": 1.8821483757979364e-06,
+      "loss": 0.9138,
+      "step": 5867
+    },
+    {
+      "epoch": 0.8075970272502064,
+      "grad_norm": 1.779761412176646,
+      "learning_rate": 1.879546241750334e-06,
+      "loss": 0.8976,
+      "step": 5868
+    },
+    {
+      "epoch": 0.8077346545554638,
+      "grad_norm": 1.9148868428212196,
+      "learning_rate": 1.8769457211257369e-06,
+      "loss": 0.9278,
+      "step": 5869
+    },
+    {
+      "epoch": 0.8078722818607211,
+      "grad_norm": 2.9562099808664204,
+      "learning_rate": 1.8743468144408328e-06,
+      "loss": 0.8945,
+      "step": 5870
+    },
+    {
+      "epoch": 0.8080099091659785,
+      "grad_norm": 1.6912473718378254,
+      "learning_rate": 1.8717495222119874e-06,
+      "loss": 0.9068,
+      "step": 5871
+    },
+    {
+      "epoch": 0.8081475364712359,
+      "grad_norm": 1.8941069544078148,
+      "learning_rate": 1.8691538449552493e-06,
+      "loss": 0.9548,
+      "step": 5872
+    },
+    {
+      "epoch": 0.8082851637764933,
+      "grad_norm": 1.7007340997479172,
+      "learning_rate": 1.8665597831863436e-06,
+      "loss": 0.9117,
+      "step": 5873
+    },
+    {
+      "epoch": 0.8084227910817506,
+      "grad_norm": 1.8550940728303436,
+      "learning_rate": 1.8639673374206735e-06,
+      "loss": 0.9315,
+      "step": 5874
+    },
+    {
+      "epoch": 0.808560418387008,
+      "grad_norm": 1.9560189926664586,
+      "learning_rate": 1.8613765081733226e-06,
+      "loss": 0.9103,
+      "step": 5875
+    },
+    {
+      "epoch": 0.8086980456922653,
+      "grad_norm": 1.857538935711028,
+      "learning_rate": 1.8587872959590548e-06,
+      "loss": 0.8306,
+      "step": 5876
+    },
+    {
+      "epoch": 0.8088356729975227,
+      "grad_norm": 1.7458577703745441,
+      "learning_rate": 1.8561997012923106e-06,
+      "loss": 0.8367,
+      "step": 5877
+    },
+    {
+      "epoch": 0.80897330030278,
+      "grad_norm": 1.6971000184472496,
+      "learning_rate": 1.8536137246872088e-06,
+      "loss": 0.9307,
+      "step": 5878
+    },
+    {
+      "epoch": 0.8091109276080375,
+      "grad_norm": 1.9779008163308227,
+      "learning_rate": 1.8510293666575453e-06,
+      "loss": 0.8744,
+      "step": 5879
+    },
+    {
+      "epoch": 0.8092485549132948,
+      "grad_norm": 1.649304634151418,
+      "learning_rate": 1.8484466277168046e-06,
+      "loss": 0.8972,
+      "step": 5880
+    },
+    {
+      "epoch": 0.8093861822185522,
+      "grad_norm": 1.7134239657030648,
+      "learning_rate": 1.8458655083781318e-06,
+      "loss": 0.9076,
+      "step": 5881
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 1.6434251866855878,
+      "learning_rate": 1.8432860091543692e-06,
+      "loss": 0.9323,
+      "step": 5882
+    },
+    {
+      "epoch": 0.8096614368290669,
+      "grad_norm": 1.7727123101705797,
+      "learning_rate": 1.8407081305580178e-06,
+      "loss": 0.9605,
+      "step": 5883
+    },
+    {
+      "epoch": 0.8097990641343242,
+      "grad_norm": 2.139855608006018,
+      "learning_rate": 1.8381318731012775e-06,
+      "loss": 0.9704,
+      "step": 5884
+    },
+    {
+      "epoch": 0.8099366914395816,
+      "grad_norm": 1.823633681444163,
+      "learning_rate": 1.8355572372960063e-06,
+      "loss": 0.892,
+      "step": 5885
+    },
+    {
+      "epoch": 0.8100743187448389,
+      "grad_norm": 1.824059112826107,
+      "learning_rate": 1.8329842236537543e-06,
+      "loss": 0.9221,
+      "step": 5886
+    },
+    {
+      "epoch": 0.8102119460500964,
+      "grad_norm": 1.9610617145350893,
+      "learning_rate": 1.8304128326857429e-06,
+      "loss": 0.8569,
+      "step": 5887
+    },
+    {
+      "epoch": 0.8103495733553537,
+      "grad_norm": 1.8453968549571853,
+      "learning_rate": 1.8278430649028734e-06,
+      "loss": 0.8999,
+      "step": 5888
+    },
+    {
+      "epoch": 0.8104872006606111,
+      "grad_norm": 2.063010382519787,
+      "learning_rate": 1.8252749208157238e-06,
+      "loss": 0.9103,
+      "step": 5889
+    },
+    {
+      "epoch": 0.8106248279658684,
+      "grad_norm": 1.5989919388871696,
+      "learning_rate": 1.822708400934542e-06,
+      "loss": 0.8899,
+      "step": 5890
+    },
+    {
+      "epoch": 0.8107624552711258,
+      "grad_norm": 1.725104448888569,
+      "learning_rate": 1.8201435057692718e-06,
+      "loss": 0.9651,
+      "step": 5891
+    },
+    {
+      "epoch": 0.8109000825763831,
+      "grad_norm": 1.8584264095018412,
+      "learning_rate": 1.8175802358295124e-06,
+      "loss": 0.8897,
+      "step": 5892
+    },
+    {
+      "epoch": 0.8110377098816405,
+      "grad_norm": 1.8391357835036597,
+      "learning_rate": 1.8150185916245565e-06,
+      "loss": 0.9717,
+      "step": 5893
+    },
+    {
+      "epoch": 0.8111753371868978,
+      "grad_norm": 1.8237961678674943,
+      "learning_rate": 1.8124585736633672e-06,
+      "loss": 0.8039,
+      "step": 5894
+    },
+    {
+      "epoch": 0.8113129644921553,
+      "grad_norm": 1.577296907577867,
+      "learning_rate": 1.8099001824545836e-06,
+      "loss": 0.9147,
+      "step": 5895
+    },
+    {
+      "epoch": 0.8114505917974126,
+      "grad_norm": 1.7997354480075671,
+      "learning_rate": 1.8073434185065232e-06,
+      "loss": 0.9461,
+      "step": 5896
+    },
+    {
+      "epoch": 0.81158821910267,
+      "grad_norm": 2.075012562113395,
+      "learning_rate": 1.8047882823271812e-06,
+      "loss": 0.9117,
+      "step": 5897
+    },
+    {
+      "epoch": 0.8117258464079273,
+      "grad_norm": 1.689439209609592,
+      "learning_rate": 1.8022347744242263e-06,
+      "loss": 0.8715,
+      "step": 5898
+    },
+    {
+      "epoch": 0.8118634737131847,
+      "grad_norm": 2.138677009035833,
+      "learning_rate": 1.7996828953050072e-06,
+      "loss": 0.9193,
+      "step": 5899
+    },
+    {
+      "epoch": 0.812001101018442,
+      "grad_norm": 1.6341215845690602,
+      "learning_rate": 1.797132645476547e-06,
+      "loss": 0.9192,
+      "step": 5900
+    },
+    {
+      "epoch": 0.8121387283236994,
+      "grad_norm": 1.7348965883010306,
+      "learning_rate": 1.7945840254455437e-06,
+      "loss": 0.894,
+      "step": 5901
+    },
+    {
+      "epoch": 0.8122763556289568,
+      "grad_norm": 1.6379400600057166,
+      "learning_rate": 1.7920370357183758e-06,
+      "loss": 0.8892,
+      "step": 5902
+    },
+    {
+      "epoch": 0.8124139829342142,
+      "grad_norm": 2.3067046533979063,
+      "learning_rate": 1.789491676801094e-06,
+      "loss": 0.8574,
+      "step": 5903
+    },
+    {
+      "epoch": 0.8125516102394715,
+      "grad_norm": 2.254981002341476,
+      "learning_rate": 1.7869479491994235e-06,
+      "loss": 0.9178,
+      "step": 5904
+    },
+    {
+      "epoch": 0.8126892375447289,
+      "grad_norm": 1.9036845189469558,
+      "learning_rate": 1.7844058534187769e-06,
+      "loss": 0.9231,
+      "step": 5905
+    },
+    {
+      "epoch": 0.8128268648499862,
+      "grad_norm": 1.7285400978119807,
+      "learning_rate": 1.7818653899642246e-06,
+      "loss": 0.8978,
+      "step": 5906
+    },
+    {
+      "epoch": 0.8129644921552436,
+      "grad_norm": 1.893218424713791,
+      "learning_rate": 1.7793265593405252e-06,
+      "loss": 0.8463,
+      "step": 5907
+    },
+    {
+      "epoch": 0.8131021194605009,
+      "grad_norm": 1.6897049268843178,
+      "learning_rate": 1.7767893620521103e-06,
+      "loss": 0.8288,
+      "step": 5908
+    },
+    {
+      "epoch": 0.8132397467657583,
+      "grad_norm": 1.7966756169892946,
+      "learning_rate": 1.7742537986030862e-06,
+      "loss": 0.9233,
+      "step": 5909
+    },
+    {
+      "epoch": 0.8133773740710157,
+      "grad_norm": 1.9169846712803489,
+      "learning_rate": 1.7717198694972326e-06,
+      "loss": 0.9985,
+      "step": 5910
+    },
+    {
+      "epoch": 0.8135150013762731,
+      "grad_norm": 1.7605628440732102,
+      "learning_rate": 1.7691875752380072e-06,
+      "loss": 0.8552,
+      "step": 5911
+    },
+    {
+      "epoch": 0.8136526286815304,
+      "grad_norm": 1.7612211027070188,
+      "learning_rate": 1.7666569163285475e-06,
+      "loss": 0.973,
+      "step": 5912
+    },
+    {
+      "epoch": 0.8137902559867878,
+      "grad_norm": 1.7426999594083206,
+      "learning_rate": 1.7641278932716499e-06,
+      "loss": 0.9729,
+      "step": 5913
+    },
+    {
+      "epoch": 0.8139278832920451,
+      "grad_norm": 1.4962393307390964,
+      "learning_rate": 1.7616005065698094e-06,
+      "loss": 0.8938,
+      "step": 5914
+    },
+    {
+      "epoch": 0.8140655105973025,
+      "grad_norm": 1.5539095638620708,
+      "learning_rate": 1.759074756725172e-06,
+      "loss": 0.899,
+      "step": 5915
+    },
+    {
+      "epoch": 0.8142031379025598,
+      "grad_norm": 1.6308362081607573,
+      "learning_rate": 1.7565506442395796e-06,
+      "loss": 0.8921,
+      "step": 5916
+    },
+    {
+      "epoch": 0.8143407652078173,
+      "grad_norm": 1.8335473936310165,
+      "learning_rate": 1.7540281696145278e-06,
+      "loss": 0.9336,
+      "step": 5917
+    },
+    {
+      "epoch": 0.8144783925130746,
+      "grad_norm": 1.8768696655085544,
+      "learning_rate": 1.7515073333512067e-06,
+      "loss": 0.926,
+      "step": 5918
+    },
+    {
+      "epoch": 0.814616019818332,
+      "grad_norm": 1.7483237710960509,
+      "learning_rate": 1.748988135950469e-06,
+      "loss": 0.8921,
+      "step": 5919
+    },
+    {
+      "epoch": 0.8147536471235893,
+      "grad_norm": 1.833444161646326,
+      "learning_rate": 1.7464705779128443e-06,
+      "loss": 0.8545,
+      "step": 5920
+    },
+    {
+      "epoch": 0.8148912744288467,
+      "grad_norm": 1.7478611116745095,
+      "learning_rate": 1.7439546597385404e-06,
+      "loss": 0.9046,
+      "step": 5921
+    },
+    {
+      "epoch": 0.815028901734104,
+      "grad_norm": 1.533946563444394,
+      "learning_rate": 1.7414403819274272e-06,
+      "loss": 0.9288,
+      "step": 5922
+    },
+    {
+      "epoch": 0.8151665290393614,
+      "grad_norm": 1.7523463045459553,
+      "learning_rate": 1.7389277449790676e-06,
+      "loss": 0.891,
+      "step": 5923
+    },
+    {
+      "epoch": 0.8153041563446187,
+      "grad_norm": 1.9146751650421423,
+      "learning_rate": 1.7364167493926786e-06,
+      "loss": 0.9489,
+      "step": 5924
+    },
+    {
+      "epoch": 0.8154417836498762,
+      "grad_norm": 1.7896290834517008,
+      "learning_rate": 1.733907395667167e-06,
+      "loss": 0.907,
+      "step": 5925
+    },
+    {
+      "epoch": 0.8155794109551335,
+      "grad_norm": 1.7487091959974983,
+      "learning_rate": 1.731399684301106e-06,
+      "loss": 0.9179,
+      "step": 5926
+    },
+    {
+      "epoch": 0.8157170382603909,
+      "grad_norm": 2.004564687643618,
+      "learning_rate": 1.7288936157927437e-06,
+      "loss": 0.8611,
+      "step": 5927
+    },
+    {
+      "epoch": 0.8158546655656482,
+      "grad_norm": 1.799723074431501,
+      "learning_rate": 1.72638919064e-06,
+      "loss": 0.911,
+      "step": 5928
+    },
+    {
+      "epoch": 0.8159922928709056,
+      "grad_norm": 1.7672423214087225,
+      "learning_rate": 1.7238864093404707e-06,
+      "loss": 0.8981,
+      "step": 5929
+    },
+    {
+      "epoch": 0.8161299201761629,
+      "grad_norm": 1.5531389445030734,
+      "learning_rate": 1.7213852723914248e-06,
+      "loss": 0.9748,
+      "step": 5930
+    },
+    {
+      "epoch": 0.8162675474814203,
+      "grad_norm": 1.9908244840292904,
+      "learning_rate": 1.7188857802898028e-06,
+      "loss": 0.9678,
+      "step": 5931
+    },
+    {
+      "epoch": 0.8164051747866776,
+      "grad_norm": 1.7794265490857712,
+      "learning_rate": 1.7163879335322198e-06,
+      "loss": 0.9183,
+      "step": 5932
+    },
+    {
+      "epoch": 0.8165428020919351,
+      "grad_norm": 1.9747284266630423,
+      "learning_rate": 1.713891732614965e-06,
+      "loss": 0.897,
+      "step": 5933
+    },
+    {
+      "epoch": 0.8166804293971924,
+      "grad_norm": 1.8136426900793992,
+      "learning_rate": 1.7113971780339978e-06,
+      "loss": 1.0084,
+      "step": 5934
+    },
+    {
+      "epoch": 0.8168180567024498,
+      "grad_norm": 1.7901329394675152,
+      "learning_rate": 1.7089042702849534e-06,
+      "loss": 0.9364,
+      "step": 5935
+    },
+    {
+      "epoch": 0.8169556840077071,
+      "grad_norm": 1.9183301483959305,
+      "learning_rate": 1.7064130098631348e-06,
+      "loss": 0.9155,
+      "step": 5936
+    },
+    {
+      "epoch": 0.8170933113129645,
+      "grad_norm": 1.6221334619433319,
+      "learning_rate": 1.7039233972635295e-06,
+      "loss": 0.8827,
+      "step": 5937
+    },
+    {
+      "epoch": 0.8172309386182218,
+      "grad_norm": 1.9957651473760132,
+      "learning_rate": 1.7014354329807815e-06,
+      "loss": 0.8503,
+      "step": 5938
+    },
+    {
+      "epoch": 0.8173685659234792,
+      "grad_norm": 2.1080582046945886,
+      "learning_rate": 1.6989491175092188e-06,
+      "loss": 0.8928,
+      "step": 5939
+    },
+    {
+      "epoch": 0.8175061932287366,
+      "grad_norm": 1.7370528409030321,
+      "learning_rate": 1.6964644513428363e-06,
+      "loss": 0.8731,
+      "step": 5940
+    },
+    {
+      "epoch": 0.817643820533994,
+      "grad_norm": 1.7421271030140655,
+      "learning_rate": 1.6939814349753049e-06,
+      "loss": 0.9438,
+      "step": 5941
+    },
+    {
+      "epoch": 0.8177814478392513,
+      "grad_norm": 1.681224383759432,
+      "learning_rate": 1.691500068899965e-06,
+      "loss": 0.8653,
+      "step": 5942
+    },
+    {
+      "epoch": 0.8179190751445087,
+      "grad_norm": 1.5965450184686145,
+      "learning_rate": 1.6890203536098282e-06,
+      "loss": 0.8535,
+      "step": 5943
+    },
+    {
+      "epoch": 0.818056702449766,
+      "grad_norm": 1.7645943608027552,
+      "learning_rate": 1.686542289597588e-06,
+      "loss": 0.9454,
+      "step": 5944
+    },
+    {
+      "epoch": 0.8181943297550234,
+      "grad_norm": 1.7627548580864085,
+      "learning_rate": 1.6840658773555896e-06,
+      "loss": 0.9023,
+      "step": 5945
+    },
+    {
+      "epoch": 0.8183319570602807,
+      "grad_norm": 1.8482797006101055,
+      "learning_rate": 1.6815911173758747e-06,
+      "loss": 0.9745,
+      "step": 5946
+    },
+    {
+      "epoch": 0.8184695843655381,
+      "grad_norm": 2.067469542196188,
+      "learning_rate": 1.6791180101501326e-06,
+      "loss": 0.857,
+      "step": 5947
+    },
+    {
+      "epoch": 0.8186072116707955,
+      "grad_norm": 2.0289724217410696,
+      "learning_rate": 1.6766465561697442e-06,
+      "loss": 0.9052,
+      "step": 5948
+    },
+    {
+      "epoch": 0.8187448389760529,
+      "grad_norm": 1.8160111192932156,
+      "learning_rate": 1.6741767559257505e-06,
+      "loss": 0.9242,
+      "step": 5949
+    },
+    {
+      "epoch": 0.8188824662813102,
+      "grad_norm": 1.9798753211157085,
+      "learning_rate": 1.6717086099088674e-06,
+      "loss": 0.9133,
+      "step": 5950
+    },
+    {
+      "epoch": 0.8190200935865676,
+      "grad_norm": 1.8904100062757685,
+      "learning_rate": 1.669242118609482e-06,
+      "loss": 0.8457,
+      "step": 5951
+    },
+    {
+      "epoch": 0.8191577208918249,
+      "grad_norm": 1.6353768802108426,
+      "learning_rate": 1.666777282517652e-06,
+      "loss": 0.8949,
+      "step": 5952
+    },
+    {
+      "epoch": 0.8192953481970823,
+      "grad_norm": 1.5649281220013609,
+      "learning_rate": 1.664314102123109e-06,
+      "loss": 0.8889,
+      "step": 5953
+    },
+    {
+      "epoch": 0.8194329755023396,
+      "grad_norm": 1.7785820148166518,
+      "learning_rate": 1.661852577915245e-06,
+      "loss": 0.9097,
+      "step": 5954
+    },
+    {
+      "epoch": 0.8195706028075971,
+      "grad_norm": 1.7502633223072503,
+      "learning_rate": 1.6593927103831399e-06,
+      "loss": 0.8888,
+      "step": 5955
+    },
+    {
+      "epoch": 0.8197082301128544,
+      "grad_norm": 1.66500490336568,
+      "learning_rate": 1.6569345000155335e-06,
+      "loss": 0.873,
+      "step": 5956
+    },
+    {
+      "epoch": 0.8198458574181118,
+      "grad_norm": 1.807777808988565,
+      "learning_rate": 1.6544779473008366e-06,
+      "loss": 0.9016,
+      "step": 5957
+    },
+    {
+      "epoch": 0.8199834847233691,
+      "grad_norm": 1.6601340473930815,
+      "learning_rate": 1.6520230527271352e-06,
+      "loss": 0.8785,
+      "step": 5958
+    },
+    {
+      "epoch": 0.8201211120286265,
+      "grad_norm": 2.059366986441786,
+      "learning_rate": 1.6495698167821805e-06,
+      "loss": 0.8933,
+      "step": 5959
+    },
+    {
+      "epoch": 0.8202587393338838,
+      "grad_norm": 1.7186924963206016,
+      "learning_rate": 1.6471182399533992e-06,
+      "loss": 0.8322,
+      "step": 5960
+    },
+    {
+      "epoch": 0.8203963666391412,
+      "grad_norm": 1.6962549222278447,
+      "learning_rate": 1.6446683227278848e-06,
+      "loss": 0.9211,
+      "step": 5961
+    },
+    {
+      "epoch": 0.8205339939443985,
+      "grad_norm": 1.7273322721931756,
+      "learning_rate": 1.6422200655924025e-06,
+      "loss": 0.8409,
+      "step": 5962
+    },
+    {
+      "epoch": 0.820671621249656,
+      "grad_norm": 1.8030875496431475,
+      "learning_rate": 1.6397734690333877e-06,
+      "loss": 0.9094,
+      "step": 5963
+    },
+    {
+      "epoch": 0.8208092485549133,
+      "grad_norm": 2.007487638080426,
+      "learning_rate": 1.637328533536946e-06,
+      "loss": 0.8514,
+      "step": 5964
+    },
+    {
+      "epoch": 0.8209468758601707,
+      "grad_norm": 1.7724447770390186,
+      "learning_rate": 1.634885259588852e-06,
+      "loss": 0.8523,
+      "step": 5965
+    },
+    {
+      "epoch": 0.821084503165428,
+      "grad_norm": 1.9355178811153453,
+      "learning_rate": 1.6324436476745487e-06,
+      "loss": 1.006,
+      "step": 5966
+    },
+    {
+      "epoch": 0.8212221304706854,
+      "grad_norm": 1.8887905745903646,
+      "learning_rate": 1.6300036982791578e-06,
+      "loss": 0.8572,
+      "step": 5967
+    },
+    {
+      "epoch": 0.8213597577759427,
+      "grad_norm": 1.978904505309744,
+      "learning_rate": 1.627565411887454e-06,
+      "loss": 0.882,
+      "step": 5968
+    },
+    {
+      "epoch": 0.8214973850812001,
+      "grad_norm": 1.8071202759003901,
+      "learning_rate": 1.6251287889839018e-06,
+      "loss": 0.889,
+      "step": 5969
+    },
+    {
+      "epoch": 0.8216350123864574,
+      "grad_norm": 1.5956031022846122,
+      "learning_rate": 1.6226938300526175e-06,
+      "loss": 0.8888,
+      "step": 5970
+    },
+    {
+      "epoch": 0.8217726396917149,
+      "grad_norm": 1.8192853400275606,
+      "learning_rate": 1.620260535577397e-06,
+      "loss": 0.8642,
+      "step": 5971
+    },
+    {
+      "epoch": 0.8219102669969722,
+      "grad_norm": 1.8444418380144318,
+      "learning_rate": 1.6178289060417007e-06,
+      "loss": 0.9,
+      "step": 5972
+    },
+    {
+      "epoch": 0.8220478943022296,
+      "grad_norm": 1.6739141534784143,
+      "learning_rate": 1.6153989419286597e-06,
+      "loss": 0.9323,
+      "step": 5973
+    },
+    {
+      "epoch": 0.8221855216074869,
+      "grad_norm": 1.620761602087883,
+      "learning_rate": 1.6129706437210812e-06,
+      "loss": 0.8267,
+      "step": 5974
+    },
+    {
+      "epoch": 0.8223231489127443,
+      "grad_norm": 1.875641359089364,
+      "learning_rate": 1.610544011901425e-06,
+      "loss": 0.9271,
+      "step": 5975
+    },
+    {
+      "epoch": 0.8224607762180016,
+      "grad_norm": 2.9390131118366716,
+      "learning_rate": 1.6081190469518393e-06,
+      "loss": 0.811,
+      "step": 5976
+    },
+    {
+      "epoch": 0.822598403523259,
+      "grad_norm": 1.8841210802653132,
+      "learning_rate": 1.6056957493541215e-06,
+      "loss": 0.8674,
+      "step": 5977
+    },
+    {
+      "epoch": 0.8227360308285164,
+      "grad_norm": 1.867670969832213,
+      "learning_rate": 1.6032741195897573e-06,
+      "loss": 0.9491,
+      "step": 5978
+    },
+    {
+      "epoch": 0.8228736581337738,
+      "grad_norm": 1.6588959027424228,
+      "learning_rate": 1.600854158139883e-06,
+      "loss": 0.9383,
+      "step": 5979
+    },
+    {
+      "epoch": 0.8230112854390311,
+      "grad_norm": 1.8549617387155988,
+      "learning_rate": 1.598435865485317e-06,
+      "loss": 0.8519,
+      "step": 5980
+    },
+    {
+      "epoch": 0.8231489127442885,
+      "grad_norm": 2.004669836985817,
+      "learning_rate": 1.5960192421065402e-06,
+      "loss": 0.9206,
+      "step": 5981
+    },
+    {
+      "epoch": 0.8232865400495458,
+      "grad_norm": 1.5988718518457394,
+      "learning_rate": 1.5936042884837022e-06,
+      "loss": 0.874,
+      "step": 5982
+    },
+    {
+      "epoch": 0.8234241673548032,
+      "grad_norm": 1.907700241593544,
+      "learning_rate": 1.5911910050966217e-06,
+      "loss": 0.8892,
+      "step": 5983
+    },
+    {
+      "epoch": 0.8235617946600605,
+      "grad_norm": 1.9557221289862043,
+      "learning_rate": 1.588779392424783e-06,
+      "loss": 0.8482,
+      "step": 5984
+    },
+    {
+      "epoch": 0.8236994219653179,
+      "grad_norm": 1.6339694592024638,
+      "learning_rate": 1.5863694509473448e-06,
+      "loss": 0.9825,
+      "step": 5985
+    },
+    {
+      "epoch": 0.8238370492705753,
+      "grad_norm": 1.602763118539975,
+      "learning_rate": 1.5839611811431222e-06,
+      "loss": 0.9375,
+      "step": 5986
+    },
+    {
+      "epoch": 0.8239746765758327,
+      "grad_norm": 1.8541876072039603,
+      "learning_rate": 1.581554583490612e-06,
+      "loss": 0.8617,
+      "step": 5987
+    },
+    {
+      "epoch": 0.82411230388109,
+      "grad_norm": 1.772090469561385,
+      "learning_rate": 1.579149658467971e-06,
+      "loss": 0.9335,
+      "step": 5988
+    },
+    {
+      "epoch": 0.8242499311863474,
+      "grad_norm": 1.8086505504839603,
+      "learning_rate": 1.5767464065530236e-06,
+      "loss": 0.9184,
+      "step": 5989
+    },
+    {
+      "epoch": 0.8243875584916047,
+      "grad_norm": 1.6400995105942466,
+      "learning_rate": 1.5743448282232632e-06,
+      "loss": 0.9227,
+      "step": 5990
+    },
+    {
+      "epoch": 0.8245251857968621,
+      "grad_norm": 1.961234479397404,
+      "learning_rate": 1.5719449239558504e-06,
+      "loss": 0.9106,
+      "step": 5991
+    },
+    {
+      "epoch": 0.8246628131021194,
+      "grad_norm": 1.7962021711435665,
+      "learning_rate": 1.5695466942276138e-06,
+      "loss": 0.8809,
+      "step": 5992
+    },
+    {
+      "epoch": 0.8248004404073769,
+      "grad_norm": 1.7425725890646862,
+      "learning_rate": 1.567150139515049e-06,
+      "loss": 0.9053,
+      "step": 5993
+    },
+    {
+      "epoch": 0.8249380677126342,
+      "grad_norm": 1.6613110819641825,
+      "learning_rate": 1.5647552602943183e-06,
+      "loss": 0.8914,
+      "step": 5994
+    },
+    {
+      "epoch": 0.8250756950178916,
+      "grad_norm": 1.7537593163092753,
+      "learning_rate": 1.5623620570412513e-06,
+      "loss": 0.894,
+      "step": 5995
+    },
+    {
+      "epoch": 0.8252133223231489,
+      "grad_norm": 1.7421062791146587,
+      "learning_rate": 1.559970530231345e-06,
+      "loss": 0.9178,
+      "step": 5996
+    },
+    {
+      "epoch": 0.8253509496284063,
+      "grad_norm": 1.785794115724862,
+      "learning_rate": 1.557580680339763e-06,
+      "loss": 0.8947,
+      "step": 5997
+    },
+    {
+      "epoch": 0.8254885769336636,
+      "grad_norm": 1.678322573360989,
+      "learning_rate": 1.5551925078413332e-06,
+      "loss": 0.9137,
+      "step": 5998
+    },
+    {
+      "epoch": 0.825626204238921,
+      "grad_norm": 1.8642530493755975,
+      "learning_rate": 1.5528060132105604e-06,
+      "loss": 0.9334,
+      "step": 5999
+    },
+    {
+      "epoch": 0.8257638315441783,
+      "grad_norm": 1.693387704353986,
+      "learning_rate": 1.5504211969215977e-06,
+      "loss": 0.8468,
+      "step": 6000
+    },
+    {
+      "epoch": 0.8259014588494358,
+      "grad_norm": 1.7635861715293604,
+      "learning_rate": 1.5480380594482868e-06,
+      "loss": 0.8664,
+      "step": 6001
+    },
+    {
+      "epoch": 0.8260390861546931,
+      "grad_norm": 1.647204212512501,
+      "learning_rate": 1.5456566012641162e-06,
+      "loss": 0.9177,
+      "step": 6002
+    },
+    {
+      "epoch": 0.8261767134599505,
+      "grad_norm": 1.8065204008396376,
+      "learning_rate": 1.5432768228422511e-06,
+      "loss": 0.8888,
+      "step": 6003
+    },
+    {
+      "epoch": 0.8263143407652078,
+      "grad_norm": 1.9338149004553775,
+      "learning_rate": 1.5408987246555207e-06,
+      "loss": 0.8264,
+      "step": 6004
+    },
+    {
+      "epoch": 0.8264519680704652,
+      "grad_norm": 1.6972476159166154,
+      "learning_rate": 1.53852230717642e-06,
+      "loss": 0.8449,
+      "step": 6005
+    },
+    {
+      "epoch": 0.8265895953757225,
+      "grad_norm": 1.7424058280356483,
+      "learning_rate": 1.5361475708771168e-06,
+      "loss": 0.8679,
+      "step": 6006
+    },
+    {
+      "epoch": 0.8267272226809799,
+      "grad_norm": 2.0524376846678276,
+      "learning_rate": 1.5337745162294281e-06,
+      "loss": 0.9167,
+      "step": 6007
+    },
+    {
+      "epoch": 0.8268648499862372,
+      "grad_norm": 1.6417447328674104,
+      "learning_rate": 1.5314031437048582e-06,
+      "loss": 0.881,
+      "step": 6008
+    },
+    {
+      "epoch": 0.8270024772914947,
+      "grad_norm": 1.5945088400528238,
+      "learning_rate": 1.5290334537745566e-06,
+      "loss": 0.8915,
+      "step": 6009
+    },
+    {
+      "epoch": 0.827140104596752,
+      "grad_norm": 1.6429579140843023,
+      "learning_rate": 1.5266654469093578e-06,
+      "loss": 0.9514,
+      "step": 6010
+    },
+    {
+      "epoch": 0.8272777319020094,
+      "grad_norm": 1.670821312578671,
+      "learning_rate": 1.5242991235797422e-06,
+      "loss": 0.8994,
+      "step": 6011
+    },
+    {
+      "epoch": 0.8274153592072667,
+      "grad_norm": 1.6570818406474768,
+      "learning_rate": 1.5219344842558735e-06,
+      "loss": 0.8667,
+      "step": 6012
+    },
+    {
+      "epoch": 0.8275529865125241,
+      "grad_norm": 1.6398390722913012,
+      "learning_rate": 1.5195715294075708e-06,
+      "loss": 0.8544,
+      "step": 6013
+    },
+    {
+      "epoch": 0.8276906138177814,
+      "grad_norm": 1.7930821626327464,
+      "learning_rate": 1.5172102595043204e-06,
+      "loss": 0.8751,
+      "step": 6014
+    },
+    {
+      "epoch": 0.8278282411230388,
+      "grad_norm": 1.87218638970103,
+      "learning_rate": 1.5148506750152748e-06,
+      "loss": 0.8796,
+      "step": 6015
+    },
+    {
+      "epoch": 0.8279658684282962,
+      "grad_norm": 1.7969397927169122,
+      "learning_rate": 1.5124927764092511e-06,
+      "loss": 0.744,
+      "step": 6016
+    },
+    {
+      "epoch": 0.8281034957335536,
+      "grad_norm": 1.8003270595762026,
+      "learning_rate": 1.5101365641547316e-06,
+      "loss": 0.9347,
+      "step": 6017
+    },
+    {
+      "epoch": 0.8282411230388109,
+      "grad_norm": 1.8119493444352799,
+      "learning_rate": 1.5077820387198627e-06,
+      "loss": 0.8865,
+      "step": 6018
+    },
+    {
+      "epoch": 0.8283787503440683,
+      "grad_norm": 1.7844322655588314,
+      "learning_rate": 1.505429200572458e-06,
+      "loss": 0.8559,
+      "step": 6019
+    },
+    {
+      "epoch": 0.8285163776493256,
+      "grad_norm": 1.8231631209107024,
+      "learning_rate": 1.5030780501799924e-06,
+      "loss": 0.9199,
+      "step": 6020
+    },
+    {
+      "epoch": 0.828654004954583,
+      "grad_norm": 1.7676342485546395,
+      "learning_rate": 1.500728588009608e-06,
+      "loss": 0.9254,
+      "step": 6021
+    },
+    {
+      "epoch": 0.8287916322598403,
+      "grad_norm": 1.5435549378672047,
+      "learning_rate": 1.4983808145281108e-06,
+      "loss": 0.8798,
+      "step": 6022
+    },
+    {
+      "epoch": 0.8289292595650977,
+      "grad_norm": 1.8045666932103723,
+      "learning_rate": 1.4960347302019718e-06,
+      "loss": 0.9228,
+      "step": 6023
+    },
+    {
+      "epoch": 0.8290668868703551,
+      "grad_norm": 1.7196811600975628,
+      "learning_rate": 1.4936903354973241e-06,
+      "loss": 0.9199,
+      "step": 6024
+    },
+    {
+      "epoch": 0.8292045141756125,
+      "grad_norm": 2.1128260571064947,
+      "learning_rate": 1.491347630879969e-06,
+      "loss": 0.8513,
+      "step": 6025
+    },
+    {
+      "epoch": 0.8293421414808698,
+      "grad_norm": 2.0650919825857206,
+      "learning_rate": 1.4890066168153695e-06,
+      "loss": 0.9354,
+      "step": 6026
+    },
+    {
+      "epoch": 0.8294797687861272,
+      "grad_norm": 1.8347582469123436,
+      "learning_rate": 1.486667293768651e-06,
+      "loss": 0.8592,
+      "step": 6027
+    },
+    {
+      "epoch": 0.8296173960913845,
+      "grad_norm": 1.8612623018733891,
+      "learning_rate": 1.4843296622046054e-06,
+      "loss": 0.9146,
+      "step": 6028
+    },
+    {
+      "epoch": 0.8297550233966419,
+      "grad_norm": 1.9485307231467388,
+      "learning_rate": 1.4819937225876935e-06,
+      "loss": 0.8136,
+      "step": 6029
+    },
+    {
+      "epoch": 0.8298926507018992,
+      "grad_norm": 1.530649891089945,
+      "learning_rate": 1.4796594753820248e-06,
+      "loss": 0.9591,
+      "step": 6030
+    },
+    {
+      "epoch": 0.8300302780071567,
+      "grad_norm": 1.7403344872108133,
+      "learning_rate": 1.4773269210513918e-06,
+      "loss": 0.9159,
+      "step": 6031
+    },
+    {
+      "epoch": 0.830167905312414,
+      "grad_norm": 2.0053602143707283,
+      "learning_rate": 1.4749960600592338e-06,
+      "loss": 0.9292,
+      "step": 6032
+    },
+    {
+      "epoch": 0.8303055326176714,
+      "grad_norm": 1.71662542031467,
+      "learning_rate": 1.4726668928686684e-06,
+      "loss": 0.8267,
+      "step": 6033
+    },
+    {
+      "epoch": 0.8304431599229287,
+      "grad_norm": 1.8465188823283023,
+      "learning_rate": 1.4703394199424615e-06,
+      "loss": 0.8344,
+      "step": 6034
+    },
+    {
+      "epoch": 0.8305807872281861,
+      "grad_norm": 1.9201293857925044,
+      "learning_rate": 1.4680136417430512e-06,
+      "loss": 0.9252,
+      "step": 6035
+    },
+    {
+      "epoch": 0.8307184145334434,
+      "grad_norm": 1.7147575710165381,
+      "learning_rate": 1.4656895587325459e-06,
+      "loss": 0.9064,
+      "step": 6036
+    },
+    {
+      "epoch": 0.8308560418387008,
+      "grad_norm": 1.7885429425135193,
+      "learning_rate": 1.4633671713726982e-06,
+      "loss": 0.931,
+      "step": 6037
+    },
+    {
+      "epoch": 0.8309936691439581,
+      "grad_norm": 1.8622770504777373,
+      "learning_rate": 1.4610464801249447e-06,
+      "loss": 0.9527,
+      "step": 6038
+    },
+    {
+      "epoch": 0.8311312964492156,
+      "grad_norm": 1.6068313407525372,
+      "learning_rate": 1.4587274854503653e-06,
+      "loss": 0.8744,
+      "step": 6039
+    },
+    {
+      "epoch": 0.8312689237544729,
+      "grad_norm": 2.1796897521838394,
+      "learning_rate": 1.4564101878097215e-06,
+      "loss": 0.9031,
+      "step": 6040
+    },
+    {
+      "epoch": 0.8314065510597303,
+      "grad_norm": 1.6578027929159391,
+      "learning_rate": 1.4540945876634194e-06,
+      "loss": 0.9256,
+      "step": 6041
+    },
+    {
+      "epoch": 0.8315441783649876,
+      "grad_norm": 1.5499973392117492,
+      "learning_rate": 1.4517806854715455e-06,
+      "loss": 0.8583,
+      "step": 6042
+    },
+    {
+      "epoch": 0.831681805670245,
+      "grad_norm": 1.716440207379264,
+      "learning_rate": 1.4494684816938364e-06,
+      "loss": 0.98,
+      "step": 6043
+    },
+    {
+      "epoch": 0.8318194329755023,
+      "grad_norm": 1.5520433952958468,
+      "learning_rate": 1.447157976789696e-06,
+      "loss": 0.9156,
+      "step": 6044
+    },
+    {
+      "epoch": 0.8319570602807597,
+      "grad_norm": 1.9394338443434653,
+      "learning_rate": 1.44484917121819e-06,
+      "loss": 0.9672,
+      "step": 6045
+    },
+    {
+      "epoch": 0.832094687586017,
+      "grad_norm": 1.5651069802077147,
+      "learning_rate": 1.4425420654380462e-06,
+      "loss": 0.9553,
+      "step": 6046
+    },
+    {
+      "epoch": 0.8322323148912745,
+      "grad_norm": 1.9230074071191414,
+      "learning_rate": 1.4402366599076546e-06,
+      "loss": 0.8945,
+      "step": 6047
+    },
+    {
+      "epoch": 0.8323699421965318,
+      "grad_norm": 2.379219331798779,
+      "learning_rate": 1.4379329550850697e-06,
+      "loss": 0.9326,
+      "step": 6048
+    },
+    {
+      "epoch": 0.8325075695017892,
+      "grad_norm": 1.6517296222712392,
+      "learning_rate": 1.4356309514280043e-06,
+      "loss": 0.8663,
+      "step": 6049
+    },
+    {
+      "epoch": 0.8326451968070465,
+      "grad_norm": 1.908762496700301,
+      "learning_rate": 1.4333306493938348e-06,
+      "loss": 0.8939,
+      "step": 6050
+    },
+    {
+      "epoch": 0.8327828241123039,
+      "grad_norm": 2.1409701002812995,
+      "learning_rate": 1.4310320494396002e-06,
+      "loss": 0.9459,
+      "step": 6051
+    },
+    {
+      "epoch": 0.8329204514175612,
+      "grad_norm": 1.8902406544057195,
+      "learning_rate": 1.4287351520220017e-06,
+      "loss": 0.9596,
+      "step": 6052
+    },
+    {
+      "epoch": 0.8330580787228186,
+      "grad_norm": 1.724051508928052,
+      "learning_rate": 1.426439957597401e-06,
+      "loss": 1.0188,
+      "step": 6053
+    },
+    {
+      "epoch": 0.833195706028076,
+      "grad_norm": 1.7958514674938604,
+      "learning_rate": 1.4241464666218207e-06,
+      "loss": 0.8938,
+      "step": 6054
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.9717124752361717,
+      "learning_rate": 1.421854679550947e-06,
+      "loss": 0.9124,
+      "step": 6055
+    },
+    {
+      "epoch": 0.8334709606385907,
+      "grad_norm": 1.7486500132746488,
+      "learning_rate": 1.419564596840126e-06,
+      "loss": 0.9308,
+      "step": 6056
+    },
+    {
+      "epoch": 0.8336085879438481,
+      "grad_norm": 1.8338332104468305,
+      "learning_rate": 1.4172762189443667e-06,
+      "loss": 0.8521,
+      "step": 6057
+    },
+    {
+      "epoch": 0.8337462152491054,
+      "grad_norm": 1.701032519197085,
+      "learning_rate": 1.4149895463183383e-06,
+      "loss": 0.8813,
+      "step": 6058
+    },
+    {
+      "epoch": 0.8338838425543628,
+      "grad_norm": 2.063316959175455,
+      "learning_rate": 1.4127045794163707e-06,
+      "loss": 0.8448,
+      "step": 6059
+    },
+    {
+      "epoch": 0.8340214698596201,
+      "grad_norm": 1.7972524092644135,
+      "learning_rate": 1.4104213186924542e-06,
+      "loss": 0.8999,
+      "step": 6060
+    },
+    {
+      "epoch": 0.8341590971648775,
+      "grad_norm": 1.794507399074356,
+      "learning_rate": 1.408139764600247e-06,
+      "loss": 0.9212,
+      "step": 6061
+    },
+    {
+      "epoch": 0.8342967244701349,
+      "grad_norm": 1.797278647567924,
+      "learning_rate": 1.4058599175930543e-06,
+      "loss": 0.8429,
+      "step": 6062
+    },
+    {
+      "epoch": 0.8344343517753923,
+      "grad_norm": 1.7912975770087685,
+      "learning_rate": 1.4035817781238603e-06,
+      "loss": 0.8991,
+      "step": 6063
+    },
+    {
+      "epoch": 0.8345719790806496,
+      "grad_norm": 1.900206946922228,
+      "learning_rate": 1.4013053466452898e-06,
+      "loss": 0.8712,
+      "step": 6064
+    },
+    {
+      "epoch": 0.834709606385907,
+      "grad_norm": 1.7026499974789298,
+      "learning_rate": 1.3990306236096474e-06,
+      "loss": 0.8511,
+      "step": 6065
+    },
+    {
+      "epoch": 0.8348472336911643,
+      "grad_norm": 1.6923000389535787,
+      "learning_rate": 1.396757609468884e-06,
+      "loss": 0.8273,
+      "step": 6066
+    },
+    {
+      "epoch": 0.8349848609964217,
+      "grad_norm": 1.5726020452464728,
+      "learning_rate": 1.394486304674616e-06,
+      "loss": 0.8615,
+      "step": 6067
+    },
+    {
+      "epoch": 0.835122488301679,
+      "grad_norm": 2.1781695674388177,
+      "learning_rate": 1.3922167096781269e-06,
+      "loss": 0.8825,
+      "step": 6068
+    },
+    {
+      "epoch": 0.8352601156069365,
+      "grad_norm": 1.8931196512359085,
+      "learning_rate": 1.3899488249303462e-06,
+      "loss": 0.8829,
+      "step": 6069
+    },
+    {
+      "epoch": 0.8353977429121938,
+      "grad_norm": 2.039192427420529,
+      "learning_rate": 1.3876826508818786e-06,
+      "loss": 0.9224,
+      "step": 6070
+    },
+    {
+      "epoch": 0.8355353702174512,
+      "grad_norm": 1.886055577250591,
+      "learning_rate": 1.3854181879829742e-06,
+      "loss": 0.9655,
+      "step": 6071
+    },
+    {
+      "epoch": 0.8356729975227085,
+      "grad_norm": 1.928488602302475,
+      "learning_rate": 1.3831554366835597e-06,
+      "loss": 0.8999,
+      "step": 6072
+    },
+    {
+      "epoch": 0.8358106248279659,
+      "grad_norm": 1.6971425320750428,
+      "learning_rate": 1.3808943974332045e-06,
+      "loss": 0.9563,
+      "step": 6073
+    },
+    {
+      "epoch": 0.8359482521332232,
+      "grad_norm": 1.7435712621514061,
+      "learning_rate": 1.3786350706811513e-06,
+      "loss": 0.9165,
+      "step": 6074
+    },
+    {
+      "epoch": 0.8360858794384806,
+      "grad_norm": 1.7310109956387163,
+      "learning_rate": 1.3763774568762956e-06,
+      "loss": 0.9168,
+      "step": 6075
+    },
+    {
+      "epoch": 0.8362235067437379,
+      "grad_norm": 1.5888331512621912,
+      "learning_rate": 1.374121556467195e-06,
+      "loss": 0.8352,
+      "step": 6076
+    },
+    {
+      "epoch": 0.8363611340489954,
+      "grad_norm": 1.9466270644967998,
+      "learning_rate": 1.3718673699020658e-06,
+      "loss": 0.9282,
+      "step": 6077
+    },
+    {
+      "epoch": 0.8364987613542527,
+      "grad_norm": 1.7857997636426537,
+      "learning_rate": 1.3696148976287838e-06,
+      "loss": 0.8978,
+      "step": 6078
+    },
+    {
+      "epoch": 0.8366363886595101,
+      "grad_norm": 2.2364335341499486,
+      "learning_rate": 1.3673641400948857e-06,
+      "loss": 0.8995,
+      "step": 6079
+    },
+    {
+      "epoch": 0.8367740159647674,
+      "grad_norm": 1.8051287150346342,
+      "learning_rate": 1.3651150977475636e-06,
+      "loss": 0.8421,
+      "step": 6080
+    },
+    {
+      "epoch": 0.8369116432700248,
+      "grad_norm": 1.6573557746162544,
+      "learning_rate": 1.3628677710336736e-06,
+      "loss": 0.9032,
+      "step": 6081
+    },
+    {
+      "epoch": 0.8370492705752821,
+      "grad_norm": 1.8074645952618298,
+      "learning_rate": 1.3606221603997272e-06,
+      "loss": 0.9146,
+      "step": 6082
+    },
+    {
+      "epoch": 0.8371868978805395,
+      "grad_norm": 1.7274708386333526,
+      "learning_rate": 1.3583782662918976e-06,
+      "loss": 0.8668,
+      "step": 6083
+    },
+    {
+      "epoch": 0.8373245251857968,
+      "grad_norm": 1.6330928082409644,
+      "learning_rate": 1.356136089156016e-06,
+      "loss": 0.9581,
+      "step": 6084
+    },
+    {
+      "epoch": 0.8374621524910543,
+      "grad_norm": 1.7054009634098362,
+      "learning_rate": 1.3538956294375715e-06,
+      "loss": 0.877,
+      "step": 6085
+    },
+    {
+      "epoch": 0.8375997797963116,
+      "grad_norm": 1.7463593761632161,
+      "learning_rate": 1.3516568875817138e-06,
+      "loss": 0.8783,
+      "step": 6086
+    },
+    {
+      "epoch": 0.837737407101569,
+      "grad_norm": 1.8605929065771427,
+      "learning_rate": 1.349419864033249e-06,
+      "loss": 0.8513,
+      "step": 6087
+    },
+    {
+      "epoch": 0.8378750344068263,
+      "grad_norm": 1.7162030312797463,
+      "learning_rate": 1.3471845592366438e-06,
+      "loss": 0.8657,
+      "step": 6088
+    },
+    {
+      "epoch": 0.8380126617120837,
+      "grad_norm": 1.8079296078158014,
+      "learning_rate": 1.3449509736360232e-06,
+      "loss": 0.9372,
+      "step": 6089
+    },
+    {
+      "epoch": 0.838150289017341,
+      "grad_norm": 1.6700336382168768,
+      "learning_rate": 1.3427191076751666e-06,
+      "loss": 0.8228,
+      "step": 6090
+    },
+    {
+      "epoch": 0.8382879163225984,
+      "grad_norm": 1.764755199441113,
+      "learning_rate": 1.3404889617975226e-06,
+      "loss": 0.8647,
+      "step": 6091
+    },
+    {
+      "epoch": 0.8384255436278558,
+      "grad_norm": 1.677207916115392,
+      "learning_rate": 1.338260536446182e-06,
+      "loss": 0.9021,
+      "step": 6092
+    },
+    {
+      "epoch": 0.8385631709331132,
+      "grad_norm": 1.8650096152289894,
+      "learning_rate": 1.3360338320639098e-06,
+      "loss": 0.9853,
+      "step": 6093
+    },
+    {
+      "epoch": 0.8387007982383705,
+      "grad_norm": 1.9181375753132912,
+      "learning_rate": 1.3338088490931144e-06,
+      "loss": 0.8916,
+      "step": 6094
+    },
+    {
+      "epoch": 0.8388384255436279,
+      "grad_norm": 1.553251292780892,
+      "learning_rate": 1.3315855879758787e-06,
+      "loss": 0.8212,
+      "step": 6095
+    },
+    {
+      "epoch": 0.8389760528488852,
+      "grad_norm": 1.610467421678477,
+      "learning_rate": 1.329364049153924e-06,
+      "loss": 0.9053,
+      "step": 6096
+    },
+    {
+      "epoch": 0.8391136801541426,
+      "grad_norm": 1.6114612195107656,
+      "learning_rate": 1.3271442330686467e-06,
+      "loss": 0.8993,
+      "step": 6097
+    },
+    {
+      "epoch": 0.8392513074593999,
+      "grad_norm": 1.6734658341730906,
+      "learning_rate": 1.3249261401610946e-06,
+      "loss": 0.8655,
+      "step": 6098
+    },
+    {
+      "epoch": 0.8393889347646573,
+      "grad_norm": 1.8339477443188514,
+      "learning_rate": 1.3227097708719649e-06,
+      "loss": 0.9351,
+      "step": 6099
+    },
+    {
+      "epoch": 0.8395265620699147,
+      "grad_norm": 2.072308235141602,
+      "learning_rate": 1.3204951256416288e-06,
+      "loss": 0.8766,
+      "step": 6100
+    },
+    {
+      "epoch": 0.8396641893751721,
+      "grad_norm": 1.6660579208588797,
+      "learning_rate": 1.3182822049100975e-06,
+      "loss": 0.8651,
+      "step": 6101
+    },
+    {
+      "epoch": 0.8398018166804294,
+      "grad_norm": 1.9730982348603476,
+      "learning_rate": 1.316071009117057e-06,
+      "loss": 0.8848,
+      "step": 6102
+    },
+    {
+      "epoch": 0.8399394439856868,
+      "grad_norm": 2.1103217808056507,
+      "learning_rate": 1.3138615387018327e-06,
+      "loss": 0.9022,
+      "step": 6103
+    },
+    {
+      "epoch": 0.8400770712909441,
+      "grad_norm": 1.7643874347913482,
+      "learning_rate": 1.3116537941034225e-06,
+      "loss": 0.9259,
+      "step": 6104
+    },
+    {
+      "epoch": 0.8402146985962015,
+      "grad_norm": 1.82608810256724,
+      "learning_rate": 1.3094477757604718e-06,
+      "loss": 0.9408,
+      "step": 6105
+    },
+    {
+      "epoch": 0.8403523259014588,
+      "grad_norm": 1.909491737390539,
+      "learning_rate": 1.307243484111288e-06,
+      "loss": 0.9644,
+      "step": 6106
+    },
+    {
+      "epoch": 0.8404899532067163,
+      "grad_norm": 1.9100606181776114,
+      "learning_rate": 1.3050409195938341e-06,
+      "loss": 0.8551,
+      "step": 6107
+    },
+    {
+      "epoch": 0.8406275805119736,
+      "grad_norm": 1.590136761381432,
+      "learning_rate": 1.3028400826457266e-06,
+      "loss": 0.8477,
+      "step": 6108
+    },
+    {
+      "epoch": 0.840765207817231,
+      "grad_norm": 1.8166123583462286,
+      "learning_rate": 1.3006409737042448e-06,
+      "loss": 0.8122,
+      "step": 6109
+    },
+    {
+      "epoch": 0.8409028351224883,
+      "grad_norm": 1.858342832182508,
+      "learning_rate": 1.29844359320632e-06,
+      "loss": 0.8927,
+      "step": 6110
+    },
+    {
+      "epoch": 0.8410404624277457,
+      "grad_norm": 1.8384373899354556,
+      "learning_rate": 1.2962479415885421e-06,
+      "loss": 0.8855,
+      "step": 6111
+    },
+    {
+      "epoch": 0.841178089733003,
+      "grad_norm": 1.9770414133472942,
+      "learning_rate": 1.2940540192871564e-06,
+      "loss": 0.857,
+      "step": 6112
+    },
+    {
+      "epoch": 0.8413157170382604,
+      "grad_norm": 1.6709901002887881,
+      "learning_rate": 1.2918618267380656e-06,
+      "loss": 0.8825,
+      "step": 6113
+    },
+    {
+      "epoch": 0.8414533443435177,
+      "grad_norm": 1.628858499422417,
+      "learning_rate": 1.2896713643768289e-06,
+      "loss": 0.8414,
+      "step": 6114
+    },
+    {
+      "epoch": 0.8415909716487752,
+      "grad_norm": 2.0096845119187727,
+      "learning_rate": 1.2874826326386601e-06,
+      "loss": 0.9519,
+      "step": 6115
+    },
+    {
+      "epoch": 0.8417285989540325,
+      "grad_norm": 1.9845448935818484,
+      "learning_rate": 1.285295631958432e-06,
+      "loss": 0.915,
+      "step": 6116
+    },
+    {
+      "epoch": 0.8418662262592899,
+      "grad_norm": 1.6497359152365212,
+      "learning_rate": 1.28311036277067e-06,
+      "loss": 0.8665,
+      "step": 6117
+    },
+    {
+      "epoch": 0.8420038535645472,
+      "grad_norm": 1.7352456052419811,
+      "learning_rate": 1.2809268255095565e-06,
+      "loss": 0.9668,
+      "step": 6118
+    },
+    {
+      "epoch": 0.8421414808698046,
+      "grad_norm": 1.7857166890053737,
+      "learning_rate": 1.2787450206089324e-06,
+      "loss": 0.8985,
+      "step": 6119
+    },
+    {
+      "epoch": 0.8422791081750619,
+      "grad_norm": 1.8961421306037025,
+      "learning_rate": 1.2765649485022923e-06,
+      "loss": 0.8759,
+      "step": 6120
+    },
+    {
+      "epoch": 0.8424167354803193,
+      "grad_norm": 1.8095611152354723,
+      "learning_rate": 1.2743866096227863e-06,
+      "loss": 0.8921,
+      "step": 6121
+    },
+    {
+      "epoch": 0.8425543627855766,
+      "grad_norm": 1.6210589012402394,
+      "learning_rate": 1.272210004403217e-06,
+      "loss": 0.9177,
+      "step": 6122
+    },
+    {
+      "epoch": 0.8426919900908341,
+      "grad_norm": 1.732450653086319,
+      "learning_rate": 1.2700351332760542e-06,
+      "loss": 0.9457,
+      "step": 6123
+    },
+    {
+      "epoch": 0.8428296173960914,
+      "grad_norm": 1.596930517048839,
+      "learning_rate": 1.2678619966734062e-06,
+      "loss": 0.8522,
+      "step": 6124
+    },
+    {
+      "epoch": 0.8429672447013488,
+      "grad_norm": 1.7366085271603748,
+      "learning_rate": 1.2656905950270537e-06,
+      "loss": 0.8737,
+      "step": 6125
+    },
+    {
+      "epoch": 0.8431048720066061,
+      "grad_norm": 1.6927379001521008,
+      "learning_rate": 1.2635209287684157e-06,
+      "loss": 0.8915,
+      "step": 6126
+    },
+    {
+      "epoch": 0.8432424993118635,
+      "grad_norm": 1.7387728033511043,
+      "learning_rate": 1.2613529983285843e-06,
+      "loss": 0.9575,
+      "step": 6127
+    },
+    {
+      "epoch": 0.8433801266171208,
+      "grad_norm": 1.8614872756380283,
+      "learning_rate": 1.2591868041382893e-06,
+      "loss": 0.924,
+      "step": 6128
+    },
+    {
+      "epoch": 0.8435177539223782,
+      "grad_norm": 2.034200146748662,
+      "learning_rate": 1.2570223466279296e-06,
+      "loss": 0.8841,
+      "step": 6129
+    },
+    {
+      "epoch": 0.8436553812276356,
+      "grad_norm": 1.632094586098541,
+      "learning_rate": 1.2548596262275526e-06,
+      "loss": 0.9254,
+      "step": 6130
+    },
+    {
+      "epoch": 0.843793008532893,
+      "grad_norm": 1.9064335207681462,
+      "learning_rate": 1.2526986433668565e-06,
+      "loss": 0.9127,
+      "step": 6131
+    },
+    {
+      "epoch": 0.8439306358381503,
+      "grad_norm": 2.125998740610062,
+      "learning_rate": 1.2505393984752068e-06,
+      "loss": 0.8841,
+      "step": 6132
+    },
+    {
+      "epoch": 0.8440682631434077,
+      "grad_norm": 1.6465219842572694,
+      "learning_rate": 1.248381891981606e-06,
+      "loss": 0.8884,
+      "step": 6133
+    },
+    {
+      "epoch": 0.844205890448665,
+      "grad_norm": 1.6813455796972567,
+      "learning_rate": 1.2462261243147322e-06,
+      "loss": 0.9313,
+      "step": 6134
+    },
+    {
+      "epoch": 0.8443435177539224,
+      "grad_norm": 1.7858511445267358,
+      "learning_rate": 1.2440720959028973e-06,
+      "loss": 0.8454,
+      "step": 6135
+    },
+    {
+      "epoch": 0.8444811450591797,
+      "grad_norm": 1.819605417144252,
+      "learning_rate": 1.2419198071740823e-06,
+      "loss": 0.8788,
+      "step": 6136
+    },
+    {
+      "epoch": 0.844618772364437,
+      "grad_norm": 1.6528858617096494,
+      "learning_rate": 1.2397692585559173e-06,
+      "loss": 0.8892,
+      "step": 6137
+    },
+    {
+      "epoch": 0.8447563996696945,
+      "grad_norm": 1.776742051552983,
+      "learning_rate": 1.2376204504756862e-06,
+      "loss": 0.8726,
+      "step": 6138
+    },
+    {
+      "epoch": 0.8448940269749519,
+      "grad_norm": 1.6741325880591398,
+      "learning_rate": 1.2354733833603272e-06,
+      "loss": 0.8917,
+      "step": 6139
+    },
+    {
+      "epoch": 0.8450316542802092,
+      "grad_norm": 1.9028593690749667,
+      "learning_rate": 1.2333280576364349e-06,
+      "loss": 0.839,
+      "step": 6140
+    },
+    {
+      "epoch": 0.8451692815854666,
+      "grad_norm": 1.7678687103518214,
+      "learning_rate": 1.2311844737302536e-06,
+      "loss": 0.9343,
+      "step": 6141
+    },
+    {
+      "epoch": 0.8453069088907239,
+      "grad_norm": 1.8638353498324214,
+      "learning_rate": 1.2290426320676852e-06,
+      "loss": 0.9142,
+      "step": 6142
+    },
+    {
+      "epoch": 0.8454445361959813,
+      "grad_norm": 1.7962682859738455,
+      "learning_rate": 1.2269025330742857e-06,
+      "loss": 0.9592,
+      "step": 6143
+    },
+    {
+      "epoch": 0.8455821635012386,
+      "grad_norm": 1.6672492017970242,
+      "learning_rate": 1.2247641771752616e-06,
+      "loss": 0.9329,
+      "step": 6144
+    },
+    {
+      "epoch": 0.8457197908064961,
+      "grad_norm": 1.8513922229830846,
+      "learning_rate": 1.2226275647954755e-06,
+      "loss": 0.9621,
+      "step": 6145
+    },
+    {
+      "epoch": 0.8458574181117534,
+      "grad_norm": 1.8075126899538192,
+      "learning_rate": 1.2204926963594433e-06,
+      "loss": 0.8541,
+      "step": 6146
+    },
+    {
+      "epoch": 0.8459950454170108,
+      "grad_norm": 1.8642184446683254,
+      "learning_rate": 1.2183595722913333e-06,
+      "loss": 0.9001,
+      "step": 6147
+    },
+    {
+      "epoch": 0.8461326727222681,
+      "grad_norm": 1.8629039170002117,
+      "learning_rate": 1.21622819301497e-06,
+      "loss": 0.9096,
+      "step": 6148
+    },
+    {
+      "epoch": 0.8462703000275255,
+      "grad_norm": 1.7683356366398046,
+      "learning_rate": 1.2140985589538269e-06,
+      "loss": 0.9099,
+      "step": 6149
+    },
+    {
+      "epoch": 0.8464079273327828,
+      "grad_norm": 1.9537421270030741,
+      "learning_rate": 1.211970670531034e-06,
+      "loss": 0.8904,
+      "step": 6150
+    },
+    {
+      "epoch": 0.8465455546380402,
+      "grad_norm": 1.705587175663193,
+      "learning_rate": 1.2098445281693748e-06,
+      "loss": 0.8336,
+      "step": 6151
+    },
+    {
+      "epoch": 0.8466831819432975,
+      "grad_norm": 1.9411358524930473,
+      "learning_rate": 1.2077201322912814e-06,
+      "loss": 0.875,
+      "step": 6152
+    },
+    {
+      "epoch": 0.846820809248555,
+      "grad_norm": 2.0164169072124647,
+      "learning_rate": 1.2055974833188488e-06,
+      "loss": 0.9274,
+      "step": 6153
+    },
+    {
+      "epoch": 0.8469584365538123,
+      "grad_norm": 1.792438573818221,
+      "learning_rate": 1.2034765816738103e-06,
+      "loss": 0.9657,
+      "step": 6154
+    },
+    {
+      "epoch": 0.8470960638590697,
+      "grad_norm": 1.8858913854586463,
+      "learning_rate": 1.2013574277775674e-06,
+      "loss": 0.8902,
+      "step": 6155
+    },
+    {
+      "epoch": 0.847233691164327,
+      "grad_norm": 2.336056892384564,
+      "learning_rate": 1.1992400220511591e-06,
+      "loss": 0.9771,
+      "step": 6156
+    },
+    {
+      "epoch": 0.8473713184695844,
+      "grad_norm": 1.7673569050817732,
+      "learning_rate": 1.197124364915293e-06,
+      "loss": 0.834,
+      "step": 6157
+    },
+    {
+      "epoch": 0.8475089457748417,
+      "grad_norm": 1.7033321142648965,
+      "learning_rate": 1.195010456790313e-06,
+      "loss": 0.8742,
+      "step": 6158
+    },
+    {
+      "epoch": 0.847646573080099,
+      "grad_norm": 1.7457197264061113,
+      "learning_rate": 1.1928982980962312e-06,
+      "loss": 0.862,
+      "step": 6159
+    },
+    {
+      "epoch": 0.8477842003853564,
+      "grad_norm": 1.808992090745284,
+      "learning_rate": 1.1907878892527003e-06,
+      "loss": 0.9053,
+      "step": 6160
+    },
+    {
+      "epoch": 0.8479218276906139,
+      "grad_norm": 1.8024166637962096,
+      "learning_rate": 1.1886792306790318e-06,
+      "loss": 0.8705,
+      "step": 6161
+    },
+    {
+      "epoch": 0.8480594549958712,
+      "grad_norm": 1.6904549719558852,
+      "learning_rate": 1.1865723227941882e-06,
+      "loss": 0.8306,
+      "step": 6162
+    },
+    {
+      "epoch": 0.8481970823011286,
+      "grad_norm": 2.0832332030525125,
+      "learning_rate": 1.1844671660167773e-06,
+      "loss": 0.9243,
+      "step": 6163
+    },
+    {
+      "epoch": 0.8483347096063859,
+      "grad_norm": 1.8260186390936586,
+      "learning_rate": 1.182363760765073e-06,
+      "loss": 0.8628,
+      "step": 6164
+    },
+    {
+      "epoch": 0.8484723369116433,
+      "grad_norm": 1.4997195278129782,
+      "learning_rate": 1.1802621074569843e-06,
+      "loss": 0.9305,
+      "step": 6165
+    },
+    {
+      "epoch": 0.8486099642169006,
+      "grad_norm": 1.7116906091568327,
+      "learning_rate": 1.1781622065100884e-06,
+      "loss": 0.8911,
+      "step": 6166
+    },
+    {
+      "epoch": 0.848747591522158,
+      "grad_norm": 1.721290941788685,
+      "learning_rate": 1.176064058341605e-06,
+      "loss": 0.9882,
+      "step": 6167
+    },
+    {
+      "epoch": 0.8488852188274154,
+      "grad_norm": 1.5974017197861239,
+      "learning_rate": 1.173967663368406e-06,
+      "loss": 0.9148,
+      "step": 6168
+    },
+    {
+      "epoch": 0.8490228461326728,
+      "grad_norm": 1.7101521735939351,
+      "learning_rate": 1.1718730220070174e-06,
+      "loss": 0.8843,
+      "step": 6169
+    },
+    {
+      "epoch": 0.8491604734379301,
+      "grad_norm": 1.7142188658696245,
+      "learning_rate": 1.1697801346736147e-06,
+      "loss": 0.917,
+      "step": 6170
+    },
+    {
+      "epoch": 0.8492981007431875,
+      "grad_norm": 1.596463143195057,
+      "learning_rate": 1.167689001784027e-06,
+      "loss": 0.9414,
+      "step": 6171
+    },
+    {
+      "epoch": 0.8494357280484448,
+      "grad_norm": 1.5805159736046763,
+      "learning_rate": 1.1655996237537337e-06,
+      "loss": 0.8933,
+      "step": 6172
+    },
+    {
+      "epoch": 0.8495733553537022,
+      "grad_norm": 2.0666283665033847,
+      "learning_rate": 1.163512000997865e-06,
+      "loss": 0.9125,
+      "step": 6173
+    },
+    {
+      "epoch": 0.8497109826589595,
+      "grad_norm": 1.993870261625524,
+      "learning_rate": 1.161426133931204e-06,
+      "loss": 0.8562,
+      "step": 6174
+    },
+    {
+      "epoch": 0.8498486099642169,
+      "grad_norm": 1.8049165722007277,
+      "learning_rate": 1.1593420229681828e-06,
+      "loss": 0.951,
+      "step": 6175
+    },
+    {
+      "epoch": 0.8499862372694743,
+      "grad_norm": 2.0571880605830795,
+      "learning_rate": 1.1572596685228865e-06,
+      "loss": 0.9509,
+      "step": 6176
+    },
+    {
+      "epoch": 0.8501238645747317,
+      "grad_norm": 1.7372973052614344,
+      "learning_rate": 1.1551790710090505e-06,
+      "loss": 0.8661,
+      "step": 6177
+    },
+    {
+      "epoch": 0.850261491879989,
+      "grad_norm": 1.7437380714988833,
+      "learning_rate": 1.1531002308400595e-06,
+      "loss": 0.8775,
+      "step": 6178
+    },
+    {
+      "epoch": 0.8503991191852464,
+      "grad_norm": 1.622455604839031,
+      "learning_rate": 1.1510231484289525e-06,
+      "loss": 0.9552,
+      "step": 6179
+    },
+    {
+      "epoch": 0.8505367464905037,
+      "grad_norm": 1.795308289686882,
+      "learning_rate": 1.1489478241884167e-06,
+      "loss": 0.8687,
+      "step": 6180
+    },
+    {
+      "epoch": 0.850674373795761,
+      "grad_norm": 1.6196311730173674,
+      "learning_rate": 1.1468742585307913e-06,
+      "loss": 0.9117,
+      "step": 6181
+    },
+    {
+      "epoch": 0.8508120011010184,
+      "grad_norm": 1.7395749033639012,
+      "learning_rate": 1.1448024518680634e-06,
+      "loss": 0.9115,
+      "step": 6182
+    },
+    {
+      "epoch": 0.8509496284062759,
+      "grad_norm": 1.579889317628041,
+      "learning_rate": 1.1427324046118738e-06,
+      "loss": 0.9077,
+      "step": 6183
+    },
+    {
+      "epoch": 0.8510872557115332,
+      "grad_norm": 1.7478329843529872,
+      "learning_rate": 1.1406641171735112e-06,
+      "loss": 0.877,
+      "step": 6184
+    },
+    {
+      "epoch": 0.8512248830167906,
+      "grad_norm": 1.597017759813138,
+      "learning_rate": 1.1385975899639212e-06,
+      "loss": 0.9397,
+      "step": 6185
+    },
+    {
+      "epoch": 0.8513625103220479,
+      "grad_norm": 1.823189759823978,
+      "learning_rate": 1.1365328233936867e-06,
+      "loss": 0.8332,
+      "step": 6186
+    },
+    {
+      "epoch": 0.8515001376273053,
+      "grad_norm": 1.749493111741998,
+      "learning_rate": 1.1344698178730574e-06,
+      "loss": 0.9116,
+      "step": 6187
+    },
+    {
+      "epoch": 0.8516377649325626,
+      "grad_norm": 1.96166600476659,
+      "learning_rate": 1.132408573811914e-06,
+      "loss": 0.9301,
+      "step": 6188
+    },
+    {
+      "epoch": 0.85177539223782,
+      "grad_norm": 1.8515830106667426,
+      "learning_rate": 1.130349091619808e-06,
+      "loss": 0.9521,
+      "step": 6189
+    },
+    {
+      "epoch": 0.8519130195430773,
+      "grad_norm": 1.8463081993784243,
+      "learning_rate": 1.1282913717059206e-06,
+      "loss": 0.8828,
+      "step": 6190
+    },
+    {
+      "epoch": 0.8520506468483348,
+      "grad_norm": 2.0521941944871753,
+      "learning_rate": 1.1262354144790999e-06,
+      "loss": 0.9125,
+      "step": 6191
+    },
+    {
+      "epoch": 0.8521882741535921,
+      "grad_norm": 1.8238912532626685,
+      "learning_rate": 1.124181220347833e-06,
+      "loss": 0.9179,
+      "step": 6192
+    },
+    {
+      "epoch": 0.8523259014588495,
+      "grad_norm": 1.94010405763025,
+      "learning_rate": 1.1221287897202604e-06,
+      "loss": 0.8766,
+      "step": 6193
+    },
+    {
+      "epoch": 0.8524635287641068,
+      "grad_norm": 1.8408612021602413,
+      "learning_rate": 1.1200781230041757e-06,
+      "loss": 0.9017,
+      "step": 6194
+    },
+    {
+      "epoch": 0.8526011560693642,
+      "grad_norm": 1.5533309872353143,
+      "learning_rate": 1.118029220607011e-06,
+      "loss": 0.8861,
+      "step": 6195
+    },
+    {
+      "epoch": 0.8527387833746215,
+      "grad_norm": 1.8491282524239647,
+      "learning_rate": 1.1159820829358626e-06,
+      "loss": 0.8449,
+      "step": 6196
+    },
+    {
+      "epoch": 0.8528764106798788,
+      "grad_norm": 1.5890951668967603,
+      "learning_rate": 1.1139367103974607e-06,
+      "loss": 0.9049,
+      "step": 6197
+    },
+    {
+      "epoch": 0.8530140379851362,
+      "grad_norm": 1.8682495051675239,
+      "learning_rate": 1.1118931033982005e-06,
+      "loss": 0.9028,
+      "step": 6198
+    },
+    {
+      "epoch": 0.8531516652903937,
+      "grad_norm": 1.6849700445351836,
+      "learning_rate": 1.109851262344116e-06,
+      "loss": 0.8539,
+      "step": 6199
+    },
+    {
+      "epoch": 0.853289292595651,
+      "grad_norm": 1.7085091748241745,
+      "learning_rate": 1.1078111876408925e-06,
+      "loss": 0.896,
+      "step": 6200
+    },
+    {
+      "epoch": 0.8534269199009084,
+      "grad_norm": 1.7627489964715688,
+      "learning_rate": 1.1057728796938649e-06,
+      "loss": 0.9103,
+      "step": 6201
+    },
+    {
+      "epoch": 0.8535645472061657,
+      "grad_norm": 1.7732265115369894,
+      "learning_rate": 1.103736338908017e-06,
+      "loss": 0.9877,
+      "step": 6202
+    },
+    {
+      "epoch": 0.853702174511423,
+      "grad_norm": 2.3317087875938443,
+      "learning_rate": 1.1017015656879826e-06,
+      "loss": 0.8577,
+      "step": 6203
+    },
+    {
+      "epoch": 0.8538398018166804,
+      "grad_norm": 1.8906089410673719,
+      "learning_rate": 1.099668560438043e-06,
+      "loss": 0.9098,
+      "step": 6204
+    },
+    {
+      "epoch": 0.8539774291219377,
+      "grad_norm": 1.7741615332785652,
+      "learning_rate": 1.0976373235621273e-06,
+      "loss": 0.9202,
+      "step": 6205
+    },
+    {
+      "epoch": 0.8541150564271952,
+      "grad_norm": 1.7795041831959177,
+      "learning_rate": 1.0956078554638161e-06,
+      "loss": 0.9812,
+      "step": 6206
+    },
+    {
+      "epoch": 0.8542526837324526,
+      "grad_norm": 1.9280939130133263,
+      "learning_rate": 1.0935801565463356e-06,
+      "loss": 0.8748,
+      "step": 6207
+    },
+    {
+      "epoch": 0.8543903110377099,
+      "grad_norm": 1.6985953902383173,
+      "learning_rate": 1.0915542272125635e-06,
+      "loss": 0.8844,
+      "step": 6208
+    },
+    {
+      "epoch": 0.8545279383429673,
+      "grad_norm": 1.7656054135347574,
+      "learning_rate": 1.0895300678650223e-06,
+      "loss": 0.8981,
+      "step": 6209
+    },
+    {
+      "epoch": 0.8546655656482246,
+      "grad_norm": 1.882449130548782,
+      "learning_rate": 1.0875076789058857e-06,
+      "loss": 0.8684,
+      "step": 6210
+    },
+    {
+      "epoch": 0.854803192953482,
+      "grad_norm": 2.0828732529177283,
+      "learning_rate": 1.0854870607369738e-06,
+      "loss": 0.8886,
+      "step": 6211
+    },
+    {
+      "epoch": 0.8549408202587393,
+      "grad_norm": 1.5762031784650956,
+      "learning_rate": 1.0834682137597575e-06,
+      "loss": 0.9016,
+      "step": 6212
+    },
+    {
+      "epoch": 0.8550784475639966,
+      "grad_norm": 1.9997176451775578,
+      "learning_rate": 1.0814511383753524e-06,
+      "loss": 0.9091,
+      "step": 6213
+    },
+    {
+      "epoch": 0.8552160748692541,
+      "grad_norm": 1.6008286782788077,
+      "learning_rate": 1.0794358349845258e-06,
+      "loss": 0.901,
+      "step": 6214
+    },
+    {
+      "epoch": 0.8553537021745115,
+      "grad_norm": 1.747513288586546,
+      "learning_rate": 1.0774223039876884e-06,
+      "loss": 0.933,
+      "step": 6215
+    },
+    {
+      "epoch": 0.8554913294797688,
+      "grad_norm": 1.691661423142964,
+      "learning_rate": 1.0754105457849007e-06,
+      "loss": 0.8802,
+      "step": 6216
+    },
+    {
+      "epoch": 0.8556289567850262,
+      "grad_norm": 2.0459210262909413,
+      "learning_rate": 1.0734005607758768e-06,
+      "loss": 0.8389,
+      "step": 6217
+    },
+    {
+      "epoch": 0.8557665840902835,
+      "grad_norm": 1.7241595170062474,
+      "learning_rate": 1.0713923493599665e-06,
+      "loss": 0.8852,
+      "step": 6218
+    },
+    {
+      "epoch": 0.8559042113955408,
+      "grad_norm": 1.760415045388538,
+      "learning_rate": 1.0693859119361805e-06,
+      "loss": 0.945,
+      "step": 6219
+    },
+    {
+      "epoch": 0.8560418387007982,
+      "grad_norm": 1.8553260892920427,
+      "learning_rate": 1.0673812489031632e-06,
+      "loss": 0.8616,
+      "step": 6220
+    },
+    {
+      "epoch": 0.8561794660060557,
+      "grad_norm": 1.8733594274032301,
+      "learning_rate": 1.0653783606592193e-06,
+      "loss": 0.9112,
+      "step": 6221
+    },
+    {
+      "epoch": 0.856317093311313,
+      "grad_norm": 1.661550675841988,
+      "learning_rate": 1.063377247602294e-06,
+      "loss": 0.9323,
+      "step": 6222
+    },
+    {
+      "epoch": 0.8564547206165704,
+      "grad_norm": 2.0148655361835837,
+      "learning_rate": 1.0613779101299794e-06,
+      "loss": 0.8318,
+      "step": 6223
+    },
+    {
+      "epoch": 0.8565923479218277,
+      "grad_norm": 1.7556268197937792,
+      "learning_rate": 1.0593803486395192e-06,
+      "loss": 0.8494,
+      "step": 6224
+    },
+    {
+      "epoch": 0.856729975227085,
+      "grad_norm": 1.824161035374066,
+      "learning_rate": 1.057384563527799e-06,
+      "loss": 0.8797,
+      "step": 6225
+    },
+    {
+      "epoch": 0.8568676025323424,
+      "grad_norm": 1.7644634000553918,
+      "learning_rate": 1.0553905551913568e-06,
+      "loss": 0.8914,
+      "step": 6226
+    },
+    {
+      "epoch": 0.8570052298375997,
+      "grad_norm": 1.8137094918409802,
+      "learning_rate": 1.05339832402637e-06,
+      "loss": 0.9026,
+      "step": 6227
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.5935991393795679,
+      "learning_rate": 1.0514078704286713e-06,
+      "loss": 0.8273,
+      "step": 6228
+    },
+    {
+      "epoch": 0.8572804844481146,
+      "grad_norm": 1.8627668337735328,
+      "learning_rate": 1.049419194793737e-06,
+      "loss": 0.8461,
+      "step": 6229
+    },
+    {
+      "epoch": 0.8574181117533719,
+      "grad_norm": 1.8243143285536125,
+      "learning_rate": 1.0474322975166872e-06,
+      "loss": 0.8747,
+      "step": 6230
+    },
+    {
+      "epoch": 0.8575557390586293,
+      "grad_norm": 1.607035697618476,
+      "learning_rate": 1.0454471789922938e-06,
+      "loss": 0.9447,
+      "step": 6231
+    },
+    {
+      "epoch": 0.8576933663638866,
+      "grad_norm": 1.8249112881971736,
+      "learning_rate": 1.043463839614971e-06,
+      "loss": 0.8677,
+      "step": 6232
+    },
+    {
+      "epoch": 0.857830993669144,
+      "grad_norm": 1.834360652593746,
+      "learning_rate": 1.0414822797787827e-06,
+      "loss": 0.8972,
+      "step": 6233
+    },
+    {
+      "epoch": 0.8579686209744013,
+      "grad_norm": 1.734085765353281,
+      "learning_rate": 1.039502499877435e-06,
+      "loss": 0.8738,
+      "step": 6234
+    },
+    {
+      "epoch": 0.8581062482796586,
+      "grad_norm": 1.7648585493823916,
+      "learning_rate": 1.0375245003042855e-06,
+      "loss": 0.9147,
+      "step": 6235
+    },
+    {
+      "epoch": 0.858243875584916,
+      "grad_norm": 1.6533946327914746,
+      "learning_rate": 1.0355482814523355e-06,
+      "loss": 0.8877,
+      "step": 6236
+    },
+    {
+      "epoch": 0.8583815028901735,
+      "grad_norm": 1.62368507915153,
+      "learning_rate": 1.0335738437142307e-06,
+      "loss": 0.8561,
+      "step": 6237
+    },
+    {
+      "epoch": 0.8585191301954308,
+      "grad_norm": 1.9459208143022695,
+      "learning_rate": 1.0316011874822673e-06,
+      "loss": 0.8548,
+      "step": 6238
+    },
+    {
+      "epoch": 0.8586567575006882,
+      "grad_norm": 1.5434341527162416,
+      "learning_rate": 1.029630313148381e-06,
+      "loss": 0.9445,
+      "step": 6239
+    },
+    {
+      "epoch": 0.8587943848059455,
+      "grad_norm": 1.6948168797812844,
+      "learning_rate": 1.0276612211041647e-06,
+      "loss": 0.8671,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8589320121112028,
+      "grad_norm": 1.655727692137088,
+      "learning_rate": 1.025693911740845e-06,
+      "loss": 0.8963,
+      "step": 6241
+    },
+    {
+      "epoch": 0.8590696394164602,
+      "grad_norm": 1.9142942378624233,
+      "learning_rate": 1.0237283854492984e-06,
+      "loss": 0.7426,
+      "step": 6242
+    },
+    {
+      "epoch": 0.8592072667217175,
+      "grad_norm": 1.7050248295069106,
+      "learning_rate": 1.0217646426200512e-06,
+      "loss": 0.8821,
+      "step": 6243
+    },
+    {
+      "epoch": 0.859344894026975,
+      "grad_norm": 1.846644966719409,
+      "learning_rate": 1.0198026836432705e-06,
+      "loss": 0.9452,
+      "step": 6244
+    },
+    {
+      "epoch": 0.8594825213322324,
+      "grad_norm": 1.589539889630898,
+      "learning_rate": 1.0178425089087707e-06,
+      "loss": 0.9316,
+      "step": 6245
+    },
+    {
+      "epoch": 0.8596201486374897,
+      "grad_norm": 1.8671977305075909,
+      "learning_rate": 1.0158841188060097e-06,
+      "loss": 0.9304,
+      "step": 6246
+    },
+    {
+      "epoch": 0.859757775942747,
+      "grad_norm": 1.6998972779955002,
+      "learning_rate": 1.0139275137241e-06,
+      "loss": 0.8841,
+      "step": 6247
+    },
+    {
+      "epoch": 0.8598954032480044,
+      "grad_norm": 1.6411868750384548,
+      "learning_rate": 1.0119726940517826e-06,
+      "loss": 0.8831,
+      "step": 6248
+    },
+    {
+      "epoch": 0.8600330305532617,
+      "grad_norm": 1.8325711136668632,
+      "learning_rate": 1.0100196601774636e-06,
+      "loss": 0.8513,
+      "step": 6249
+    },
+    {
+      "epoch": 0.8601706578585191,
+      "grad_norm": 1.673958912207221,
+      "learning_rate": 1.0080684124891737e-06,
+      "loss": 0.8326,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8603082851637764,
+      "grad_norm": 1.6930924073611227,
+      "learning_rate": 1.0061189513746084e-06,
+      "loss": 0.8934,
+      "step": 6251
+    },
+    {
+      "epoch": 0.8604459124690339,
+      "grad_norm": 1.6526857250595426,
+      "learning_rate": 1.00417127722109e-06,
+      "loss": 0.8588,
+      "step": 6252
+    },
+    {
+      "epoch": 0.8605835397742913,
+      "grad_norm": 1.6392924387162284,
+      "learning_rate": 1.0022253904156021e-06,
+      "loss": 0.9076,
+      "step": 6253
+    },
+    {
+      "epoch": 0.8607211670795486,
+      "grad_norm": 1.9012759811366886,
+      "learning_rate": 1.0002812913447623e-06,
+      "loss": 0.9007,
+      "step": 6254
+    },
+    {
+      "epoch": 0.860858794384806,
+      "grad_norm": 1.6634464242122433,
+      "learning_rate": 9.983389803948374e-07,
+      "loss": 0.9173,
+      "step": 6255
+    },
+    {
+      "epoch": 0.8609964216900633,
+      "grad_norm": 1.5711386034534134,
+      "learning_rate": 9.963984579517382e-07,
+      "loss": 0.8745,
+      "step": 6256
+    },
+    {
+      "epoch": 0.8611340489953206,
+      "grad_norm": 1.8215760569065331,
+      "learning_rate": 9.94459724401019e-07,
+      "loss": 0.8029,
+      "step": 6257
+    },
+    {
+      "epoch": 0.861271676300578,
+      "grad_norm": 2.254793095089383,
+      "learning_rate": 9.925227801278826e-07,
+      "loss": 0.9907,
+      "step": 6258
+    },
+    {
+      "epoch": 0.8614093036058355,
+      "grad_norm": 1.6885256395454302,
+      "learning_rate": 9.905876255171665e-07,
+      "loss": 0.8028,
+      "step": 6259
+    },
+    {
+      "epoch": 0.8615469309110928,
+      "grad_norm": 1.9243112534056797,
+      "learning_rate": 9.886542609533668e-07,
+      "loss": 0.8267,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8616845582163501,
+      "grad_norm": 1.9010430236512281,
+      "learning_rate": 9.867226868206127e-07,
+      "loss": 0.8641,
+      "step": 6261
+    },
+    {
+      "epoch": 0.8618221855216075,
+      "grad_norm": 1.9111698643083475,
+      "learning_rate": 9.847929035026826e-07,
+      "loss": 0.9461,
+      "step": 6262
+    },
+    {
+      "epoch": 0.8619598128268648,
+      "grad_norm": 2.8209566271193287,
+      "learning_rate": 9.82864911382997e-07,
+      "loss": 0.9151,
+      "step": 6263
+    },
+    {
+      "epoch": 0.8620974401321222,
+      "grad_norm": 1.7380561692099294,
+      "learning_rate": 9.809387108446234e-07,
+      "loss": 0.8923,
+      "step": 6264
+    },
+    {
+      "epoch": 0.8622350674373795,
+      "grad_norm": 1.7854983960326407,
+      "learning_rate": 9.790143022702703e-07,
+      "loss": 0.9246,
+      "step": 6265
+    },
+    {
+      "epoch": 0.8623726947426369,
+      "grad_norm": 1.8401999725636915,
+      "learning_rate": 9.770916860422907e-07,
+      "loss": 0.9202,
+      "step": 6266
+    },
+    {
+      "epoch": 0.8625103220478944,
+      "grad_norm": 1.9592330923428067,
+      "learning_rate": 9.751708625426836e-07,
+      "loss": 0.9526,
+      "step": 6267
+    },
+    {
+      "epoch": 0.8626479493531517,
+      "grad_norm": 1.8987182144016828,
+      "learning_rate": 9.732518321530904e-07,
+      "loss": 0.9473,
+      "step": 6268
+    },
+    {
+      "epoch": 0.862785576658409,
+      "grad_norm": 2.079129585369098,
+      "learning_rate": 9.713345952547949e-07,
+      "loss": 0.9041,
+      "step": 6269
+    },
+    {
+      "epoch": 0.8629232039636664,
+      "grad_norm": 1.9200819556924664,
+      "learning_rate": 9.694191522287256e-07,
+      "loss": 0.938,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8630608312689237,
+      "grad_norm": 1.8636943318134989,
+      "learning_rate": 9.675055034554547e-07,
+      "loss": 0.9252,
+      "step": 6271
+    },
+    {
+      "epoch": 0.8631984585741811,
+      "grad_norm": 1.5840490516828283,
+      "learning_rate": 9.655936493152018e-07,
+      "loss": 0.7854,
+      "step": 6272
+    },
+    {
+      "epoch": 0.8633360858794384,
+      "grad_norm": 1.9078694551675215,
+      "learning_rate": 9.636835901878217e-07,
+      "loss": 0.8825,
+      "step": 6273
+    },
+    {
+      "epoch": 0.8634737131846958,
+      "grad_norm": 1.9859282786300518,
+      "learning_rate": 9.617753264528184e-07,
+      "loss": 1.0039,
+      "step": 6274
+    },
+    {
+      "epoch": 0.8636113404899532,
+      "grad_norm": 1.8353425757594175,
+      "learning_rate": 9.59868858489338e-07,
+      "loss": 0.8419,
+      "step": 6275
+    },
+    {
+      "epoch": 0.8637489677952106,
+      "grad_norm": 1.6029197278096232,
+      "learning_rate": 9.579641866761702e-07,
+      "loss": 0.9292,
+      "step": 6276
+    },
+    {
+      "epoch": 0.863886595100468,
+      "grad_norm": 1.7337445203655124,
+      "learning_rate": 9.560613113917462e-07,
+      "loss": 0.9725,
+      "step": 6277
+    },
+    {
+      "epoch": 0.8640242224057253,
+      "grad_norm": 1.5568334729503015,
+      "learning_rate": 9.541602330141386e-07,
+      "loss": 0.8627,
+      "step": 6278
+    },
+    {
+      "epoch": 0.8641618497109826,
+      "grad_norm": 1.8046165228091906,
+      "learning_rate": 9.522609519210735e-07,
+      "loss": 0.8959,
+      "step": 6279
+    },
+    {
+      "epoch": 0.86429947701624,
+      "grad_norm": 1.839020720605846,
+      "learning_rate": 9.503634684899033e-07,
+      "loss": 0.8946,
+      "step": 6280
+    },
+    {
+      "epoch": 0.8644371043214973,
+      "grad_norm": 1.744491745807985,
+      "learning_rate": 9.484677830976396e-07,
+      "loss": 0.8513,
+      "step": 6281
+    },
+    {
+      "epoch": 0.8645747316267548,
+      "grad_norm": 1.7600007823660464,
+      "learning_rate": 9.465738961209214e-07,
+      "loss": 0.8734,
+      "step": 6282
+    },
+    {
+      "epoch": 0.8647123589320121,
+      "grad_norm": 2.097317385804689,
+      "learning_rate": 9.446818079360465e-07,
+      "loss": 0.8637,
+      "step": 6283
+    },
+    {
+      "epoch": 0.8648499862372695,
+      "grad_norm": 1.792349322173358,
+      "learning_rate": 9.427915189189385e-07,
+      "loss": 0.92,
+      "step": 6284
+    },
+    {
+      "epoch": 0.8649876135425268,
+      "grad_norm": 1.6838078052423404,
+      "learning_rate": 9.409030294451771e-07,
+      "loss": 0.8759,
+      "step": 6285
+    },
+    {
+      "epoch": 0.8651252408477842,
+      "grad_norm": 1.8965747909653592,
+      "learning_rate": 9.390163398899788e-07,
+      "loss": 0.8594,
+      "step": 6286
+    },
+    {
+      "epoch": 0.8652628681530415,
+      "grad_norm": 1.7187042335812959,
+      "learning_rate": 9.371314506282026e-07,
+      "loss": 0.8658,
+      "step": 6287
+    },
+    {
+      "epoch": 0.8654004954582989,
+      "grad_norm": 1.736646926580443,
+      "learning_rate": 9.3524836203435e-07,
+      "loss": 0.903,
+      "step": 6288
+    },
+    {
+      "epoch": 0.8655381227635562,
+      "grad_norm": 1.6405004807193584,
+      "learning_rate": 9.333670744825652e-07,
+      "loss": 0.8413,
+      "step": 6289
+    },
+    {
+      "epoch": 0.8656757500688137,
+      "grad_norm": 1.9928999627402626,
+      "learning_rate": 9.314875883466335e-07,
+      "loss": 0.9367,
+      "step": 6290
+    },
+    {
+      "epoch": 0.865813377374071,
+      "grad_norm": 2.015746288041542,
+      "learning_rate": 9.29609903999984e-07,
+      "loss": 0.9,
+      "step": 6291
+    },
+    {
+      "epoch": 0.8659510046793284,
+      "grad_norm": 1.7682184098481524,
+      "learning_rate": 9.27734021815686e-07,
+      "loss": 0.8982,
+      "step": 6292
+    },
+    {
+      "epoch": 0.8660886319845857,
+      "grad_norm": 1.9331544655860518,
+      "learning_rate": 9.258599421664527e-07,
+      "loss": 0.8793,
+      "step": 6293
+    },
+    {
+      "epoch": 0.8662262592898431,
+      "grad_norm": 1.8157555149059545,
+      "learning_rate": 9.239876654246371e-07,
+      "loss": 0.9145,
+      "step": 6294
+    },
+    {
+      "epoch": 0.8663638865951004,
+      "grad_norm": 1.8900435670286593,
+      "learning_rate": 9.221171919622352e-07,
+      "loss": 0.8649,
+      "step": 6295
+    },
+    {
+      "epoch": 0.8665015139003578,
+      "grad_norm": 2.109691307902643,
+      "learning_rate": 9.20248522150885e-07,
+      "loss": 0.8854,
+      "step": 6296
+    },
+    {
+      "epoch": 0.8666391412056152,
+      "grad_norm": 1.8004300274607514,
+      "learning_rate": 9.183816563618641e-07,
+      "loss": 0.932,
+      "step": 6297
+    },
+    {
+      "epoch": 0.8667767685108726,
+      "grad_norm": 1.7325947887059283,
+      "learning_rate": 9.165165949660959e-07,
+      "loss": 0.9331,
+      "step": 6298
+    },
+    {
+      "epoch": 0.86691439581613,
+      "grad_norm": 1.634354875671112,
+      "learning_rate": 9.146533383341394e-07,
+      "loss": 0.9443,
+      "step": 6299
+    },
+    {
+      "epoch": 0.8670520231213873,
+      "grad_norm": 1.7494753398255085,
+      "learning_rate": 9.127918868362006e-07,
+      "loss": 0.8638,
+      "step": 6300
+    },
+    {
+      "epoch": 0.8671896504266446,
+      "grad_norm": 1.733450030302839,
+      "learning_rate": 9.109322408421217e-07,
+      "loss": 0.8891,
+      "step": 6301
+    },
+    {
+      "epoch": 0.867327277731902,
+      "grad_norm": 1.584905250717093,
+      "learning_rate": 9.090744007213948e-07,
+      "loss": 0.8938,
+      "step": 6302
+    },
+    {
+      "epoch": 0.8674649050371593,
+      "grad_norm": 1.8141678255413258,
+      "learning_rate": 9.072183668431389e-07,
+      "loss": 0.8805,
+      "step": 6303
+    },
+    {
+      "epoch": 0.8676025323424167,
+      "grad_norm": 1.9020457554564945,
+      "learning_rate": 9.053641395761303e-07,
+      "loss": 0.9193,
+      "step": 6304
+    },
+    {
+      "epoch": 0.8677401596476741,
+      "grad_norm": 1.6497987739186861,
+      "learning_rate": 9.035117192887743e-07,
+      "loss": 0.8922,
+      "step": 6305
+    },
+    {
+      "epoch": 0.8678777869529315,
+      "grad_norm": 1.6889959614906276,
+      "learning_rate": 9.01661106349121e-07,
+      "loss": 0.9396,
+      "step": 6306
+    },
+    {
+      "epoch": 0.8680154142581888,
+      "grad_norm": 1.8666693448828282,
+      "learning_rate": 8.998123011248639e-07,
+      "loss": 0.8792,
+      "step": 6307
+    },
+    {
+      "epoch": 0.8681530415634462,
+      "grad_norm": 1.6628947515863657,
+      "learning_rate": 8.979653039833336e-07,
+      "loss": 0.9292,
+      "step": 6308
+    },
+    {
+      "epoch": 0.8682906688687035,
+      "grad_norm": 1.7178438640142257,
+      "learning_rate": 8.961201152915067e-07,
+      "loss": 0.9303,
+      "step": 6309
+    },
+    {
+      "epoch": 0.8684282961739609,
+      "grad_norm": 2.1109822306351296,
+      "learning_rate": 8.94276735415991e-07,
+      "loss": 0.9046,
+      "step": 6310
+    },
+    {
+      "epoch": 0.8685659234792182,
+      "grad_norm": 1.8667267425314287,
+      "learning_rate": 8.924351647230479e-07,
+      "loss": 0.8646,
+      "step": 6311
+    },
+    {
+      "epoch": 0.8687035507844756,
+      "grad_norm": 1.789102127393617,
+      "learning_rate": 8.905954035785646e-07,
+      "loss": 0.8941,
+      "step": 6312
+    },
+    {
+      "epoch": 0.868841178089733,
+      "grad_norm": 1.7309625894204395,
+      "learning_rate": 8.887574523480847e-07,
+      "loss": 0.8685,
+      "step": 6313
+    },
+    {
+      "epoch": 0.8689788053949904,
+      "grad_norm": 1.810638738665834,
+      "learning_rate": 8.869213113967756e-07,
+      "loss": 0.855,
+      "step": 6314
+    },
+    {
+      "epoch": 0.8691164327002477,
+      "grad_norm": 1.8431368190404611,
+      "learning_rate": 8.850869810894591e-07,
+      "loss": 0.8699,
+      "step": 6315
+    },
+    {
+      "epoch": 0.8692540600055051,
+      "grad_norm": 1.7187531972697576,
+      "learning_rate": 8.832544617905902e-07,
+      "loss": 0.8786,
+      "step": 6316
+    },
+    {
+      "epoch": 0.8693916873107624,
+      "grad_norm": 1.6521170750378698,
+      "learning_rate": 8.814237538642656e-07,
+      "loss": 0.8979,
+      "step": 6317
+    },
+    {
+      "epoch": 0.8695293146160198,
+      "grad_norm": 1.645183724165715,
+      "learning_rate": 8.795948576742209e-07,
+      "loss": 0.916,
+      "step": 6318
+    },
+    {
+      "epoch": 0.8696669419212771,
+      "grad_norm": 1.76123362687622,
+      "learning_rate": 8.77767773583833e-07,
+      "loss": 0.9087,
+      "step": 6319
+    },
+    {
+      "epoch": 0.8698045692265346,
+      "grad_norm": 1.9093273651390985,
+      "learning_rate": 8.759425019561196e-07,
+      "loss": 0.9791,
+      "step": 6320
+    },
+    {
+      "epoch": 0.869942196531792,
+      "grad_norm": 1.780088933989605,
+      "learning_rate": 8.74119043153735e-07,
+      "loss": 0.9689,
+      "step": 6321
+    },
+    {
+      "epoch": 0.8700798238370493,
+      "grad_norm": 1.7879835357412397,
+      "learning_rate": 8.722973975389782e-07,
+      "loss": 0.9233,
+      "step": 6322
+    },
+    {
+      "epoch": 0.8702174511423066,
+      "grad_norm": 1.5389503308348593,
+      "learning_rate": 8.704775654737829e-07,
+      "loss": 0.8581,
+      "step": 6323
+    },
+    {
+      "epoch": 0.870355078447564,
+      "grad_norm": 1.753843554425785,
+      "learning_rate": 8.686595473197257e-07,
+      "loss": 0.9103,
+      "step": 6324
+    },
+    {
+      "epoch": 0.8704927057528213,
+      "grad_norm": 1.6995973295819178,
+      "learning_rate": 8.66843343438023e-07,
+      "loss": 0.8744,
+      "step": 6325
+    },
+    {
+      "epoch": 0.8706303330580787,
+      "grad_norm": 1.8521250832114422,
+      "learning_rate": 8.650289541895285e-07,
+      "loss": 0.9923,
+      "step": 6326
+    },
+    {
+      "epoch": 0.870767960363336,
+      "grad_norm": 1.6785191208659946,
+      "learning_rate": 8.63216379934736e-07,
+      "loss": 0.9399,
+      "step": 6327
+    },
+    {
+      "epoch": 0.8709055876685935,
+      "grad_norm": 1.7242708645722309,
+      "learning_rate": 8.614056210337795e-07,
+      "loss": 0.9523,
+      "step": 6328
+    },
+    {
+      "epoch": 0.8710432149738508,
+      "grad_norm": 1.5861174351691378,
+      "learning_rate": 8.595966778464338e-07,
+      "loss": 0.7835,
+      "step": 6329
+    },
+    {
+      "epoch": 0.8711808422791082,
+      "grad_norm": 1.6840028909608171,
+      "learning_rate": 8.577895507321088e-07,
+      "loss": 0.9035,
+      "step": 6330
+    },
+    {
+      "epoch": 0.8713184695843655,
+      "grad_norm": 1.7191411712837283,
+      "learning_rate": 8.559842400498564e-07,
+      "loss": 0.879,
+      "step": 6331
+    },
+    {
+      "epoch": 0.8714560968896229,
+      "grad_norm": 1.7825350412061591,
+      "learning_rate": 8.541807461583684e-07,
+      "loss": 0.8704,
+      "step": 6332
+    },
+    {
+      "epoch": 0.8715937241948802,
+      "grad_norm": 1.5307914260203452,
+      "learning_rate": 8.523790694159717e-07,
+      "loss": 0.8741,
+      "step": 6333
+    },
+    {
+      "epoch": 0.8717313515001376,
+      "grad_norm": 1.5895739537259068,
+      "learning_rate": 8.505792101806399e-07,
+      "loss": 0.9052,
+      "step": 6334
+    },
+    {
+      "epoch": 0.871868978805395,
+      "grad_norm": 1.6085163080113656,
+      "learning_rate": 8.487811688099734e-07,
+      "loss": 0.8694,
+      "step": 6335
+    },
+    {
+      "epoch": 0.8720066061106524,
+      "grad_norm": 1.8569943216146498,
+      "learning_rate": 8.469849456612267e-07,
+      "loss": 0.8521,
+      "step": 6336
+    },
+    {
+      "epoch": 0.8721442334159097,
+      "grad_norm": 1.6871735656310338,
+      "learning_rate": 8.451905410912775e-07,
+      "loss": 0.8608,
+      "step": 6337
+    },
+    {
+      "epoch": 0.8722818607211671,
+      "grad_norm": 1.5621967849273881,
+      "learning_rate": 8.433979554566518e-07,
+      "loss": 0.9175,
+      "step": 6338
+    },
+    {
+      "epoch": 0.8724194880264244,
+      "grad_norm": 1.991136048183134,
+      "learning_rate": 8.416071891135124e-07,
+      "loss": 0.8564,
+      "step": 6339
+    },
+    {
+      "epoch": 0.8725571153316818,
+      "grad_norm": 1.7755791811167518,
+      "learning_rate": 8.398182424176582e-07,
+      "loss": 0.938,
+      "step": 6340
+    },
+    {
+      "epoch": 0.8726947426369391,
+      "grad_norm": 1.717335521906552,
+      "learning_rate": 8.380311157245335e-07,
+      "loss": 0.8686,
+      "step": 6341
+    },
+    {
+      "epoch": 0.8728323699421965,
+      "grad_norm": 1.9847187413732654,
+      "learning_rate": 8.36245809389209e-07,
+      "loss": 0.9475,
+      "step": 6342
+    },
+    {
+      "epoch": 0.8729699972474539,
+      "grad_norm": 1.6617495762541499,
+      "learning_rate": 8.344623237664073e-07,
+      "loss": 0.9122,
+      "step": 6343
+    },
+    {
+      "epoch": 0.8731076245527113,
+      "grad_norm": 1.7830617518430338,
+      "learning_rate": 8.326806592104764e-07,
+      "loss": 0.9082,
+      "step": 6344
+    },
+    {
+      "epoch": 0.8732452518579686,
+      "grad_norm": 1.7477154403518669,
+      "learning_rate": 8.309008160754151e-07,
+      "loss": 0.887,
+      "step": 6345
+    },
+    {
+      "epoch": 0.873382879163226,
+      "grad_norm": 2.0417345218428284,
+      "learning_rate": 8.291227947148462e-07,
+      "loss": 0.9705,
+      "step": 6346
+    },
+    {
+      "epoch": 0.8735205064684833,
+      "grad_norm": 1.7391080974950275,
+      "learning_rate": 8.273465954820436e-07,
+      "loss": 0.889,
+      "step": 6347
+    },
+    {
+      "epoch": 0.8736581337737407,
+      "grad_norm": 2.014714504012215,
+      "learning_rate": 8.255722187299142e-07,
+      "loss": 0.9744,
+      "step": 6348
+    },
+    {
+      "epoch": 0.873795761078998,
+      "grad_norm": 1.7514079570638046,
+      "learning_rate": 8.237996648109991e-07,
+      "loss": 0.9581,
+      "step": 6349
+    },
+    {
+      "epoch": 0.8739333883842554,
+      "grad_norm": 1.8316563545438742,
+      "learning_rate": 8.220289340774835e-07,
+      "loss": 0.8527,
+      "step": 6350
+    },
+    {
+      "epoch": 0.8740710156895128,
+      "grad_norm": 1.988474765344503,
+      "learning_rate": 8.202600268811844e-07,
+      "loss": 0.9276,
+      "step": 6351
+    },
+    {
+      "epoch": 0.8742086429947702,
+      "grad_norm": 1.8268774279820574,
+      "learning_rate": 8.184929435735622e-07,
+      "loss": 0.8775,
+      "step": 6352
+    },
+    {
+      "epoch": 0.8743462703000275,
+      "grad_norm": 1.891187046561271,
+      "learning_rate": 8.167276845057104e-07,
+      "loss": 0.9322,
+      "step": 6353
+    },
+    {
+      "epoch": 0.8744838976052849,
+      "grad_norm": 1.8819324364304035,
+      "learning_rate": 8.149642500283627e-07,
+      "loss": 0.9324,
+      "step": 6354
+    },
+    {
+      "epoch": 0.8746215249105422,
+      "grad_norm": 1.9641177679425152,
+      "learning_rate": 8.132026404918891e-07,
+      "loss": 0.9039,
+      "step": 6355
+    },
+    {
+      "epoch": 0.8747591522157996,
+      "grad_norm": 1.6889952085972064,
+      "learning_rate": 8.114428562462973e-07,
+      "loss": 0.9126,
+      "step": 6356
+    },
+    {
+      "epoch": 0.8748967795210569,
+      "grad_norm": 1.7317647726110619,
+      "learning_rate": 8.096848976412319e-07,
+      "loss": 0.8739,
+      "step": 6357
+    },
+    {
+      "epoch": 0.8750344068263144,
+      "grad_norm": 1.761934660674444,
+      "learning_rate": 8.079287650259759e-07,
+      "loss": 0.9663,
+      "step": 6358
+    },
+    {
+      "epoch": 0.8751720341315717,
+      "grad_norm": 1.983540032809982,
+      "learning_rate": 8.061744587494491e-07,
+      "loss": 0.8636,
+      "step": 6359
+    },
+    {
+      "epoch": 0.8753096614368291,
+      "grad_norm": 1.875032675908984,
+      "learning_rate": 8.04421979160207e-07,
+      "loss": 0.9254,
+      "step": 6360
+    },
+    {
+      "epoch": 0.8754472887420864,
+      "grad_norm": 1.794436931006903,
+      "learning_rate": 8.026713266064434e-07,
+      "loss": 0.8982,
+      "step": 6361
+    },
+    {
+      "epoch": 0.8755849160473438,
+      "grad_norm": 2.0245174167597124,
+      "learning_rate": 8.0092250143599e-07,
+      "loss": 0.9641,
+      "step": 6362
+    },
+    {
+      "epoch": 0.8757225433526011,
+      "grad_norm": 1.664289405988689,
+      "learning_rate": 7.991755039963134e-07,
+      "loss": 0.8719,
+      "step": 6363
+    },
+    {
+      "epoch": 0.8758601706578585,
+      "grad_norm": 1.584552844868047,
+      "learning_rate": 7.974303346345202e-07,
+      "loss": 0.9307,
+      "step": 6364
+    },
+    {
+      "epoch": 0.8759977979631158,
+      "grad_norm": 1.991380471171234,
+      "learning_rate": 7.956869936973477e-07,
+      "loss": 0.9301,
+      "step": 6365
+    },
+    {
+      "epoch": 0.8761354252683733,
+      "grad_norm": 2.214259244239434,
+      "learning_rate": 7.939454815311787e-07,
+      "loss": 0.9309,
+      "step": 6366
+    },
+    {
+      "epoch": 0.8762730525736306,
+      "grad_norm": 1.89229127387758,
+      "learning_rate": 7.92205798482023e-07,
+      "loss": 0.9328,
+      "step": 6367
+    },
+    {
+      "epoch": 0.876410679878888,
+      "grad_norm": 1.7676159587421614,
+      "learning_rate": 7.904679448955366e-07,
+      "loss": 0.8914,
+      "step": 6368
+    },
+    {
+      "epoch": 0.8765483071841453,
+      "grad_norm": 1.8367416268357815,
+      "learning_rate": 7.887319211170042e-07,
+      "loss": 0.9715,
+      "step": 6369
+    },
+    {
+      "epoch": 0.8766859344894027,
+      "grad_norm": 1.6900339129382613,
+      "learning_rate": 7.869977274913476e-07,
+      "loss": 0.8871,
+      "step": 6370
+    },
+    {
+      "epoch": 0.87682356179466,
+      "grad_norm": 1.649013007781255,
+      "learning_rate": 7.852653643631347e-07,
+      "loss": 0.8966,
+      "step": 6371
+    },
+    {
+      "epoch": 0.8769611890999174,
+      "grad_norm": 1.7585566679889797,
+      "learning_rate": 7.835348320765546e-07,
+      "loss": 0.8506,
+      "step": 6372
+    },
+    {
+      "epoch": 0.8770988164051748,
+      "grad_norm": 1.902847572757343,
+      "learning_rate": 7.818061309754465e-07,
+      "loss": 0.8945,
+      "step": 6373
+    },
+    {
+      "epoch": 0.8772364437104322,
+      "grad_norm": 1.7229573270950396,
+      "learning_rate": 7.800792614032738e-07,
+      "loss": 0.9047,
+      "step": 6374
+    },
+    {
+      "epoch": 0.8773740710156895,
+      "grad_norm": 1.7390330253904451,
+      "learning_rate": 7.783542237031483e-07,
+      "loss": 0.8453,
+      "step": 6375
+    },
+    {
+      "epoch": 0.8775116983209469,
+      "grad_norm": 1.771730135659942,
+      "learning_rate": 7.766310182178038e-07,
+      "loss": 0.9133,
+      "step": 6376
+    },
+    {
+      "epoch": 0.8776493256262042,
+      "grad_norm": 1.8284534533947143,
+      "learning_rate": 7.749096452896243e-07,
+      "loss": 0.8712,
+      "step": 6377
+    },
+    {
+      "epoch": 0.8777869529314616,
+      "grad_norm": 1.6582122102078258,
+      "learning_rate": 7.731901052606206e-07,
+      "loss": 0.8298,
+      "step": 6378
+    },
+    {
+      "epoch": 0.8779245802367189,
+      "grad_norm": 1.7878471537398035,
+      "learning_rate": 7.714723984724415e-07,
+      "loss": 0.8937,
+      "step": 6379
+    },
+    {
+      "epoch": 0.8780622075419763,
+      "grad_norm": 1.6685627449282436,
+      "learning_rate": 7.697565252663719e-07,
+      "loss": 0.8694,
+      "step": 6380
+    },
+    {
+      "epoch": 0.8781998348472337,
+      "grad_norm": 1.807688598708972,
+      "learning_rate": 7.680424859833325e-07,
+      "loss": 1.0326,
+      "step": 6381
+    },
+    {
+      "epoch": 0.8783374621524911,
+      "grad_norm": 1.8627588549183183,
+      "learning_rate": 7.663302809638795e-07,
+      "loss": 0.9078,
+      "step": 6382
+    },
+    {
+      "epoch": 0.8784750894577484,
+      "grad_norm": 1.610877059058267,
+      "learning_rate": 7.646199105482043e-07,
+      "loss": 0.8238,
+      "step": 6383
+    },
+    {
+      "epoch": 0.8786127167630058,
+      "grad_norm": 1.85234750961446,
+      "learning_rate": 7.629113750761352e-07,
+      "loss": 0.9746,
+      "step": 6384
+    },
+    {
+      "epoch": 0.8787503440682631,
+      "grad_norm": 1.7961471682033971,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.9441,
+      "step": 6385
+    },
+    {
+      "epoch": 0.8788879713735205,
+      "grad_norm": 1.7439640481186351,
+      "learning_rate": 7.594998103202967e-07,
+      "loss": 0.9372,
+      "step": 6386
+    },
+    {
+      "epoch": 0.8790255986787778,
+      "grad_norm": 1.666755666368937,
+      "learning_rate": 7.577967817143595e-07,
+      "loss": 0.8755,
+      "step": 6387
+    },
+    {
+      "epoch": 0.8791632259840352,
+      "grad_norm": 1.8588723254955457,
+      "learning_rate": 7.560955894076894e-07,
+      "loss": 0.9137,
+      "step": 6388
+    },
+    {
+      "epoch": 0.8793008532892926,
+      "grad_norm": 1.6512079211921402,
+      "learning_rate": 7.543962337382915e-07,
+      "loss": 0.8697,
+      "step": 6389
+    },
+    {
+      "epoch": 0.87943848059455,
+      "grad_norm": 1.5663438714845923,
+      "learning_rate": 7.526987150438025e-07,
+      "loss": 0.9217,
+      "step": 6390
+    },
+    {
+      "epoch": 0.8795761078998073,
+      "grad_norm": 1.8061798966086877,
+      "learning_rate": 7.51003033661497e-07,
+      "loss": 0.8471,
+      "step": 6391
+    },
+    {
+      "epoch": 0.8797137352050647,
+      "grad_norm": 1.5708147991688186,
+      "learning_rate": 7.493091899282845e-07,
+      "loss": 0.9186,
+      "step": 6392
+    },
+    {
+      "epoch": 0.879851362510322,
+      "grad_norm": 1.6590333758088234,
+      "learning_rate": 7.476171841807078e-07,
+      "loss": 0.9595,
+      "step": 6393
+    },
+    {
+      "epoch": 0.8799889898155794,
+      "grad_norm": 1.694780532764805,
+      "learning_rate": 7.459270167549448e-07,
+      "loss": 0.9009,
+      "step": 6394
+    },
+    {
+      "epoch": 0.8801266171208367,
+      "grad_norm": 1.625017323761249,
+      "learning_rate": 7.442386879868069e-07,
+      "loss": 0.9063,
+      "step": 6395
+    },
+    {
+      "epoch": 0.8802642444260942,
+      "grad_norm": 1.9261840238073933,
+      "learning_rate": 7.425521982117489e-07,
+      "loss": 0.8558,
+      "step": 6396
+    },
+    {
+      "epoch": 0.8804018717313515,
+      "grad_norm": 2.299072595871546,
+      "learning_rate": 7.408675477648442e-07,
+      "loss": 0.9506,
+      "step": 6397
+    },
+    {
+      "epoch": 0.8805394990366089,
+      "grad_norm": 1.6093723065269614,
+      "learning_rate": 7.39184736980817e-07,
+      "loss": 0.9176,
+      "step": 6398
+    },
+    {
+      "epoch": 0.8806771263418662,
+      "grad_norm": 1.7452446315608874,
+      "learning_rate": 7.375037661940132e-07,
+      "loss": 0.911,
+      "step": 6399
+    },
+    {
+      "epoch": 0.8808147536471236,
+      "grad_norm": 1.611112393288828,
+      "learning_rate": 7.358246357384247e-07,
+      "loss": 0.9607,
+      "step": 6400
+    },
+    {
+      "epoch": 0.8809523809523809,
+      "grad_norm": 1.644251578042853,
+      "learning_rate": 7.341473459476645e-07,
+      "loss": 0.9208,
+      "step": 6401
+    },
+    {
+      "epoch": 0.8810900082576383,
+      "grad_norm": 1.8027601113899416,
+      "learning_rate": 7.324718971549926e-07,
+      "loss": 0.887,
+      "step": 6402
+    },
+    {
+      "epoch": 0.8812276355628956,
+      "grad_norm": 1.6692914484855061,
+      "learning_rate": 7.307982896932986e-07,
+      "loss": 0.8873,
+      "step": 6403
+    },
+    {
+      "epoch": 0.8813652628681531,
+      "grad_norm": 1.770575001284231,
+      "learning_rate": 7.291265238950984e-07,
+      "loss": 0.9359,
+      "step": 6404
+    },
+    {
+      "epoch": 0.8815028901734104,
+      "grad_norm": 1.685006217012473,
+      "learning_rate": 7.274566000925575e-07,
+      "loss": 0.9147,
+      "step": 6405
+    },
+    {
+      "epoch": 0.8816405174786678,
+      "grad_norm": 1.8131165110999612,
+      "learning_rate": 7.257885186174596e-07,
+      "loss": 0.887,
+      "step": 6406
+    },
+    {
+      "epoch": 0.8817781447839251,
+      "grad_norm": 2.2749024770192783,
+      "learning_rate": 7.241222798012359e-07,
+      "loss": 0.8377,
+      "step": 6407
+    },
+    {
+      "epoch": 0.8819157720891825,
+      "grad_norm": 2.2328916164174273,
+      "learning_rate": 7.224578839749386e-07,
+      "loss": 0.9509,
+      "step": 6408
+    },
+    {
+      "epoch": 0.8820533993944398,
+      "grad_norm": 1.769521910442314,
+      "learning_rate": 7.207953314692673e-07,
+      "loss": 0.866,
+      "step": 6409
+    },
+    {
+      "epoch": 0.8821910266996972,
+      "grad_norm": 1.8185570386594185,
+      "learning_rate": 7.191346226145435e-07,
+      "loss": 0.8848,
+      "step": 6410
+    },
+    {
+      "epoch": 0.8823286540049546,
+      "grad_norm": 1.705985114440559,
+      "learning_rate": 7.174757577407298e-07,
+      "loss": 0.9407,
+      "step": 6411
+    },
+    {
+      "epoch": 0.882466281310212,
+      "grad_norm": 1.5383726606934731,
+      "learning_rate": 7.158187371774183e-07,
+      "loss": 0.8851,
+      "step": 6412
+    },
+    {
+      "epoch": 0.8826039086154693,
+      "grad_norm": 1.5539039197212894,
+      "learning_rate": 7.141635612538378e-07,
+      "loss": 0.8647,
+      "step": 6413
+    },
+    {
+      "epoch": 0.8827415359207267,
+      "grad_norm": 1.5523027222060135,
+      "learning_rate": 7.125102302988485e-07,
+      "loss": 0.9209,
+      "step": 6414
+    },
+    {
+      "epoch": 0.882879163225984,
+      "grad_norm": 2.015737098231876,
+      "learning_rate": 7.108587446409443e-07,
+      "loss": 0.9087,
+      "step": 6415
+    },
+    {
+      "epoch": 0.8830167905312414,
+      "grad_norm": 1.8150814702452889,
+      "learning_rate": 7.092091046082538e-07,
+      "loss": 0.8421,
+      "step": 6416
+    },
+    {
+      "epoch": 0.8831544178364987,
+      "grad_norm": 1.848789992413101,
+      "learning_rate": 7.075613105285361e-07,
+      "loss": 0.9596,
+      "step": 6417
+    },
+    {
+      "epoch": 0.8832920451417561,
+      "grad_norm": 1.9201010090851653,
+      "learning_rate": 7.059153627291871e-07,
+      "loss": 0.883,
+      "step": 6418
+    },
+    {
+      "epoch": 0.8834296724470135,
+      "grad_norm": 2.005900188495301,
+      "learning_rate": 7.042712615372327e-07,
+      "loss": 0.9078,
+      "step": 6419
+    },
+    {
+      "epoch": 0.8835672997522709,
+      "grad_norm": 2.048393098564577,
+      "learning_rate": 7.02629007279334e-07,
+      "loss": 0.9436,
+      "step": 6420
+    },
+    {
+      "epoch": 0.8837049270575282,
+      "grad_norm": 1.9853746294734698,
+      "learning_rate": 7.009886002817856e-07,
+      "loss": 0.8772,
+      "step": 6421
+    },
+    {
+      "epoch": 0.8838425543627856,
+      "grad_norm": 1.7725021607751894,
+      "learning_rate": 6.993500408705112e-07,
+      "loss": 0.9129,
+      "step": 6422
+    },
+    {
+      "epoch": 0.8839801816680429,
+      "grad_norm": 1.620304211930674,
+      "learning_rate": 6.977133293710725e-07,
+      "loss": 0.8819,
+      "step": 6423
+    },
+    {
+      "epoch": 0.8841178089733003,
+      "grad_norm": 1.9642300616043322,
+      "learning_rate": 6.960784661086606e-07,
+      "loss": 0.8956,
+      "step": 6424
+    },
+    {
+      "epoch": 0.8842554362785576,
+      "grad_norm": 1.6405778621070912,
+      "learning_rate": 6.94445451408099e-07,
+      "loss": 0.8629,
+      "step": 6425
+    },
+    {
+      "epoch": 0.884393063583815,
+      "grad_norm": 1.6084542656993244,
+      "learning_rate": 6.928142855938513e-07,
+      "loss": 0.8574,
+      "step": 6426
+    },
+    {
+      "epoch": 0.8845306908890724,
+      "grad_norm": 1.6392540744553863,
+      "learning_rate": 6.911849689899985e-07,
+      "loss": 0.8935,
+      "step": 6427
+    },
+    {
+      "epoch": 0.8846683181943298,
+      "grad_norm": 1.8897373748705015,
+      "learning_rate": 6.895575019202727e-07,
+      "loss": 0.8816,
+      "step": 6428
+    },
+    {
+      "epoch": 0.8848059454995871,
+      "grad_norm": 1.940786978969368,
+      "learning_rate": 6.879318847080229e-07,
+      "loss": 0.935,
+      "step": 6429
+    },
+    {
+      "epoch": 0.8849435728048445,
+      "grad_norm": 1.8321974051191918,
+      "learning_rate": 6.863081176762409e-07,
+      "loss": 0.8916,
+      "step": 6430
+    },
+    {
+      "epoch": 0.8850812001101018,
+      "grad_norm": 2.0176252624408,
+      "learning_rate": 6.84686201147543e-07,
+      "loss": 0.8429,
+      "step": 6431
+    },
+    {
+      "epoch": 0.8852188274153592,
+      "grad_norm": 1.7645600873921015,
+      "learning_rate": 6.830661354441858e-07,
+      "loss": 0.897,
+      "step": 6432
+    },
+    {
+      "epoch": 0.8853564547206165,
+      "grad_norm": 1.6067030412374046,
+      "learning_rate": 6.81447920888052e-07,
+      "loss": 0.9046,
+      "step": 6433
+    },
+    {
+      "epoch": 0.885494082025874,
+      "grad_norm": 1.641955020927479,
+      "learning_rate": 6.798315578006597e-07,
+      "loss": 0.8355,
+      "step": 6434
+    },
+    {
+      "epoch": 0.8856317093311313,
+      "grad_norm": 1.9703812498182551,
+      "learning_rate": 6.782170465031601e-07,
+      "loss": 1.0185,
+      "step": 6435
+    },
+    {
+      "epoch": 0.8857693366363887,
+      "grad_norm": 1.8242606371538628,
+      "learning_rate": 6.766043873163275e-07,
+      "loss": 0.8507,
+      "step": 6436
+    },
+    {
+      "epoch": 0.885906963941646,
+      "grad_norm": 2.01352940475627,
+      "learning_rate": 6.749935805605833e-07,
+      "loss": 0.8617,
+      "step": 6437
+    },
+    {
+      "epoch": 0.8860445912469034,
+      "grad_norm": 2.049467317992396,
+      "learning_rate": 6.733846265559663e-07,
+      "loss": 0.8898,
+      "step": 6438
+    },
+    {
+      "epoch": 0.8861822185521607,
+      "grad_norm": 1.58098666551709,
+      "learning_rate": 6.717775256221582e-07,
+      "loss": 0.8349,
+      "step": 6439
+    },
+    {
+      "epoch": 0.8863198458574181,
+      "grad_norm": 1.8406874726848967,
+      "learning_rate": 6.701722780784658e-07,
+      "loss": 0.9221,
+      "step": 6440
+    },
+    {
+      "epoch": 0.8864574731626754,
+      "grad_norm": 2.3589597139645315,
+      "learning_rate": 6.68568884243831e-07,
+      "loss": 0.8771,
+      "step": 6441
+    },
+    {
+      "epoch": 0.8865951004679329,
+      "grad_norm": 1.5908031430634777,
+      "learning_rate": 6.669673444368252e-07,
+      "loss": 0.8677,
+      "step": 6442
+    },
+    {
+      "epoch": 0.8867327277731902,
+      "grad_norm": 2.0651342130525876,
+      "learning_rate": 6.653676589756519e-07,
+      "loss": 0.9312,
+      "step": 6443
+    },
+    {
+      "epoch": 0.8868703550784476,
+      "grad_norm": 1.6486704535154544,
+      "learning_rate": 6.637698281781491e-07,
+      "loss": 0.9149,
+      "step": 6444
+    },
+    {
+      "epoch": 0.8870079823837049,
+      "grad_norm": 2.449555071958951,
+      "learning_rate": 6.621738523617815e-07,
+      "loss": 0.9483,
+      "step": 6445
+    },
+    {
+      "epoch": 0.8871456096889623,
+      "grad_norm": 1.7431737535788985,
+      "learning_rate": 6.605797318436491e-07,
+      "loss": 0.9202,
+      "step": 6446
+    },
+    {
+      "epoch": 0.8872832369942196,
+      "grad_norm": 2.734494514258936,
+      "learning_rate": 6.589874669404805e-07,
+      "loss": 0.8957,
+      "step": 6447
+    },
+    {
+      "epoch": 0.887420864299477,
+      "grad_norm": 1.8748408386995095,
+      "learning_rate": 6.573970579686384e-07,
+      "loss": 0.9096,
+      "step": 6448
+    },
+    {
+      "epoch": 0.8875584916047344,
+      "grad_norm": 1.6687726414456006,
+      "learning_rate": 6.558085052441132e-07,
+      "loss": 0.8552,
+      "step": 6449
+    },
+    {
+      "epoch": 0.8876961189099918,
+      "grad_norm": 1.7932994132205653,
+      "learning_rate": 6.542218090825314e-07,
+      "loss": 0.8306,
+      "step": 6450
+    },
+    {
+      "epoch": 0.8878337462152491,
+      "grad_norm": 1.7570804728183498,
+      "learning_rate": 6.526369697991463e-07,
+      "loss": 0.8671,
+      "step": 6451
+    },
+    {
+      "epoch": 0.8879713735205065,
+      "grad_norm": 1.7647289643909725,
+      "learning_rate": 6.510539877088429e-07,
+      "loss": 0.9161,
+      "step": 6452
+    },
+    {
+      "epoch": 0.8881090008257638,
+      "grad_norm": 1.936996257939106,
+      "learning_rate": 6.494728631261405e-07,
+      "loss": 0.9673,
+      "step": 6453
+    },
+    {
+      "epoch": 0.8882466281310212,
+      "grad_norm": 1.8594045279895188,
+      "learning_rate": 6.478935963651844e-07,
+      "loss": 0.8677,
+      "step": 6454
+    },
+    {
+      "epoch": 0.8883842554362785,
+      "grad_norm": 1.8227155626760876,
+      "learning_rate": 6.463161877397539e-07,
+      "loss": 0.9431,
+      "step": 6455
+    },
+    {
+      "epoch": 0.8885218827415359,
+      "grad_norm": 1.877559899528615,
+      "learning_rate": 6.447406375632592e-07,
+      "loss": 0.8641,
+      "step": 6456
+    },
+    {
+      "epoch": 0.8886595100467933,
+      "grad_norm": 1.9494281376945986,
+      "learning_rate": 6.431669461487389e-07,
+      "loss": 0.939,
+      "step": 6457
+    },
+    {
+      "epoch": 0.8887971373520507,
+      "grad_norm": 1.7092941486354354,
+      "learning_rate": 6.415951138088694e-07,
+      "loss": 0.94,
+      "step": 6458
+    },
+    {
+      "epoch": 0.888934764657308,
+      "grad_norm": 1.8321943127031395,
+      "learning_rate": 6.400251408559443e-07,
+      "loss": 0.8427,
+      "step": 6459
+    },
+    {
+      "epoch": 0.8890723919625654,
+      "grad_norm": 1.9229410203008541,
+      "learning_rate": 6.38457027601902e-07,
+      "loss": 0.895,
+      "step": 6460
+    },
+    {
+      "epoch": 0.8892100192678227,
+      "grad_norm": 1.445636448719149,
+      "learning_rate": 6.368907743583008e-07,
+      "loss": 0.8834,
+      "step": 6461
+    },
+    {
+      "epoch": 0.8893476465730801,
+      "grad_norm": 1.663931414796164,
+      "learning_rate": 6.353263814363375e-07,
+      "loss": 0.8976,
+      "step": 6462
+    },
+    {
+      "epoch": 0.8894852738783374,
+      "grad_norm": 4.914740749767287,
+      "learning_rate": 6.337638491468323e-07,
+      "loss": 0.8938,
+      "step": 6463
+    },
+    {
+      "epoch": 0.8896229011835948,
+      "grad_norm": 1.6062424724252355,
+      "learning_rate": 6.322031778002413e-07,
+      "loss": 0.8563,
+      "step": 6464
+    },
+    {
+      "epoch": 0.8897605284888522,
+      "grad_norm": 1.8240505572592254,
+      "learning_rate": 6.306443677066476e-07,
+      "loss": 0.8081,
+      "step": 6465
+    },
+    {
+      "epoch": 0.8898981557941096,
+      "grad_norm": 1.6403812662549373,
+      "learning_rate": 6.290874191757656e-07,
+      "loss": 0.9013,
+      "step": 6466
+    },
+    {
+      "epoch": 0.8900357830993669,
+      "grad_norm": 2.1965582163033024,
+      "learning_rate": 6.275323325169403e-07,
+      "loss": 0.9684,
+      "step": 6467
+    },
+    {
+      "epoch": 0.8901734104046243,
+      "grad_norm": 1.636816816862865,
+      "learning_rate": 6.259791080391431e-07,
+      "loss": 0.9072,
+      "step": 6468
+    },
+    {
+      "epoch": 0.8903110377098816,
+      "grad_norm": 1.9809936772996666,
+      "learning_rate": 6.24427746050984e-07,
+      "loss": 0.9171,
+      "step": 6469
+    },
+    {
+      "epoch": 0.890448665015139,
+      "grad_norm": 1.861002482375235,
+      "learning_rate": 6.228782468606909e-07,
+      "loss": 0.9036,
+      "step": 6470
+    },
+    {
+      "epoch": 0.8905862923203963,
+      "grad_norm": 1.5489148953570382,
+      "learning_rate": 6.21330610776133e-07,
+      "loss": 0.8569,
+      "step": 6471
+    },
+    {
+      "epoch": 0.8907239196256538,
+      "grad_norm": 1.9892753746223564,
+      "learning_rate": 6.197848381048033e-07,
+      "loss": 0.8269,
+      "step": 6472
+    },
+    {
+      "epoch": 0.8908615469309111,
+      "grad_norm": 1.8385620898065118,
+      "learning_rate": 6.182409291538238e-07,
+      "loss": 0.8437,
+      "step": 6473
+    },
+    {
+      "epoch": 0.8909991742361685,
+      "grad_norm": 1.755218217609776,
+      "learning_rate": 6.166988842299504e-07,
+      "loss": 0.9044,
+      "step": 6474
+    },
+    {
+      "epoch": 0.8911368015414258,
+      "grad_norm": 1.7390119752546944,
+      "learning_rate": 6.151587036395656e-07,
+      "loss": 0.94,
+      "step": 6475
+    },
+    {
+      "epoch": 0.8912744288466832,
+      "grad_norm": 1.8919811094019443,
+      "learning_rate": 6.136203876886803e-07,
+      "loss": 0.8448,
+      "step": 6476
+    },
+    {
+      "epoch": 0.8914120561519405,
+      "grad_norm": 1.62869687066339,
+      "learning_rate": 6.120839366829401e-07,
+      "loss": 0.8602,
+      "step": 6477
+    },
+    {
+      "epoch": 0.8915496834571979,
+      "grad_norm": 1.7149783406023593,
+      "learning_rate": 6.105493509276139e-07,
+      "loss": 0.9188,
+      "step": 6478
+    },
+    {
+      "epoch": 0.8916873107624552,
+      "grad_norm": 2.2772423936115986,
+      "learning_rate": 6.090166307276047e-07,
+      "loss": 0.8516,
+      "step": 6479
+    },
+    {
+      "epoch": 0.8918249380677127,
+      "grad_norm": 2.082448804519124,
+      "learning_rate": 6.074857763874431e-07,
+      "loss": 0.838,
+      "step": 6480
+    },
+    {
+      "epoch": 0.89196256537297,
+      "grad_norm": 1.7899594499955493,
+      "learning_rate": 6.059567882112871e-07,
+      "loss": 0.8448,
+      "step": 6481
+    },
+    {
+      "epoch": 0.8921001926782274,
+      "grad_norm": 1.8206715252367143,
+      "learning_rate": 6.04429666502927e-07,
+      "loss": 0.8179,
+      "step": 6482
+    },
+    {
+      "epoch": 0.8922378199834847,
+      "grad_norm": 1.8154475150126275,
+      "learning_rate": 6.029044115657812e-07,
+      "loss": 0.9081,
+      "step": 6483
+    },
+    {
+      "epoch": 0.8923754472887421,
+      "grad_norm": 2.051830224255475,
+      "learning_rate": 6.013810237028972e-07,
+      "loss": 0.8667,
+      "step": 6484
+    },
+    {
+      "epoch": 0.8925130745939994,
+      "grad_norm": 1.9637870769196732,
+      "learning_rate": 5.998595032169496e-07,
+      "loss": 0.861,
+      "step": 6485
+    },
+    {
+      "epoch": 0.8926507018992568,
+      "grad_norm": 1.816180860691923,
+      "learning_rate": 5.983398504102456e-07,
+      "loss": 0.7809,
+      "step": 6486
+    },
+    {
+      "epoch": 0.8927883292045142,
+      "grad_norm": 1.530421870800164,
+      "learning_rate": 5.968220655847167e-07,
+      "loss": 0.8821,
+      "step": 6487
+    },
+    {
+      "epoch": 0.8929259565097716,
+      "grad_norm": 1.7413773067831742,
+      "learning_rate": 5.95306149041931e-07,
+      "loss": 0.8828,
+      "step": 6488
+    },
+    {
+      "epoch": 0.8930635838150289,
+      "grad_norm": 1.7877887698130557,
+      "learning_rate": 5.937921010830738e-07,
+      "loss": 0.8731,
+      "step": 6489
+    },
+    {
+      "epoch": 0.8932012111202863,
+      "grad_norm": 1.9499295015221687,
+      "learning_rate": 5.922799220089726e-07,
+      "loss": 0.9063,
+      "step": 6490
+    },
+    {
+      "epoch": 0.8933388384255436,
+      "grad_norm": 1.8620475071039337,
+      "learning_rate": 5.907696121200701e-07,
+      "loss": 0.8926,
+      "step": 6491
+    },
+    {
+      "epoch": 0.893476465730801,
+      "grad_norm": 1.771179601397548,
+      "learning_rate": 5.892611717164498e-07,
+      "loss": 0.9292,
+      "step": 6492
+    },
+    {
+      "epoch": 0.8936140930360583,
+      "grad_norm": 1.4933871262496432,
+      "learning_rate": 5.877546010978142e-07,
+      "loss": 0.9897,
+      "step": 6493
+    },
+    {
+      "epoch": 0.8937517203413157,
+      "grad_norm": 1.8174475111573583,
+      "learning_rate": 5.862499005635003e-07,
+      "loss": 0.9226,
+      "step": 6494
+    },
+    {
+      "epoch": 0.8938893476465731,
+      "grad_norm": 1.8019261540055285,
+      "learning_rate": 5.847470704124703e-07,
+      "loss": 0.9148,
+      "step": 6495
+    },
+    {
+      "epoch": 0.8940269749518305,
+      "grad_norm": 1.8400245643491624,
+      "learning_rate": 5.832461109433174e-07,
+      "loss": 0.9034,
+      "step": 6496
+    },
+    {
+      "epoch": 0.8941646022570878,
+      "grad_norm": 1.7588803585853499,
+      "learning_rate": 5.81747022454261e-07,
+      "loss": 0.9313,
+      "step": 6497
+    },
+    {
+      "epoch": 0.8943022295623452,
+      "grad_norm": 1.8443099176665407,
+      "learning_rate": 5.802498052431493e-07,
+      "loss": 0.8383,
+      "step": 6498
+    },
+    {
+      "epoch": 0.8944398568676025,
+      "grad_norm": 1.751664613122743,
+      "learning_rate": 5.787544596074613e-07,
+      "loss": 0.9357,
+      "step": 6499
+    },
+    {
+      "epoch": 0.8945774841728599,
+      "grad_norm": 1.8561395911336072,
+      "learning_rate": 5.772609858442946e-07,
+      "loss": 0.8961,
+      "step": 6500
+    },
+    {
+      "epoch": 0.8947151114781172,
+      "grad_norm": 1.5885039219810266,
+      "learning_rate": 5.757693842503898e-07,
+      "loss": 0.9413,
+      "step": 6501
+    },
+    {
+      "epoch": 0.8948527387833746,
+      "grad_norm": 1.7939590814932405,
+      "learning_rate": 5.742796551221042e-07,
+      "loss": 0.8626,
+      "step": 6502
+    },
+    {
+      "epoch": 0.894990366088632,
+      "grad_norm": 1.9411485752127589,
+      "learning_rate": 5.727917987554266e-07,
+      "loss": 0.9683,
+      "step": 6503
+    },
+    {
+      "epoch": 0.8951279933938894,
+      "grad_norm": 1.8222550125872903,
+      "learning_rate": 5.71305815445975e-07,
+      "loss": 0.8859,
+      "step": 6504
+    },
+    {
+      "epoch": 0.8952656206991467,
+      "grad_norm": 1.54451157826374,
+      "learning_rate": 5.698217054889921e-07,
+      "loss": 0.8742,
+      "step": 6505
+    },
+    {
+      "epoch": 0.8954032480044041,
+      "grad_norm": 1.8822293308043638,
+      "learning_rate": 5.68339469179352e-07,
+      "loss": 0.9062,
+      "step": 6506
+    },
+    {
+      "epoch": 0.8955408753096614,
+      "grad_norm": 1.79554558514194,
+      "learning_rate": 5.668591068115536e-07,
+      "loss": 0.9112,
+      "step": 6507
+    },
+    {
+      "epoch": 0.8956785026149188,
+      "grad_norm": 1.948484488289092,
+      "learning_rate": 5.653806186797262e-07,
+      "loss": 0.9484,
+      "step": 6508
+    },
+    {
+      "epoch": 0.8958161299201761,
+      "grad_norm": 1.8407065427371514,
+      "learning_rate": 5.639040050776223e-07,
+      "loss": 0.9034,
+      "step": 6509
+    },
+    {
+      "epoch": 0.8959537572254336,
+      "grad_norm": 1.7359488427788166,
+      "learning_rate": 5.624292662986275e-07,
+      "loss": 1.0032,
+      "step": 6510
+    },
+    {
+      "epoch": 0.8960913845306909,
+      "grad_norm": 1.6428633158115507,
+      "learning_rate": 5.609564026357517e-07,
+      "loss": 0.9344,
+      "step": 6511
+    },
+    {
+      "epoch": 0.8962290118359483,
+      "grad_norm": 1.6470189540858218,
+      "learning_rate": 5.594854143816298e-07,
+      "loss": 0.8314,
+      "step": 6512
+    },
+    {
+      "epoch": 0.8963666391412056,
+      "grad_norm": 1.8032430998251854,
+      "learning_rate": 5.580163018285323e-07,
+      "loss": 0.8015,
+      "step": 6513
+    },
+    {
+      "epoch": 0.896504266446463,
+      "grad_norm": 1.6037953052878011,
+      "learning_rate": 5.565490652683481e-07,
+      "loss": 0.8968,
+      "step": 6514
+    },
+    {
+      "epoch": 0.8966418937517203,
+      "grad_norm": 1.8551423989590532,
+      "learning_rate": 5.550837049925984e-07,
+      "loss": 0.9818,
+      "step": 6515
+    },
+    {
+      "epoch": 0.8967795210569777,
+      "grad_norm": 1.8045038102060473,
+      "learning_rate": 5.536202212924291e-07,
+      "loss": 0.9151,
+      "step": 6516
+    },
+    {
+      "epoch": 0.896917148362235,
+      "grad_norm": 1.8950969321366096,
+      "learning_rate": 5.521586144586144e-07,
+      "loss": 0.8661,
+      "step": 6517
+    },
+    {
+      "epoch": 0.8970547756674925,
+      "grad_norm": 1.7504246122810019,
+      "learning_rate": 5.506988847815564e-07,
+      "loss": 0.8152,
+      "step": 6518
+    },
+    {
+      "epoch": 0.8971924029727498,
+      "grad_norm": 1.6141600866720605,
+      "learning_rate": 5.49241032551282e-07,
+      "loss": 0.8249,
+      "step": 6519
+    },
+    {
+      "epoch": 0.8973300302780072,
+      "grad_norm": 1.7202165545868682,
+      "learning_rate": 5.477850580574506e-07,
+      "loss": 0.903,
+      "step": 6520
+    },
+    {
+      "epoch": 0.8974676575832645,
+      "grad_norm": 1.6087666737187465,
+      "learning_rate": 5.463309615893387e-07,
+      "loss": 0.9334,
+      "step": 6521
+    },
+    {
+      "epoch": 0.8976052848885219,
+      "grad_norm": 1.7519328044161546,
+      "learning_rate": 5.448787434358604e-07,
+      "loss": 0.8884,
+      "step": 6522
+    },
+    {
+      "epoch": 0.8977429121937792,
+      "grad_norm": 1.6882940409760934,
+      "learning_rate": 5.434284038855475e-07,
+      "loss": 0.864,
+      "step": 6523
+    },
+    {
+      "epoch": 0.8978805394990366,
+      "grad_norm": 1.7296911630054215,
+      "learning_rate": 5.41979943226566e-07,
+      "loss": 0.8739,
+      "step": 6524
+    },
+    {
+      "epoch": 0.898018166804294,
+      "grad_norm": 1.8246021185485055,
+      "learning_rate": 5.405333617467012e-07,
+      "loss": 0.9371,
+      "step": 6525
+    },
+    {
+      "epoch": 0.8981557941095514,
+      "grad_norm": 1.6346340198394016,
+      "learning_rate": 5.390886597333733e-07,
+      "loss": 0.9056,
+      "step": 6526
+    },
+    {
+      "epoch": 0.8982934214148087,
+      "grad_norm": 1.771731218858257,
+      "learning_rate": 5.376458374736227e-07,
+      "loss": 0.8966,
+      "step": 6527
+    },
+    {
+      "epoch": 0.8984310487200661,
+      "grad_norm": 2.039338060693509,
+      "learning_rate": 5.362048952541176e-07,
+      "loss": 0.9419,
+      "step": 6528
+    },
+    {
+      "epoch": 0.8985686760253234,
+      "grad_norm": 1.649239631302988,
+      "learning_rate": 5.347658333611561e-07,
+      "loss": 0.8356,
+      "step": 6529
+    },
+    {
+      "epoch": 0.8987063033305808,
+      "grad_norm": 2.2454122330319044,
+      "learning_rate": 5.33328652080658e-07,
+      "loss": 0.8263,
+      "step": 6530
+    },
+    {
+      "epoch": 0.8988439306358381,
+      "grad_norm": 1.690810632231821,
+      "learning_rate": 5.318933516981728e-07,
+      "loss": 0.877,
+      "step": 6531
+    },
+    {
+      "epoch": 0.8989815579410955,
+      "grad_norm": 1.6766251105945962,
+      "learning_rate": 5.304599324988724e-07,
+      "loss": 0.893,
+      "step": 6532
+    },
+    {
+      "epoch": 0.8991191852463529,
+      "grad_norm": 1.7273685942227646,
+      "learning_rate": 5.290283947675612e-07,
+      "loss": 0.7914,
+      "step": 6533
+    },
+    {
+      "epoch": 0.8992568125516103,
+      "grad_norm": 1.8820816369764046,
+      "learning_rate": 5.275987387886638e-07,
+      "loss": 0.8305,
+      "step": 6534
+    },
+    {
+      "epoch": 0.8993944398568676,
+      "grad_norm": 1.6419479409456463,
+      "learning_rate": 5.261709648462344e-07,
+      "loss": 0.8891,
+      "step": 6535
+    },
+    {
+      "epoch": 0.899532067162125,
+      "grad_norm": 1.6177035326997828,
+      "learning_rate": 5.247450732239512e-07,
+      "loss": 0.9873,
+      "step": 6536
+    },
+    {
+      "epoch": 0.8996696944673823,
+      "grad_norm": 2.0802440436101244,
+      "learning_rate": 5.2332106420512e-07,
+      "loss": 0.8938,
+      "step": 6537
+    },
+    {
+      "epoch": 0.8998073217726397,
+      "grad_norm": 1.9184231674953178,
+      "learning_rate": 5.218989380726725e-07,
+      "loss": 0.9194,
+      "step": 6538
+    },
+    {
+      "epoch": 0.899944949077897,
+      "grad_norm": 1.741997863632587,
+      "learning_rate": 5.204786951091645e-07,
+      "loss": 0.9334,
+      "step": 6539
+    },
+    {
+      "epoch": 0.9000825763831544,
+      "grad_norm": 1.6562565353807313,
+      "learning_rate": 5.190603355967793e-07,
+      "loss": 0.938,
+      "step": 6540
+    },
+    {
+      "epoch": 0.9002202036884118,
+      "grad_norm": 1.6806233156899084,
+      "learning_rate": 5.176438598173251e-07,
+      "loss": 0.9268,
+      "step": 6541
+    },
+    {
+      "epoch": 0.9003578309936692,
+      "grad_norm": 1.6122806132307825,
+      "learning_rate": 5.162292680522363e-07,
+      "loss": 0.8818,
+      "step": 6542
+    },
+    {
+      "epoch": 0.9004954582989265,
+      "grad_norm": 1.7468473576364618,
+      "learning_rate": 5.148165605825739e-07,
+      "loss": 0.9561,
+      "step": 6543
+    },
+    {
+      "epoch": 0.9006330856041839,
+      "grad_norm": 1.8286724928426037,
+      "learning_rate": 5.134057376890222e-07,
+      "loss": 0.8924,
+      "step": 6544
+    },
+    {
+      "epoch": 0.9007707129094412,
+      "grad_norm": 1.54532617431928,
+      "learning_rate": 5.119967996518948e-07,
+      "loss": 0.9292,
+      "step": 6545
+    },
+    {
+      "epoch": 0.9009083402146986,
+      "grad_norm": 1.6685094442604222,
+      "learning_rate": 5.105897467511256e-07,
+      "loss": 0.9587,
+      "step": 6546
+    },
+    {
+      "epoch": 0.9010459675199559,
+      "grad_norm": 1.7635740070767605,
+      "learning_rate": 5.091845792662775e-07,
+      "loss": 0.8965,
+      "step": 6547
+    },
+    {
+      "epoch": 0.9011835948252134,
+      "grad_norm": 1.6007294737907005,
+      "learning_rate": 5.077812974765395e-07,
+      "loss": 0.8552,
+      "step": 6548
+    },
+    {
+      "epoch": 0.9013212221304707,
+      "grad_norm": 1.533965175058837,
+      "learning_rate": 5.063799016607218e-07,
+      "loss": 0.879,
+      "step": 6549
+    },
+    {
+      "epoch": 0.9014588494357281,
+      "grad_norm": 1.8042562024634408,
+      "learning_rate": 5.049803920972652e-07,
+      "loss": 0.8919,
+      "step": 6550
+    },
+    {
+      "epoch": 0.9015964767409854,
+      "grad_norm": 1.8070898356916127,
+      "learning_rate": 5.035827690642303e-07,
+      "loss": 0.9123,
+      "step": 6551
+    },
+    {
+      "epoch": 0.9017341040462428,
+      "grad_norm": 1.8737460169104463,
+      "learning_rate": 5.021870328393108e-07,
+      "loss": 0.8766,
+      "step": 6552
+    },
+    {
+      "epoch": 0.9018717313515001,
+      "grad_norm": 1.6068687612614978,
+      "learning_rate": 5.007931836998136e-07,
+      "loss": 0.9058,
+      "step": 6553
+    },
+    {
+      "epoch": 0.9020093586567575,
+      "grad_norm": 1.5647655450688793,
+      "learning_rate": 4.99401221922684e-07,
+      "loss": 0.9407,
+      "step": 6554
+    },
+    {
+      "epoch": 0.9021469859620148,
+      "grad_norm": 1.7292845833284105,
+      "learning_rate": 4.980111477844796e-07,
+      "loss": 0.9001,
+      "step": 6555
+    },
+    {
+      "epoch": 0.9022846132672723,
+      "grad_norm": 1.7460313749169862,
+      "learning_rate": 4.966229615613927e-07,
+      "loss": 0.9419,
+      "step": 6556
+    },
+    {
+      "epoch": 0.9024222405725296,
+      "grad_norm": 1.7949317549858668,
+      "learning_rate": 4.952366635292371e-07,
+      "loss": 0.9027,
+      "step": 6557
+    },
+    {
+      "epoch": 0.902559867877787,
+      "grad_norm": 1.753305123783024,
+      "learning_rate": 4.938522539634505e-07,
+      "loss": 0.8999,
+      "step": 6558
+    },
+    {
+      "epoch": 0.9026974951830443,
+      "grad_norm": 2.290788703166616,
+      "learning_rate": 4.92469733139096e-07,
+      "loss": 0.8811,
+      "step": 6559
+    },
+    {
+      "epoch": 0.9028351224883017,
+      "grad_norm": 1.8858108772066675,
+      "learning_rate": 4.910891013308617e-07,
+      "loss": 0.8626,
+      "step": 6560
+    },
+    {
+      "epoch": 0.902972749793559,
+      "grad_norm": 2.0487260009049444,
+      "learning_rate": 4.897103588130591e-07,
+      "loss": 0.8643,
+      "step": 6561
+    },
+    {
+      "epoch": 0.9031103770988164,
+      "grad_norm": 1.8731328089032253,
+      "learning_rate": 4.883335058596284e-07,
+      "loss": 0.9016,
+      "step": 6562
+    },
+    {
+      "epoch": 0.9032480044040738,
+      "grad_norm": 1.7335271641120151,
+      "learning_rate": 4.869585427441281e-07,
+      "loss": 0.9036,
+      "step": 6563
+    },
+    {
+      "epoch": 0.9033856317093312,
+      "grad_norm": 1.8864747923351008,
+      "learning_rate": 4.855854697397467e-07,
+      "loss": 0.8174,
+      "step": 6564
+    },
+    {
+      "epoch": 0.9035232590145885,
+      "grad_norm": 1.8963899210084483,
+      "learning_rate": 4.842142871192934e-07,
+      "loss": 0.8826,
+      "step": 6565
+    },
+    {
+      "epoch": 0.9036608863198459,
+      "grad_norm": 1.646440538421722,
+      "learning_rate": 4.82844995155205e-07,
+      "loss": 0.9325,
+      "step": 6566
+    },
+    {
+      "epoch": 0.9037985136251032,
+      "grad_norm": 1.977717039678017,
+      "learning_rate": 4.814775941195404e-07,
+      "loss": 0.8555,
+      "step": 6567
+    },
+    {
+      "epoch": 0.9039361409303606,
+      "grad_norm": 1.9414915727587263,
+      "learning_rate": 4.801120842839834e-07,
+      "loss": 0.8718,
+      "step": 6568
+    },
+    {
+      "epoch": 0.9040737682356179,
+      "grad_norm": 1.882564153265852,
+      "learning_rate": 4.787484659198427e-07,
+      "loss": 0.9298,
+      "step": 6569
+    },
+    {
+      "epoch": 0.9042113955408753,
+      "grad_norm": 1.8674386775015053,
+      "learning_rate": 4.773867392980491e-07,
+      "loss": 0.8205,
+      "step": 6570
+    },
+    {
+      "epoch": 0.9043490228461327,
+      "grad_norm": 1.6384417469931487,
+      "learning_rate": 4.760269046891608e-07,
+      "loss": 0.8777,
+      "step": 6571
+    },
+    {
+      "epoch": 0.9044866501513901,
+      "grad_norm": 1.8653101773206573,
+      "learning_rate": 4.7466896236335823e-07,
+      "loss": 0.9435,
+      "step": 6572
+    },
+    {
+      "epoch": 0.9046242774566474,
+      "grad_norm": 1.6540119127835107,
+      "learning_rate": 4.7331291259044345e-07,
+      "loss": 0.8822,
+      "step": 6573
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 1.6567127234929107,
+      "learning_rate": 4.719587556398464e-07,
+      "loss": 0.8459,
+      "step": 6574
+    },
+    {
+      "epoch": 0.9048995320671621,
+      "grad_norm": 1.724847266600291,
+      "learning_rate": 4.70606491780623e-07,
+      "loss": 0.8452,
+      "step": 6575
+    },
+    {
+      "epoch": 0.9050371593724195,
+      "grad_norm": 1.9040297228230547,
+      "learning_rate": 4.692561212814439e-07,
+      "loss": 0.8167,
+      "step": 6576
+    },
+    {
+      "epoch": 0.9051747866776768,
+      "grad_norm": 1.6101297233610008,
+      "learning_rate": 4.6790764441061345e-07,
+      "loss": 0.9209,
+      "step": 6577
+    },
+    {
+      "epoch": 0.9053124139829342,
+      "grad_norm": 1.839525351460534,
+      "learning_rate": 4.665610614360527e-07,
+      "loss": 0.7991,
+      "step": 6578
+    },
+    {
+      "epoch": 0.9054500412881916,
+      "grad_norm": 1.6563624770820888,
+      "learning_rate": 4.6521637262531117e-07,
+      "loss": 0.9147,
+      "step": 6579
+    },
+    {
+      "epoch": 0.905587668593449,
+      "grad_norm": 1.8234873772590174,
+      "learning_rate": 4.6387357824555945e-07,
+      "loss": 0.7843,
+      "step": 6580
+    },
+    {
+      "epoch": 0.9057252958987063,
+      "grad_norm": 2.0193544014437013,
+      "learning_rate": 4.6253267856359087e-07,
+      "loss": 0.8302,
+      "step": 6581
+    },
+    {
+      "epoch": 0.9058629232039637,
+      "grad_norm": 2.014872590370536,
+      "learning_rate": 4.6119367384582783e-07,
+      "loss": 0.9115,
+      "step": 6582
+    },
+    {
+      "epoch": 0.906000550509221,
+      "grad_norm": 2.11212818691487,
+      "learning_rate": 4.598565643583075e-07,
+      "loss": 0.9088,
+      "step": 6583
+    },
+    {
+      "epoch": 0.9061381778144784,
+      "grad_norm": 2.108754784364996,
+      "learning_rate": 4.5852135036669965e-07,
+      "loss": 0.8706,
+      "step": 6584
+    },
+    {
+      "epoch": 0.9062758051197357,
+      "grad_norm": 1.6985852446316199,
+      "learning_rate": 4.571880321362887e-07,
+      "loss": 0.8444,
+      "step": 6585
+    },
+    {
+      "epoch": 0.9064134324249932,
+      "grad_norm": 1.8210663379800185,
+      "learning_rate": 4.5585660993199166e-07,
+      "loss": 0.8954,
+      "step": 6586
+    },
+    {
+      "epoch": 0.9065510597302505,
+      "grad_norm": 1.6887238460973126,
+      "learning_rate": 4.54527084018338e-07,
+      "loss": 0.8618,
+      "step": 6587
+    },
+    {
+      "epoch": 0.9066886870355079,
+      "grad_norm": 1.4555231728404738,
+      "learning_rate": 4.531994546594909e-07,
+      "loss": 0.8832,
+      "step": 6588
+    },
+    {
+      "epoch": 0.9068263143407652,
+      "grad_norm": 1.7671927289371971,
+      "learning_rate": 4.518737221192304e-07,
+      "loss": 0.9437,
+      "step": 6589
+    },
+    {
+      "epoch": 0.9069639416460226,
+      "grad_norm": 1.9954060623643808,
+      "learning_rate": 4.5054988666096255e-07,
+      "loss": 0.8603,
+      "step": 6590
+    },
+    {
+      "epoch": 0.9071015689512799,
+      "grad_norm": 1.7106584993534777,
+      "learning_rate": 4.4922794854771355e-07,
+      "loss": 0.8496,
+      "step": 6591
+    },
+    {
+      "epoch": 0.9072391962565373,
+      "grad_norm": 1.81272496832236,
+      "learning_rate": 4.479079080421356e-07,
+      "loss": 0.9054,
+      "step": 6592
+    },
+    {
+      "epoch": 0.9073768235617946,
+      "grad_norm": 2.0798587406957223,
+      "learning_rate": 4.465897654065032e-07,
+      "loss": 0.8819,
+      "step": 6593
+    },
+    {
+      "epoch": 0.9075144508670521,
+      "grad_norm": 1.9205973601469513,
+      "learning_rate": 4.4527352090271147e-07,
+      "loss": 0.9219,
+      "step": 6594
+    },
+    {
+      "epoch": 0.9076520781723094,
+      "grad_norm": 1.8955465109042444,
+      "learning_rate": 4.4395917479228225e-07,
+      "loss": 0.8912,
+      "step": 6595
+    },
+    {
+      "epoch": 0.9077897054775668,
+      "grad_norm": 1.8527383773608623,
+      "learning_rate": 4.426467273363566e-07,
+      "loss": 0.8703,
+      "step": 6596
+    },
+    {
+      "epoch": 0.9079273327828241,
+      "grad_norm": 1.7415456547412604,
+      "learning_rate": 4.413361787956993e-07,
+      "loss": 0.8579,
+      "step": 6597
+    },
+    {
+      "epoch": 0.9080649600880815,
+      "grad_norm": 1.7758426039912696,
+      "learning_rate": 4.4002752943070105e-07,
+      "loss": 0.8935,
+      "step": 6598
+    },
+    {
+      "epoch": 0.9082025873933388,
+      "grad_norm": 1.8530171865645382,
+      "learning_rate": 4.3872077950136925e-07,
+      "loss": 0.8698,
+      "step": 6599
+    },
+    {
+      "epoch": 0.9083402146985962,
+      "grad_norm": 1.596771878669106,
+      "learning_rate": 4.3741592926734076e-07,
+      "loss": 0.9345,
+      "step": 6600
+    },
+    {
+      "epoch": 0.9084778420038536,
+      "grad_norm": 1.6112671912522478,
+      "learning_rate": 4.3611297898786817e-07,
+      "loss": 0.8113,
+      "step": 6601
+    },
+    {
+      "epoch": 0.908615469309111,
+      "grad_norm": 1.9419853892114927,
+      "learning_rate": 4.348119289218322e-07,
+      "loss": 0.7748,
+      "step": 6602
+    },
+    {
+      "epoch": 0.9087530966143683,
+      "grad_norm": 1.801264363217741,
+      "learning_rate": 4.335127793277316e-07,
+      "loss": 0.977,
+      "step": 6603
+    },
+    {
+      "epoch": 0.9088907239196257,
+      "grad_norm": 1.9890863003779633,
+      "learning_rate": 4.322155304636921e-07,
+      "loss": 0.887,
+      "step": 6604
+    },
+    {
+      "epoch": 0.909028351224883,
+      "grad_norm": 1.717762295220119,
+      "learning_rate": 4.309201825874576e-07,
+      "loss": 0.8402,
+      "step": 6605
+    },
+    {
+      "epoch": 0.9091659785301404,
+      "grad_norm": 1.8232544832949633,
+      "learning_rate": 4.2962673595639435e-07,
+      "loss": 0.8873,
+      "step": 6606
+    },
+    {
+      "epoch": 0.9093036058353977,
+      "grad_norm": 1.6639896450540121,
+      "learning_rate": 4.28335190827498e-07,
+      "loss": 0.8065,
+      "step": 6607
+    },
+    {
+      "epoch": 0.9094412331406551,
+      "grad_norm": 1.8037209367726625,
+      "learning_rate": 4.2704554745737534e-07,
+      "loss": 0.907,
+      "step": 6608
+    },
+    {
+      "epoch": 0.9095788604459125,
+      "grad_norm": 1.8621254840145887,
+      "learning_rate": 4.2575780610226494e-07,
+      "loss": 0.9288,
+      "step": 6609
+    },
+    {
+      "epoch": 0.9097164877511699,
+      "grad_norm": 1.574332402170311,
+      "learning_rate": 4.2447196701801976e-07,
+      "loss": 0.9324,
+      "step": 6610
+    },
+    {
+      "epoch": 0.9098541150564272,
+      "grad_norm": 1.7674899905187675,
+      "learning_rate": 4.231880304601199e-07,
+      "loss": 0.9331,
+      "step": 6611
+    },
+    {
+      "epoch": 0.9099917423616846,
+      "grad_norm": 1.706511890907948,
+      "learning_rate": 4.2190599668366584e-07,
+      "loss": 0.8878,
+      "step": 6612
+    },
+    {
+      "epoch": 0.9101293696669419,
+      "grad_norm": 1.833179119890394,
+      "learning_rate": 4.206258659433804e-07,
+      "loss": 0.8856,
+      "step": 6613
+    },
+    {
+      "epoch": 0.9102669969721993,
+      "grad_norm": 1.5151215350581662,
+      "learning_rate": 4.1934763849361016e-07,
+      "loss": 0.9007,
+      "step": 6614
+    },
+    {
+      "epoch": 0.9104046242774566,
+      "grad_norm": 1.7594633829797883,
+      "learning_rate": 4.180713145883164e-07,
+      "loss": 0.8985,
+      "step": 6615
+    },
+    {
+      "epoch": 0.910542251582714,
+      "grad_norm": 1.975519087189147,
+      "learning_rate": 4.1679689448109294e-07,
+      "loss": 0.8899,
+      "step": 6616
+    },
+    {
+      "epoch": 0.9106798788879714,
+      "grad_norm": 1.9699918505680831,
+      "learning_rate": 4.1552437842514395e-07,
+      "loss": 0.8378,
+      "step": 6617
+    },
+    {
+      "epoch": 0.9108175061932288,
+      "grad_norm": 1.9591865936143695,
+      "learning_rate": 4.1425376667330596e-07,
+      "loss": 0.9208,
+      "step": 6618
+    },
+    {
+      "epoch": 0.9109551334984861,
+      "grad_norm": 1.6895012658487327,
+      "learning_rate": 4.1298505947802825e-07,
+      "loss": 0.8609,
+      "step": 6619
+    },
+    {
+      "epoch": 0.9110927608037435,
+      "grad_norm": 1.6723245518917729,
+      "learning_rate": 4.1171825709138804e-07,
+      "loss": 0.9109,
+      "step": 6620
+    },
+    {
+      "epoch": 0.9112303881090008,
+      "grad_norm": 1.7093450591013928,
+      "learning_rate": 4.104533597650817e-07,
+      "loss": 0.8829,
+      "step": 6621
+    },
+    {
+      "epoch": 0.9113680154142582,
+      "grad_norm": 1.7379208229181309,
+      "learning_rate": 4.0919036775042496e-07,
+      "loss": 0.9415,
+      "step": 6622
+    },
+    {
+      "epoch": 0.9115056427195155,
+      "grad_norm": 1.9339672998004858,
+      "learning_rate": 4.0792928129835927e-07,
+      "loss": 0.9521,
+      "step": 6623
+    },
+    {
+      "epoch": 0.911643270024773,
+      "grad_norm": 1.6733579771648264,
+      "learning_rate": 4.0667010065944423e-07,
+      "loss": 0.8637,
+      "step": 6624
+    },
+    {
+      "epoch": 0.9117808973300303,
+      "grad_norm": 1.7416226244790092,
+      "learning_rate": 4.0541282608386077e-07,
+      "loss": 0.863,
+      "step": 6625
+    },
+    {
+      "epoch": 0.9119185246352877,
+      "grad_norm": 1.8401733008096999,
+      "learning_rate": 4.041574578214147e-07,
+      "loss": 0.8833,
+      "step": 6626
+    },
+    {
+      "epoch": 0.912056151940545,
+      "grad_norm": 1.7730339800730637,
+      "learning_rate": 4.0290399612152755e-07,
+      "loss": 0.9953,
+      "step": 6627
+    },
+    {
+      "epoch": 0.9121937792458024,
+      "grad_norm": 1.6269898459706014,
+      "learning_rate": 4.0165244123324674e-07,
+      "loss": 0.8787,
+      "step": 6628
+    },
+    {
+      "epoch": 0.9123314065510597,
+      "grad_norm": 1.750878474384059,
+      "learning_rate": 4.00402793405239e-07,
+      "loss": 0.8903,
+      "step": 6629
+    },
+    {
+      "epoch": 0.9124690338563171,
+      "grad_norm": 1.6926773769276244,
+      "learning_rate": 3.9915505288579236e-07,
+      "loss": 0.8295,
+      "step": 6630
+    },
+    {
+      "epoch": 0.9126066611615744,
+      "grad_norm": 1.780408957544683,
+      "learning_rate": 3.9790921992281515e-07,
+      "loss": 0.9826,
+      "step": 6631
+    },
+    {
+      "epoch": 0.9127442884668319,
+      "grad_norm": 1.6870287344489507,
+      "learning_rate": 3.966652947638383e-07,
+      "loss": 0.9341,
+      "step": 6632
+    },
+    {
+      "epoch": 0.9128819157720892,
+      "grad_norm": 1.6532776105631557,
+      "learning_rate": 3.9542327765601183e-07,
+      "loss": 0.9941,
+      "step": 6633
+    },
+    {
+      "epoch": 0.9130195430773466,
+      "grad_norm": 1.8311052463862865,
+      "learning_rate": 3.941831688461073e-07,
+      "loss": 0.8675,
+      "step": 6634
+    },
+    {
+      "epoch": 0.9131571703826039,
+      "grad_norm": 1.7801054480401524,
+      "learning_rate": 3.9294496858051757e-07,
+      "loss": 0.9045,
+      "step": 6635
+    },
+    {
+      "epoch": 0.9132947976878613,
+      "grad_norm": 1.7328620085983233,
+      "learning_rate": 3.9170867710525583e-07,
+      "loss": 0.9292,
+      "step": 6636
+    },
+    {
+      "epoch": 0.9134324249931186,
+      "grad_norm": 1.8854547764240481,
+      "learning_rate": 3.90474294665959e-07,
+      "loss": 0.8593,
+      "step": 6637
+    },
+    {
+      "epoch": 0.913570052298376,
+      "grad_norm": 1.6695318283187075,
+      "learning_rate": 3.8924182150787635e-07,
+      "loss": 0.8711,
+      "step": 6638
+    },
+    {
+      "epoch": 0.9137076796036334,
+      "grad_norm": 1.714850573071997,
+      "learning_rate": 3.88011257875891e-07,
+      "loss": 0.9459,
+      "step": 6639
+    },
+    {
+      "epoch": 0.9138453069088908,
+      "grad_norm": 1.7400816330670235,
+      "learning_rate": 3.867826040144906e-07,
+      "loss": 0.8727,
+      "step": 6640
+    },
+    {
+      "epoch": 0.9139829342141481,
+      "grad_norm": 2.0030781093942145,
+      "learning_rate": 3.855558601678e-07,
+      "loss": 0.9388,
+      "step": 6641
+    },
+    {
+      "epoch": 0.9141205615194055,
+      "grad_norm": 1.7510381620771789,
+      "learning_rate": 3.843310265795508e-07,
+      "loss": 0.8248,
+      "step": 6642
+    },
+    {
+      "epoch": 0.9142581888246628,
+      "grad_norm": 1.9172835490329863,
+      "learning_rate": 3.8310810349310165e-07,
+      "loss": 0.8942,
+      "step": 6643
+    },
+    {
+      "epoch": 0.9143958161299202,
+      "grad_norm": 2.0032379166871634,
+      "learning_rate": 3.8188709115143384e-07,
+      "loss": 0.8625,
+      "step": 6644
+    },
+    {
+      "epoch": 0.9145334434351775,
+      "grad_norm": 1.9548096070771217,
+      "learning_rate": 3.8066798979714106e-07,
+      "loss": 0.9453,
+      "step": 6645
+    },
+    {
+      "epoch": 0.9146710707404349,
+      "grad_norm": 1.9540109225545668,
+      "learning_rate": 3.7945079967244613e-07,
+      "loss": 0.9001,
+      "step": 6646
+    },
+    {
+      "epoch": 0.9148086980456923,
+      "grad_norm": 1.7875528064076698,
+      "learning_rate": 3.782355210191846e-07,
+      "loss": 0.9243,
+      "step": 6647
+    },
+    {
+      "epoch": 0.9149463253509497,
+      "grad_norm": 1.740506784024848,
+      "learning_rate": 3.770221540788188e-07,
+      "loss": 0.9466,
+      "step": 6648
+    },
+    {
+      "epoch": 0.915083952656207,
+      "grad_norm": 1.9094809839627789,
+      "learning_rate": 3.75810699092426e-07,
+      "loss": 0.8493,
+      "step": 6649
+    },
+    {
+      "epoch": 0.9152215799614644,
+      "grad_norm": 1.6571270209215923,
+      "learning_rate": 3.746011563007057e-07,
+      "loss": 0.9009,
+      "step": 6650
+    },
+    {
+      "epoch": 0.9153592072667217,
+      "grad_norm": 1.879873722723251,
+      "learning_rate": 3.733935259439803e-07,
+      "loss": 0.8319,
+      "step": 6651
+    },
+    {
+      "epoch": 0.9154968345719791,
+      "grad_norm": 1.7502571279791121,
+      "learning_rate": 3.721878082621866e-07,
+      "loss": 0.926,
+      "step": 6652
+    },
+    {
+      "epoch": 0.9156344618772364,
+      "grad_norm": 1.5251944349329831,
+      "learning_rate": 3.7098400349488415e-07,
+      "loss": 0.8703,
+      "step": 6653
+    },
+    {
+      "epoch": 0.9157720891824938,
+      "grad_norm": 1.8625417040164118,
+      "learning_rate": 3.6978211188125504e-07,
+      "loss": 0.8786,
+      "step": 6654
+    },
+    {
+      "epoch": 0.9159097164877512,
+      "grad_norm": 1.6554450799381786,
+      "learning_rate": 3.685821336600959e-07,
+      "loss": 0.8419,
+      "step": 6655
+    },
+    {
+      "epoch": 0.9160473437930086,
+      "grad_norm": 1.910404918622632,
+      "learning_rate": 3.6738406906982716e-07,
+      "loss": 0.867,
+      "step": 6656
+    },
+    {
+      "epoch": 0.9161849710982659,
+      "grad_norm": 1.7115333544832927,
+      "learning_rate": 3.6618791834848734e-07,
+      "loss": 0.8915,
+      "step": 6657
+    },
+    {
+      "epoch": 0.9163225984035233,
+      "grad_norm": 1.853647120727828,
+      "learning_rate": 3.649936817337363e-07,
+      "loss": 0.8883,
+      "step": 6658
+    },
+    {
+      "epoch": 0.9164602257087806,
+      "grad_norm": 1.6383060504556373,
+      "learning_rate": 3.63801359462852e-07,
+      "loss": 0.8606,
+      "step": 6659
+    },
+    {
+      "epoch": 0.916597853014038,
+      "grad_norm": 1.7642863664843762,
+      "learning_rate": 3.6261095177273164e-07,
+      "loss": 0.8689,
+      "step": 6660
+    },
+    {
+      "epoch": 0.9167354803192953,
+      "grad_norm": 1.9386432165065688,
+      "learning_rate": 3.614224588998938e-07,
+      "loss": 0.8728,
+      "step": 6661
+    },
+    {
+      "epoch": 0.9168731076245528,
+      "grad_norm": 1.6334616354537612,
+      "learning_rate": 3.6023588108047627e-07,
+      "loss": 0.8911,
+      "step": 6662
+    },
+    {
+      "epoch": 0.9170107349298101,
+      "grad_norm": 1.812030793003262,
+      "learning_rate": 3.590512185502348e-07,
+      "loss": 0.8758,
+      "step": 6663
+    },
+    {
+      "epoch": 0.9171483622350675,
+      "grad_norm": 1.6046301523042072,
+      "learning_rate": 3.5786847154454576e-07,
+      "loss": 0.8928,
+      "step": 6664
+    },
+    {
+      "epoch": 0.9172859895403248,
+      "grad_norm": 1.8652323603229102,
+      "learning_rate": 3.5668764029840543e-07,
+      "loss": 0.9667,
+      "step": 6665
+    },
+    {
+      "epoch": 0.9174236168455822,
+      "grad_norm": 1.6119432651216818,
+      "learning_rate": 3.5550872504642844e-07,
+      "loss": 0.9323,
+      "step": 6666
+    },
+    {
+      "epoch": 0.9175612441508395,
+      "grad_norm": 1.6850927496028545,
+      "learning_rate": 3.543317260228485e-07,
+      "loss": 0.9707,
+      "step": 6667
+    },
+    {
+      "epoch": 0.9176988714560969,
+      "grad_norm": 1.6147412596159219,
+      "learning_rate": 3.5315664346151857e-07,
+      "loss": 0.9153,
+      "step": 6668
+    },
+    {
+      "epoch": 0.9178364987613542,
+      "grad_norm": 1.8546617511959536,
+      "learning_rate": 3.5198347759591634e-07,
+      "loss": 0.873,
+      "step": 6669
+    },
+    {
+      "epoch": 0.9179741260666117,
+      "grad_norm": 1.924301868998822,
+      "learning_rate": 3.5081222865912645e-07,
+      "loss": 0.9002,
+      "step": 6670
+    },
+    {
+      "epoch": 0.918111753371869,
+      "grad_norm": 1.7908707947828946,
+      "learning_rate": 3.496428968838661e-07,
+      "loss": 0.9398,
+      "step": 6671
+    },
+    {
+      "epoch": 0.9182493806771264,
+      "grad_norm": 1.7311379218446268,
+      "learning_rate": 3.4847548250246167e-07,
+      "loss": 0.9166,
+      "step": 6672
+    },
+    {
+      "epoch": 0.9183870079823837,
+      "grad_norm": 1.738987796179036,
+      "learning_rate": 3.473099857468665e-07,
+      "loss": 0.8442,
+      "step": 6673
+    },
+    {
+      "epoch": 0.9185246352876411,
+      "grad_norm": 1.5105149515565026,
+      "learning_rate": 3.461464068486442e-07,
+      "loss": 0.8594,
+      "step": 6674
+    },
+    {
+      "epoch": 0.9186622625928984,
+      "grad_norm": 1.7604936033848666,
+      "learning_rate": 3.4498474603898323e-07,
+      "loss": 0.8679,
+      "step": 6675
+    },
+    {
+      "epoch": 0.9187998898981558,
+      "grad_norm": 1.7241475854773618,
+      "learning_rate": 3.438250035486923e-07,
+      "loss": 0.8838,
+      "step": 6676
+    },
+    {
+      "epoch": 0.9189375172034132,
+      "grad_norm": 1.8471638937604435,
+      "learning_rate": 3.426671796081926e-07,
+      "loss": 0.8193,
+      "step": 6677
+    },
+    {
+      "epoch": 0.9190751445086706,
+      "grad_norm": 1.7662535974844213,
+      "learning_rate": 3.415112744475313e-07,
+      "loss": 0.8909,
+      "step": 6678
+    },
+    {
+      "epoch": 0.9192127718139279,
+      "grad_norm": 1.4666376711190297,
+      "learning_rate": 3.4035728829636796e-07,
+      "loss": 0.9573,
+      "step": 6679
+    },
+    {
+      "epoch": 0.9193503991191853,
+      "grad_norm": 1.6793110481500233,
+      "learning_rate": 3.3920522138398805e-07,
+      "loss": 0.8592,
+      "step": 6680
+    },
+    {
+      "epoch": 0.9194880264244426,
+      "grad_norm": 1.63743169968516,
+      "learning_rate": 3.380550739392863e-07,
+      "loss": 0.8868,
+      "step": 6681
+    },
+    {
+      "epoch": 0.9196256537297,
+      "grad_norm": 1.8810930160267187,
+      "learning_rate": 3.369068461907843e-07,
+      "loss": 0.87,
+      "step": 6682
+    },
+    {
+      "epoch": 0.9197632810349573,
+      "grad_norm": 1.8177284947069905,
+      "learning_rate": 3.357605383666185e-07,
+      "loss": 0.9294,
+      "step": 6683
+    },
+    {
+      "epoch": 0.9199009083402147,
+      "grad_norm": 1.8911852417396828,
+      "learning_rate": 3.3461615069454555e-07,
+      "loss": 0.9342,
+      "step": 6684
+    },
+    {
+      "epoch": 0.9200385356454721,
+      "grad_norm": 1.64706066456111,
+      "learning_rate": 3.334736834019381e-07,
+      "loss": 0.9067,
+      "step": 6685
+    },
+    {
+      "epoch": 0.9201761629507295,
+      "grad_norm": 1.5695035191908084,
+      "learning_rate": 3.3233313671579e-07,
+      "loss": 0.877,
+      "step": 6686
+    },
+    {
+      "epoch": 0.9203137902559868,
+      "grad_norm": 1.7427359490335155,
+      "learning_rate": 3.3119451086271125e-07,
+      "loss": 0.9123,
+      "step": 6687
+    },
+    {
+      "epoch": 0.9204514175612442,
+      "grad_norm": 1.5918768816296056,
+      "learning_rate": 3.3005780606893187e-07,
+      "loss": 0.8355,
+      "step": 6688
+    },
+    {
+      "epoch": 0.9205890448665015,
+      "grad_norm": 1.8127723040383803,
+      "learning_rate": 3.2892302256029907e-07,
+      "loss": 0.9059,
+      "step": 6689
+    },
+    {
+      "epoch": 0.9207266721717589,
+      "grad_norm": 1.7057368540740396,
+      "learning_rate": 3.277901605622802e-07,
+      "loss": 0.8904,
+      "step": 6690
+    },
+    {
+      "epoch": 0.9208642994770162,
+      "grad_norm": 2.253848317789013,
+      "learning_rate": 3.2665922029995635e-07,
+      "loss": 0.8578,
+      "step": 6691
+    },
+    {
+      "epoch": 0.9210019267822736,
+      "grad_norm": 1.7667600616031471,
+      "learning_rate": 3.255302019980333e-07,
+      "loss": 0.8533,
+      "step": 6692
+    },
+    {
+      "epoch": 0.921139554087531,
+      "grad_norm": 1.6974616079507912,
+      "learning_rate": 3.2440310588082946e-07,
+      "loss": 0.9021,
+      "step": 6693
+    },
+    {
+      "epoch": 0.9212771813927884,
+      "grad_norm": 2.084911902778498,
+      "learning_rate": 3.232779321722845e-07,
+      "loss": 0.848,
+      "step": 6694
+    },
+    {
+      "epoch": 0.9214148086980457,
+      "grad_norm": 1.9454637788301041,
+      "learning_rate": 3.2215468109595416e-07,
+      "loss": 0.9145,
+      "step": 6695
+    },
+    {
+      "epoch": 0.921552436003303,
+      "grad_norm": 1.7208332923753595,
+      "learning_rate": 3.2103335287501314e-07,
+      "loss": 0.8818,
+      "step": 6696
+    },
+    {
+      "epoch": 0.9216900633085604,
+      "grad_norm": 1.835177135267487,
+      "learning_rate": 3.1991394773225435e-07,
+      "loss": 0.8931,
+      "step": 6697
+    },
+    {
+      "epoch": 0.9218276906138178,
+      "grad_norm": 1.8319827998143843,
+      "learning_rate": 3.1879646589008775e-07,
+      "loss": 0.8974,
+      "step": 6698
+    },
+    {
+      "epoch": 0.9219653179190751,
+      "grad_norm": 2.004390978715643,
+      "learning_rate": 3.176809075705445e-07,
+      "loss": 0.9169,
+      "step": 6699
+    },
+    {
+      "epoch": 0.9221029452243326,
+      "grad_norm": 1.8122984673102467,
+      "learning_rate": 3.1656727299526513e-07,
+      "loss": 0.8577,
+      "step": 6700
+    },
+    {
+      "epoch": 0.9222405725295899,
+      "grad_norm": 1.7913896494943586,
+      "learning_rate": 3.154555623855193e-07,
+      "loss": 0.9805,
+      "step": 6701
+    },
+    {
+      "epoch": 0.9223781998348473,
+      "grad_norm": 1.7884600144595337,
+      "learning_rate": 3.1434577596218483e-07,
+      "loss": 0.9343,
+      "step": 6702
+    },
+    {
+      "epoch": 0.9225158271401046,
+      "grad_norm": 1.7956755109017832,
+      "learning_rate": 3.132379139457642e-07,
+      "loss": 0.8523,
+      "step": 6703
+    },
+    {
+      "epoch": 0.922653454445362,
+      "grad_norm": 1.733126742744374,
+      "learning_rate": 3.121319765563702e-07,
+      "loss": 0.9207,
+      "step": 6704
+    },
+    {
+      "epoch": 0.9227910817506193,
+      "grad_norm": 1.8734521486985727,
+      "learning_rate": 3.110279640137415e-07,
+      "loss": 0.8499,
+      "step": 6705
+    },
+    {
+      "epoch": 0.9229287090558767,
+      "grad_norm": 1.8803686463013634,
+      "learning_rate": 3.099258765372304e-07,
+      "loss": 0.9017,
+      "step": 6706
+    },
+    {
+      "epoch": 0.923066336361134,
+      "grad_norm": 1.8920744331639814,
+      "learning_rate": 3.088257143458018e-07,
+      "loss": 0.8457,
+      "step": 6707
+    },
+    {
+      "epoch": 0.9232039636663915,
+      "grad_norm": 1.5580227608560706,
+      "learning_rate": 3.0772747765804854e-07,
+      "loss": 0.9205,
+      "step": 6708
+    },
+    {
+      "epoch": 0.9233415909716488,
+      "grad_norm": 1.9287959286744265,
+      "learning_rate": 3.0663116669217064e-07,
+      "loss": 0.9555,
+      "step": 6709
+    },
+    {
+      "epoch": 0.9234792182769062,
+      "grad_norm": 1.7840050128312877,
+      "learning_rate": 3.055367816659949e-07,
+      "loss": 0.9155,
+      "step": 6710
+    },
+    {
+      "epoch": 0.9236168455821635,
+      "grad_norm": 1.6611551553663357,
+      "learning_rate": 3.0444432279695514e-07,
+      "loss": 0.9608,
+      "step": 6711
+    },
+    {
+      "epoch": 0.9237544728874209,
+      "grad_norm": 1.8760277839256523,
+      "learning_rate": 3.033537903021111e-07,
+      "loss": 0.8878,
+      "step": 6712
+    },
+    {
+      "epoch": 0.9238921001926782,
+      "grad_norm": 1.9739036425922312,
+      "learning_rate": 3.022651843981372e-07,
+      "loss": 0.9726,
+      "step": 6713
+    },
+    {
+      "epoch": 0.9240297274979355,
+      "grad_norm": 2.187773574199657,
+      "learning_rate": 3.0117850530132385e-07,
+      "loss": 0.8506,
+      "step": 6714
+    },
+    {
+      "epoch": 0.924167354803193,
+      "grad_norm": 1.784754402597568,
+      "learning_rate": 3.0009375322757826e-07,
+      "loss": 0.8557,
+      "step": 6715
+    },
+    {
+      "epoch": 0.9243049821084504,
+      "grad_norm": 1.8168110741645787,
+      "learning_rate": 2.99010928392427e-07,
+      "loss": 0.8784,
+      "step": 6716
+    },
+    {
+      "epoch": 0.9244426094137077,
+      "grad_norm": 2.0265989117451118,
+      "learning_rate": 2.9793003101101225e-07,
+      "loss": 0.9467,
+      "step": 6717
+    },
+    {
+      "epoch": 0.924580236718965,
+      "grad_norm": 1.6499394612165763,
+      "learning_rate": 2.968510612980935e-07,
+      "loss": 0.8682,
+      "step": 6718
+    },
+    {
+      "epoch": 0.9247178640242224,
+      "grad_norm": 1.7920370365755391,
+      "learning_rate": 2.957740194680481e-07,
+      "loss": 0.9218,
+      "step": 6719
+    },
+    {
+      "epoch": 0.9248554913294798,
+      "grad_norm": 1.7402510327042713,
+      "learning_rate": 2.9469890573486705e-07,
+      "loss": 0.8526,
+      "step": 6720
+    },
+    {
+      "epoch": 0.9249931186347371,
+      "grad_norm": 1.8590331059899194,
+      "learning_rate": 2.93625720312164e-07,
+      "loss": 0.8838,
+      "step": 6721
+    },
+    {
+      "epoch": 0.9251307459399944,
+      "grad_norm": 2.057331600625596,
+      "learning_rate": 2.925544634131638e-07,
+      "loss": 0.9378,
+      "step": 6722
+    },
+    {
+      "epoch": 0.9252683732452519,
+      "grad_norm": 2.0154618278172243,
+      "learning_rate": 2.9148513525071085e-07,
+      "loss": 0.8736,
+      "step": 6723
+    },
+    {
+      "epoch": 0.9254060005505093,
+      "grad_norm": 1.7521081650127621,
+      "learning_rate": 2.904177360372662e-07,
+      "loss": 0.9283,
+      "step": 6724
+    },
+    {
+      "epoch": 0.9255436278557666,
+      "grad_norm": 2.037714302100558,
+      "learning_rate": 2.8935226598490904e-07,
+      "loss": 0.8737,
+      "step": 6725
+    },
+    {
+      "epoch": 0.925681255161024,
+      "grad_norm": 2.4180617062426237,
+      "learning_rate": 2.8828872530533127e-07,
+      "loss": 0.8403,
+      "step": 6726
+    },
+    {
+      "epoch": 0.9258188824662813,
+      "grad_norm": 1.7423897982131689,
+      "learning_rate": 2.872271142098448e-07,
+      "loss": 0.925,
+      "step": 6727
+    },
+    {
+      "epoch": 0.9259565097715386,
+      "grad_norm": 1.761866097084584,
+      "learning_rate": 2.861674329093778e-07,
+      "loss": 0.8629,
+      "step": 6728
+    },
+    {
+      "epoch": 0.926094137076796,
+      "grad_norm": 2.1137081289410853,
+      "learning_rate": 2.8510968161447386e-07,
+      "loss": 0.9382,
+      "step": 6729
+    },
+    {
+      "epoch": 0.9262317643820533,
+      "grad_norm": 1.8521920591215564,
+      "learning_rate": 2.840538605352927e-07,
+      "loss": 0.9803,
+      "step": 6730
+    },
+    {
+      "epoch": 0.9263693916873108,
+      "grad_norm": 1.8039333554448262,
+      "learning_rate": 2.829999698816155e-07,
+      "loss": 0.9215,
+      "step": 6731
+    },
+    {
+      "epoch": 0.9265070189925682,
+      "grad_norm": 1.8825457669069858,
+      "learning_rate": 2.819480098628302e-07,
+      "loss": 0.8621,
+      "step": 6732
+    },
+    {
+      "epoch": 0.9266446462978255,
+      "grad_norm": 1.7522793000762182,
+      "learning_rate": 2.808979806879519e-07,
+      "loss": 0.8051,
+      "step": 6733
+    },
+    {
+      "epoch": 0.9267822736030829,
+      "grad_norm": 1.620167762810065,
+      "learning_rate": 2.7984988256560265e-07,
+      "loss": 0.9609,
+      "step": 6734
+    },
+    {
+      "epoch": 0.9269199009083402,
+      "grad_norm": 4.753660088830378,
+      "learning_rate": 2.788037157040302e-07,
+      "loss": 0.8456,
+      "step": 6735
+    },
+    {
+      "epoch": 0.9270575282135975,
+      "grad_norm": 1.734463601719378,
+      "learning_rate": 2.7775948031108835e-07,
+      "loss": 0.8798,
+      "step": 6736
+    },
+    {
+      "epoch": 0.9271951555188549,
+      "grad_norm": 2.0480650194414745,
+      "learning_rate": 2.767171765942556e-07,
+      "loss": 0.9232,
+      "step": 6737
+    },
+    {
+      "epoch": 0.9273327828241124,
+      "grad_norm": 1.6693304190490281,
+      "learning_rate": 2.7567680476062287e-07,
+      "loss": 0.9126,
+      "step": 6738
+    },
+    {
+      "epoch": 0.9274704101293697,
+      "grad_norm": 1.7766777369505644,
+      "learning_rate": 2.746383650168971e-07,
+      "loss": 0.8125,
+      "step": 6739
+    },
+    {
+      "epoch": 0.927608037434627,
+      "grad_norm": 1.6354296490214693,
+      "learning_rate": 2.7360185756940326e-07,
+      "loss": 0.9499,
+      "step": 6740
+    },
+    {
+      "epoch": 0.9277456647398844,
+      "grad_norm": 2.1053599947431305,
+      "learning_rate": 2.7256728262407884e-07,
+      "loss": 0.7553,
+      "step": 6741
+    },
+    {
+      "epoch": 0.9278832920451417,
+      "grad_norm": 1.7647441018626433,
+      "learning_rate": 2.715346403864838e-07,
+      "loss": 0.8812,
+      "step": 6742
+    },
+    {
+      "epoch": 0.9280209193503991,
+      "grad_norm": 1.67035945811495,
+      "learning_rate": 2.705039310617852e-07,
+      "loss": 0.9077,
+      "step": 6743
+    },
+    {
+      "epoch": 0.9281585466556564,
+      "grad_norm": 1.70150900897486,
+      "learning_rate": 2.694751548547736e-07,
+      "loss": 0.9253,
+      "step": 6744
+    },
+    {
+      "epoch": 0.9282961739609138,
+      "grad_norm": 1.6407886142683044,
+      "learning_rate": 2.684483119698533e-07,
+      "loss": 0.8626,
+      "step": 6745
+    },
+    {
+      "epoch": 0.9284338012661713,
+      "grad_norm": 1.6808680263612912,
+      "learning_rate": 2.6742340261104226e-07,
+      "loss": 0.9122,
+      "step": 6746
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 3.153135252003833,
+      "learning_rate": 2.6640042698197754e-07,
+      "loss": 0.9082,
+      "step": 6747
+    },
+    {
+      "epoch": 0.928709055876686,
+      "grad_norm": 1.7994744391412743,
+      "learning_rate": 2.653793852859088e-07,
+      "loss": 0.9158,
+      "step": 6748
+    },
+    {
+      "epoch": 0.9288466831819433,
+      "grad_norm": 1.7590305572227787,
+      "learning_rate": 2.643602777257037e-07,
+      "loss": 0.8904,
+      "step": 6749
+    },
+    {
+      "epoch": 0.9289843104872006,
+      "grad_norm": 1.6574442553348636,
+      "learning_rate": 2.633431045038448e-07,
+      "loss": 0.9064,
+      "step": 6750
+    },
+    {
+      "epoch": 0.929121937792458,
+      "grad_norm": 1.8289226024863323,
+      "learning_rate": 2.6232786582243154e-07,
+      "loss": 0.8852,
+      "step": 6751
+    },
+    {
+      "epoch": 0.9292595650977153,
+      "grad_norm": 1.9218483386695762,
+      "learning_rate": 2.6131456188317693e-07,
+      "loss": 0.9728,
+      "step": 6752
+    },
+    {
+      "epoch": 0.9293971924029728,
+      "grad_norm": 1.806320292935636,
+      "learning_rate": 2.6030319288741e-07,
+      "loss": 0.8817,
+      "step": 6753
+    },
+    {
+      "epoch": 0.9295348197082302,
+      "grad_norm": 1.7984514680359818,
+      "learning_rate": 2.592937590360778e-07,
+      "loss": 0.8991,
+      "step": 6754
+    },
+    {
+      "epoch": 0.9296724470134875,
+      "grad_norm": 1.8492724272917933,
+      "learning_rate": 2.582862605297398e-07,
+      "loss": 0.8819,
+      "step": 6755
+    },
+    {
+      "epoch": 0.9298100743187449,
+      "grad_norm": 1.9723252453530116,
+      "learning_rate": 2.572806975685715e-07,
+      "loss": 0.8688,
+      "step": 6756
+    },
+    {
+      "epoch": 0.9299477016240022,
+      "grad_norm": 1.8624228459278565,
+      "learning_rate": 2.562770703523665e-07,
+      "loss": 0.9136,
+      "step": 6757
+    },
+    {
+      "epoch": 0.9300853289292595,
+      "grad_norm": 1.9245564314205306,
+      "learning_rate": 2.5527537908052955e-07,
+      "loss": 0.9283,
+      "step": 6758
+    },
+    {
+      "epoch": 0.9302229562345169,
+      "grad_norm": 1.6357061165864268,
+      "learning_rate": 2.542756239520838e-07,
+      "loss": 0.9027,
+      "step": 6759
+    },
+    {
+      "epoch": 0.9303605835397742,
+      "grad_norm": 1.8474170957099505,
+      "learning_rate": 2.53277805165667e-07,
+      "loss": 0.9135,
+      "step": 6760
+    },
+    {
+      "epoch": 0.9304982108450317,
+      "grad_norm": 1.7750893802721934,
+      "learning_rate": 2.522819229195339e-07,
+      "loss": 0.8972,
+      "step": 6761
+    },
+    {
+      "epoch": 0.930635838150289,
+      "grad_norm": 1.9631761810096797,
+      "learning_rate": 2.5128797741154953e-07,
+      "loss": 0.8484,
+      "step": 6762
+    },
+    {
+      "epoch": 0.9307734654555464,
+      "grad_norm": 1.7482420845792697,
+      "learning_rate": 2.502959688392004e-07,
+      "loss": 0.8023,
+      "step": 6763
+    },
+    {
+      "epoch": 0.9309110927608037,
+      "grad_norm": 1.940705987395037,
+      "learning_rate": 2.49305897399581e-07,
+      "loss": 0.9226,
+      "step": 6764
+    },
+    {
+      "epoch": 0.9310487200660611,
+      "grad_norm": 1.9222557913282416,
+      "learning_rate": 2.483177632894096e-07,
+      "loss": 0.7704,
+      "step": 6765
+    },
+    {
+      "epoch": 0.9311863473713184,
+      "grad_norm": 1.8166968851934966,
+      "learning_rate": 2.4733156670501024e-07,
+      "loss": 0.9132,
+      "step": 6766
+    },
+    {
+      "epoch": 0.9313239746765758,
+      "grad_norm": 1.7524205401489148,
+      "learning_rate": 2.4634730784233065e-07,
+      "loss": 0.8915,
+      "step": 6767
+    },
+    {
+      "epoch": 0.9314616019818331,
+      "grad_norm": 1.9535002696157346,
+      "learning_rate": 2.453649868969277e-07,
+      "loss": 0.9533,
+      "step": 6768
+    },
+    {
+      "epoch": 0.9315992292870906,
+      "grad_norm": 1.5964397472559173,
+      "learning_rate": 2.4438460406397523e-07,
+      "loss": 0.8872,
+      "step": 6769
+    },
+    {
+      "epoch": 0.931736856592348,
+      "grad_norm": 2.533761718768302,
+      "learning_rate": 2.434061595382631e-07,
+      "loss": 0.9668,
+      "step": 6770
+    },
+    {
+      "epoch": 0.9318744838976053,
+      "grad_norm": 1.8085988263727473,
+      "learning_rate": 2.4242965351419235e-07,
+      "loss": 0.9081,
+      "step": 6771
+    },
+    {
+      "epoch": 0.9320121112028626,
+      "grad_norm": 1.7762197657678571,
+      "learning_rate": 2.414550861857845e-07,
+      "loss": 0.9095,
+      "step": 6772
+    },
+    {
+      "epoch": 0.93214973850812,
+      "grad_norm": 1.6881265047663423,
+      "learning_rate": 2.4048245774667025e-07,
+      "loss": 0.9605,
+      "step": 6773
+    },
+    {
+      "epoch": 0.9322873658133773,
+      "grad_norm": 1.7050383224414745,
+      "learning_rate": 2.395117683900983e-07,
+      "loss": 0.8747,
+      "step": 6774
+    },
+    {
+      "epoch": 0.9324249931186347,
+      "grad_norm": 1.8098885347307578,
+      "learning_rate": 2.3854301830893103e-07,
+      "loss": 0.9256,
+      "step": 6775
+    },
+    {
+      "epoch": 0.9325626204238922,
+      "grad_norm": 1.6812604516262217,
+      "learning_rate": 2.375762076956467e-07,
+      "loss": 0.8757,
+      "step": 6776
+    },
+    {
+      "epoch": 0.9327002477291495,
+      "grad_norm": 1.6718316606730066,
+      "learning_rate": 2.3661133674233727e-07,
+      "loss": 0.8363,
+      "step": 6777
+    },
+    {
+      "epoch": 0.9328378750344068,
+      "grad_norm": 1.830815890317496,
+      "learning_rate": 2.3564840564070933e-07,
+      "loss": 0.9076,
+      "step": 6778
+    },
+    {
+      "epoch": 0.9329755023396642,
+      "grad_norm": 2.1090006586241468,
+      "learning_rate": 2.3468741458208322e-07,
+      "loss": 0.9122,
+      "step": 6779
+    },
+    {
+      "epoch": 0.9331131296449215,
+      "grad_norm": 1.8956518754324183,
+      "learning_rate": 2.3372836375739506e-07,
+      "loss": 0.8679,
+      "step": 6780
+    },
+    {
+      "epoch": 0.9332507569501789,
+      "grad_norm": 2.0224689082816236,
+      "learning_rate": 2.3277125335719576e-07,
+      "loss": 0.8495,
+      "step": 6781
+    },
+    {
+      "epoch": 0.9333883842554362,
+      "grad_norm": 1.6435425807621866,
+      "learning_rate": 2.3181608357164874e-07,
+      "loss": 0.8987,
+      "step": 6782
+    },
+    {
+      "epoch": 0.9335260115606936,
+      "grad_norm": 1.8556749691602903,
+      "learning_rate": 2.3086285459053448e-07,
+      "loss": 0.9048,
+      "step": 6783
+    },
+    {
+      "epoch": 0.933663638865951,
+      "grad_norm": 1.661868194075282,
+      "learning_rate": 2.2991156660324698e-07,
+      "loss": 0.8208,
+      "step": 6784
+    },
+    {
+      "epoch": 0.9338012661712084,
+      "grad_norm": 2.059331357189913,
+      "learning_rate": 2.2896221979879173e-07,
+      "loss": 0.9161,
+      "step": 6785
+    },
+    {
+      "epoch": 0.9339388934764657,
+      "grad_norm": 2.171894164108577,
+      "learning_rate": 2.2801481436579344e-07,
+      "loss": 0.9523,
+      "step": 6786
+    },
+    {
+      "epoch": 0.9340765207817231,
+      "grad_norm": 2.1251543423482584,
+      "learning_rate": 2.2706935049248812e-07,
+      "loss": 0.8402,
+      "step": 6787
+    },
+    {
+      "epoch": 0.9342141480869804,
+      "grad_norm": 1.8816532150290932,
+      "learning_rate": 2.261258283667256e-07,
+      "loss": 0.9228,
+      "step": 6788
+    },
+    {
+      "epoch": 0.9343517753922378,
+      "grad_norm": 2.0154914803720088,
+      "learning_rate": 2.251842481759703e-07,
+      "loss": 0.9016,
+      "step": 6789
+    },
+    {
+      "epoch": 0.9344894026974951,
+      "grad_norm": 1.8329452619946158,
+      "learning_rate": 2.242446101073037e-07,
+      "loss": 0.8955,
+      "step": 6790
+    },
+    {
+      "epoch": 0.9346270300027526,
+      "grad_norm": 1.6874954454685946,
+      "learning_rate": 2.2330691434741647e-07,
+      "loss": 0.8578,
+      "step": 6791
+    },
+    {
+      "epoch": 0.93476465730801,
+      "grad_norm": 1.611040839581695,
+      "learning_rate": 2.2237116108261626e-07,
+      "loss": 0.8631,
+      "step": 6792
+    },
+    {
+      "epoch": 0.9349022846132673,
+      "grad_norm": 1.9036183638110515,
+      "learning_rate": 2.2143735049882765e-07,
+      "loss": 0.98,
+      "step": 6793
+    },
+    {
+      "epoch": 0.9350399119185246,
+      "grad_norm": 1.6988433964126208,
+      "learning_rate": 2.2050548278158113e-07,
+      "loss": 0.8362,
+      "step": 6794
+    },
+    {
+      "epoch": 0.935177539223782,
+      "grad_norm": 1.618138428940883,
+      "learning_rate": 2.1957555811603083e-07,
+      "loss": 0.9037,
+      "step": 6795
+    },
+    {
+      "epoch": 0.9353151665290393,
+      "grad_norm": 1.760559718276881,
+      "learning_rate": 2.186475766869356e-07,
+      "loss": 0.9204,
+      "step": 6796
+    },
+    {
+      "epoch": 0.9354527938342967,
+      "grad_norm": 1.7973875944856716,
+      "learning_rate": 2.1772153867867574e-07,
+      "loss": 0.8035,
+      "step": 6797
+    },
+    {
+      "epoch": 0.935590421139554,
+      "grad_norm": 1.6015590836416564,
+      "learning_rate": 2.1679744427524074e-07,
+      "loss": 0.8744,
+      "step": 6798
+    },
+    {
+      "epoch": 0.9357280484448115,
+      "grad_norm": 1.5994055198518262,
+      "learning_rate": 2.1587529366023597e-07,
+      "loss": 0.8914,
+      "step": 6799
+    },
+    {
+      "epoch": 0.9358656757500688,
+      "grad_norm": 1.719984044174885,
+      "learning_rate": 2.149550870168815e-07,
+      "loss": 0.8953,
+      "step": 6800
+    },
+    {
+      "epoch": 0.9360033030553262,
+      "grad_norm": 1.9498908135327602,
+      "learning_rate": 2.1403682452800778e-07,
+      "loss": 0.9363,
+      "step": 6801
+    },
+    {
+      "epoch": 0.9361409303605835,
+      "grad_norm": 1.9873876841070839,
+      "learning_rate": 2.131205063760633e-07,
+      "loss": 0.885,
+      "step": 6802
+    },
+    {
+      "epoch": 0.9362785576658409,
+      "grad_norm": 2.095703520387624,
+      "learning_rate": 2.1220613274310353e-07,
+      "loss": 0.8993,
+      "step": 6803
+    },
+    {
+      "epoch": 0.9364161849710982,
+      "grad_norm": 1.8325264793310525,
+      "learning_rate": 2.112937038108076e-07,
+      "loss": 0.9272,
+      "step": 6804
+    },
+    {
+      "epoch": 0.9365538122763556,
+      "grad_norm": 1.6836455140490378,
+      "learning_rate": 2.1038321976045828e-07,
+      "loss": 0.8428,
+      "step": 6805
+    },
+    {
+      "epoch": 0.9366914395816129,
+      "grad_norm": 1.5674543537096617,
+      "learning_rate": 2.0947468077295864e-07,
+      "loss": 0.8383,
+      "step": 6806
+    },
+    {
+      "epoch": 0.9368290668868704,
+      "grad_norm": 1.6639053428604236,
+      "learning_rate": 2.0856808702882203e-07,
+      "loss": 0.8581,
+      "step": 6807
+    },
+    {
+      "epoch": 0.9369666941921277,
+      "grad_norm": 1.8406505240881303,
+      "learning_rate": 2.076634387081766e-07,
+      "loss": 0.8853,
+      "step": 6808
+    },
+    {
+      "epoch": 0.9371043214973851,
+      "grad_norm": 1.6226107877185898,
+      "learning_rate": 2.0676073599076308e-07,
+      "loss": 0.9263,
+      "step": 6809
+    },
+    {
+      "epoch": 0.9372419488026424,
+      "grad_norm": 1.884552336688101,
+      "learning_rate": 2.058599790559368e-07,
+      "loss": 0.8819,
+      "step": 6810
+    },
+    {
+      "epoch": 0.9373795761078998,
+      "grad_norm": 1.6703037581392963,
+      "learning_rate": 2.0496116808266465e-07,
+      "loss": 0.821,
+      "step": 6811
+    },
+    {
+      "epoch": 0.9375172034131571,
+      "grad_norm": 1.6695697513911147,
+      "learning_rate": 2.040643032495304e-07,
+      "loss": 0.9441,
+      "step": 6812
+    },
+    {
+      "epoch": 0.9376548307184145,
+      "grad_norm": 1.965974359909565,
+      "learning_rate": 2.0316938473472601e-07,
+      "loss": 0.7742,
+      "step": 6813
+    },
+    {
+      "epoch": 0.937792458023672,
+      "grad_norm": 1.8621729865719097,
+      "learning_rate": 2.0227641271606147e-07,
+      "loss": 0.9026,
+      "step": 6814
+    },
+    {
+      "epoch": 0.9379300853289293,
+      "grad_norm": 2.252289326331772,
+      "learning_rate": 2.0138538737095813e-07,
+      "loss": 0.9264,
+      "step": 6815
+    },
+    {
+      "epoch": 0.9380677126341866,
+      "grad_norm": 1.7380675462760236,
+      "learning_rate": 2.004963088764489e-07,
+      "loss": 0.8842,
+      "step": 6816
+    },
+    {
+      "epoch": 0.938205339939444,
+      "grad_norm": 1.8169649370835967,
+      "learning_rate": 1.996091774091824e-07,
+      "loss": 0.8819,
+      "step": 6817
+    },
+    {
+      "epoch": 0.9383429672447013,
+      "grad_norm": 1.906652865819821,
+      "learning_rate": 1.98723993145421e-07,
+      "loss": 0.9148,
+      "step": 6818
+    },
+    {
+      "epoch": 0.9384805945499587,
+      "grad_norm": 1.6423592066571813,
+      "learning_rate": 1.9784075626103628e-07,
+      "loss": 0.8795,
+      "step": 6819
+    },
+    {
+      "epoch": 0.938618221855216,
+      "grad_norm": 1.7795390507722146,
+      "learning_rate": 1.9695946693151668e-07,
+      "loss": 0.8855,
+      "step": 6820
+    },
+    {
+      "epoch": 0.9387558491604734,
+      "grad_norm": 1.725285942699275,
+      "learning_rate": 1.9608012533196108e-07,
+      "loss": 0.9265,
+      "step": 6821
+    },
+    {
+      "epoch": 0.9388934764657308,
+      "grad_norm": 1.694721896140251,
+      "learning_rate": 1.9520273163708414e-07,
+      "loss": 0.9468,
+      "step": 6822
+    },
+    {
+      "epoch": 0.9390311037709882,
+      "grad_norm": 1.8458326033733117,
+      "learning_rate": 1.9432728602121198e-07,
+      "loss": 0.9037,
+      "step": 6823
+    },
+    {
+      "epoch": 0.9391687310762455,
+      "grad_norm": 1.802677502222908,
+      "learning_rate": 1.9345378865828212e-07,
+      "loss": 0.9096,
+      "step": 6824
+    },
+    {
+      "epoch": 0.9393063583815029,
+      "grad_norm": 1.6893783687251545,
+      "learning_rate": 1.9258223972184798e-07,
+      "loss": 0.9373,
+      "step": 6825
+    },
+    {
+      "epoch": 0.9394439856867602,
+      "grad_norm": 1.6661542399146276,
+      "learning_rate": 1.9171263938507323e-07,
+      "loss": 0.8329,
+      "step": 6826
+    },
+    {
+      "epoch": 0.9395816129920176,
+      "grad_norm": 1.87368604806992,
+      "learning_rate": 1.9084498782073746e-07,
+      "loss": 0.8011,
+      "step": 6827
+    },
+    {
+      "epoch": 0.9397192402972749,
+      "grad_norm": 2.4287459631488155,
+      "learning_rate": 1.8997928520122833e-07,
+      "loss": 0.8895,
+      "step": 6828
+    },
+    {
+      "epoch": 0.9398568676025324,
+      "grad_norm": 2.11521151978369,
+      "learning_rate": 1.8911553169855268e-07,
+      "loss": 0.928,
+      "step": 6829
+    },
+    {
+      "epoch": 0.9399944949077897,
+      "grad_norm": 1.8997163815764124,
+      "learning_rate": 1.8825372748432325e-07,
+      "loss": 0.8996,
+      "step": 6830
+    },
+    {
+      "epoch": 0.9401321222130471,
+      "grad_norm": 1.5883085310461664,
+      "learning_rate": 1.873938727297697e-07,
+      "loss": 0.9394,
+      "step": 6831
+    },
+    {
+      "epoch": 0.9402697495183044,
+      "grad_norm": 1.6952534693097505,
+      "learning_rate": 1.8653596760573434e-07,
+      "loss": 0.9016,
+      "step": 6832
+    },
+    {
+      "epoch": 0.9404073768235618,
+      "grad_norm": 1.5906014988442498,
+      "learning_rate": 1.8568001228267075e-07,
+      "loss": 0.9547,
+      "step": 6833
+    },
+    {
+      "epoch": 0.9405450041288191,
+      "grad_norm": 1.7242234500542897,
+      "learning_rate": 1.8482600693064512e-07,
+      "loss": 0.941,
+      "step": 6834
+    },
+    {
+      "epoch": 0.9406826314340765,
+      "grad_norm": 1.930076913120841,
+      "learning_rate": 1.8397395171933618e-07,
+      "loss": 0.9152,
+      "step": 6835
+    },
+    {
+      "epoch": 0.9408202587393338,
+      "grad_norm": 1.780982631877666,
+      "learning_rate": 1.831238468180363e-07,
+      "loss": 0.9401,
+      "step": 6836
+    },
+    {
+      "epoch": 0.9409578860445913,
+      "grad_norm": 1.7441465485561063,
+      "learning_rate": 1.8227569239565034e-07,
+      "loss": 0.8982,
+      "step": 6837
+    },
+    {
+      "epoch": 0.9410955133498486,
+      "grad_norm": 1.9075067610843672,
+      "learning_rate": 1.8142948862069465e-07,
+      "loss": 0.9356,
+      "step": 6838
+    },
+    {
+      "epoch": 0.941233140655106,
+      "grad_norm": 1.888594453459353,
+      "learning_rate": 1.8058523566129694e-07,
+      "loss": 0.7884,
+      "step": 6839
+    },
+    {
+      "epoch": 0.9413707679603633,
+      "grad_norm": 1.9188049500760371,
+      "learning_rate": 1.7974293368520078e-07,
+      "loss": 0.8901,
+      "step": 6840
+    },
+    {
+      "epoch": 0.9415083952656207,
+      "grad_norm": 2.0468617212023146,
+      "learning_rate": 1.789025828597579e-07,
+      "loss": 0.8479,
+      "step": 6841
+    },
+    {
+      "epoch": 0.941646022570878,
+      "grad_norm": 1.7100144286067072,
+      "learning_rate": 1.7806418335193698e-07,
+      "loss": 0.9088,
+      "step": 6842
+    },
+    {
+      "epoch": 0.9417836498761354,
+      "grad_norm": 1.7944488655465505,
+      "learning_rate": 1.7722773532831362e-07,
+      "loss": 0.8843,
+      "step": 6843
+    },
+    {
+      "epoch": 0.9419212771813927,
+      "grad_norm": 1.6462022125060438,
+      "learning_rate": 1.7639323895508153e-07,
+      "loss": 0.7841,
+      "step": 6844
+    },
+    {
+      "epoch": 0.9420589044866502,
+      "grad_norm": 1.6377588720319618,
+      "learning_rate": 1.7556069439804148e-07,
+      "loss": 0.9321,
+      "step": 6845
+    },
+    {
+      "epoch": 0.9421965317919075,
+      "grad_norm": 1.651221842504661,
+      "learning_rate": 1.7473010182260997e-07,
+      "loss": 0.9496,
+      "step": 6846
+    },
+    {
+      "epoch": 0.9423341590971649,
+      "grad_norm": 1.6424145879414012,
+      "learning_rate": 1.739014613938128e-07,
+      "loss": 0.9082,
+      "step": 6847
+    },
+    {
+      "epoch": 0.9424717864024222,
+      "grad_norm": 1.8134316529454835,
+      "learning_rate": 1.7307477327629273e-07,
+      "loss": 0.8738,
+      "step": 6848
+    },
+    {
+      "epoch": 0.9426094137076796,
+      "grad_norm": 1.6032671795878075,
+      "learning_rate": 1.7225003763429726e-07,
+      "loss": 0.8873,
+      "step": 6849
+    },
+    {
+      "epoch": 0.9427470410129369,
+      "grad_norm": 1.8897536107345827,
+      "learning_rate": 1.7142725463169307e-07,
+      "loss": 0.8788,
+      "step": 6850
+    },
+    {
+      "epoch": 0.9428846683181943,
+      "grad_norm": 1.6484570428276837,
+      "learning_rate": 1.7060642443195498e-07,
+      "loss": 0.871,
+      "step": 6851
+    },
+    {
+      "epoch": 0.9430222956234517,
+      "grad_norm": 2.713572682890576,
+      "learning_rate": 1.6978754719817025e-07,
+      "loss": 0.8783,
+      "step": 6852
+    },
+    {
+      "epoch": 0.9431599229287091,
+      "grad_norm": 1.8591786908805175,
+      "learning_rate": 1.6897062309303992e-07,
+      "loss": 0.8542,
+      "step": 6853
+    },
+    {
+      "epoch": 0.9432975502339664,
+      "grad_norm": 1.7822794633437862,
+      "learning_rate": 1.68155652278873e-07,
+      "loss": 0.8841,
+      "step": 6854
+    },
+    {
+      "epoch": 0.9434351775392238,
+      "grad_norm": 1.7357737931361794,
+      "learning_rate": 1.673426349175955e-07,
+      "loss": 0.9237,
+      "step": 6855
+    },
+    {
+      "epoch": 0.9435728048444811,
+      "grad_norm": 1.8219701820555858,
+      "learning_rate": 1.665315711707416e-07,
+      "loss": 0.9221,
+      "step": 6856
+    },
+    {
+      "epoch": 0.9437104321497385,
+      "grad_norm": 1.4587533711022453,
+      "learning_rate": 1.6572246119946012e-07,
+      "loss": 0.913,
+      "step": 6857
+    },
+    {
+      "epoch": 0.9438480594549958,
+      "grad_norm": 1.9511745537913943,
+      "learning_rate": 1.6491530516450805e-07,
+      "loss": 0.8174,
+      "step": 6858
+    },
+    {
+      "epoch": 0.9439856867602532,
+      "grad_norm": 1.93786226890944,
+      "learning_rate": 1.641101032262582e-07,
+      "loss": 0.884,
+      "step": 6859
+    },
+    {
+      "epoch": 0.9441233140655106,
+      "grad_norm": 1.7090203449006098,
+      "learning_rate": 1.6330685554469038e-07,
+      "loss": 0.8366,
+      "step": 6860
+    },
+    {
+      "epoch": 0.944260941370768,
+      "grad_norm": 1.9480260550299469,
+      "learning_rate": 1.625055622794014e-07,
+      "loss": 0.9221,
+      "step": 6861
+    },
+    {
+      "epoch": 0.9443985686760253,
+      "grad_norm": 1.7514886934050018,
+      "learning_rate": 1.617062235895972e-07,
+      "loss": 0.9019,
+      "step": 6862
+    },
+    {
+      "epoch": 0.9445361959812827,
+      "grad_norm": 1.9031940664476543,
+      "learning_rate": 1.6090883963409522e-07,
+      "loss": 0.9087,
+      "step": 6863
+    },
+    {
+      "epoch": 0.94467382328654,
+      "grad_norm": 1.599433700658534,
+      "learning_rate": 1.6011341057132312e-07,
+      "loss": 0.8505,
+      "step": 6864
+    },
+    {
+      "epoch": 0.9448114505917974,
+      "grad_norm": 1.6609440991956699,
+      "learning_rate": 1.5931993655932342e-07,
+      "loss": 0.9288,
+      "step": 6865
+    },
+    {
+      "epoch": 0.9449490778970547,
+      "grad_norm": 1.985269901145207,
+      "learning_rate": 1.585284177557489e-07,
+      "loss": 0.8587,
+      "step": 6866
+    },
+    {
+      "epoch": 0.9450867052023122,
+      "grad_norm": 1.74941215028353,
+      "learning_rate": 1.577388543178615e-07,
+      "loss": 0.9051,
+      "step": 6867
+    },
+    {
+      "epoch": 0.9452243325075695,
+      "grad_norm": 1.6527080430527137,
+      "learning_rate": 1.5695124640253801e-07,
+      "loss": 0.9933,
+      "step": 6868
+    },
+    {
+      "epoch": 0.9453619598128269,
+      "grad_norm": 1.7057586260970554,
+      "learning_rate": 1.5616559416626654e-07,
+      "loss": 0.863,
+      "step": 6869
+    },
+    {
+      "epoch": 0.9454995871180842,
+      "grad_norm": 2.021323598905727,
+      "learning_rate": 1.5538189776514223e-07,
+      "loss": 0.9071,
+      "step": 6870
+    },
+    {
+      "epoch": 0.9456372144233416,
+      "grad_norm": 1.8610819063118123,
+      "learning_rate": 1.5460015735487833e-07,
+      "loss": 0.9208,
+      "step": 6871
+    },
+    {
+      "epoch": 0.9457748417285989,
+      "grad_norm": 1.7209252707659366,
+      "learning_rate": 1.5382037309079279e-07,
+      "loss": 0.8907,
+      "step": 6872
+    },
+    {
+      "epoch": 0.9459124690338563,
+      "grad_norm": 1.7049072112952166,
+      "learning_rate": 1.5304254512782057e-07,
+      "loss": 0.8763,
+      "step": 6873
+    },
+    {
+      "epoch": 0.9460500963391136,
+      "grad_norm": 1.8114946610084695,
+      "learning_rate": 1.5226667362050473e-07,
+      "loss": 0.9097,
+      "step": 6874
+    },
+    {
+      "epoch": 0.9461877236443711,
+      "grad_norm": 1.8290428067948414,
+      "learning_rate": 1.5149275872299973e-07,
+      "loss": 0.8354,
+      "step": 6875
+    },
+    {
+      "epoch": 0.9463253509496284,
+      "grad_norm": 1.8362736926075185,
+      "learning_rate": 1.5072080058907145e-07,
+      "loss": 0.9482,
+      "step": 6876
+    },
+    {
+      "epoch": 0.9464629782548858,
+      "grad_norm": 1.7598012381585473,
+      "learning_rate": 1.4995079937209943e-07,
+      "loss": 0.8973,
+      "step": 6877
+    },
+    {
+      "epoch": 0.9466006055601431,
+      "grad_norm": 1.9346401065048422,
+      "learning_rate": 1.4918275522507018e-07,
+      "loss": 0.9604,
+      "step": 6878
+    },
+    {
+      "epoch": 0.9467382328654005,
+      "grad_norm": 1.7481026930345087,
+      "learning_rate": 1.4841666830058387e-07,
+      "loss": 0.8927,
+      "step": 6879
+    },
+    {
+      "epoch": 0.9468758601706578,
+      "grad_norm": 1.703252507385437,
+      "learning_rate": 1.4765253875085429e-07,
+      "loss": 0.8628,
+      "step": 6880
+    },
+    {
+      "epoch": 0.9470134874759152,
+      "grad_norm": 1.6467941019533234,
+      "learning_rate": 1.4689036672769886e-07,
+      "loss": 0.8061,
+      "step": 6881
+    },
+    {
+      "epoch": 0.9471511147811725,
+      "grad_norm": 1.742991072709727,
+      "learning_rate": 1.4613015238255536e-07,
+      "loss": 0.9639,
+      "step": 6882
+    },
+    {
+      "epoch": 0.94728874208643,
+      "grad_norm": 1.604378255148692,
+      "learning_rate": 1.4537189586646516e-07,
+      "loss": 0.9046,
+      "step": 6883
+    },
+    {
+      "epoch": 0.9474263693916873,
+      "grad_norm": 1.8039167060632009,
+      "learning_rate": 1.4461559733008334e-07,
+      "loss": 0.8897,
+      "step": 6884
+    },
+    {
+      "epoch": 0.9475639966969447,
+      "grad_norm": 1.5982083816414294,
+      "learning_rate": 1.4386125692367746e-07,
+      "loss": 0.9095,
+      "step": 6885
+    },
+    {
+      "epoch": 0.947701624002202,
+      "grad_norm": 1.7404152677773783,
+      "learning_rate": 1.431088747971232e-07,
+      "loss": 0.9093,
+      "step": 6886
+    },
+    {
+      "epoch": 0.9478392513074594,
+      "grad_norm": 1.6271280808373305,
+      "learning_rate": 1.4235845109990987e-07,
+      "loss": 0.8674,
+      "step": 6887
+    },
+    {
+      "epoch": 0.9479768786127167,
+      "grad_norm": 1.5477127685447425,
+      "learning_rate": 1.4160998598113595e-07,
+      "loss": 0.9662,
+      "step": 6888
+    },
+    {
+      "epoch": 0.9481145059179741,
+      "grad_norm": 1.611945638519736,
+      "learning_rate": 1.408634795895114e-07,
+      "loss": 0.9062,
+      "step": 6889
+    },
+    {
+      "epoch": 0.9482521332232315,
+      "grad_norm": 1.6921620485760902,
+      "learning_rate": 1.4011893207335536e-07,
+      "loss": 0.8839,
+      "step": 6890
+    },
+    {
+      "epoch": 0.9483897605284889,
+      "grad_norm": 1.6752551530284163,
+      "learning_rate": 1.393763435806028e-07,
+      "loss": 0.9184,
+      "step": 6891
+    },
+    {
+      "epoch": 0.9485273878337462,
+      "grad_norm": 1.6260459481826721,
+      "learning_rate": 1.386357142587913e-07,
+      "loss": 0.8516,
+      "step": 6892
+    },
+    {
+      "epoch": 0.9486650151390036,
+      "grad_norm": 1.635264907323393,
+      "learning_rate": 1.3789704425507644e-07,
+      "loss": 0.9061,
+      "step": 6893
+    },
+    {
+      "epoch": 0.9488026424442609,
+      "grad_norm": 1.7348693046302868,
+      "learning_rate": 1.3716033371622195e-07,
+      "loss": 0.8708,
+      "step": 6894
+    },
+    {
+      "epoch": 0.9489402697495183,
+      "grad_norm": 1.838946065465115,
+      "learning_rate": 1.3642558278860186e-07,
+      "loss": 0.9347,
+      "step": 6895
+    },
+    {
+      "epoch": 0.9490778970547756,
+      "grad_norm": 1.9035343401570166,
+      "learning_rate": 1.356927916182005e-07,
+      "loss": 0.9378,
+      "step": 6896
+    },
+    {
+      "epoch": 0.949215524360033,
+      "grad_norm": 1.7097838268781598,
+      "learning_rate": 1.349619603506136e-07,
+      "loss": 0.8581,
+      "step": 6897
+    },
+    {
+      "epoch": 0.9493531516652904,
+      "grad_norm": 1.9035869275975081,
+      "learning_rate": 1.3423308913104726e-07,
+      "loss": 0.8088,
+      "step": 6898
+    },
+    {
+      "epoch": 0.9494907789705478,
+      "grad_norm": 1.768536041605859,
+      "learning_rate": 1.3350617810431898e-07,
+      "loss": 0.8914,
+      "step": 6899
+    },
+    {
+      "epoch": 0.9496284062758051,
+      "grad_norm": 1.5628857172004424,
+      "learning_rate": 1.3278122741485544e-07,
+      "loss": 0.9246,
+      "step": 6900
+    },
+    {
+      "epoch": 0.9497660335810625,
+      "grad_norm": 1.7206375989574216,
+      "learning_rate": 1.3205823720669474e-07,
+      "loss": 0.8669,
+      "step": 6901
+    },
+    {
+      "epoch": 0.9499036608863198,
+      "grad_norm": 1.9548474546284382,
+      "learning_rate": 1.3133720762348535e-07,
+      "loss": 0.8804,
+      "step": 6902
+    },
+    {
+      "epoch": 0.9500412881915772,
+      "grad_norm": 1.799202762265101,
+      "learning_rate": 1.3061813880848483e-07,
+      "loss": 0.8902,
+      "step": 6903
+    },
+    {
+      "epoch": 0.9501789154968345,
+      "grad_norm": 1.957096562528851,
+      "learning_rate": 1.299010309045623e-07,
+      "loss": 0.8651,
+      "step": 6904
+    },
+    {
+      "epoch": 0.950316542802092,
+      "grad_norm": 1.722148664865578,
+      "learning_rate": 1.291858840541993e-07,
+      "loss": 0.8284,
+      "step": 6905
+    },
+    {
+      "epoch": 0.9504541701073493,
+      "grad_norm": 1.6793670397722704,
+      "learning_rate": 1.284726983994833e-07,
+      "loss": 0.8894,
+      "step": 6906
+    },
+    {
+      "epoch": 0.9505917974126067,
+      "grad_norm": 1.5468472951583538,
+      "learning_rate": 1.2776147408211648e-07,
+      "loss": 0.9088,
+      "step": 6907
+    },
+    {
+      "epoch": 0.950729424717864,
+      "grad_norm": 1.767197458812727,
+      "learning_rate": 1.2705221124340806e-07,
+      "loss": 0.888,
+      "step": 6908
+    },
+    {
+      "epoch": 0.9508670520231214,
+      "grad_norm": 1.8940533878713974,
+      "learning_rate": 1.2634491002427972e-07,
+      "loss": 0.8935,
+      "step": 6909
+    },
+    {
+      "epoch": 0.9510046793283787,
+      "grad_norm": 2.0538176723179764,
+      "learning_rate": 1.2563957056526244e-07,
+      "loss": 0.8878,
+      "step": 6910
+    },
+    {
+      "epoch": 0.9511423066336361,
+      "grad_norm": 1.5601413797606019,
+      "learning_rate": 1.2493619300649628e-07,
+      "loss": 0.9163,
+      "step": 6911
+    },
+    {
+      "epoch": 0.9512799339388934,
+      "grad_norm": 1.7847668279717401,
+      "learning_rate": 1.2423477748773394e-07,
+      "loss": 0.9424,
+      "step": 6912
+    },
+    {
+      "epoch": 0.9514175612441509,
+      "grad_norm": 1.6621132179279845,
+      "learning_rate": 1.2353532414833724e-07,
+      "loss": 0.9495,
+      "step": 6913
+    },
+    {
+      "epoch": 0.9515551885494082,
+      "grad_norm": 1.7643640954635127,
+      "learning_rate": 1.2283783312727727e-07,
+      "loss": 0.8714,
+      "step": 6914
+    },
+    {
+      "epoch": 0.9516928158546656,
+      "grad_norm": 1.6621149191507572,
+      "learning_rate": 1.2214230456313646e-07,
+      "loss": 0.8449,
+      "step": 6915
+    },
+    {
+      "epoch": 0.9518304431599229,
+      "grad_norm": 2.3067853768737465,
+      "learning_rate": 1.2144873859410655e-07,
+      "loss": 0.91,
+      "step": 6916
+    },
+    {
+      "epoch": 0.9519680704651803,
+      "grad_norm": 1.8187932142648133,
+      "learning_rate": 1.207571353579895e-07,
+      "loss": 0.9583,
+      "step": 6917
+    },
+    {
+      "epoch": 0.9521056977704376,
+      "grad_norm": 1.7698829258459328,
+      "learning_rate": 1.2006749499219762e-07,
+      "loss": 0.9204,
+      "step": 6918
+    },
+    {
+      "epoch": 0.952243325075695,
+      "grad_norm": 1.8413161658003894,
+      "learning_rate": 1.193798176337535e-07,
+      "loss": 0.907,
+      "step": 6919
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 1.8226692976326315,
+      "learning_rate": 1.1869410341928678e-07,
+      "loss": 0.8351,
+      "step": 6920
+    },
+    {
+      "epoch": 0.9525185796862098,
+      "grad_norm": 1.677338772043695,
+      "learning_rate": 1.18010352485044e-07,
+      "loss": 0.9658,
+      "step": 6921
+    },
+    {
+      "epoch": 0.9526562069914671,
+      "grad_norm": 1.8381883995940593,
+      "learning_rate": 1.1732856496687207e-07,
+      "loss": 0.9598,
+      "step": 6922
+    },
+    {
+      "epoch": 0.9527938342967245,
+      "grad_norm": 2.175476063743664,
+      "learning_rate": 1.1664874100023704e-07,
+      "loss": 0.9099,
+      "step": 6923
+    },
+    {
+      "epoch": 0.9529314616019818,
+      "grad_norm": 1.7078863106447115,
+      "learning_rate": 1.1597088072020757e-07,
+      "loss": 0.9198,
+      "step": 6924
+    },
+    {
+      "epoch": 0.9530690889072392,
+      "grad_norm": 1.801628346378403,
+      "learning_rate": 1.1529498426146701e-07,
+      "loss": 0.9548,
+      "step": 6925
+    },
+    {
+      "epoch": 0.9532067162124965,
+      "grad_norm": 1.8010725461775434,
+      "learning_rate": 1.1462105175830685e-07,
+      "loss": 0.8791,
+      "step": 6926
+    },
+    {
+      "epoch": 0.9533443435177539,
+      "grad_norm": 1.7402721188306143,
+      "learning_rate": 1.1394908334462662e-07,
+      "loss": 0.8926,
+      "step": 6927
+    },
+    {
+      "epoch": 0.9534819708230113,
+      "grad_norm": 1.8363826000219154,
+      "learning_rate": 1.1327907915393843e-07,
+      "loss": 0.8772,
+      "step": 6928
+    },
+    {
+      "epoch": 0.9536195981282687,
+      "grad_norm": 1.8741681065445746,
+      "learning_rate": 1.1261103931936357e-07,
+      "loss": 0.872,
+      "step": 6929
+    },
+    {
+      "epoch": 0.953757225433526,
+      "grad_norm": 1.7090532301161931,
+      "learning_rate": 1.1194496397363142e-07,
+      "loss": 0.868,
+      "step": 6930
+    },
+    {
+      "epoch": 0.9538948527387834,
+      "grad_norm": 1.9148420407390783,
+      "learning_rate": 1.1128085324908167e-07,
+      "loss": 0.917,
+      "step": 6931
+    },
+    {
+      "epoch": 0.9540324800440407,
+      "grad_norm": 1.5286601520457046,
+      "learning_rate": 1.1061870727766432e-07,
+      "loss": 0.9382,
+      "step": 6932
+    },
+    {
+      "epoch": 0.9541701073492981,
+      "grad_norm": 1.6787249411567307,
+      "learning_rate": 1.0995852619093972e-07,
+      "loss": 0.9363,
+      "step": 6933
+    },
+    {
+      "epoch": 0.9543077346545554,
+      "grad_norm": 1.737895652636811,
+      "learning_rate": 1.0930031012007513e-07,
+      "loss": 0.898,
+      "step": 6934
+    },
+    {
+      "epoch": 0.9544453619598128,
+      "grad_norm": 1.6957826630059414,
+      "learning_rate": 1.0864405919585042e-07,
+      "loss": 0.8123,
+      "step": 6935
+    },
+    {
+      "epoch": 0.9545829892650702,
+      "grad_norm": 1.7927120064184359,
+      "learning_rate": 1.0798977354865348e-07,
+      "loss": 0.9282,
+      "step": 6936
+    },
+    {
+      "epoch": 0.9547206165703276,
+      "grad_norm": 1.7648377881144928,
+      "learning_rate": 1.0733745330848144e-07,
+      "loss": 0.9037,
+      "step": 6937
+    },
+    {
+      "epoch": 0.9548582438755849,
+      "grad_norm": 1.62374679435468,
+      "learning_rate": 1.0668709860494175e-07,
+      "loss": 0.8616,
+      "step": 6938
+    },
+    {
+      "epoch": 0.9549958711808423,
+      "grad_norm": 1.7772697920250673,
+      "learning_rate": 1.0603870956725105e-07,
+      "loss": 0.871,
+      "step": 6939
+    },
+    {
+      "epoch": 0.9551334984860996,
+      "grad_norm": 1.6234404492675212,
+      "learning_rate": 1.0539228632423515e-07,
+      "loss": 0.866,
+      "step": 6940
+    },
+    {
+      "epoch": 0.955271125791357,
+      "grad_norm": 2.952890658428871,
+      "learning_rate": 1.0474782900432911e-07,
+      "loss": 0.9761,
+      "step": 6941
+    },
+    {
+      "epoch": 0.9554087530966143,
+      "grad_norm": 1.7791744873658384,
+      "learning_rate": 1.041053377355783e-07,
+      "loss": 0.9494,
+      "step": 6942
+    },
+    {
+      "epoch": 0.9555463804018718,
+      "grad_norm": 1.779703124382904,
+      "learning_rate": 1.0346481264563723e-07,
+      "loss": 0.9354,
+      "step": 6943
+    },
+    {
+      "epoch": 0.9556840077071291,
+      "grad_norm": 1.8196610948124943,
+      "learning_rate": 1.0282625386176858e-07,
+      "loss": 0.9467,
+      "step": 6944
+    },
+    {
+      "epoch": 0.9558216350123865,
+      "grad_norm": 1.6573816023128902,
+      "learning_rate": 1.0218966151084531e-07,
+      "loss": 0.8288,
+      "step": 6945
+    },
+    {
+      "epoch": 0.9559592623176438,
+      "grad_norm": 1.8216858888705667,
+      "learning_rate": 1.0155503571935176e-07,
+      "loss": 0.9409,
+      "step": 6946
+    },
+    {
+      "epoch": 0.9560968896229012,
+      "grad_norm": 1.7303447376220138,
+      "learning_rate": 1.0092237661337712e-07,
+      "loss": 0.9139,
+      "step": 6947
+    },
+    {
+      "epoch": 0.9562345169281585,
+      "grad_norm": 1.7605000367523342,
+      "learning_rate": 1.002916843186208e-07,
+      "loss": 0.8481,
+      "step": 6948
+    },
+    {
+      "epoch": 0.9563721442334159,
+      "grad_norm": 1.7386245902049096,
+      "learning_rate": 9.966295896039702e-08,
+      "loss": 0.9163,
+      "step": 6949
+    },
+    {
+      "epoch": 0.9565097715386732,
+      "grad_norm": 1.7256887638584524,
+      "learning_rate": 9.90362006636203e-08,
+      "loss": 0.9646,
+      "step": 6950
+    },
+    {
+      "epoch": 0.9566473988439307,
+      "grad_norm": 1.9453443968185502,
+      "learning_rate": 9.841140955282325e-08,
+      "loss": 0.8962,
+      "step": 6951
+    },
+    {
+      "epoch": 0.956785026149188,
+      "grad_norm": 1.7132711639508447,
+      "learning_rate": 9.778858575213989e-08,
+      "loss": 0.9286,
+      "step": 6952
+    },
+    {
+      "epoch": 0.9569226534544454,
+      "grad_norm": 1.6954225061418475,
+      "learning_rate": 9.716772938531793e-08,
+      "loss": 0.926,
+      "step": 6953
+    },
+    {
+      "epoch": 0.9570602807597027,
+      "grad_norm": 1.8555176837014946,
+      "learning_rate": 9.654884057571312e-08,
+      "loss": 0.8675,
+      "step": 6954
+    },
+    {
+      "epoch": 0.9571979080649601,
+      "grad_norm": 1.6174415267741171,
+      "learning_rate": 9.593191944629044e-08,
+      "loss": 0.9499,
+      "step": 6955
+    },
+    {
+      "epoch": 0.9573355353702174,
+      "grad_norm": 1.6253734363025731,
+      "learning_rate": 9.531696611962404e-08,
+      "loss": 0.914,
+      "step": 6956
+    },
+    {
+      "epoch": 0.9574731626754748,
+      "grad_norm": 1.7932315131048848,
+      "learning_rate": 9.47039807178951e-08,
+      "loss": 0.8301,
+      "step": 6957
+    },
+    {
+      "epoch": 0.9576107899807321,
+      "grad_norm": 1.8963226926498309,
+      "learning_rate": 9.409296336289731e-08,
+      "loss": 0.9464,
+      "step": 6958
+    },
+    {
+      "epoch": 0.9577484172859896,
+      "grad_norm": 1.8406455576947864,
+      "learning_rate": 9.34839141760302e-08,
+      "loss": 0.8934,
+      "step": 6959
+    },
+    {
+      "epoch": 0.9578860445912469,
+      "grad_norm": 1.7782969157227637,
+      "learning_rate": 9.287683327830365e-08,
+      "loss": 0.8568,
+      "step": 6960
+    },
+    {
+      "epoch": 0.9580236718965043,
+      "grad_norm": 1.8705540028502288,
+      "learning_rate": 9.227172079033675e-08,
+      "loss": 0.9157,
+      "step": 6961
+    },
+    {
+      "epoch": 0.9581612992017616,
+      "grad_norm": 2.284878545650477,
+      "learning_rate": 9.166857683235775e-08,
+      "loss": 0.8699,
+      "step": 6962
+    },
+    {
+      "epoch": 0.958298926507019,
+      "grad_norm": 1.648912133150703,
+      "learning_rate": 9.10674015242019e-08,
+      "loss": 0.8736,
+      "step": 6963
+    },
+    {
+      "epoch": 0.9584365538122763,
+      "grad_norm": 1.804291699317642,
+      "learning_rate": 9.046819498531367e-08,
+      "loss": 0.866,
+      "step": 6964
+    },
+    {
+      "epoch": 0.9585741811175337,
+      "grad_norm": 1.7908943194680291,
+      "learning_rate": 8.987095733475004e-08,
+      "loss": 0.7528,
+      "step": 6965
+    },
+    {
+      "epoch": 0.9587118084227911,
+      "grad_norm": 1.7643524380595477,
+      "learning_rate": 8.927568869117165e-08,
+      "loss": 0.8525,
+      "step": 6966
+    },
+    {
+      "epoch": 0.9588494357280485,
+      "grad_norm": 1.5671598855667779,
+      "learning_rate": 8.868238917285054e-08,
+      "loss": 0.9166,
+      "step": 6967
+    },
+    {
+      "epoch": 0.9589870630333058,
+      "grad_norm": 2.2313684624255723,
+      "learning_rate": 8.8091058897668e-08,
+      "loss": 0.8821,
+      "step": 6968
+    },
+    {
+      "epoch": 0.9591246903385632,
+      "grad_norm": 1.8898481906538556,
+      "learning_rate": 8.750169798311226e-08,
+      "loss": 0.909,
+      "step": 6969
+    },
+    {
+      "epoch": 0.9592623176438205,
+      "grad_norm": 1.8090622495547521,
+      "learning_rate": 8.691430654628186e-08,
+      "loss": 0.8543,
+      "step": 6970
+    },
+    {
+      "epoch": 0.9593999449490779,
+      "grad_norm": 1.7445148777120731,
+      "learning_rate": 8.632888470388345e-08,
+      "loss": 0.8649,
+      "step": 6971
+    },
+    {
+      "epoch": 0.9595375722543352,
+      "grad_norm": 2.031451180271669,
+      "learning_rate": 8.574543257223289e-08,
+      "loss": 0.8948,
+      "step": 6972
+    },
+    {
+      "epoch": 0.9596751995595926,
+      "grad_norm": 1.7517091162860967,
+      "learning_rate": 8.516395026725299e-08,
+      "loss": 0.8267,
+      "step": 6973
+    },
+    {
+      "epoch": 0.95981282686485,
+      "grad_norm": 1.8254721777825589,
+      "learning_rate": 8.45844379044769e-08,
+      "loss": 0.8579,
+      "step": 6974
+    },
+    {
+      "epoch": 0.9599504541701074,
+      "grad_norm": 1.8825797925639514,
+      "learning_rate": 8.400689559904585e-08,
+      "loss": 0.8035,
+      "step": 6975
+    },
+    {
+      "epoch": 0.9600880814753647,
+      "grad_norm": 2.073357047620673,
+      "learning_rate": 8.343132346571025e-08,
+      "loss": 0.8853,
+      "step": 6976
+    },
+    {
+      "epoch": 0.9602257087806221,
+      "grad_norm": 1.8073888233335662,
+      "learning_rate": 8.285772161882644e-08,
+      "loss": 0.8852,
+      "step": 6977
+    },
+    {
+      "epoch": 0.9603633360858794,
+      "grad_norm": 1.7165820075084854,
+      "learning_rate": 8.228609017236211e-08,
+      "loss": 0.8977,
+      "step": 6978
+    },
+    {
+      "epoch": 0.9605009633911368,
+      "grad_norm": 1.9913822552626756,
+      "learning_rate": 8.17164292398942e-08,
+      "loss": 0.9043,
+      "step": 6979
+    },
+    {
+      "epoch": 0.9606385906963941,
+      "grad_norm": 2.1315839148885685,
+      "learning_rate": 8.11487389346044e-08,
+      "loss": 0.872,
+      "step": 6980
+    },
+    {
+      "epoch": 0.9607762180016516,
+      "grad_norm": 2.0705572994385975,
+      "learning_rate": 8.058301936928691e-08,
+      "loss": 0.8246,
+      "step": 6981
+    },
+    {
+      "epoch": 0.9609138453069089,
+      "grad_norm": 1.8625344413489078,
+      "learning_rate": 8.001927065634075e-08,
+      "loss": 0.9051,
+      "step": 6982
+    },
+    {
+      "epoch": 0.9610514726121663,
+      "grad_norm": 1.5114800185758899,
+      "learning_rate": 7.945749290777627e-08,
+      "loss": 0.8641,
+      "step": 6983
+    },
+    {
+      "epoch": 0.9611890999174236,
+      "grad_norm": 2.023599806503092,
+      "learning_rate": 7.889768623520977e-08,
+      "loss": 0.9091,
+      "step": 6984
+    },
+    {
+      "epoch": 0.961326727222681,
+      "grad_norm": 1.618965657476201,
+      "learning_rate": 7.833985074986782e-08,
+      "loss": 0.8833,
+      "step": 6985
+    },
+    {
+      "epoch": 0.9614643545279383,
+      "grad_norm": 1.6148255733306132,
+      "learning_rate": 7.778398656258624e-08,
+      "loss": 0.9217,
+      "step": 6986
+    },
+    {
+      "epoch": 0.9616019818331957,
+      "grad_norm": 1.9215345961213792,
+      "learning_rate": 7.723009378380442e-08,
+      "loss": 0.8955,
+      "step": 6987
+    },
+    {
+      "epoch": 0.961739609138453,
+      "grad_norm": 1.7833526329607186,
+      "learning_rate": 7.667817252357656e-08,
+      "loss": 0.9959,
+      "step": 6988
+    },
+    {
+      "epoch": 0.9618772364437105,
+      "grad_norm": 1.8493407618206463,
+      "learning_rate": 7.612822289155941e-08,
+      "loss": 0.9831,
+      "step": 6989
+    },
+    {
+      "epoch": 0.9620148637489678,
+      "grad_norm": 1.629222252590816,
+      "learning_rate": 7.55802449970211e-08,
+      "loss": 0.8592,
+      "step": 6990
+    },
+    {
+      "epoch": 0.9621524910542252,
+      "grad_norm": 1.995894256611542,
+      "learning_rate": 7.503423894883899e-08,
+      "loss": 0.9448,
+      "step": 6991
+    },
+    {
+      "epoch": 0.9622901183594825,
+      "grad_norm": 2.0485114804737243,
+      "learning_rate": 7.449020485549408e-08,
+      "loss": 0.8779,
+      "step": 6992
+    },
+    {
+      "epoch": 0.9624277456647399,
+      "grad_norm": 1.6756352730392277,
+      "learning_rate": 7.394814282508101e-08,
+      "loss": 0.8326,
+      "step": 6993
+    },
+    {
+      "epoch": 0.9625653729699972,
+      "grad_norm": 1.5076506885932646,
+      "learning_rate": 7.340805296529918e-08,
+      "loss": 0.9111,
+      "step": 6994
+    },
+    {
+      "epoch": 0.9627030002752546,
+      "grad_norm": 1.7752676933032443,
+      "learning_rate": 7.286993538345611e-08,
+      "loss": 0.9274,
+      "step": 6995
+    },
+    {
+      "epoch": 0.9628406275805119,
+      "grad_norm": 1.6653512229654122,
+      "learning_rate": 7.233379018646958e-08,
+      "loss": 0.8586,
+      "step": 6996
+    },
+    {
+      "epoch": 0.9629782548857694,
+      "grad_norm": 1.6178332517512912,
+      "learning_rate": 7.17996174808644e-08,
+      "loss": 0.8028,
+      "step": 6997
+    },
+    {
+      "epoch": 0.9631158821910267,
+      "grad_norm": 1.644701178168763,
+      "learning_rate": 7.126741737277343e-08,
+      "loss": 0.8919,
+      "step": 6998
+    },
+    {
+      "epoch": 0.9632535094962841,
+      "grad_norm": 1.6406366587468344,
+      "learning_rate": 7.073718996793544e-08,
+      "loss": 0.908,
+      "step": 6999
+    },
+    {
+      "epoch": 0.9633911368015414,
+      "grad_norm": 1.6196882894827915,
+      "learning_rate": 7.020893537170281e-08,
+      "loss": 0.9098,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9635287641067988,
+      "grad_norm": 1.764158510497846,
+      "learning_rate": 6.968265368902938e-08,
+      "loss": 0.8192,
+      "step": 7001
+    },
+    {
+      "epoch": 0.9636663914120561,
+      "grad_norm": 1.783624002239502,
+      "learning_rate": 6.91583450244826e-08,
+      "loss": 0.8553,
+      "step": 7002
+    },
+    {
+      "epoch": 0.9638040187173135,
+      "grad_norm": 1.7430201499350384,
+      "learning_rate": 6.863600948223248e-08,
+      "loss": 0.8184,
+      "step": 7003
+    },
+    {
+      "epoch": 0.9639416460225709,
+      "grad_norm": 1.6454871583395467,
+      "learning_rate": 6.811564716606378e-08,
+      "loss": 0.9221,
+      "step": 7004
+    },
+    {
+      "epoch": 0.9640792733278283,
+      "grad_norm": 1.6234864712663126,
+      "learning_rate": 6.759725817936158e-08,
+      "loss": 0.9052,
+      "step": 7005
+    },
+    {
+      "epoch": 0.9642169006330856,
+      "grad_norm": 1.8693921712753214,
+      "learning_rate": 6.708084262512571e-08,
+      "loss": 0.9744,
+      "step": 7006
+    },
+    {
+      "epoch": 0.964354527938343,
+      "grad_norm": 1.770293438978404,
+      "learning_rate": 6.656640060595854e-08,
+      "loss": 0.9326,
+      "step": 7007
+    },
+    {
+      "epoch": 0.9644921552436003,
+      "grad_norm": 1.6874844657827082,
+      "learning_rate": 6.605393222407496e-08,
+      "loss": 0.9154,
+      "step": 7008
+    },
+    {
+      "epoch": 0.9646297825488577,
+      "grad_norm": 1.567401810734057,
+      "learning_rate": 6.554343758129244e-08,
+      "loss": 0.9151,
+      "step": 7009
+    },
+    {
+      "epoch": 0.964767409854115,
+      "grad_norm": 1.799476901678856,
+      "learning_rate": 6.503491677904206e-08,
+      "loss": 1.002,
+      "step": 7010
+    },
+    {
+      "epoch": 0.9649050371593724,
+      "grad_norm": 1.9390278608736544,
+      "learning_rate": 6.452836991835964e-08,
+      "loss": 0.9198,
+      "step": 7011
+    },
+    {
+      "epoch": 0.9650426644646298,
+      "grad_norm": 1.7870153168920437,
+      "learning_rate": 6.402379709988693e-08,
+      "loss": 0.9473,
+      "step": 7012
+    },
+    {
+      "epoch": 0.9651802917698872,
+      "grad_norm": 1.9248514625385063,
+      "learning_rate": 6.352119842387816e-08,
+      "loss": 0.9746,
+      "step": 7013
+    },
+    {
+      "epoch": 0.9653179190751445,
+      "grad_norm": 2.1967987735886494,
+      "learning_rate": 6.302057399019012e-08,
+      "loss": 0.9037,
+      "step": 7014
+    },
+    {
+      "epoch": 0.9654555463804019,
+      "grad_norm": 2.6477680284718708,
+      "learning_rate": 6.252192389829326e-08,
+      "loss": 0.9229,
+      "step": 7015
+    },
+    {
+      "epoch": 0.9655931736856592,
+      "grad_norm": 2.159799236518315,
+      "learning_rate": 6.202524824725941e-08,
+      "loss": 0.8118,
+      "step": 7016
+    },
+    {
+      "epoch": 0.9657308009909166,
+      "grad_norm": 1.9141925848759096,
+      "learning_rate": 6.1530547135773e-08,
+      "loss": 0.798,
+      "step": 7017
+    },
+    {
+      "epoch": 0.9658684282961739,
+      "grad_norm": 2.0899251558882392,
+      "learning_rate": 6.103782066212427e-08,
+      "loss": 0.899,
+      "step": 7018
+    },
+    {
+      "epoch": 0.9660060556014314,
+      "grad_norm": 1.6179207512221074,
+      "learning_rate": 6.054706892421159e-08,
+      "loss": 0.9552,
+      "step": 7019
+    },
+    {
+      "epoch": 0.9661436829066887,
+      "grad_norm": 1.959348144592709,
+      "learning_rate": 6.005829201954028e-08,
+      "loss": 0.9204,
+      "step": 7020
+    },
+    {
+      "epoch": 0.9662813102119461,
+      "grad_norm": 1.7572820185439972,
+      "learning_rate": 5.957149004522267e-08,
+      "loss": 0.8896,
+      "step": 7021
+    },
+    {
+      "epoch": 0.9664189375172034,
+      "grad_norm": 1.5475420228345198,
+      "learning_rate": 5.908666309798028e-08,
+      "loss": 0.8669,
+      "step": 7022
+    },
+    {
+      "epoch": 0.9665565648224608,
+      "grad_norm": 1.7126660355640158,
+      "learning_rate": 5.860381127414161e-08,
+      "loss": 0.8697,
+      "step": 7023
+    },
+    {
+      "epoch": 0.9666941921277181,
+      "grad_norm": 1.627670364016188,
+      "learning_rate": 5.812293466964325e-08,
+      "loss": 0.8701,
+      "step": 7024
+    },
+    {
+      "epoch": 0.9668318194329755,
+      "grad_norm": 1.657946058586508,
+      "learning_rate": 5.764403338002766e-08,
+      "loss": 0.8259,
+      "step": 7025
+    },
+    {
+      "epoch": 0.9669694467382328,
+      "grad_norm": 1.6565077518102114,
+      "learning_rate": 5.7167107500446517e-08,
+      "loss": 0.8696,
+      "step": 7026
+    },
+    {
+      "epoch": 0.9671070740434903,
+      "grad_norm": 1.9254474491903764,
+      "learning_rate": 5.6692157125659564e-08,
+      "loss": 0.8874,
+      "step": 7027
+    },
+    {
+      "epoch": 0.9672447013487476,
+      "grad_norm": 1.910671640416932,
+      "learning_rate": 5.6219182350031323e-08,
+      "loss": 0.9159,
+      "step": 7028
+    },
+    {
+      "epoch": 0.967382328654005,
+      "grad_norm": 1.7187620750712238,
+      "learning_rate": 5.574818326753551e-08,
+      "loss": 0.8941,
+      "step": 7029
+    },
+    {
+      "epoch": 0.9675199559592623,
+      "grad_norm": 1.9654077114318835,
+      "learning_rate": 5.527915997175504e-08,
+      "loss": 0.9295,
+      "step": 7030
+    },
+    {
+      "epoch": 0.9676575832645197,
+      "grad_norm": 1.950631943657173,
+      "learning_rate": 5.4812112555876486e-08,
+      "loss": 0.9025,
+      "step": 7031
+    },
+    {
+      "epoch": 0.967795210569777,
+      "grad_norm": 1.8138899446435934,
+      "learning_rate": 5.434704111269562e-08,
+      "loss": 0.8916,
+      "step": 7032
+    },
+    {
+      "epoch": 0.9679328378750344,
+      "grad_norm": 1.497912686583352,
+      "learning_rate": 5.38839457346163e-08,
+      "loss": 0.8431,
+      "step": 7033
+    },
+    {
+      "epoch": 0.9680704651802917,
+      "grad_norm": 1.8072146789943926,
+      "learning_rate": 5.3422826513651604e-08,
+      "loss": 0.8971,
+      "step": 7034
+    },
+    {
+      "epoch": 0.9682080924855492,
+      "grad_norm": 1.857030784314519,
+      "learning_rate": 5.29636835414149e-08,
+      "loss": 0.8701,
+      "step": 7035
+    },
+    {
+      "epoch": 0.9683457197908065,
+      "grad_norm": 1.7185076013725726,
+      "learning_rate": 5.250651690913655e-08,
+      "loss": 0.9123,
+      "step": 7036
+    },
+    {
+      "epoch": 0.9684833470960639,
+      "grad_norm": 1.8004951915791596,
+      "learning_rate": 5.205132670764501e-08,
+      "loss": 0.9014,
+      "step": 7037
+    },
+    {
+      "epoch": 0.9686209744013212,
+      "grad_norm": 1.7832089529766129,
+      "learning_rate": 5.159811302738349e-08,
+      "loss": 0.8038,
+      "step": 7038
+    },
+    {
+      "epoch": 0.9687586017065786,
+      "grad_norm": 1.607315490247505,
+      "learning_rate": 5.1146875958397733e-08,
+      "loss": 0.8913,
+      "step": 7039
+    },
+    {
+      "epoch": 0.9688962290118359,
+      "grad_norm": 1.6073677551355436,
+      "learning_rate": 5.06976155903427e-08,
+      "loss": 0.9062,
+      "step": 7040
+    },
+    {
+      "epoch": 0.9690338563170933,
+      "grad_norm": 1.7291127339334689,
+      "learning_rate": 5.0250332012480306e-08,
+      "loss": 0.9029,
+      "step": 7041
+    },
+    {
+      "epoch": 0.9691714836223507,
+      "grad_norm": 2.2705236371704043,
+      "learning_rate": 4.980502531367948e-08,
+      "loss": 0.8555,
+      "step": 7042
+    },
+    {
+      "epoch": 0.9693091109276081,
+      "grad_norm": 1.6377248069593295,
+      "learning_rate": 4.936169558241832e-08,
+      "loss": 0.9233,
+      "step": 7043
+    },
+    {
+      "epoch": 0.9694467382328654,
+      "grad_norm": 1.6983329803823561,
+      "learning_rate": 4.8920342906776385e-08,
+      "loss": 0.9009,
+      "step": 7044
+    },
+    {
+      "epoch": 0.9695843655381228,
+      "grad_norm": 1.930205798657553,
+      "learning_rate": 4.8480967374449074e-08,
+      "loss": 0.8876,
+      "step": 7045
+    },
+    {
+      "epoch": 0.9697219928433801,
+      "grad_norm": 1.6604617054859634,
+      "learning_rate": 4.8043569072731e-08,
+      "loss": 0.8088,
+      "step": 7046
+    },
+    {
+      "epoch": 0.9698596201486375,
+      "grad_norm": 1.6435404385680668,
+      "learning_rate": 4.7608148088528204e-08,
+      "loss": 0.9228,
+      "step": 7047
+    },
+    {
+      "epoch": 0.9699972474538948,
+      "grad_norm": 1.878680655909647,
+      "learning_rate": 4.717470450835371e-08,
+      "loss": 0.9732,
+      "step": 7048
+    },
+    {
+      "epoch": 0.9701348747591522,
+      "grad_norm": 1.7178339084567391,
+      "learning_rate": 4.674323841832529e-08,
+      "loss": 0.8686,
+      "step": 7049
+    },
+    {
+      "epoch": 0.9702725020644096,
+      "grad_norm": 1.9310697815721647,
+      "learning_rate": 4.6313749904171033e-08,
+      "loss": 0.8774,
+      "step": 7050
+    },
+    {
+      "epoch": 0.970410129369667,
+      "grad_norm": 1.7564715812605578,
+      "learning_rate": 4.5886239051222694e-08,
+      "loss": 0.8046,
+      "step": 7051
+    },
+    {
+      "epoch": 0.9705477566749243,
+      "grad_norm": 1.7360743681770483,
+      "learning_rate": 4.5460705944422315e-08,
+      "loss": 0.863,
+      "step": 7052
+    },
+    {
+      "epoch": 0.9706853839801817,
+      "grad_norm": 1.732559229584392,
+      "learning_rate": 4.503715066831671e-08,
+      "loss": 0.8172,
+      "step": 7053
+    },
+    {
+      "epoch": 0.970823011285439,
+      "grad_norm": 1.9029916280153099,
+      "learning_rate": 4.461557330705968e-08,
+      "loss": 0.9006,
+      "step": 7054
+    },
+    {
+      "epoch": 0.9709606385906964,
+      "grad_norm": 1.7351383194836687,
+      "learning_rate": 4.419597394441533e-08,
+      "loss": 0.889,
+      "step": 7055
+    },
+    {
+      "epoch": 0.9710982658959537,
+      "grad_norm": 1.9810484868613516,
+      "learning_rate": 4.37783526637503e-08,
+      "loss": 0.8931,
+      "step": 7056
+    },
+    {
+      "epoch": 0.9712358932012112,
+      "grad_norm": 1.6340388016239356,
+      "learning_rate": 4.336270954804045e-08,
+      "loss": 0.8946,
+      "step": 7057
+    },
+    {
+      "epoch": 0.9713735205064685,
+      "grad_norm": 2.814674002625467,
+      "learning_rate": 4.2949044679868603e-08,
+      "loss": 0.8515,
+      "step": 7058
+    },
+    {
+      "epoch": 0.9715111478117259,
+      "grad_norm": 1.670225065981339,
+      "learning_rate": 4.253735814142568e-08,
+      "loss": 0.9069,
+      "step": 7059
+    },
+    {
+      "epoch": 0.9716487751169832,
+      "grad_norm": 1.707548224833195,
+      "learning_rate": 4.212765001450514e-08,
+      "loss": 0.9165,
+      "step": 7060
+    },
+    {
+      "epoch": 0.9717864024222406,
+      "grad_norm": 1.9274105244583501,
+      "learning_rate": 4.171992038051298e-08,
+      "loss": 0.7479,
+      "step": 7061
+    },
+    {
+      "epoch": 0.9719240297274979,
+      "grad_norm": 1.6990494038235444,
+      "learning_rate": 4.131416932045884e-08,
+      "loss": 0.81,
+      "step": 7062
+    },
+    {
+      "epoch": 0.9720616570327553,
+      "grad_norm": 1.779826395743542,
+      "learning_rate": 4.091039691495935e-08,
+      "loss": 0.9026,
+      "step": 7063
+    },
+    {
+      "epoch": 0.9721992843380126,
+      "grad_norm": 1.5873756799538337,
+      "learning_rate": 4.050860324423922e-08,
+      "loss": 0.8841,
+      "step": 7064
+    },
+    {
+      "epoch": 0.9723369116432701,
+      "grad_norm": 1.7713466072247923,
+      "learning_rate": 4.0108788388129036e-08,
+      "loss": 0.9259,
+      "step": 7065
+    },
+    {
+      "epoch": 0.9724745389485274,
+      "grad_norm": 1.562599255512465,
+      "learning_rate": 3.9710952426066376e-08,
+      "loss": 0.834,
+      "step": 7066
+    },
+    {
+      "epoch": 0.9726121662537848,
+      "grad_norm": 1.7775577202185444,
+      "learning_rate": 3.9315095437095774e-08,
+      "loss": 0.8822,
+      "step": 7067
+    },
+    {
+      "epoch": 0.9727497935590421,
+      "grad_norm": 1.8095496986083404,
+      "learning_rate": 3.892121749986988e-08,
+      "loss": 0.8947,
+      "step": 7068
+    },
+    {
+      "epoch": 0.9728874208642995,
+      "grad_norm": 1.9787130731560385,
+      "learning_rate": 3.852931869264387e-08,
+      "loss": 0.9374,
+      "step": 7069
+    },
+    {
+      "epoch": 0.9730250481695568,
+      "grad_norm": 1.7208339392159893,
+      "learning_rate": 3.8139399093286564e-08,
+      "loss": 0.9056,
+      "step": 7070
+    },
+    {
+      "epoch": 0.9731626754748142,
+      "grad_norm": 2.0487655381513807,
+      "learning_rate": 3.7751458779266e-08,
+      "loss": 0.9751,
+      "step": 7071
+    },
+    {
+      "epoch": 0.9733003027800715,
+      "grad_norm": 1.757429371628704,
+      "learning_rate": 3.736549782766275e-08,
+      "loss": 0.9429,
+      "step": 7072
+    },
+    {
+      "epoch": 0.973437930085329,
+      "grad_norm": 1.6676552724924816,
+      "learning_rate": 3.698151631516211e-08,
+      "loss": 0.9467,
+      "step": 7073
+    },
+    {
+      "epoch": 0.9735755573905863,
+      "grad_norm": 2.3134057320639854,
+      "learning_rate": 3.6599514318054195e-08,
+      "loss": 0.8625,
+      "step": 7074
+    },
+    {
+      "epoch": 0.9737131846958437,
+      "grad_norm": 1.873842007538208,
+      "learning_rate": 3.6219491912240504e-08,
+      "loss": 0.8584,
+      "step": 7075
+    },
+    {
+      "epoch": 0.973850812001101,
+      "grad_norm": 1.7468240122028678,
+      "learning_rate": 3.5841449173222854e-08,
+      "loss": 0.8672,
+      "step": 7076
+    },
+    {
+      "epoch": 0.9739884393063584,
+      "grad_norm": 1.855062850565374,
+      "learning_rate": 3.546538617611561e-08,
+      "loss": 0.8361,
+      "step": 7077
+    },
+    {
+      "epoch": 0.9741260666116157,
+      "grad_norm": 1.9802498697079032,
+      "learning_rate": 3.5091302995635676e-08,
+      "loss": 0.9438,
+      "step": 7078
+    },
+    {
+      "epoch": 0.9742636939168731,
+      "grad_norm": 1.6309589036400995,
+      "learning_rate": 3.471919970610915e-08,
+      "loss": 0.8868,
+      "step": 7079
+    },
+    {
+      "epoch": 0.9744013212221305,
+      "grad_norm": 1.6353552114617786,
+      "learning_rate": 3.434907638146911e-08,
+      "loss": 0.9223,
+      "step": 7080
+    },
+    {
+      "epoch": 0.9745389485273879,
+      "grad_norm": 1.981548368938733,
+      "learning_rate": 3.398093309525119e-08,
+      "loss": 0.8385,
+      "step": 7081
+    },
+    {
+      "epoch": 0.9746765758326452,
+      "grad_norm": 1.9207662262798872,
+      "learning_rate": 3.361476992060242e-08,
+      "loss": 0.8541,
+      "step": 7082
+    },
+    {
+      "epoch": 0.9748142031379026,
+      "grad_norm": 2.002015667002389,
+      "learning_rate": 3.325058693027461e-08,
+      "loss": 0.9624,
+      "step": 7083
+    },
+    {
+      "epoch": 0.9749518304431599,
+      "grad_norm": 1.7864833524666346,
+      "learning_rate": 3.2888384196624325e-08,
+      "loss": 0.9298,
+      "step": 7084
+    },
+    {
+      "epoch": 0.9750894577484173,
+      "grad_norm": 1.93100028770135,
+      "learning_rate": 3.252816179161733e-08,
+      "loss": 0.8986,
+      "step": 7085
+    },
+    {
+      "epoch": 0.9752270850536746,
+      "grad_norm": 2.0246020688298962,
+      "learning_rate": 3.216991978682526e-08,
+      "loss": 0.9944,
+      "step": 7086
+    },
+    {
+      "epoch": 0.975364712358932,
+      "grad_norm": 1.737797327729844,
+      "learning_rate": 3.181365825342564e-08,
+      "loss": 0.9388,
+      "step": 7087
+    },
+    {
+      "epoch": 0.9755023396641894,
+      "grad_norm": 1.651356637157781,
+      "learning_rate": 3.145937726220183e-08,
+      "loss": 0.8931,
+      "step": 7088
+    },
+    {
+      "epoch": 0.9756399669694468,
+      "grad_norm": 2.0748649500398098,
+      "learning_rate": 3.110707688354531e-08,
+      "loss": 0.8894,
+      "step": 7089
+    },
+    {
+      "epoch": 0.9757775942747041,
+      "grad_norm": 1.6741498669703077,
+      "learning_rate": 3.0756757187453415e-08,
+      "loss": 0.8863,
+      "step": 7090
+    },
+    {
+      "epoch": 0.9759152215799615,
+      "grad_norm": 1.7821527478448937,
+      "learning_rate": 3.040841824353047e-08,
+      "loss": 0.9133,
+      "step": 7091
+    },
+    {
+      "epoch": 0.9760528488852188,
+      "grad_norm": 2.080579411181176,
+      "learning_rate": 3.006206012098556e-08,
+      "loss": 0.8973,
+      "step": 7092
+    },
+    {
+      "epoch": 0.9761904761904762,
+      "grad_norm": 1.6463141622308273,
+      "learning_rate": 2.9717682888635857e-08,
+      "loss": 0.8887,
+      "step": 7093
+    },
+    {
+      "epoch": 0.9763281034957335,
+      "grad_norm": 1.884228711354478,
+      "learning_rate": 2.93752866149033e-08,
+      "loss": 0.9382,
+      "step": 7094
+    },
+    {
+      "epoch": 0.976465730800991,
+      "grad_norm": 1.832278164879594,
+      "learning_rate": 2.9034871367819018e-08,
+      "loss": 0.8391,
+      "step": 7095
+    },
+    {
+      "epoch": 0.9766033581062483,
+      "grad_norm": 1.845464799891437,
+      "learning_rate": 2.8696437215018915e-08,
+      "loss": 0.8733,
+      "step": 7096
+    },
+    {
+      "epoch": 0.9767409854115057,
+      "grad_norm": 1.968923791616318,
+      "learning_rate": 2.8359984223743643e-08,
+      "loss": 0.8854,
+      "step": 7097
+    },
+    {
+      "epoch": 0.976878612716763,
+      "grad_norm": 1.877231737483089,
+      "learning_rate": 2.8025512460843064e-08,
+      "loss": 0.8104,
+      "step": 7098
+    },
+    {
+      "epoch": 0.9770162400220204,
+      "grad_norm": 2.022533927179004,
+      "learning_rate": 2.7693021992770685e-08,
+      "loss": 0.9126,
+      "step": 7099
+    },
+    {
+      "epoch": 0.9771538673272777,
+      "grad_norm": 1.9363774928982873,
+      "learning_rate": 2.736251288558922e-08,
+      "loss": 0.9292,
+      "step": 7100
+    },
+    {
+      "epoch": 0.9772914946325351,
+      "grad_norm": 1.8129414646002855,
+      "learning_rate": 2.703398520496614e-08,
+      "loss": 0.8727,
+      "step": 7101
+    },
+    {
+      "epoch": 0.9774291219377924,
+      "grad_norm": 1.7222336543085603,
+      "learning_rate": 2.6707439016175897e-08,
+      "loss": 0.8765,
+      "step": 7102
+    },
+    {
+      "epoch": 0.9775667492430499,
+      "grad_norm": 2.0180193711113543,
+      "learning_rate": 2.638287438409659e-08,
+      "loss": 0.9492,
+      "step": 7103
+    },
+    {
+      "epoch": 0.9777043765483072,
+      "grad_norm": 1.7221709607586126,
+      "learning_rate": 2.606029137321775e-08,
+      "loss": 0.9094,
+      "step": 7104
+    },
+    {
+      "epoch": 0.9778420038535646,
+      "grad_norm": 1.767951368711889,
+      "learning_rate": 2.5739690047629216e-08,
+      "loss": 0.8179,
+      "step": 7105
+    },
+    {
+      "epoch": 0.9779796311588219,
+      "grad_norm": 1.6224241246194628,
+      "learning_rate": 2.542107047103226e-08,
+      "loss": 0.8257,
+      "step": 7106
+    },
+    {
+      "epoch": 0.9781172584640793,
+      "grad_norm": 1.6655550828321986,
+      "learning_rate": 2.5104432706731796e-08,
+      "loss": 0.9045,
+      "step": 7107
+    },
+    {
+      "epoch": 0.9782548857693366,
+      "grad_norm": 1.6619768555904941,
+      "learning_rate": 2.478977681763861e-08,
+      "loss": 0.8761,
+      "step": 7108
+    },
+    {
+      "epoch": 0.978392513074594,
+      "grad_norm": 1.600156895390456,
+      "learning_rate": 2.4477102866271584e-08,
+      "loss": 0.8956,
+      "step": 7109
+    },
+    {
+      "epoch": 0.9785301403798513,
+      "grad_norm": 1.9348779299398633,
+      "learning_rate": 2.4166410914753246e-08,
+      "loss": 0.8323,
+      "step": 7110
+    },
+    {
+      "epoch": 0.9786677676851088,
+      "grad_norm": 1.6900605750041795,
+      "learning_rate": 2.3857701024815327e-08,
+      "loss": 0.9341,
+      "step": 7111
+    },
+    {
+      "epoch": 0.9788053949903661,
+      "grad_norm": 1.9220245514876881,
+      "learning_rate": 2.3550973257794317e-08,
+      "loss": 0.8694,
+      "step": 7112
+    },
+    {
+      "epoch": 0.9789430222956235,
+      "grad_norm": 1.819841654352463,
+      "learning_rate": 2.3246227674631472e-08,
+      "loss": 0.8453,
+      "step": 7113
+    },
+    {
+      "epoch": 0.9790806496008808,
+      "grad_norm": 1.7882061470919681,
+      "learning_rate": 2.294346433587613e-08,
+      "loss": 0.8872,
+      "step": 7114
+    },
+    {
+      "epoch": 0.9792182769061382,
+      "grad_norm": 1.6107359036725846,
+      "learning_rate": 2.264268330168351e-08,
+      "loss": 0.8894,
+      "step": 7115
+    },
+    {
+      "epoch": 0.9793559042113955,
+      "grad_norm": 2.0282243889102793,
+      "learning_rate": 2.2343884631814694e-08,
+      "loss": 0.8053,
+      "step": 7116
+    },
+    {
+      "epoch": 0.9794935315166529,
+      "grad_norm": 1.5875665012265534,
+      "learning_rate": 2.2047068385636638e-08,
+      "loss": 0.9109,
+      "step": 7117
+    },
+    {
+      "epoch": 0.9796311588219103,
+      "grad_norm": 1.695998402398623,
+      "learning_rate": 2.175223462212328e-08,
+      "loss": 0.9461,
+      "step": 7118
+    },
+    {
+      "epoch": 0.9797687861271677,
+      "grad_norm": 1.8006566939648083,
+      "learning_rate": 2.1459383399853318e-08,
+      "loss": 0.9442,
+      "step": 7119
+    },
+    {
+      "epoch": 0.979906413432425,
+      "grad_norm": 1.7083784965100186,
+      "learning_rate": 2.116851477701243e-08,
+      "loss": 0.8962,
+      "step": 7120
+    },
+    {
+      "epoch": 0.9800440407376824,
+      "grad_norm": 1.6297415725852664,
+      "learning_rate": 2.087962881139216e-08,
+      "loss": 0.8785,
+      "step": 7121
+    },
+    {
+      "epoch": 0.9801816680429397,
+      "grad_norm": 2.0103854470664295,
+      "learning_rate": 2.059272556039105e-08,
+      "loss": 0.8635,
+      "step": 7122
+    },
+    {
+      "epoch": 0.9803192953481971,
+      "grad_norm": 1.9178600599825455,
+      "learning_rate": 2.0307805081011266e-08,
+      "loss": 0.9392,
+      "step": 7123
+    },
+    {
+      "epoch": 0.9804569226534544,
+      "grad_norm": 1.8588559567919347,
+      "learning_rate": 2.0024867429864204e-08,
+      "loss": 0.9162,
+      "step": 7124
+    },
+    {
+      "epoch": 0.9805945499587118,
+      "grad_norm": 1.7798291085729843,
+      "learning_rate": 1.9743912663164888e-08,
+      "loss": 0.8889,
+      "step": 7125
+    },
+    {
+      "epoch": 0.9807321772639692,
+      "grad_norm": 1.6385197147716186,
+      "learning_rate": 1.9464940836735336e-08,
+      "loss": 0.854,
+      "step": 7126
+    },
+    {
+      "epoch": 0.9808698045692266,
+      "grad_norm": 1.9030545231460738,
+      "learning_rate": 1.918795200600343e-08,
+      "loss": 0.9209,
+      "step": 7127
+    },
+    {
+      "epoch": 0.9810074318744839,
+      "grad_norm": 1.7393012268092334,
+      "learning_rate": 1.8912946226002927e-08,
+      "loss": 0.929,
+      "step": 7128
+    },
+    {
+      "epoch": 0.9811450591797413,
+      "grad_norm": 1.7645949434336508,
+      "learning_rate": 1.8639923551373452e-08,
+      "loss": 0.802,
+      "step": 7129
+    },
+    {
+      "epoch": 0.9812826864849986,
+      "grad_norm": 1.6210850927119227,
+      "learning_rate": 1.8368884036361612e-08,
+      "loss": 0.916,
+      "step": 7130
+    },
+    {
+      "epoch": 0.981420313790256,
+      "grad_norm": 1.6870470902510133,
+      "learning_rate": 1.8099827734818775e-08,
+      "loss": 0.9073,
+      "step": 7131
+    },
+    {
+      "epoch": 0.9815579410955133,
+      "grad_norm": 1.723240574149357,
+      "learning_rate": 1.7832754700203292e-08,
+      "loss": 0.8968,
+      "step": 7132
+    },
+    {
+      "epoch": 0.9816955684007708,
+      "grad_norm": 1.7463134039853085,
+      "learning_rate": 1.756766498557827e-08,
+      "loss": 0.9474,
+      "step": 7133
+    },
+    {
+      "epoch": 0.9818331957060281,
+      "grad_norm": 1.5540838318403067,
+      "learning_rate": 1.730455864361269e-08,
+      "loss": 0.8886,
+      "step": 7134
+    },
+    {
+      "epoch": 0.9819708230112855,
+      "grad_norm": 2.046228938546936,
+      "learning_rate": 1.7043435726583624e-08,
+      "loss": 0.9232,
+      "step": 7135
+    },
+    {
+      "epoch": 0.9821084503165428,
+      "grad_norm": 1.5883942130631237,
+      "learning_rate": 1.6784296286372904e-08,
+      "loss": 0.949,
+      "step": 7136
+    },
+    {
+      "epoch": 0.9822460776218002,
+      "grad_norm": 1.7508680171451814,
+      "learning_rate": 1.652714037446601e-08,
+      "loss": 0.8526,
+      "step": 7137
+    },
+    {
+      "epoch": 0.9823837049270575,
+      "grad_norm": 1.6884182567313746,
+      "learning_rate": 1.627196804195874e-08,
+      "loss": 0.9074,
+      "step": 7138
+    },
+    {
+      "epoch": 0.9825213322323149,
+      "grad_norm": 1.8071955475154486,
+      "learning_rate": 1.601877933954832e-08,
+      "loss": 0.9622,
+      "step": 7139
+    },
+    {
+      "epoch": 0.9826589595375722,
+      "grad_norm": 1.812399178990143,
+      "learning_rate": 1.5767574317540058e-08,
+      "loss": 0.9188,
+      "step": 7140
+    },
+    {
+      "epoch": 0.9827965868428297,
+      "grad_norm": 1.7091239674428909,
+      "learning_rate": 1.5518353025846255e-08,
+      "loss": 0.9243,
+      "step": 7141
+    },
+    {
+      "epoch": 0.982934214148087,
+      "grad_norm": 1.8651624804345133,
+      "learning_rate": 1.527111551398286e-08,
+      "loss": 0.8821,
+      "step": 7142
+    },
+    {
+      "epoch": 0.9830718414533444,
+      "grad_norm": 1.6731486557083872,
+      "learning_rate": 1.5025861831072795e-08,
+      "loss": 0.9148,
+      "step": 7143
+    },
+    {
+      "epoch": 0.9832094687586017,
+      "grad_norm": 1.8797866104253402,
+      "learning_rate": 1.4782592025844866e-08,
+      "loss": 0.8485,
+      "step": 7144
+    },
+    {
+      "epoch": 0.9833470960638591,
+      "grad_norm": 1.6941139921541544,
+      "learning_rate": 1.4541306146633737e-08,
+      "loss": 0.8823,
+      "step": 7145
+    },
+    {
+      "epoch": 0.9834847233691164,
+      "grad_norm": 1.7277284702614353,
+      "learning_rate": 1.4302004241377731e-08,
+      "loss": 0.8649,
+      "step": 7146
+    },
+    {
+      "epoch": 0.9836223506743738,
+      "grad_norm": 1.8878176929955857,
+      "learning_rate": 1.406468635762437e-08,
+      "loss": 0.7993,
+      "step": 7147
+    },
+    {
+      "epoch": 0.9837599779796311,
+      "grad_norm": 1.7430252944041433,
+      "learning_rate": 1.3829352542525932e-08,
+      "loss": 0.848,
+      "step": 7148
+    },
+    {
+      "epoch": 0.9838976052848886,
+      "grad_norm": 1.6787876598970166,
+      "learning_rate": 1.3596002842839461e-08,
+      "loss": 0.9098,
+      "step": 7149
+    },
+    {
+      "epoch": 0.9840352325901459,
+      "grad_norm": 1.802135022013292,
+      "learning_rate": 1.3364637304927875e-08,
+      "loss": 0.938,
+      "step": 7150
+    },
+    {
+      "epoch": 0.9841728598954033,
+      "grad_norm": 1.7066248869045448,
+      "learning_rate": 1.3135255974759952e-08,
+      "loss": 0.9316,
+      "step": 7151
+    },
+    {
+      "epoch": 0.9843104872006606,
+      "grad_norm": 1.688818959217137,
+      "learning_rate": 1.2907858897912574e-08,
+      "loss": 0.845,
+      "step": 7152
+    },
+    {
+      "epoch": 0.984448114505918,
+      "grad_norm": 1.6738683910176764,
+      "learning_rate": 1.268244611956404e-08,
+      "loss": 0.9179,
+      "step": 7153
+    },
+    {
+      "epoch": 0.9845857418111753,
+      "grad_norm": 1.841478877326216,
+      "learning_rate": 1.2459017684502971e-08,
+      "loss": 0.8658,
+      "step": 7154
+    },
+    {
+      "epoch": 0.9847233691164327,
+      "grad_norm": 1.6639405162849878,
+      "learning_rate": 1.2237573637120526e-08,
+      "loss": 0.8854,
+      "step": 7155
+    },
+    {
+      "epoch": 0.9848609964216901,
+      "grad_norm": 1.7656578461971746,
+      "learning_rate": 1.201811402141484e-08,
+      "loss": 0.9147,
+      "step": 7156
+    },
+    {
+      "epoch": 0.9849986237269475,
+      "grad_norm": 1.671497392496297,
+      "learning_rate": 1.1800638880988814e-08,
+      "loss": 0.9047,
+      "step": 7157
+    },
+    {
+      "epoch": 0.9851362510322048,
+      "grad_norm": 1.7568125475738678,
+      "learning_rate": 1.1585148259052326e-08,
+      "loss": 0.8614,
+      "step": 7158
+    },
+    {
+      "epoch": 0.9852738783374622,
+      "grad_norm": 1.8364603475377745,
+      "learning_rate": 1.1371642198420018e-08,
+      "loss": 0.9752,
+      "step": 7159
+    },
+    {
+      "epoch": 0.9854115056427195,
+      "grad_norm": 1.5406872133853837,
+      "learning_rate": 1.1160120741513514e-08,
+      "loss": 0.8888,
+      "step": 7160
+    },
+    {
+      "epoch": 0.9855491329479769,
+      "grad_norm": 1.9034274589405134,
+      "learning_rate": 1.0950583930358082e-08,
+      "loss": 0.9222,
+      "step": 7161
+    },
+    {
+      "epoch": 0.9856867602532342,
+      "grad_norm": 2.0903873546796805,
+      "learning_rate": 1.074303180658709e-08,
+      "loss": 0.8383,
+      "step": 7162
+    },
+    {
+      "epoch": 0.9858243875584916,
+      "grad_norm": 1.8757467134829442,
+      "learning_rate": 1.0537464411437548e-08,
+      "loss": 0.8267,
+      "step": 7163
+    },
+    {
+      "epoch": 0.985962014863749,
+      "grad_norm": 1.6811245777192023,
+      "learning_rate": 1.0333881785753452e-08,
+      "loss": 0.8863,
+      "step": 7164
+    },
+    {
+      "epoch": 0.9860996421690064,
+      "grad_norm": 1.7987369059245906,
+      "learning_rate": 1.0132283969982448e-08,
+      "loss": 0.9106,
+      "step": 7165
+    },
+    {
+      "epoch": 0.9862372694742637,
+      "grad_norm": 1.8398364755139685,
+      "learning_rate": 9.932671004180273e-09,
+      "loss": 0.8475,
+      "step": 7166
+    },
+    {
+      "epoch": 0.9863748967795211,
+      "grad_norm": 1.6356891369890763,
+      "learning_rate": 9.73504292800742e-09,
+      "loss": 0.8665,
+      "step": 7167
+    },
+    {
+      "epoch": 0.9865125240847784,
+      "grad_norm": 1.8368386893276074,
+      "learning_rate": 9.539399780730262e-09,
+      "loss": 0.8093,
+      "step": 7168
+    },
+    {
+      "epoch": 0.9866501513900358,
+      "grad_norm": 1.7693342834125982,
+      "learning_rate": 9.345741601219926e-09,
+      "loss": 0.8748,
+      "step": 7169
+    },
+    {
+      "epoch": 0.9867877786952931,
+      "grad_norm": 1.7533188654178768,
+      "learning_rate": 9.154068427952301e-09,
+      "loss": 0.8254,
+      "step": 7170
+    },
+    {
+      "epoch": 0.9869254060005506,
+      "grad_norm": 1.627192571262827,
+      "learning_rate": 8.964380299012476e-09,
+      "loss": 0.8866,
+      "step": 7171
+    },
+    {
+      "epoch": 0.9870630333058079,
+      "grad_norm": 1.7538591528372371,
+      "learning_rate": 8.776677252088084e-09,
+      "loss": 0.848,
+      "step": 7172
+    },
+    {
+      "epoch": 0.9872006606110653,
+      "grad_norm": 1.6214170946739905,
+      "learning_rate": 8.590959324472626e-09,
+      "loss": 0.8923,
+      "step": 7173
+    },
+    {
+      "epoch": 0.9873382879163226,
+      "grad_norm": 2.1878000251190852,
+      "learning_rate": 8.407226553066583e-09,
+      "loss": 0.8922,
+      "step": 7174
+    },
+    {
+      "epoch": 0.98747591522158,
+      "grad_norm": 1.914370840493567,
+      "learning_rate": 8.22547897437409e-09,
+      "loss": 0.9315,
+      "step": 7175
+    },
+    {
+      "epoch": 0.9876135425268373,
+      "grad_norm": 1.697390313345376,
+      "learning_rate": 8.045716624506262e-09,
+      "loss": 0.9452,
+      "step": 7176
+    },
+    {
+      "epoch": 0.9877511698320947,
+      "grad_norm": 1.717139225600654,
+      "learning_rate": 7.86793953918008e-09,
+      "loss": 0.893,
+      "step": 7177
+    },
+    {
+      "epoch": 0.987888797137352,
+      "grad_norm": 1.6663640520102418,
+      "learning_rate": 7.692147753717294e-09,
+      "loss": 0.8751,
+      "step": 7178
+    },
+    {
+      "epoch": 0.9880264244426095,
+      "grad_norm": 2.354178759618853,
+      "learning_rate": 7.518341303043298e-09,
+      "loss": 0.8489,
+      "step": 7179
+    },
+    {
+      "epoch": 0.9881640517478668,
+      "grad_norm": 1.871022062912179,
+      "learning_rate": 7.346520221693798e-09,
+      "loss": 0.8801,
+      "step": 7180
+    },
+    {
+      "epoch": 0.9883016790531242,
+      "grad_norm": 1.6714705860707322,
+      "learning_rate": 7.176684543807044e-09,
+      "loss": 0.8728,
+      "step": 7181
+    },
+    {
+      "epoch": 0.9884393063583815,
+      "grad_norm": 1.6633120166078814,
+      "learning_rate": 7.008834303124934e-09,
+      "loss": 0.9347,
+      "step": 7182
+    },
+    {
+      "epoch": 0.9885769336636389,
+      "grad_norm": 2.3732573888755057,
+      "learning_rate": 6.842969532998567e-09,
+      "loss": 0.7813,
+      "step": 7183
+    },
+    {
+      "epoch": 0.9887145609688962,
+      "grad_norm": 1.930358242580649,
+      "learning_rate": 6.679090266382693e-09,
+      "loss": 0.8848,
+      "step": 7184
+    },
+    {
+      "epoch": 0.9888521882741536,
+      "grad_norm": 1.6646359585491275,
+      "learning_rate": 6.517196535839043e-09,
+      "loss": 0.8924,
+      "step": 7185
+    },
+    {
+      "epoch": 0.9889898155794109,
+      "grad_norm": 1.9205596433668473,
+      "learning_rate": 6.357288373531889e-09,
+      "loss": 0.8764,
+      "step": 7186
+    },
+    {
+      "epoch": 0.9891274428846684,
+      "grad_norm": 1.7452758425073653,
+      "learning_rate": 6.199365811232483e-09,
+      "loss": 0.885,
+      "step": 7187
+    },
+    {
+      "epoch": 0.9892650701899257,
+      "grad_norm": 1.6915464698187117,
+      "learning_rate": 6.043428880320168e-09,
+      "loss": 0.9049,
+      "step": 7188
+    },
+    {
+      "epoch": 0.9894026974951831,
+      "grad_norm": 1.8021505276916292,
+      "learning_rate": 5.889477611775718e-09,
+      "loss": 0.8834,
+      "step": 7189
+    },
+    {
+      "epoch": 0.9895403248004404,
+      "grad_norm": 1.6676111625334,
+      "learning_rate": 5.7375120361868876e-09,
+      "loss": 0.9045,
+      "step": 7190
+    },
+    {
+      "epoch": 0.9896779521056978,
+      "grad_norm": 2.0276735160386994,
+      "learning_rate": 5.587532183748412e-09,
+      "loss": 0.8431,
+      "step": 7191
+    },
+    {
+      "epoch": 0.9898155794109551,
+      "grad_norm": 1.7840647250241717,
+      "learning_rate": 5.439538084259788e-09,
+      "loss": 0.877,
+      "step": 7192
+    },
+    {
+      "epoch": 0.9899532067162125,
+      "grad_norm": 1.6591409980739253,
+      "learning_rate": 5.293529767123051e-09,
+      "loss": 0.9372,
+      "step": 7193
+    },
+    {
+      "epoch": 0.9900908340214699,
+      "grad_norm": 1.7355841443800708,
+      "learning_rate": 5.14950726134944e-09,
+      "loss": 0.9776,
+      "step": 7194
+    },
+    {
+      "epoch": 0.9902284613267273,
+      "grad_norm": 1.6009324818743722,
+      "learning_rate": 5.007470595553843e-09,
+      "loss": 0.9175,
+      "step": 7195
+    },
+    {
+      "epoch": 0.9903660886319846,
+      "grad_norm": 1.5085758678570025,
+      "learning_rate": 4.867419797958128e-09,
+      "loss": 0.886,
+      "step": 7196
+    },
+    {
+      "epoch": 0.990503715937242,
+      "grad_norm": 1.8269963812930436,
+      "learning_rate": 4.729354896387817e-09,
+      "loss": 0.9232,
+      "step": 7197
+    },
+    {
+      "epoch": 0.9906413432424993,
+      "grad_norm": 2.0486648439703714,
+      "learning_rate": 4.593275918273188e-09,
+      "loss": 0.926,
+      "step": 7198
+    },
+    {
+      "epoch": 0.9907789705477567,
+      "grad_norm": 1.7664048782002812,
+      "learning_rate": 4.4591828906537235e-09,
+      "loss": 0.983,
+      "step": 7199
+    },
+    {
+      "epoch": 0.990916597853014,
+      "grad_norm": 1.7876032568880444,
+      "learning_rate": 4.327075840169226e-09,
+      "loss": 0.9074,
+      "step": 7200
+    },
+    {
+      "epoch": 0.9910542251582714,
+      "grad_norm": 1.4863395063889362,
+      "learning_rate": 4.196954793069807e-09,
+      "loss": 0.8924,
+      "step": 7201
+    },
+    {
+      "epoch": 0.9911918524635288,
+      "grad_norm": 2.4224974981447023,
+      "learning_rate": 4.068819775208122e-09,
+      "loss": 0.8228,
+      "step": 7202
+    },
+    {
+      "epoch": 0.9913294797687862,
+      "grad_norm": 1.8670057646834823,
+      "learning_rate": 3.942670812041583e-09,
+      "loss": 0.8476,
+      "step": 7203
+    },
+    {
+      "epoch": 0.9914671070740435,
+      "grad_norm": 1.5374751589277778,
+      "learning_rate": 3.818507928635695e-09,
+      "loss": 0.9175,
+      "step": 7204
+    },
+    {
+      "epoch": 0.9916047343793009,
+      "grad_norm": 1.9344898542373,
+      "learning_rate": 3.6963311496596153e-09,
+      "loss": 0.8507,
+      "step": 7205
+    },
+    {
+      "epoch": 0.9917423616845582,
+      "grad_norm": 1.866868848122515,
+      "learning_rate": 3.5761404993883696e-09,
+      "loss": 0.9843,
+      "step": 7206
+    },
+    {
+      "epoch": 0.9918799889898156,
+      "grad_norm": 1.6443382459047127,
+      "learning_rate": 3.4579360017006347e-09,
+      "loss": 0.9257,
+      "step": 7207
+    },
+    {
+      "epoch": 0.9920176162950729,
+      "grad_norm": 2.0249260568886043,
+      "learning_rate": 3.341717680084289e-09,
+      "loss": 0.9139,
+      "step": 7208
+    },
+    {
+      "epoch": 0.9921552436003304,
+      "grad_norm": 1.6176446174493126,
+      "learning_rate": 3.2274855576297504e-09,
+      "loss": 0.9083,
+      "step": 7209
+    },
+    {
+      "epoch": 0.9922928709055877,
+      "grad_norm": 1.8328209013930994,
+      "learning_rate": 3.1152396570321986e-09,
+      "loss": 0.9116,
+      "step": 7210
+    },
+    {
+      "epoch": 0.9924304982108451,
+      "grad_norm": 1.7733833238101095,
+      "learning_rate": 3.0049800005937934e-09,
+      "loss": 0.8739,
+      "step": 7211
+    },
+    {
+      "epoch": 0.9925681255161024,
+      "grad_norm": 1.773101807202807,
+      "learning_rate": 2.896706610222566e-09,
+      "loss": 0.913,
+      "step": 7212
+    },
+    {
+      "epoch": 0.9927057528213598,
+      "grad_norm": 1.7250848979914113,
+      "learning_rate": 2.7904195074301976e-09,
+      "loss": 0.9128,
+      "step": 7213
+    },
+    {
+      "epoch": 0.9928433801266171,
+      "grad_norm": 1.5710101725478094,
+      "learning_rate": 2.6861187133342402e-09,
+      "loss": 0.8911,
+      "step": 7214
+    },
+    {
+      "epoch": 0.9929810074318745,
+      "grad_norm": 1.607077490723191,
+      "learning_rate": 2.5838042486581172e-09,
+      "loss": 0.9104,
+      "step": 7215
+    },
+    {
+      "epoch": 0.9931186347371318,
+      "grad_norm": 1.7471667402672224,
+      "learning_rate": 2.483476133730012e-09,
+      "loss": 0.9068,
+      "step": 7216
+    },
+    {
+      "epoch": 0.9932562620423893,
+      "grad_norm": 1.6388262276604368,
+      "learning_rate": 2.3851343884850885e-09,
+      "loss": 0.8962,
+      "step": 7217
+    },
+    {
+      "epoch": 0.9933938893476466,
+      "grad_norm": 1.6416106440544962,
+      "learning_rate": 2.2887790324610526e-09,
+      "loss": 0.9077,
+      "step": 7218
+    },
+    {
+      "epoch": 0.993531516652904,
+      "grad_norm": 1.8303266911909823,
+      "learning_rate": 2.1944100848025896e-09,
+      "loss": 0.9481,
+      "step": 7219
+    },
+    {
+      "epoch": 0.9936691439581613,
+      "grad_norm": 1.9798627145381948,
+      "learning_rate": 2.102027564260256e-09,
+      "loss": 0.8361,
+      "step": 7220
+    },
+    {
+      "epoch": 0.9938067712634187,
+      "grad_norm": 1.8522145317892034,
+      "learning_rate": 2.0116314891893694e-09,
+      "loss": 0.8778,
+      "step": 7221
+    },
+    {
+      "epoch": 0.993944398568676,
+      "grad_norm": 2.555581321355594,
+      "learning_rate": 1.9232218775488976e-09,
+      "loss": 0.9617,
+      "step": 7222
+    },
+    {
+      "epoch": 0.9940820258739334,
+      "grad_norm": 1.8090208334745586,
+      "learning_rate": 1.836798746905899e-09,
+      "loss": 0.9101,
+      "step": 7223
+    },
+    {
+      "epoch": 0.9942196531791907,
+      "grad_norm": 1.8893466543300572,
+      "learning_rate": 1.752362114429973e-09,
+      "loss": 0.9297,
+      "step": 7224
+    },
+    {
+      "epoch": 0.9943572804844482,
+      "grad_norm": 1.9970676149799507,
+      "learning_rate": 1.66991199689992e-09,
+      "loss": 0.9157,
+      "step": 7225
+    },
+    {
+      "epoch": 0.9944949077897055,
+      "grad_norm": 1.8116763067699089,
+      "learning_rate": 1.5894484106959706e-09,
+      "loss": 0.8939,
+      "step": 7226
+    },
+    {
+      "epoch": 0.9946325350949629,
+      "grad_norm": 1.682717522225496,
+      "learning_rate": 1.5109713718042262e-09,
+      "loss": 0.9293,
+      "step": 7227
+    },
+    {
+      "epoch": 0.9947701624002202,
+      "grad_norm": 1.6954144217200076,
+      "learning_rate": 1.4344808958188794e-09,
+      "loss": 0.9107,
+      "step": 7228
+    },
+    {
+      "epoch": 0.9949077897054776,
+      "grad_norm": 1.6244666882231236,
+      "learning_rate": 1.3599769979366629e-09,
+      "loss": 0.7895,
+      "step": 7229
+    },
+    {
+      "epoch": 0.9950454170107349,
+      "grad_norm": 1.9278337293288486,
+      "learning_rate": 1.28745969296018e-09,
+      "loss": 0.9036,
+      "step": 7230
+    },
+    {
+      "epoch": 0.9951830443159922,
+      "grad_norm": 1.423921089428278,
+      "learning_rate": 1.2169289952990159e-09,
+      "loss": 0.888,
+      "step": 7231
+    },
+    {
+      "epoch": 0.9953206716212497,
+      "grad_norm": 1.6769835867933292,
+      "learning_rate": 1.148384918964185e-09,
+      "loss": 0.8599,
+      "step": 7232
+    },
+    {
+      "epoch": 0.9954582989265071,
+      "grad_norm": 1.975109384235802,
+      "learning_rate": 1.0818274775770132e-09,
+      "loss": 0.8535,
+      "step": 7233
+    },
+    {
+      "epoch": 0.9955959262317644,
+      "grad_norm": 2.183982952391194,
+      "learning_rate": 1.0172566843602571e-09,
+      "loss": 0.9815,
+      "step": 7234
+    },
+    {
+      "epoch": 0.9957335535370218,
+      "grad_norm": 1.6946357384558168,
+      "learning_rate": 9.546725521425437e-10,
+      "loss": 0.9489,
+      "step": 7235
+    },
+    {
+      "epoch": 0.9958711808422791,
+      "grad_norm": 1.7365714638028211,
+      "learning_rate": 8.940750933605912e-10,
+      "loss": 0.9244,
+      "step": 7236
+    },
+    {
+      "epoch": 0.9960088081475365,
+      "grad_norm": 1.7723060904959356,
+      "learning_rate": 8.35464320052548e-10,
+      "loss": 0.8988,
+      "step": 7237
+    },
+    {
+      "epoch": 0.9961464354527938,
+      "grad_norm": 2.4314326930622556,
+      "learning_rate": 7.788402438646536e-10,
+      "loss": 0.8811,
+      "step": 7238
+    },
+    {
+      "epoch": 0.9962840627580511,
+      "grad_norm": 1.7308191369354555,
+      "learning_rate": 7.242028760456876e-10,
+      "loss": 0.8604,
+      "step": 7239
+    },
+    {
+      "epoch": 0.9964216900633086,
+      "grad_norm": 1.8478710866104708,
+      "learning_rate": 6.715522274536312e-10,
+      "loss": 0.8873,
+      "step": 7240
+    },
+    {
+      "epoch": 0.996559317368566,
+      "grad_norm": 2.726233529420824,
+      "learning_rate": 6.208883085467854e-10,
+      "loss": 0.9578,
+      "step": 7241
+    },
+    {
+      "epoch": 0.9966969446738233,
+      "grad_norm": 1.702964764368626,
+      "learning_rate": 5.722111293937626e-10,
+      "loss": 0.8969,
+      "step": 7242
+    },
+    {
+      "epoch": 0.9968345719790807,
+      "grad_norm": 1.627391315594623,
+      "learning_rate": 5.255206996657158e-10,
+      "loss": 0.9395,
+      "step": 7243
+    },
+    {
+      "epoch": 0.996972199284338,
+      "grad_norm": 1.6857255285555028,
+      "learning_rate": 4.80817028637448e-10,
+      "loss": 0.8893,
+      "step": 7244
+    },
+    {
+      "epoch": 0.9971098265895953,
+      "grad_norm": 1.8055265930315514,
+      "learning_rate": 4.381001251940742e-10,
+      "loss": 0.7719,
+      "step": 7245
+    },
+    {
+      "epoch": 0.9972474538948527,
+      "grad_norm": 1.8320218215201705,
+      "learning_rate": 3.973699978199186e-10,
+      "loss": 0.847,
+      "step": 7246
+    },
+    {
+      "epoch": 0.9973850812001102,
+      "grad_norm": 1.8760038641452677,
+      "learning_rate": 3.586266546085071e-10,
+      "loss": 0.9716,
+      "step": 7247
+    },
+    {
+      "epoch": 0.9975227085053675,
+      "grad_norm": 1.7235155603671224,
+      "learning_rate": 3.2187010325812617e-10,
+      "loss": 0.8192,
+      "step": 7248
+    },
+    {
+      "epoch": 0.9976603358106249,
+      "grad_norm": 1.7155100422710021,
+      "learning_rate": 2.871003510718229e-10,
+      "loss": 0.8443,
+      "step": 7249
+    },
+    {
+      "epoch": 0.9977979631158822,
+      "grad_norm": 2.109932552655822,
+      "learning_rate": 2.543174049574049e-10,
+      "loss": 0.9018,
+      "step": 7250
+    },
+    {
+      "epoch": 0.9979355904211396,
+      "grad_norm": 2.0033343001823662,
+      "learning_rate": 2.2352127142855063e-10,
+      "loss": 0.8536,
+      "step": 7251
+    },
+    {
+      "epoch": 0.9980732177263969,
+      "grad_norm": 1.719209198321801,
+      "learning_rate": 1.9471195660369925e-10,
+      "loss": 0.8949,
+      "step": 7252
+    },
+    {
+      "epoch": 0.9982108450316542,
+      "grad_norm": 1.9271712515601436,
+      "learning_rate": 1.6788946620716063e-10,
+      "loss": 0.9116,
+      "step": 7253
+    },
+    {
+      "epoch": 0.9983484723369116,
+      "grad_norm": 1.876080886910594,
+      "learning_rate": 1.4305380556800529e-10,
+      "loss": 0.8995,
+      "step": 7254
+    },
+    {
+      "epoch": 0.9984860996421691,
+      "grad_norm": 1.8420027205658958,
+      "learning_rate": 1.202049796211746e-10,
+      "loss": 0.9426,
+      "step": 7255
+    },
+    {
+      "epoch": 0.9986237269474264,
+      "grad_norm": 1.6243810005406427,
+      "learning_rate": 9.93429929063705e-11,
+      "loss": 0.9549,
+      "step": 7256
+    },
+    {
+      "epoch": 0.9987613542526838,
+      "grad_norm": 1.8848285743983697,
+      "learning_rate": 8.046784956805553e-11,
+      "loss": 0.9488,
+      "step": 7257
+    },
+    {
+      "epoch": 0.9988989815579411,
+      "grad_norm": 1.6938589434188889,
+      "learning_rate": 6.357955335767329e-11,
+      "loss": 0.9334,
+      "step": 7258
+    },
+    {
+      "epoch": 0.9990366088631984,
+      "grad_norm": 1.6163266701171075,
+      "learning_rate": 4.867810762920755e-11,
+      "loss": 0.8548,
+      "step": 7259
+    },
+    {
+      "epoch": 0.9991742361684558,
+      "grad_norm": 2.0646338993543925,
+      "learning_rate": 3.57635153436231e-11,
+      "loss": 0.8011,
+      "step": 7260
+    },
+    {
+      "epoch": 0.9993118634737131,
+      "grad_norm": 1.7544564251939918,
+      "learning_rate": 2.4835779067755583e-11,
+      "loss": 0.8468,
+      "step": 7261
+    },
+    {
+      "epoch": 0.9994494907789705,
+      "grad_norm": 1.9342070082883098,
+      "learning_rate": 1.5894900972091008e-11,
+      "loss": 0.9321,
+      "step": 7262
+    },
+    {
+      "epoch": 0.999587118084228,
+      "grad_norm": 1.8564244616293597,
+      "learning_rate": 8.940882832986219e-12,
+      "loss": 0.9073,
+      "step": 7263
+    },
+    {
+      "epoch": 0.9997247453894853,
+      "grad_norm": 1.5797573733611097,
+      "learning_rate": 3.973726032668879e-12,
+      "loss": 0.8586,
+      "step": 7264
+    },
+    {
+      "epoch": 0.9998623726947427,
+      "grad_norm": 1.688711086309991,
+      "learning_rate": 9.934315570170327e-13,
+      "loss": 0.9282,
+      "step": 7265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6864384435827913,
+      "learning_rate": 0.0,
+      "loss": 0.8906,
+      "step": 7266
+    },
+    {
+      "epoch": 1.0,
+      "step": 7266,
+      "total_flos": 3.5387205742913126e+18,
+      "train_loss": 1.064733874492661,
+      "train_runtime": 10499.2866,
+      "train_samples_per_second": 88.582,
+      "train_steps_per_second": 0.692
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7266,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.5387205742913126e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}