diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,121877 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 17405,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0507381496559225,
+      "learning_rate": 3.824091778202677e-08,
+      "loss": 0.2942,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8070182416007133,
+      "learning_rate": 7.648183556405354e-08,
+      "loss": 0.3872,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.80041143859621,
+      "learning_rate": 1.1472275334608032e-07,
+      "loss": 0.5905,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8120071191832767,
+      "learning_rate": 1.5296367112810708e-07,
+      "loss": 0.2618,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5949392601041631,
+      "learning_rate": 1.9120458891013387e-07,
+      "loss": 0.2642,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.055790990377875,
+      "learning_rate": 2.2944550669216063e-07,
+      "loss": 0.7541,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7613072929404066,
+      "learning_rate": 2.676864244741874e-07,
+      "loss": 0.2649,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.304037252213996,
+      "learning_rate": 3.0592734225621416e-07,
+      "loss": 0.5504,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9141654125014961,
+      "learning_rate": 3.441682600382409e-07,
+      "loss": 0.3869,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7609269558117047,
+      "learning_rate": 3.8240917782026774e-07,
+      "loss": 0.2751,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7955995450796811,
+      "learning_rate": 4.206500956022945e-07,
+      "loss": 0.1885,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8612661473785725,
+      "learning_rate": 4.5889101338432127e-07,
+      "loss": 0.585,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7748274547715579,
+      "learning_rate": 4.97131931166348e-07,
+      "loss": 0.3195,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8639653689032786,
+      "learning_rate": 5.353728489483748e-07,
+      "loss": 0.3883,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5884579720855796,
+      "learning_rate": 5.736137667304016e-07,
+      "loss": 0.785,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8703771449042337,
+      "learning_rate": 6.118546845124283e-07,
+      "loss": 0.3993,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6961613428356354,
+      "learning_rate": 6.500956022944552e-07,
+      "loss": 0.2686,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.971276714643732,
+      "learning_rate": 6.883365200764818e-07,
+      "loss": 0.3745,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4283398141820636,
+      "learning_rate": 7.265774378585087e-07,
+      "loss": 0.4011,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9170881209785016,
+      "learning_rate": 7.648183556405355e-07,
+      "loss": 0.3542,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9135336035316859,
+      "learning_rate": 8.030592734225621e-07,
+      "loss": 0.3697,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0476506879512788,
+      "learning_rate": 8.41300191204589e-07,
+      "loss": 0.3371,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6115452493182032,
+      "learning_rate": 8.795411089866157e-07,
+      "loss": 0.2105,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.67676855881491,
+      "learning_rate": 9.177820267686425e-07,
+      "loss": 0.5603,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7202407335460682,
+      "learning_rate": 9.560229445506693e-07,
+      "loss": 0.3293,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9865131037247048,
+      "learning_rate": 9.94263862332696e-07,
+      "loss": 0.41,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4535881440886347,
+      "learning_rate": 1.0325047801147228e-06,
+      "loss": 0.5634,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0947257185323476,
+      "learning_rate": 1.0707456978967496e-06,
+      "loss": 0.3149,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0129147168499806,
+      "learning_rate": 1.1089866156787763e-06,
+      "loss": 0.3681,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6576020756810448,
+      "learning_rate": 1.1472275334608031e-06,
+      "loss": 0.2731,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9047392383713737,
+      "learning_rate": 1.1854684512428299e-06,
+      "loss": 0.3647,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4541026438021523,
+      "learning_rate": 1.2237093690248566e-06,
+      "loss": 0.4479,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8653286599987811,
+      "learning_rate": 1.2619502868068834e-06,
+      "loss": 0.3584,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.121457616803094,
+      "learning_rate": 1.3001912045889104e-06,
+      "loss": 0.3729,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8277941941005295,
+      "learning_rate": 1.3384321223709371e-06,
+      "loss": 0.3191,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7443085270016412,
+      "learning_rate": 1.3766730401529637e-06,
+      "loss": 0.3493,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1585945317118969,
+      "learning_rate": 1.4149139579349905e-06,
+      "loss": 0.3186,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5269750513842424,
+      "learning_rate": 1.4531548757170174e-06,
+      "loss": 0.3761,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.837589017463728,
+      "learning_rate": 1.4913957934990442e-06,
+      "loss": 0.7611,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1361050627404332,
+      "learning_rate": 1.529636711281071e-06,
+      "loss": 0.1478,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9717312245736287,
+      "learning_rate": 1.5678776290630975e-06,
+      "loss": 0.3778,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6502427723791129,
+      "learning_rate": 1.6061185468451243e-06,
+      "loss": 0.5467,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6014184946161949,
+      "learning_rate": 1.6443594646271512e-06,
+      "loss": 0.1958,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7347266365889769,
+      "learning_rate": 1.682600382409178e-06,
+      "loss": 0.4705,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9784708341611575,
+      "learning_rate": 1.7208413001912048e-06,
+      "loss": 0.3925,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8155632192584802,
+      "learning_rate": 1.7590822179732313e-06,
+      "loss": 0.2767,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6297697128120328,
+      "learning_rate": 1.7973231357552585e-06,
+      "loss": 0.4902,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0192981344155907,
+      "learning_rate": 1.835564053537285e-06,
+      "loss": 0.5705,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6176671121632008,
+      "learning_rate": 1.8738049713193118e-06,
+      "loss": 0.2417,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.715246719483139,
+      "learning_rate": 1.9120458891013386e-06,
+      "loss": 0.2814,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0156791314075595,
+      "learning_rate": 1.950286806883365e-06,
+      "loss": 0.8249,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.656560840965785,
+      "learning_rate": 1.988527724665392e-06,
+      "loss": 0.5013,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1577287645486485,
+      "learning_rate": 2.026768642447419e-06,
+      "loss": 0.2761,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0509099785445395,
+      "learning_rate": 2.0650095602294456e-06,
+      "loss": 0.6112,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9025595249195338,
+      "learning_rate": 2.103250478011472e-06,
+      "loss": 0.1648,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9417125486809438,
+      "learning_rate": 2.141491395793499e-06,
+      "loss": 0.325,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9816339235493463,
+      "learning_rate": 2.179732313575526e-06,
+      "loss": 0.39,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7398045274330625,
+      "learning_rate": 2.2179732313575527e-06,
+      "loss": 0.495,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1507513410092913,
+      "learning_rate": 2.2562141491395797e-06,
+      "loss": 0.3634,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.505952386635836,
+      "learning_rate": 2.2944550669216062e-06,
+      "loss": 0.7398,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.706667753638737,
+      "learning_rate": 2.332695984703633e-06,
+      "loss": 0.2919,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.702597844899516,
+      "learning_rate": 2.3709369024856597e-06,
+      "loss": 0.2103,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.783725003259625,
+      "learning_rate": 2.4091778202676867e-06,
+      "loss": 0.853,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3779057437038236,
+      "learning_rate": 2.4474187380497133e-06,
+      "loss": 0.4548,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4971913730088304,
+      "learning_rate": 2.4856596558317402e-06,
+      "loss": 0.3769,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.077225117758281,
+      "learning_rate": 2.523900573613767e-06,
+      "loss": 0.4513,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7469593011509847,
+      "learning_rate": 2.5621414913957938e-06,
+      "loss": 0.1754,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1701958149020584,
+      "learning_rate": 2.6003824091778207e-06,
+      "loss": 0.381,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0667732033492074,
+      "learning_rate": 2.6386233269598473e-06,
+      "loss": 0.365,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.805103712753199,
+      "learning_rate": 2.6768642447418743e-06,
+      "loss": 0.5996,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1103519370245432,
+      "learning_rate": 2.7151051625239004e-06,
+      "loss": 0.2977,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9741205038133375,
+      "learning_rate": 2.7533460803059274e-06,
+      "loss": 0.3441,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3389588104138266,
+      "learning_rate": 2.7915869980879544e-06,
+      "loss": 0.4686,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1622466550114607,
+      "learning_rate": 2.829827915869981e-06,
+      "loss": 0.2293,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8095204360677337,
+      "learning_rate": 2.868068833652008e-06,
+      "loss": 0.4982,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.216426651097327,
+      "learning_rate": 2.906309751434035e-06,
+      "loss": 0.4559,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7137062684636206,
+      "learning_rate": 2.9445506692160614e-06,
+      "loss": 0.3426,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.186090472631493,
+      "learning_rate": 2.9827915869980884e-06,
+      "loss": 0.6149,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2457594513024053,
+      "learning_rate": 3.021032504780115e-06,
+      "loss": 0.2545,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3233143755012595,
+      "learning_rate": 3.059273422562142e-06,
+      "loss": 0.3427,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9962990513631782,
+      "learning_rate": 3.097514340344169e-06,
+      "loss": 0.4061,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1748881777328464,
+      "learning_rate": 3.135755258126195e-06,
+      "loss": 0.4804,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1349511100419203,
+      "learning_rate": 3.173996175908222e-06,
+      "loss": 0.3407,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8325230383368618,
+      "learning_rate": 3.2122370936902485e-06,
+      "loss": 0.2679,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4409698819778003,
+      "learning_rate": 3.2504780114722755e-06,
+      "loss": 0.2756,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7717083869066103,
+      "learning_rate": 3.2887189292543025e-06,
+      "loss": 0.2771,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0396106154008105,
+      "learning_rate": 3.326959847036329e-06,
+      "loss": 0.5461,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8349352537444616,
+      "learning_rate": 3.365200764818356e-06,
+      "loss": 0.3778,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8634730104834349,
+      "learning_rate": 3.4034416826003826e-06,
+      "loss": 0.26,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8073245544233498,
+      "learning_rate": 3.4416826003824095e-06,
+      "loss": 0.3411,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7074868343906857,
+      "learning_rate": 3.4799235181644365e-06,
+      "loss": 0.7505,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4009581352720348,
+      "learning_rate": 3.5181644359464626e-06,
+      "loss": 0.3037,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7581392814988335,
+      "learning_rate": 3.5564053537284896e-06,
+      "loss": 0.4342,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.586880035707701,
+      "learning_rate": 3.594646271510517e-06,
+      "loss": 0.95,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9134436895106094,
+      "learning_rate": 3.632887189292543e-06,
+      "loss": 0.1859,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5368674428633557,
+      "learning_rate": 3.67112810707457e-06,
+      "loss": 0.4042,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.695940105136588,
+      "learning_rate": 3.7093690248565967e-06,
+      "loss": 0.4439,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4933772907830718,
+      "learning_rate": 3.7476099426386236e-06,
+      "loss": 0.3166,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.929994478426711,
+      "learning_rate": 3.7858508604206506e-06,
+      "loss": 0.6366,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0588657191516118,
+      "learning_rate": 3.824091778202677e-06,
+      "loss": 0.3998,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8299018542448147,
+      "learning_rate": 3.862332695984704e-06,
+      "loss": 0.3014,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.894972655187633,
+      "learning_rate": 3.90057361376673e-06,
+      "loss": 0.152,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0056612950240744,
+      "learning_rate": 3.938814531548758e-06,
+      "loss": 0.5743,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9439871853559727,
+      "learning_rate": 3.977055449330784e-06,
+      "loss": 0.292,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0959590718704717,
+      "learning_rate": 4.015296367112811e-06,
+      "loss": 0.3727,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2912607966735963,
+      "learning_rate": 4.053537284894838e-06,
+      "loss": 0.7559,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9171050201317897,
+      "learning_rate": 4.091778202676865e-06,
+      "loss": 0.314,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6284229206544971,
+      "learning_rate": 4.130019120458891e-06,
+      "loss": 0.2165,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.212927916169031,
+      "learning_rate": 4.168260038240919e-06,
+      "loss": 0.6106,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8099374882393748,
+      "learning_rate": 4.206500956022944e-06,
+      "loss": 0.3248,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.655364301227268,
+      "learning_rate": 4.244741873804972e-06,
+      "loss": 0.5664,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4636213484496585,
+      "learning_rate": 4.282982791586998e-06,
+      "loss": 0.393,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0063990268097915,
+      "learning_rate": 4.321223709369025e-06,
+      "loss": 0.3236,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6371885352644112,
+      "learning_rate": 4.359464627151052e-06,
+      "loss": 0.1923,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.490580634292843,
+      "learning_rate": 4.397705544933079e-06,
+      "loss": 0.5172,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8466201174343196,
+      "learning_rate": 4.435946462715105e-06,
+      "loss": 0.3246,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5069924896763471,
+      "learning_rate": 4.474187380497133e-06,
+      "loss": 0.3987,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.72546443790111,
+      "learning_rate": 4.512428298279159e-06,
+      "loss": 0.6617,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2270235639211564,
+      "learning_rate": 4.550669216061186e-06,
+      "loss": 0.3173,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.611271995853063,
+      "learning_rate": 4.5889101338432124e-06,
+      "loss": 0.2566,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1591179991848053,
+      "learning_rate": 4.627151051625239e-06,
+      "loss": 0.4692,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8409326010763974,
+      "learning_rate": 4.665391969407266e-06,
+      "loss": 0.2925,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6635869940410941,
+      "learning_rate": 4.703632887189293e-06,
+      "loss": 0.4966,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8489652670343713,
+      "learning_rate": 4.7418738049713195e-06,
+      "loss": 0.3289,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7948026475678698,
+      "learning_rate": 4.780114722753346e-06,
+      "loss": 0.3079,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.62879378549951,
+      "learning_rate": 4.8183556405353734e-06,
+      "loss": 0.2514,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9267329931460127,
+      "learning_rate": 4.8565965583174e-06,
+      "loss": 0.575,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6930803429078386,
+      "learning_rate": 4.8948374760994265e-06,
+      "loss": 0.2828,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.567288278653447,
+      "learning_rate": 4.933078393881454e-06,
+      "loss": 0.5037,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4508660819457764,
+      "learning_rate": 4.9713193116634805e-06,
+      "loss": 0.8513,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7264682519312051,
+      "learning_rate": 5.009560229445507e-06,
+      "loss": 0.2465,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.826742636113808,
+      "learning_rate": 5.047801147227534e-06,
+      "loss": 0.3116,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1392276722179755,
+      "learning_rate": 5.086042065009561e-06,
+      "loss": 0.3726,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0376511756469218,
+      "learning_rate": 5.1242829827915875e-06,
+      "loss": 0.2853,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1718548855836737,
+      "learning_rate": 5.162523900573614e-06,
+      "loss": 0.8045,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8519765854258867,
+      "learning_rate": 5.2007648183556415e-06,
+      "loss": 0.4227,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8055693275531942,
+      "learning_rate": 5.239005736137668e-06,
+      "loss": 0.2194,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7267804162988287,
+      "learning_rate": 5.277246653919695e-06,
+      "loss": 0.4669,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5457146032547153,
+      "learning_rate": 5.315487571701722e-06,
+      "loss": 0.2702,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8774674725605524,
+      "learning_rate": 5.3537284894837486e-06,
+      "loss": 0.3322,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.150336747102144,
+      "learning_rate": 5.391969407265774e-06,
+      "loss": 0.3492,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1991018274354253,
+      "learning_rate": 5.430210325047801e-06,
+      "loss": 0.7824,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2228654899985612,
+      "learning_rate": 5.468451242829829e-06,
+      "loss": 0.4072,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7038931034059297,
+      "learning_rate": 5.506692160611855e-06,
+      "loss": 0.3396,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.745883442870367,
+      "learning_rate": 5.544933078393881e-06,
+      "loss": 0.6713,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5065291039901288,
+      "learning_rate": 5.583173996175909e-06,
+      "loss": 0.2111,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.89516023920091,
+      "learning_rate": 5.621414913957935e-06,
+      "loss": 0.3344,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5215200346193276,
+      "learning_rate": 5.659655831739962e-06,
+      "loss": 0.4512,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0561066764454057,
+      "learning_rate": 5.697896749521989e-06,
+      "loss": 0.348,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7077571805921623,
+      "learning_rate": 5.736137667304016e-06,
+      "loss": 0.44,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7032590857528236,
+      "learning_rate": 5.774378585086042e-06,
+      "loss": 0.3231,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6939036942549888,
+      "learning_rate": 5.81261950286807e-06,
+      "loss": 0.2651,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.23769484190234,
+      "learning_rate": 5.850860420650096e-06,
+      "loss": 0.4505,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.57505503947449,
+      "learning_rate": 5.889101338432123e-06,
+      "loss": 0.5387,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4496022682391447,
+      "learning_rate": 5.927342256214149e-06,
+      "loss": 0.4709,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9171040848212857,
+      "learning_rate": 5.965583173996177e-06,
+      "loss": 0.3718,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3430140069069527,
+      "learning_rate": 6.003824091778203e-06,
+      "loss": 0.2232,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7080554277096157,
+      "learning_rate": 6.04206500956023e-06,
+      "loss": 0.2103,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2894404724722592,
+      "learning_rate": 6.080305927342257e-06,
+      "loss": 0.3918,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1163562975594825,
+      "learning_rate": 6.118546845124284e-06,
+      "loss": 0.4489,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8601545568123237,
+      "learning_rate": 6.15678776290631e-06,
+      "loss": 0.427,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9819232651964342,
+      "learning_rate": 6.195028680688338e-06,
+      "loss": 0.4107,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0574104719168467,
+      "learning_rate": 6.233269598470364e-06,
+      "loss": 0.3337,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7953279060143272,
+      "learning_rate": 6.27151051625239e-06,
+      "loss": 0.2119,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0482843536337256,
+      "learning_rate": 6.3097514340344166e-06,
+      "loss": 0.3817,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1370186037298056,
+      "learning_rate": 6.347992351816444e-06,
+      "loss": 0.6306,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6915570774478592,
+      "learning_rate": 6.3862332695984705e-06,
+      "loss": 0.3499,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9773007045626713,
+      "learning_rate": 6.424474187380497e-06,
+      "loss": 0.369,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6951767139469633,
+      "learning_rate": 6.4627151051625245e-06,
+      "loss": 0.7238,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6679450780611562,
+      "learning_rate": 6.500956022944551e-06,
+      "loss": 0.1775,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2440364112811422,
+      "learning_rate": 6.539196940726578e-06,
+      "loss": 0.4651,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4532282158517489,
+      "learning_rate": 6.577437858508605e-06,
+      "loss": 0.4368,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0484087595525342,
+      "learning_rate": 6.6156787762906315e-06,
+      "loss": 0.2313,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2959855414896784,
+      "learning_rate": 6.653919694072658e-06,
+      "loss": 0.3893,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.149727668633913,
+      "learning_rate": 6.6921606118546855e-06,
+      "loss": 0.4034,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3990532753289666,
+      "learning_rate": 6.730401529636712e-06,
+      "loss": 0.7575,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7216334237931397,
+      "learning_rate": 6.768642447418739e-06,
+      "loss": 0.2945,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6587270290307061,
+      "learning_rate": 6.806883365200765e-06,
+      "loss": 0.5882,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7186727610904583,
+      "learning_rate": 6.8451242829827925e-06,
+      "loss": 0.2788,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7580278664452965,
+      "learning_rate": 6.883365200764819e-06,
+      "loss": 0.2313,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3484131561448582,
+      "learning_rate": 6.921606118546846e-06,
+      "loss": 0.7148,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2480514471359154,
+      "learning_rate": 6.959847036328873e-06,
+      "loss": 0.5018,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.853569911541321,
+      "learning_rate": 6.9980879541109e-06,
+      "loss": 0.334,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5935129539425128,
+      "learning_rate": 7.036328871892925e-06,
+      "loss": 0.484,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8264510735195749,
+      "learning_rate": 7.0745697896749535e-06,
+      "loss": 0.2185,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9399844092870507,
+      "learning_rate": 7.112810707456979e-06,
+      "loss": 0.3235,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7297204562340768,
+      "learning_rate": 7.151051625239006e-06,
+      "loss": 0.3819,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2104810033539928,
+      "learning_rate": 7.189292543021034e-06,
+      "loss": 0.4596,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.931742528076528,
+      "learning_rate": 7.22753346080306e-06,
+      "loss": 0.4061,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.490075769645383,
+      "learning_rate": 7.265774378585086e-06,
+      "loss": 0.5611,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8152941819336589,
+      "learning_rate": 7.304015296367113e-06,
+      "loss": 0.3674,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.710166833510888,
+      "learning_rate": 7.34225621414914e-06,
+      "loss": 0.2668,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.568404819733602,
+      "learning_rate": 7.380497131931167e-06,
+      "loss": 0.1746,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2671387794567825,
+      "learning_rate": 7.418738049713193e-06,
+      "loss": 0.519,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7997462545674416,
+      "learning_rate": 7.456978967495221e-06,
+      "loss": 0.3353,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9409506841315287,
+      "learning_rate": 7.495219885277247e-06,
+      "loss": 0.3744,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0262626181568872,
+      "learning_rate": 7.533460803059274e-06,
+      "loss": 0.1996,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6961696439810705,
+      "learning_rate": 7.571701720841301e-06,
+      "loss": 0.3293,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.805884804921434,
+      "learning_rate": 7.609942638623328e-06,
+      "loss": 0.3308,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.94997387801345,
+      "learning_rate": 7.648183556405354e-06,
+      "loss": 0.3929,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7291239386927754,
+      "learning_rate": 7.686424474187381e-06,
+      "loss": 0.3934,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.004028030890788,
+      "learning_rate": 7.724665391969407e-06,
+      "loss": 0.8334,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7999130816704142,
+      "learning_rate": 7.762906309751434e-06,
+      "loss": 0.3051,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5985667826912873,
+      "learning_rate": 7.80114722753346e-06,
+      "loss": 0.2605,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1179195289862098,
+      "learning_rate": 7.839388145315489e-06,
+      "loss": 0.2994,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0573596138813663,
+      "learning_rate": 7.877629063097515e-06,
+      "loss": 0.4686,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.024477703497844,
+      "learning_rate": 7.915869980879542e-06,
+      "loss": 0.3344,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8805912272772863,
+      "learning_rate": 7.954110898661568e-06,
+      "loss": 0.4153,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8453276392969327,
+      "learning_rate": 7.992351816443595e-06,
+      "loss": 0.3134,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6630727990199321,
+      "learning_rate": 8.030592734225622e-06,
+      "loss": 0.278,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.593524418186559,
+      "learning_rate": 8.06883365200765e-06,
+      "loss": 0.3308,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5867712659122044,
+      "learning_rate": 8.107074569789676e-06,
+      "loss": 0.5022,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.858433894348274,
+      "learning_rate": 8.145315487571703e-06,
+      "loss": 0.3147,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9048694714090977,
+      "learning_rate": 8.18355640535373e-06,
+      "loss": 0.8704,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.795882586881069,
+      "learning_rate": 8.221797323135756e-06,
+      "loss": 0.3508,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2363337807852626,
+      "learning_rate": 8.260038240917783e-06,
+      "loss": 0.2234,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2961495565006664,
+      "learning_rate": 8.298279158699809e-06,
+      "loss": 0.3844,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7121762384248471,
+      "learning_rate": 8.336520076481837e-06,
+      "loss": 0.6158,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7273021014242879,
+      "learning_rate": 8.374760994263862e-06,
+      "loss": 0.2802,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6273244509030462,
+      "learning_rate": 8.413001912045889e-06,
+      "loss": 0.7777,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4955213875868227,
+      "learning_rate": 8.451242829827917e-06,
+      "loss": 0.7094,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0844734774581914,
+      "learning_rate": 8.489483747609944e-06,
+      "loss": 0.2128,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7637627517102057,
+      "learning_rate": 8.52772466539197e-06,
+      "loss": 0.416,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5085068537732527,
+      "learning_rate": 8.565965583173997e-06,
+      "loss": 0.4051,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6779193752161682,
+      "learning_rate": 8.604206500956023e-06,
+      "loss": 0.248,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.324560462924011,
+      "learning_rate": 8.64244741873805e-06,
+      "loss": 0.823,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7247870815802524,
+      "learning_rate": 8.680688336520076e-06,
+      "loss": 0.3926,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.874007650243524,
+      "learning_rate": 8.718929254302105e-06,
+      "loss": 0.3379,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4143629669283038,
+      "learning_rate": 8.757170172084131e-06,
+      "loss": 0.3663,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.164683195417105,
+      "learning_rate": 8.795411089866158e-06,
+      "loss": 0.2845,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5604708590835067,
+      "learning_rate": 8.833652007648184e-06,
+      "loss": 0.3083,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.196489898146485,
+      "learning_rate": 8.87189292543021e-06,
+      "loss": 0.4396,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3549774616852077,
+      "learning_rate": 8.910133843212237e-06,
+      "loss": 0.7806,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8069917514152943,
+      "learning_rate": 8.948374760994266e-06,
+      "loss": 0.3283,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8060819660517314,
+      "learning_rate": 8.986615678776292e-06,
+      "loss": 0.4036,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6268257383806288,
+      "learning_rate": 9.024856596558319e-06,
+      "loss": 0.2297,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8055104715664961,
+      "learning_rate": 9.063097514340345e-06,
+      "loss": 0.325,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.922017240792222,
+      "learning_rate": 9.101338432122372e-06,
+      "loss": 0.6324,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1570165829521175,
+      "learning_rate": 9.139579349904398e-06,
+      "loss": 0.4164,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0882648871617964,
+      "learning_rate": 9.177820267686425e-06,
+      "loss": 0.3779,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0650365945242295,
+      "learning_rate": 9.216061185468453e-06,
+      "loss": 0.6302,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7876721236295158,
+      "learning_rate": 9.254302103250478e-06,
+      "loss": 0.2161,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.877348325878266,
+      "learning_rate": 9.292543021032505e-06,
+      "loss": 0.3256,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2294123749330266,
+      "learning_rate": 9.330783938814533e-06,
+      "loss": 0.4789,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0693063329445263,
+      "learning_rate": 9.36902485659656e-06,
+      "loss": 0.5469,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8369746366963294,
+      "learning_rate": 9.407265774378586e-06,
+      "loss": 0.342,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9630683957243528,
+      "learning_rate": 9.445506692160612e-06,
+      "loss": 0.3704,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5500498119249109,
+      "learning_rate": 9.483747609942639e-06,
+      "loss": 0.0887,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7378523176182469,
+      "learning_rate": 9.521988527724666e-06,
+      "loss": 0.3357,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2327162595508403,
+      "learning_rate": 9.560229445506692e-06,
+      "loss": 0.4636,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9886430366490976,
+      "learning_rate": 9.59847036328872e-06,
+      "loss": 0.416,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9459014328163804,
+      "learning_rate": 9.636711281070747e-06,
+      "loss": 0.3347,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4714918256305243,
+      "learning_rate": 9.674952198852773e-06,
+      "loss": 0.6434,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6094277496484368,
+      "learning_rate": 9.7131931166348e-06,
+      "loss": 0.2764,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6710565760041877,
+      "learning_rate": 9.751434034416827e-06,
+      "loss": 0.242,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.07823406563584,
+      "learning_rate": 9.789674952198853e-06,
+      "loss": 0.8126,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7712955553118728,
+      "learning_rate": 9.827915869980881e-06,
+      "loss": 0.5811,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8150028793715761,
+      "learning_rate": 9.866156787762908e-06,
+      "loss": 0.2496,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7771432411433113,
+      "learning_rate": 9.904397705544934e-06,
+      "loss": 0.3765,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6160523614998065,
+      "learning_rate": 9.942638623326961e-06,
+      "loss": 0.2061,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7046157767914728,
+      "learning_rate": 9.980879541108988e-06,
+      "loss": 0.232,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.149024983560915,
+      "learning_rate": 1.0019120458891014e-05,
+      "loss": 0.4943,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2286307024685774,
+      "learning_rate": 1.005736137667304e-05,
+      "loss": 0.4477,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.097112552920304,
+      "learning_rate": 1.0095602294455067e-05,
+      "loss": 0.3751,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7614533976101915,
+      "learning_rate": 1.0133843212237095e-05,
+      "loss": 0.3311,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7984309540612244,
+      "learning_rate": 1.0172084130019122e-05,
+      "loss": 0.3827,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.564765501949692,
+      "learning_rate": 1.0210325047801149e-05,
+      "loss": 0.213,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.75929373969915,
+      "learning_rate": 1.0248565965583175e-05,
+      "loss": 0.3951,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.166353414260506,
+      "learning_rate": 1.0286806883365202e-05,
+      "loss": 0.6514,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49891390595289997,
+      "learning_rate": 1.0325047801147228e-05,
+      "loss": 0.2669,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7155232121133612,
+      "learning_rate": 1.0363288718929255e-05,
+      "loss": 0.3295,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4276693546072683,
+      "learning_rate": 1.0401529636711283e-05,
+      "loss": 0.8387,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7231598135654922,
+      "learning_rate": 1.043977055449331e-05,
+      "loss": 0.3431,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7399633483574997,
+      "learning_rate": 1.0478011472275336e-05,
+      "loss": 0.4369,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0433473187882643,
+      "learning_rate": 1.0516252390057363e-05,
+      "loss": 0.4318,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4793959174200006,
+      "learning_rate": 1.055449330783939e-05,
+      "loss": 0.2016,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1899034899997953,
+      "learning_rate": 1.0592734225621416e-05,
+      "loss": 0.4827,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7622063676502636,
+      "learning_rate": 1.0630975143403444e-05,
+      "loss": 0.3748,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0017165788816116,
+      "learning_rate": 1.066921606118547e-05,
+      "loss": 0.4185,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8309753387936472,
+      "learning_rate": 1.0707456978967497e-05,
+      "loss": 0.4142,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6895111232491048,
+      "learning_rate": 1.0745697896749524e-05,
+      "loss": 0.211,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6991545205239938,
+      "learning_rate": 1.0783938814531549e-05,
+      "loss": 0.3342,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8430434837260139,
+      "learning_rate": 1.0822179732313575e-05,
+      "loss": 0.3487,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4751180199401044,
+      "learning_rate": 1.0860420650095602e-05,
+      "loss": 0.5592,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1958526321793117,
+      "learning_rate": 1.0898661567877632e-05,
+      "loss": 0.4458,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.893083908512126,
+      "learning_rate": 1.0936902485659658e-05,
+      "loss": 0.3897,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1128175053171323,
+      "learning_rate": 1.0975143403441683e-05,
+      "loss": 0.3729,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7310765740074888,
+      "learning_rate": 1.101338432122371e-05,
+      "loss": 0.1865,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.146099557909946,
+      "learning_rate": 1.1051625239005736e-05,
+      "loss": 0.3986,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8860784783630814,
+      "learning_rate": 1.1089866156787763e-05,
+      "loss": 0.4463,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7447706908798476,
+      "learning_rate": 1.1128107074569791e-05,
+      "loss": 0.3275,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.202436710568476,
+      "learning_rate": 1.1166347992351817e-05,
+      "loss": 0.5282,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7152994728714328,
+      "learning_rate": 1.1204588910133844e-05,
+      "loss": 0.7841,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5435181164029416,
+      "learning_rate": 1.124282982791587e-05,
+      "loss": 0.2422,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6497594633770308,
+      "learning_rate": 1.1281070745697897e-05,
+      "loss": 0.315,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4737099800720082,
+      "learning_rate": 1.1319311663479924e-05,
+      "loss": 0.6034,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.60322244237985,
+      "learning_rate": 1.135755258126195e-05,
+      "loss": 0.3813,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7002085464698281,
+      "learning_rate": 1.1395793499043978e-05,
+      "loss": 0.3664,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.951906463609564,
+      "learning_rate": 1.1434034416826005e-05,
+      "loss": 0.4852,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44217514977603756,
+      "learning_rate": 1.1472275334608032e-05,
+      "loss": 0.1206,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6930363078552553,
+      "learning_rate": 1.1510516252390058e-05,
+      "loss": 0.3479,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5346406725473178,
+      "learning_rate": 1.1548757170172085e-05,
+      "loss": 0.4308,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.83215837021332,
+      "learning_rate": 1.1586998087954111e-05,
+      "loss": 0.5532,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6617004606774994,
+      "learning_rate": 1.162523900573614e-05,
+      "loss": 0.2918,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7033213061876957,
+      "learning_rate": 1.1663479923518166e-05,
+      "loss": 0.8174,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5124769881897817,
+      "learning_rate": 1.1701720841300193e-05,
+      "loss": 0.2797,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5573421129903319,
+      "learning_rate": 1.1739961759082219e-05,
+      "loss": 0.2435,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.936707015756441,
+      "learning_rate": 1.1778202676864246e-05,
+      "loss": 0.587,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.101229402299148,
+      "learning_rate": 1.1816443594646272e-05,
+      "loss": 0.5648,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5718313616620087,
+      "learning_rate": 1.1854684512428299e-05,
+      "loss": 0.2873,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9736432054583405,
+      "learning_rate": 1.1892925430210327e-05,
+      "loss": 0.4525,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5300437727042384,
+      "learning_rate": 1.1931166347992354e-05,
+      "loss": 0.1757,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6818941883625804,
+      "learning_rate": 1.196940726577438e-05,
+      "loss": 0.2652,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0449214685551322,
+      "learning_rate": 1.2007648183556407e-05,
+      "loss": 0.4432,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3218968880735233,
+      "learning_rate": 1.2045889101338433e-05,
+      "loss": 0.5656,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.675172558207587,
+      "learning_rate": 1.208413001912046e-05,
+      "loss": 0.3515,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9376915018087,
+      "learning_rate": 1.2122370936902486e-05,
+      "loss": 0.6026,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7527951764348736,
+      "learning_rate": 1.2160611854684515e-05,
+      "loss": 0.39,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6271551344942625,
+      "learning_rate": 1.2198852772466541e-05,
+      "loss": 0.3249,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9520118512963162,
+      "learning_rate": 1.2237093690248568e-05,
+      "loss": 0.2872,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9489835344276659,
+      "learning_rate": 1.2275334608030594e-05,
+      "loss": 0.4268,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7608542124414743,
+      "learning_rate": 1.231357552581262e-05,
+      "loss": 0.3252,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.06054704906878,
+      "learning_rate": 1.2351816443594646e-05,
+      "loss": 0.692,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0376904064881367,
+      "learning_rate": 1.2390057361376676e-05,
+      "loss": 0.4878,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7936001862726354,
+      "learning_rate": 1.2428298279158702e-05,
+      "loss": 0.3846,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7655996437533465,
+      "learning_rate": 1.2466539196940729e-05,
+      "loss": 0.2969,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6075196309967965,
+      "learning_rate": 1.2504780114722753e-05,
+      "loss": 0.322,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8333976096028658,
+      "learning_rate": 1.254302103250478e-05,
+      "loss": 0.3605,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.395249474296435,
+      "learning_rate": 1.2581261950286807e-05,
+      "loss": 0.8674,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9426907172159265,
+      "learning_rate": 1.2619502868068833e-05,
+      "loss": 0.3545,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6597426619503529,
+      "learning_rate": 1.2657743785850863e-05,
+      "loss": 0.3462,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9277589366404783,
+      "learning_rate": 1.2695984703632888e-05,
+      "loss": 0.4374,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49886983704319643,
+      "learning_rate": 1.2734225621414914e-05,
+      "loss": 0.2108,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6541306009678697,
+      "learning_rate": 1.2772466539196941e-05,
+      "loss": 0.3265,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7330589458418397,
+      "learning_rate": 1.2810707456978968e-05,
+      "loss": 0.765,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9292808339282705,
+      "learning_rate": 1.2848948374760994e-05,
+      "loss": 0.5198,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8238173474277146,
+      "learning_rate": 1.2887189292543022e-05,
+      "loss": 0.313,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7391319260856145,
+      "learning_rate": 1.2925430210325049e-05,
+      "loss": 0.3138,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8080684350177465,
+      "learning_rate": 1.2963671128107076e-05,
+      "loss": 0.3086,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7345914153185755,
+      "learning_rate": 1.3001912045889102e-05,
+      "loss": 0.2706,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9255928748149223,
+      "learning_rate": 1.3040152963671129e-05,
+      "loss": 0.7527,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2808433824506942,
+      "learning_rate": 1.3078393881453155e-05,
+      "loss": 0.4384,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7058183166356123,
+      "learning_rate": 1.3116634799235182e-05,
+      "loss": 0.3182,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.226462284510022,
+      "learning_rate": 1.315487571701721e-05,
+      "loss": 0.7395,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6742925329478396,
+      "learning_rate": 1.3193116634799237e-05,
+      "loss": 0.2795,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7439587116812453,
+      "learning_rate": 1.3231357552581263e-05,
+      "loss": 0.2749,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4361449975745808,
+      "learning_rate": 1.326959847036329e-05,
+      "loss": 0.7106,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2439492090019573,
+      "learning_rate": 1.3307839388145316e-05,
+      "loss": 0.589,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7435609340200998,
+      "learning_rate": 1.3346080305927343e-05,
+      "loss": 0.2519,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7689761834441273,
+      "learning_rate": 1.3384321223709371e-05,
+      "loss": 0.3878,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.604279488549852,
+      "learning_rate": 1.3422562141491398e-05,
+      "loss": 0.2728,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6789890550766914,
+      "learning_rate": 1.3460803059273424e-05,
+      "loss": 0.3179,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.115636152118617,
+      "learning_rate": 1.349904397705545e-05,
+      "loss": 0.4138,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6232254061695315,
+      "learning_rate": 1.3537284894837477e-05,
+      "loss": 0.3321,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6478378710624862,
+      "learning_rate": 1.3575525812619504e-05,
+      "loss": 0.2918,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6760264548742129,
+      "learning_rate": 1.361376673040153e-05,
+      "loss": 0.2258,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8021332779117495,
+      "learning_rate": 1.3652007648183559e-05,
+      "loss": 0.4408,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6232210579121872,
+      "learning_rate": 1.3690248565965585e-05,
+      "loss": 0.3153,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.737205410962497,
+      "learning_rate": 1.3728489483747612e-05,
+      "loss": 0.5192,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.882470126979341,
+      "learning_rate": 1.3766730401529638e-05,
+      "loss": 0.428,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6041133303431246,
+      "learning_rate": 1.3804971319311665e-05,
+      "loss": 0.3065,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8041421990760149,
+      "learning_rate": 1.3843212237093691e-05,
+      "loss": 0.3698,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7265607031963566,
+      "learning_rate": 1.388145315487572e-05,
+      "loss": 0.7857,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8204545366840523,
+      "learning_rate": 1.3919694072657746e-05,
+      "loss": 0.3723,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8724560011222986,
+      "learning_rate": 1.3957934990439773e-05,
+      "loss": 0.579,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49392242007190124,
+      "learning_rate": 1.39961759082218e-05,
+      "loss": 0.2412,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6500233155836509,
+      "learning_rate": 1.4034416826003826e-05,
+      "loss": 0.3158,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0068899962526328,
+      "learning_rate": 1.407265774378585e-05,
+      "loss": 0.3929,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1908103503569025,
+      "learning_rate": 1.4110898661567877e-05,
+      "loss": 0.3623,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6938542623801782,
+      "learning_rate": 1.4149139579349907e-05,
+      "loss": 0.3063,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1507116005912006,
+      "learning_rate": 1.4187380497131934e-05,
+      "loss": 0.5147,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7129151557497264,
+      "learning_rate": 1.4225621414913958e-05,
+      "loss": 0.1442,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5561459063539664,
+      "learning_rate": 1.4263862332695985e-05,
+      "loss": 0.3316,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7795311581768043,
+      "learning_rate": 1.4302103250478012e-05,
+      "loss": 0.3971,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1573702444835594,
+      "learning_rate": 1.4340344168260038e-05,
+      "loss": 0.5285,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8271672714492758,
+      "learning_rate": 1.4378585086042068e-05,
+      "loss": 0.346,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.22345199872424,
+      "learning_rate": 1.4416826003824093e-05,
+      "loss": 0.4845,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6311654742572094,
+      "learning_rate": 1.445506692160612e-05,
+      "loss": 0.3056,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48625752015949253,
+      "learning_rate": 1.4493307839388146e-05,
+      "loss": 0.2126,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9799198948538195,
+      "learning_rate": 1.4531548757170173e-05,
+      "loss": 0.7895,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8516318849459286,
+      "learning_rate": 1.4569789674952199e-05,
+      "loss": 0.4181,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.014567135790311,
+      "learning_rate": 1.4608030592734226e-05,
+      "loss": 0.3468,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8579154771118827,
+      "learning_rate": 1.4646271510516254e-05,
+      "loss": 0.4599,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6148793445042039,
+      "learning_rate": 1.468451242829828e-05,
+      "loss": 0.234,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5488181146615484,
+      "learning_rate": 1.4722753346080307e-05,
+      "loss": 0.2141,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6799560781486946,
+      "learning_rate": 1.4760994263862334e-05,
+      "loss": 0.3486,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3813615427326338,
+      "learning_rate": 1.479923518164436e-05,
+      "loss": 0.5956,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6414359509532044,
+      "learning_rate": 1.4837476099426387e-05,
+      "loss": 0.2277,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.586932100761519,
+      "learning_rate": 1.4875717017208415e-05,
+      "loss": 0.3545,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7838674501697059,
+      "learning_rate": 1.4913957934990441e-05,
+      "loss": 0.2907,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5658830728766918,
+      "learning_rate": 1.4952198852772468e-05,
+      "loss": 0.2235,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8068619501663652,
+      "learning_rate": 1.4990439770554495e-05,
+      "loss": 0.344,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6596708569982442,
+      "learning_rate": 1.5028680688336521e-05,
+      "loss": 0.3774,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8793793597200517,
+      "learning_rate": 1.5066921606118548e-05,
+      "loss": 0.4466,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0600744438699607,
+      "learning_rate": 1.5105162523900574e-05,
+      "loss": 0.3287,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7919519835895109,
+      "learning_rate": 1.5143403441682602e-05,
+      "loss": 0.3511,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5578491432818139,
+      "learning_rate": 1.5181644359464629e-05,
+      "loss": 0.2575,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8105475118217892,
+      "learning_rate": 1.5219885277246656e-05,
+      "loss": 0.3245,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7809252479896669,
+      "learning_rate": 1.5258126195028682e-05,
+      "loss": 0.5919,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6766528930681027,
+      "learning_rate": 1.529636711281071e-05,
+      "loss": 0.3405,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6551308795259978,
+      "learning_rate": 1.5334608030592735e-05,
+      "loss": 0.3199,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6205927231666513,
+      "learning_rate": 1.5372848948374762e-05,
+      "loss": 0.2926,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6434102165393524,
+      "learning_rate": 1.541108986615679e-05,
+      "loss": 0.2168,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7035727497301221,
+      "learning_rate": 1.5449330783938815e-05,
+      "loss": 0.3766,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6351981901932477,
+      "learning_rate": 1.548757170172084e-05,
+      "loss": 0.4042,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0733492412828631,
+      "learning_rate": 1.5525812619502868e-05,
+      "loss": 0.3453,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6931124633556552,
+      "learning_rate": 1.5564053537284895e-05,
+      "loss": 0.4334,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.344580603509804,
+      "learning_rate": 1.560229445506692e-05,
+      "loss": 0.3942,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48351949036648106,
+      "learning_rate": 1.564053537284895e-05,
+      "loss": 0.2785,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5724541069062071,
+      "learning_rate": 1.5678776290630978e-05,
+      "loss": 0.3334,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0690880110758512,
+      "learning_rate": 1.5717017208413004e-05,
+      "loss": 0.5032,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8955745333838802,
+      "learning_rate": 1.575525812619503e-05,
+      "loss": 0.4174,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6489343273710856,
+      "learning_rate": 1.5793499043977057e-05,
+      "loss": 0.354,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.00140719177433,
+      "learning_rate": 1.5831739961759084e-05,
+      "loss": 0.6871,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9468441119751057,
+      "learning_rate": 1.586998087954111e-05,
+      "loss": 0.3439,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6811167520988824,
+      "learning_rate": 1.5908221797323137e-05,
+      "loss": 0.4035,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5067638312443203,
+      "learning_rate": 1.5946462715105163e-05,
+      "loss": 0.2935,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7214352543145547,
+      "learning_rate": 1.598470363288719e-05,
+      "loss": 0.2881,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0508862135463248,
+      "learning_rate": 1.6022944550669217e-05,
+      "loss": 0.4995,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8285100746936782,
+      "learning_rate": 1.6061185468451243e-05,
+      "loss": 0.6615,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5708789432768756,
+      "learning_rate": 1.609942638623327e-05,
+      "loss": 0.3237,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7903225965702565,
+      "learning_rate": 1.61376673040153e-05,
+      "loss": 0.3891,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5582795431672651,
+      "learning_rate": 1.6175908221797326e-05,
+      "loss": 0.2004,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8407191661050721,
+      "learning_rate": 1.6214149139579353e-05,
+      "loss": 0.312,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6198087613324328,
+      "learning_rate": 1.625239005736138e-05,
+      "loss": 0.3276,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8207898222982781,
+      "learning_rate": 1.6290630975143406e-05,
+      "loss": 0.7767,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.330115693764209,
+      "learning_rate": 1.6328871892925432e-05,
+      "loss": 0.5159,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6387056248865568,
+      "learning_rate": 1.636711281070746e-05,
+      "loss": 0.3419,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6215273109764146,
+      "learning_rate": 1.6405353728489485e-05,
+      "loss": 0.3152,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9514248403782344,
+      "learning_rate": 1.6443594646271512e-05,
+      "loss": 0.2815,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9310589290777825,
+      "learning_rate": 1.648183556405354e-05,
+      "loss": 0.3134,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.858149168694772,
+      "learning_rate": 1.6520076481835565e-05,
+      "loss": 0.5817,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7691230463430312,
+      "learning_rate": 1.655831739961759e-05,
+      "loss": 0.3825,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7411479882760116,
+      "learning_rate": 1.6596558317399618e-05,
+      "loss": 0.2704,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1171597084891796,
+      "learning_rate": 1.6634799235181648e-05,
+      "loss": 0.1792,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0198428305122083,
+      "learning_rate": 1.6673040152963675e-05,
+      "loss": 0.4587,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8646954669117713,
+      "learning_rate": 1.67112810707457e-05,
+      "loss": 0.3177,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2919448984679605,
+      "learning_rate": 1.6749521988527724e-05,
+      "loss": 0.6591,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.385420612343266,
+      "learning_rate": 1.678776290630975e-05,
+      "loss": 0.5364,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8656076895746075,
+      "learning_rate": 1.6826003824091778e-05,
+      "loss": 0.3426,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8693095612262008,
+      "learning_rate": 1.6864244741873804e-05,
+      "loss": 0.2902,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.203916729584415,
+      "learning_rate": 1.6902485659655834e-05,
+      "loss": 0.9356,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6401954440906067,
+      "learning_rate": 1.694072657743786e-05,
+      "loss": 0.2328,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2732766268188778,
+      "learning_rate": 1.6978967495219887e-05,
+      "loss": 0.4489,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0743702431986437,
+      "learning_rate": 1.7017208413001914e-05,
+      "loss": 0.4417,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7493088642138847,
+      "learning_rate": 1.705544933078394e-05,
+      "loss": 0.3438,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.872577130088869,
+      "learning_rate": 1.7093690248565967e-05,
+      "loss": 0.5092,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6800381906510835,
+      "learning_rate": 1.7131931166347993e-05,
+      "loss": 0.3587,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6966844101543027,
+      "learning_rate": 1.717017208413002e-05,
+      "loss": 0.3088,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2041520310637208,
+      "learning_rate": 1.7208413001912046e-05,
+      "loss": 0.5373,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1907325057349623,
+      "learning_rate": 1.7246653919694073e-05,
+      "loss": 0.3502,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9197893826285918,
+      "learning_rate": 1.72848948374761e-05,
+      "loss": 0.3349,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8202204456029969,
+      "learning_rate": 1.7323135755258126e-05,
+      "loss": 0.391,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0635885255387458,
+      "learning_rate": 1.7361376673040153e-05,
+      "loss": 0.4402,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7312803167255396,
+      "learning_rate": 1.7399617590822183e-05,
+      "loss": 0.341,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7672739780107467,
+      "learning_rate": 1.743785850860421e-05,
+      "loss": 0.3814,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7107829663539708,
+      "learning_rate": 1.7476099426386236e-05,
+      "loss": 0.2131,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7640513015983967,
+      "learning_rate": 1.7514340344168262e-05,
+      "loss": 0.3358,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4799590833631586,
+      "learning_rate": 1.755258126195029e-05,
+      "loss": 0.7856,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6436605240448509,
+      "learning_rate": 1.7590822179732315e-05,
+      "loss": 0.3497,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7464987117639431,
+      "learning_rate": 1.7629063097514342e-05,
+      "loss": 0.3435,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7696930422997337,
+      "learning_rate": 1.766730401529637e-05,
+      "loss": 0.4322,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6025265599658696,
+      "learning_rate": 1.7705544933078395e-05,
+      "loss": 0.1584,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6378754667711887,
+      "learning_rate": 1.774378585086042e-05,
+      "loss": 0.3131,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6729866894407964,
+      "learning_rate": 1.7782026768642448e-05,
+      "loss": 0.3856,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.935070447059843,
+      "learning_rate": 1.7820267686424475e-05,
+      "loss": 0.6001,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5333343068447588,
+      "learning_rate": 1.78585086042065e-05,
+      "loss": 0.3228,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9091879424576893,
+      "learning_rate": 1.789674952198853e-05,
+      "loss": 0.4292,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5387821924240094,
+      "learning_rate": 1.7934990439770558e-05,
+      "loss": 0.2639,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5311447953582863,
+      "learning_rate": 1.7973231357552584e-05,
+      "loss": 0.2444,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.395307363682624,
+      "learning_rate": 1.801147227533461e-05,
+      "loss": 0.7412,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6504582290964441,
+      "learning_rate": 1.8049713193116637e-05,
+      "loss": 0.4336,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5715139902093684,
+      "learning_rate": 1.8087954110898664e-05,
+      "loss": 0.2398,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.125542602171137,
+      "learning_rate": 1.812619502868069e-05,
+      "loss": 0.6667,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4982371633361562,
+      "learning_rate": 1.8164435946462717e-05,
+      "loss": 0.237,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5691266935810649,
+      "learning_rate": 1.8202676864244744e-05,
+      "loss": 0.2451,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7030820342200949,
+      "learning_rate": 1.824091778202677e-05,
+      "loss": 0.4042,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3941896853612525,
+      "learning_rate": 1.8279158699808797e-05,
+      "loss": 0.5841,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6743046891263709,
+      "learning_rate": 1.8317399617590823e-05,
+      "loss": 0.3182,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.750355710951829,
+      "learning_rate": 1.835564053537285e-05,
+      "loss": 0.3391,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46562694341915173,
+      "learning_rate": 1.839388145315488e-05,
+      "loss": 0.2027,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6383543781218285,
+      "learning_rate": 1.8432122370936906e-05,
+      "loss": 0.3138,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.128120166212137,
+      "learning_rate": 1.847036328871893e-05,
+      "loss": 0.5683,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7605870557660157,
+      "learning_rate": 1.8508604206500956e-05,
+      "loss": 0.4408,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6354556583007445,
+      "learning_rate": 1.8546845124282983e-05,
+      "loss": 0.3147,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9816014322679436,
+      "learning_rate": 1.858508604206501e-05,
+      "loss": 0.5888,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5289682974012722,
+      "learning_rate": 1.8623326959847036e-05,
+      "loss": 0.2015,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6236586560186679,
+      "learning_rate": 1.8661567877629066e-05,
+      "loss": 0.3256,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6494916866609775,
+      "learning_rate": 1.8699808795411092e-05,
+      "loss": 0.3929,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1871808128906545,
+      "learning_rate": 1.873804971319312e-05,
+      "loss": 0.6403,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7214128916285852,
+      "learning_rate": 1.8776290630975145e-05,
+      "loss": 0.3347,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6602266191900171,
+      "learning_rate": 1.8814531548757172e-05,
+      "loss": 0.3601,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.694604817799548,
+      "learning_rate": 1.88527724665392e-05,
+      "loss": 0.1621,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7076032930955318,
+      "learning_rate": 1.8891013384321225e-05,
+      "loss": 0.3279,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8666075990621492,
+      "learning_rate": 1.892925430210325e-05,
+      "loss": 0.4989,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7678193345063739,
+      "learning_rate": 1.8967495219885278e-05,
+      "loss": 0.3314,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.555716909316726,
+      "learning_rate": 1.9005736137667305e-05,
+      "loss": 0.3502,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8106958346923805,
+      "learning_rate": 1.904397705544933e-05,
+      "loss": 0.4847,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5347451642599353,
+      "learning_rate": 1.9082217973231358e-05,
+      "loss": 0.1286,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6613152718234083,
+      "learning_rate": 1.9120458891013384e-05,
+      "loss": 0.3141,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9116473642098698,
+      "learning_rate": 1.9158699808795414e-05,
+      "loss": 0.4658,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.199682601869777,
+      "learning_rate": 1.919694072657744e-05,
+      "loss": 0.6147,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7314781117930744,
+      "learning_rate": 1.9235181644359467e-05,
+      "loss": 0.2652,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8325395903217065,
+      "learning_rate": 1.9273422562141494e-05,
+      "loss": 0.3995,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5062483839186255,
+      "learning_rate": 1.931166347992352e-05,
+      "loss": 0.3077,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5967051191367326,
+      "learning_rate": 1.9349904397705547e-05,
+      "loss": 0.3258,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8272368056777742,
+      "learning_rate": 1.9388145315487573e-05,
+      "loss": 0.3995,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9304286527087062,
+      "learning_rate": 1.94263862332696e-05,
+      "loss": 0.3921,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5272587543969978,
+      "learning_rate": 1.9464627151051627e-05,
+      "loss": 0.2309,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8486834700401805,
+      "learning_rate": 1.9502868068833653e-05,
+      "loss": 0.5258,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6631668030453906,
+      "learning_rate": 1.954110898661568e-05,
+      "loss": 0.4544,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.763621465095935,
+      "learning_rate": 1.9579349904397706e-05,
+      "loss": 0.2702,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5529369486422537,
+      "learning_rate": 1.9617590822179733e-05,
+      "loss": 0.3651,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.648527669233175,
+      "learning_rate": 1.9655831739961763e-05,
+      "loss": 0.3392,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8333363603111317,
+      "learning_rate": 1.969407265774379e-05,
+      "loss": 0.4271,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9578624303870017,
+      "learning_rate": 1.9732313575525816e-05,
+      "loss": 0.3442,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9338089452068773,
+      "learning_rate": 1.9770554493307842e-05,
+      "loss": 0.5805,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7261028256374711,
+      "learning_rate": 1.980879541108987e-05,
+      "loss": 0.4576,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6206482079476885,
+      "learning_rate": 1.9847036328871892e-05,
+      "loss": 0.2308,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5731488669023147,
+      "learning_rate": 1.9885277246653922e-05,
+      "loss": 0.3147,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5715431597204902,
+      "learning_rate": 1.992351816443595e-05,
+      "loss": 0.3604,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7894908009717267,
+      "learning_rate": 1.9961759082217975e-05,
+      "loss": 0.3063,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8229915510438305,
+      "learning_rate": 2e-05,
+      "loss": 0.4539,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6027096957912251,
+      "learning_rate": 1.9999999826850218e-05,
+      "loss": 0.3497,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9160865998931952,
+      "learning_rate": 1.999999930740087e-05,
+      "loss": 0.2652,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4656131298936089,
+      "learning_rate": 1.9999998441651974e-05,
+      "loss": 0.224,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0868401681216302,
+      "learning_rate": 1.999999722960356e-05,
+      "loss": 0.5024,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0309672847824662,
+      "learning_rate": 1.9999995671255675e-05,
+      "loss": 0.3227,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9537024074650399,
+      "learning_rate": 1.999999376660837e-05,
+      "loss": 0.5217,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7989227071893219,
+      "learning_rate": 1.9999991515661712e-05,
+      "loss": 0.3248,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5587548459372644,
+      "learning_rate": 1.9999988918415777e-05,
+      "loss": 0.2575,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6038947163659312,
+      "learning_rate": 1.9999985974870653e-05,
+      "loss": 0.2639,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9005065511749109,
+      "learning_rate": 1.999998268502645e-05,
+      "loss": 0.424,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7791223727687504,
+      "learning_rate": 1.9999979048883275e-05,
+      "loss": 0.3973,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.049952558614156,
+      "learning_rate": 1.999997506644125e-05,
+      "loss": 0.402,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.654247607069547,
+      "learning_rate": 1.9999970737700526e-05,
+      "loss": 0.3599,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.794701933156306,
+      "learning_rate": 1.999996606266124e-05,
+      "loss": 0.3839,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7277135121566177,
+      "learning_rate": 1.999996104132356e-05,
+      "loss": 0.235,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4981718868298558,
+      "learning_rate": 1.9999955673687663e-05,
+      "loss": 0.2761,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7402595774332386,
+      "learning_rate": 1.999994995975373e-05,
+      "loss": 0.3425,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5576818896204434,
+      "learning_rate": 1.9999943899521955e-05,
+      "loss": 0.587,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5357370858034539,
+      "learning_rate": 1.9999937492992558e-05,
+      "loss": 0.2118,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8512678621652553,
+      "learning_rate": 1.9999930740165755e-05,
+      "loss": 0.3902,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8553424623445812,
+      "learning_rate": 1.999992364104178e-05,
+      "loss": 0.3927,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5979085965888817,
+      "learning_rate": 1.9999916195620875e-05,
+      "loss": 0.3931,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7901713482230103,
+      "learning_rate": 1.9999908403903307e-05,
+      "loss": 0.3896,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.394859846631288,
+      "learning_rate": 1.999990026588934e-05,
+      "loss": 0.2883,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7640357579682913,
+      "learning_rate": 1.9999891781579256e-05,
+      "loss": 0.3455,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6769055305092657,
+      "learning_rate": 1.9999882950973352e-05,
+      "loss": 0.3481,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9577307589064314,
+      "learning_rate": 1.999987377407193e-05,
+      "loss": 0.7917,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9439456813870318,
+      "learning_rate": 1.9999864250875305e-05,
+      "loss": 0.2947,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8175729706645952,
+      "learning_rate": 1.9999854381383817e-05,
+      "loss": 0.3907,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4732231542439032,
+      "learning_rate": 1.99998441655978e-05,
+      "loss": 0.2284,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.238787974808822,
+      "learning_rate": 1.999983360351761e-05,
+      "loss": 0.5136,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7050079922045303,
+      "learning_rate": 1.999982269514361e-05,
+      "loss": 0.3216,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8573120183532486,
+      "learning_rate": 1.9999811440476182e-05,
+      "loss": 0.3806,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2215743690492187,
+      "learning_rate": 1.999979983951571e-05,
+      "loss": 0.5398,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6446022868528241,
+      "learning_rate": 1.9999787892262605e-05,
+      "loss": 0.3098,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5002653329984211,
+      "learning_rate": 1.9999775598717276e-05,
+      "loss": 0.2903,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5895344073161108,
+      "learning_rate": 1.9999762958880145e-05,
+      "loss": 0.2155,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8267429050787638,
+      "learning_rate": 1.9999749972751653e-05,
+      "loss": 0.3351,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.353467415655348,
+      "learning_rate": 1.999973664033225e-05,
+      "loss": 0.5793,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7361197465689773,
+      "learning_rate": 1.9999722961622395e-05,
+      "loss": 0.3886,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5447246086308706,
+      "learning_rate": 1.9999708936622564e-05,
+      "loss": 0.2858,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5984252294712458,
+      "learning_rate": 1.9999694565333246e-05,
+      "loss": 0.1902,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.482940990942458,
+      "learning_rate": 1.999967984775493e-05,
+      "loss": 0.724,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6741795289929788,
+      "learning_rate": 1.9999664783888138e-05,
+      "loss": 0.3405,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9182428191621349,
+      "learning_rate": 1.9999649373733377e-05,
+      "loss": 0.3617,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6389573056553957,
+      "learning_rate": 1.9999633617291196e-05,
+      "loss": 0.6102,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6475024165740241,
+      "learning_rate": 1.9999617514562125e-05,
+      "loss": 0.3249,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7182517478278182,
+      "learning_rate": 1.9999601065546733e-05,
+      "loss": 0.3586,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6687765532978037,
+      "learning_rate": 1.9999584270245588e-05,
+      "loss": 0.206,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7089634046898335,
+      "learning_rate": 1.9999567128659267e-05,
+      "loss": 0.3192,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.260767215457327,
+      "learning_rate": 1.9999549640788368e-05,
+      "loss": 0.5278,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6480748864054804,
+      "learning_rate": 1.9999531806633493e-05,
+      "loss": 0.3919,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.675673550540879,
+      "learning_rate": 1.9999513626195265e-05,
+      "loss": 0.3158,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6538901519919962,
+      "learning_rate": 1.9999495099474306e-05,
+      "loss": 0.831,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5293004891675508,
+      "learning_rate": 1.9999476226471265e-05,
+      "loss": 0.1482,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9542678819942969,
+      "learning_rate": 1.999945700718679e-05,
+      "loss": 0.3594,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3111617705092098,
+      "learning_rate": 1.9999437441621547e-05,
+      "loss": 0.4239,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0926709671371593,
+      "learning_rate": 1.9999417529776218e-05,
+      "loss": 0.5737,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6708490869863095,
+      "learning_rate": 1.999939727165149e-05,
+      "loss": 0.3331,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7622221883483998,
+      "learning_rate": 1.9999376667248068e-05,
+      "loss": 0.3782,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7441804908580606,
+      "learning_rate": 1.9999355716566655e-05,
+      "loss": 0.1226,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.899774384946839,
+      "learning_rate": 1.999933441960799e-05,
+      "loss": 0.3422,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4082682727776334,
+      "learning_rate": 1.9999312776372798e-05,
+      "loss": 0.6145,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9534436469990148,
+      "learning_rate": 1.9999290786861837e-05,
+      "loss": 0.3507,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7050693202744416,
+      "learning_rate": 1.999926845107587e-05,
+      "loss": 0.3421,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7025023466205644,
+      "learning_rate": 1.9999245769015663e-05,
+      "loss": 0.7674,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7588867665045335,
+      "learning_rate": 1.9999222740682004e-05,
+      "loss": 0.2988,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8053056561388221,
+      "learning_rate": 1.9999199366075694e-05,
+      "loss": 0.279,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7201821145047025,
+      "learning_rate": 1.9999175645197537e-05,
+      "loss": 0.6306,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6922097540301058,
+      "learning_rate": 1.9999151578048357e-05,
+      "loss": 0.806,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5154629765050703,
+      "learning_rate": 1.9999127164628992e-05,
+      "loss": 0.2397,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6208505950553166,
+      "learning_rate": 1.999910240494028e-05,
+      "loss": 0.3727,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6884024496186021,
+      "learning_rate": 1.9999077298983084e-05,
+      "loss": 0.2778,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7202762821630867,
+      "learning_rate": 1.9999051846758267e-05,
+      "loss": 0.2434,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2942921881997282,
+      "learning_rate": 1.999902604826672e-05,
+      "loss": 0.7034,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.772582124815898,
+      "learning_rate": 1.9998999903509326e-05,
+      "loss": 0.4107,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6564284208707677,
+      "learning_rate": 1.9998973412487e-05,
+      "loss": 0.2485,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5218916723217415,
+      "learning_rate": 1.9998946575200652e-05,
+      "loss": 0.6996,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2246507040544916,
+      "learning_rate": 1.9998919391651214e-05,
+      "loss": 0.5521,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6937961374682375,
+      "learning_rate": 1.9998891861839627e-05,
+      "loss": 0.3197,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0318409532385129,
+      "learning_rate": 1.9998863985766845e-05,
+      "loss": 0.2606,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.722572887274754,
+      "learning_rate": 1.9998835763433836e-05,
+      "loss": 0.8884,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6300069061706091,
+      "learning_rate": 1.999880719484157e-05,
+      "loss": 0.3101,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6983482017923085,
+      "learning_rate": 1.9998778279991042e-05,
+      "loss": 0.3086,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0707704364945454,
+      "learning_rate": 1.999874901888325e-05,
+      "loss": 0.6135,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5749567826655014,
+      "learning_rate": 1.9998719411519215e-05,
+      "loss": 0.3335,
+      "step": 609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5182988783531453,
+      "learning_rate": 1.9998689457899955e-05,
+      "loss": 0.1555,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7617833054443183,
+      "learning_rate": 1.999865915802651e-05,
+      "loss": 0.4108,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1526576959541612,
+      "learning_rate": 1.9998628511899925e-05,
+      "loss": 0.4166,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6571422467528628,
+      "learning_rate": 1.999859751952127e-05,
+      "loss": 0.3332,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7513551490147077,
+      "learning_rate": 1.9998566180891606e-05,
+      "loss": 0.4341,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5806459633657569,
+      "learning_rate": 1.9998534496012026e-05,
+      "loss": 0.3158,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6232848400792341,
+      "learning_rate": 1.9998502464883632e-05,
+      "loss": 0.2476,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5384024602749533,
+      "learning_rate": 1.9998470087507522e-05,
+      "loss": 0.6709,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6685859795334497,
+      "learning_rate": 1.9998437363884825e-05,
+      "loss": 0.3515,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5741696913240106,
+      "learning_rate": 1.999840429401667e-05,
+      "loss": 0.3636,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5049238219312997,
+      "learning_rate": 1.9998370877904208e-05,
+      "loss": 0.5154,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.672421234183969,
+      "learning_rate": 1.9998337115548588e-05,
+      "loss": 0.2922,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2798101602612184,
+      "learning_rate": 1.999830300695099e-05,
+      "loss": 0.3201,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6970423165330458,
+      "learning_rate": 1.9998268552112586e-05,
+      "loss": 0.3156,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4610523510184624,
+      "learning_rate": 1.999823375103457e-05,
+      "loss": 0.2628,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7809585438545255,
+      "learning_rate": 1.9998198603718148e-05,
+      "loss": 0.6929,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7009702901425052,
+      "learning_rate": 1.9998163110164543e-05,
+      "loss": 0.4038,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7154276273247365,
+      "learning_rate": 1.9998127270374975e-05,
+      "loss": 0.3343,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6401189026507259,
+      "learning_rate": 1.999809108435069e-05,
+      "loss": 0.2824,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1197570166445445,
+      "learning_rate": 1.9998054552092943e-05,
+      "loss": 0.3704,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6645133821988366,
+      "learning_rate": 1.9998017673602996e-05,
+      "loss": 0.2726,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6906799629741556,
+      "learning_rate": 1.9997980448882125e-05,
+      "loss": 0.3588,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.227238559775126,
+      "learning_rate": 1.9997942877931624e-05,
+      "loss": 0.6336,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7603717426281402,
+      "learning_rate": 1.9997904960752794e-05,
+      "loss": 0.2539,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7780626216278012,
+      "learning_rate": 1.999786669734694e-05,
+      "loss": 0.4715,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5482206816162322,
+      "learning_rate": 1.999782808771539e-05,
+      "loss": 0.3319,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7702010294483638,
+      "learning_rate": 1.999778913185949e-05,
+      "loss": 0.2772,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5466923355560228,
+      "learning_rate": 1.9997749829780577e-05,
+      "loss": 0.3879,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7200638224194983,
+      "learning_rate": 1.9997710181480018e-05,
+      "loss": 0.381,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47588589352893357,
+      "learning_rate": 1.9997670186959187e-05,
+      "loss": 0.288,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8159078143196873,
+      "learning_rate": 1.999762984621947e-05,
+      "loss": 0.4299,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.764774529343326,
+      "learning_rate": 1.9997589159262255e-05,
+      "loss": 0.3122,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5095595388066965,
+      "learning_rate": 1.9997548126088963e-05,
+      "loss": 0.3581,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5936175913635908,
+      "learning_rate": 1.9997506746701006e-05,
+      "loss": 0.3294,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5773573972412904,
+      "learning_rate": 1.9997465021099818e-05,
+      "loss": 0.4026,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5790983443211878,
+      "learning_rate": 1.9997422949286852e-05,
+      "loss": 0.3016,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5202482266217976,
+      "learning_rate": 1.9997380531263555e-05,
+      "loss": 0.3345,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5326923128194373,
+      "learning_rate": 1.99973377670314e-05,
+      "loss": 0.4049,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6871631996017432,
+      "learning_rate": 1.999729465659187e-05,
+      "loss": 0.4462,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5156786584531907,
+      "learning_rate": 1.9997251199946456e-05,
+      "loss": 0.2645,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.523957794870545,
+      "learning_rate": 1.999720739709666e-05,
+      "loss": 0.3247,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1519934314835216,
+      "learning_rate": 1.9997163248044008e-05,
+      "loss": 0.4499,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.714868485827706,
+      "learning_rate": 1.9997118752790016e-05,
+      "loss": 0.2821,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.732881875664929,
+      "learning_rate": 1.9997073911336234e-05,
+      "loss": 0.5607,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6503986849828309,
+      "learning_rate": 1.9997028723684213e-05,
+      "loss": 0.3507,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5726787275180806,
+      "learning_rate": 1.999698318983552e-05,
+      "loss": 0.3602,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5050293657924345,
+      "learning_rate": 1.9996937309791722e-05,
+      "loss": 0.1513,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8105239826073464,
+      "learning_rate": 1.999689108355442e-05,
+      "loss": 0.4691,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5513704345716917,
+      "learning_rate": 1.9996844511125205e-05,
+      "loss": 0.3814,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6100721289861497,
+      "learning_rate": 1.9996797592505703e-05,
+      "loss": 0.4041,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6755649661583865,
+      "learning_rate": 1.9996750327697523e-05,
+      "loss": 0.4135,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5909747328001143,
+      "learning_rate": 1.999670271670231e-05,
+      "loss": 0.3755,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5530198259517983,
+      "learning_rate": 1.9996654759521713e-05,
+      "loss": 0.2282,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0659525428191399,
+      "learning_rate": 1.999660645615739e-05,
+      "loss": 0.2963,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6884011137041578,
+      "learning_rate": 1.9996557806611017e-05,
+      "loss": 0.375,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2005303148264228,
+      "learning_rate": 1.9996508810884277e-05,
+      "loss": 0.5657,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7004073880089609,
+      "learning_rate": 1.9996459468978865e-05,
+      "loss": 0.2783,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6849488869877197,
+      "learning_rate": 1.9996409780896495e-05,
+      "loss": 0.3791,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9692201762247729,
+      "learning_rate": 1.999635974663888e-05,
+      "loss": 0.2391,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3591328381927894,
+      "learning_rate": 1.999630936620776e-05,
+      "loss": 0.3121,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6988627891771934,
+      "learning_rate": 1.9996258639604874e-05,
+      "loss": 0.3539,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9556786113936547,
+      "learning_rate": 1.999620756683198e-05,
+      "loss": 0.4806,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4342440011179771,
+      "learning_rate": 1.999615614789085e-05,
+      "loss": 0.3508,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6267491039851717,
+      "learning_rate": 1.9996104382783266e-05,
+      "loss": 0.3627,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6388301817445797,
+      "learning_rate": 1.9996052271511012e-05,
+      "loss": 0.2443,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6126860481430475,
+      "learning_rate": 1.99959998140759e-05,
+      "loss": 0.2574,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6335639107595146,
+      "learning_rate": 1.9995947010479744e-05,
+      "loss": 0.4149,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.009173507203124,
+      "learning_rate": 1.999589386072437e-05,
+      "loss": 0.6275,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5334355696559099,
+      "learning_rate": 1.9995840364811627e-05,
+      "loss": 0.3687,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5264094885639136,
+      "learning_rate": 1.999578652274336e-05,
+      "loss": 0.2631,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7406091669303169,
+      "learning_rate": 1.9995732334521432e-05,
+      "loss": 0.2581,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9294914520950267,
+      "learning_rate": 1.9995677800147728e-05,
+      "loss": 0.5048,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5268582006547304,
+      "learning_rate": 1.9995622919624127e-05,
+      "loss": 0.2644,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.848890819064537,
+      "learning_rate": 1.9995567692952537e-05,
+      "loss": 0.4564,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9587797278208916,
+      "learning_rate": 1.9995512120134867e-05,
+      "loss": 0.5373,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5326515080037058,
+      "learning_rate": 1.9995456201173044e-05,
+      "loss": 0.2536,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4332630514103188,
+      "learning_rate": 1.9995399936069e-05,
+      "loss": 0.2911,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8241672572613798,
+      "learning_rate": 1.9995343324824686e-05,
+      "loss": 0.3667,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5755331586063459,
+      "learning_rate": 1.9995286367442062e-05,
+      "loss": 0.3133,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.406266965308756,
+      "learning_rate": 1.9995229063923104e-05,
+      "loss": 0.8357,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5461542233249763,
+      "learning_rate": 1.9995171414269793e-05,
+      "loss": 0.3926,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5332452855766027,
+      "learning_rate": 1.999511341848412e-05,
+      "loss": 0.3142,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5558375093163075,
+      "learning_rate": 1.9995055076568107e-05,
+      "loss": 0.1537,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5096980092738956,
+      "learning_rate": 1.999499638852376e-05,
+      "loss": 0.3855,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.585061570533692,
+      "learning_rate": 1.999493735435312e-05,
+      "loss": 0.3167,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3571760544795732,
+      "learning_rate": 1.999487797405823e-05,
+      "loss": 0.5282,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1375980474442728,
+      "learning_rate": 1.9994818247641147e-05,
+      "loss": 0.5273,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6087244752271485,
+      "learning_rate": 1.9994758175103935e-05,
+      "loss": 0.3229,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6091437037486682,
+      "learning_rate": 1.999469775644868e-05,
+      "loss": 0.316,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0768666202960535,
+      "learning_rate": 1.999463699167747e-05,
+      "loss": 0.5091,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6652825079939793,
+      "learning_rate": 1.999457588079241e-05,
+      "loss": 0.3182,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5327514124177343,
+      "learning_rate": 1.9994514423795616e-05,
+      "loss": 0.1642,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6094186828307505,
+      "learning_rate": 1.9994452620689218e-05,
+      "loss": 0.3472,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5620751646923052,
+      "learning_rate": 1.999439047147536e-05,
+      "loss": 0.3312,
+      "step": 703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9069035528958392,
+      "learning_rate": 1.9994327976156184e-05,
+      "loss": 0.6049,
+      "step": 704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6506383954312657,
+      "learning_rate": 1.9994265134733862e-05,
+      "loss": 0.397,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45557859200370937,
+      "learning_rate": 1.999420194721057e-05,
+      "loss": 0.3243,
+      "step": 706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48790557123093586,
+      "learning_rate": 1.999413841358849e-05,
+      "loss": 0.2508,
+      "step": 707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.776312246851432,
+      "learning_rate": 1.999407453386983e-05,
+      "loss": 0.3304,
+      "step": 708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4977790934365835,
+      "learning_rate": 1.99940103080568e-05,
+      "loss": 0.3172,
+      "step": 709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6252880128089249,
+      "learning_rate": 1.999394573615162e-05,
+      "loss": 0.3972,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1498258098687408,
+      "learning_rate": 1.999388081815653e-05,
+      "loss": 0.5164,
+      "step": 711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.892056556618054,
+      "learning_rate": 1.9993815554073778e-05,
+      "loss": 0.2629,
+      "step": 712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6940538088084591,
+      "learning_rate": 1.999374994390562e-05,
+      "loss": 0.7976,
+      "step": 713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.500147764955215,
+      "learning_rate": 1.9993683987654333e-05,
+      "loss": 0.2667,
+      "step": 714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.560974892337304,
+      "learning_rate": 1.99936176853222e-05,
+      "loss": 0.3075,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.183879772416323,
+      "learning_rate": 1.9993551036911514e-05,
+      "loss": 0.5875,
+      "step": 716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5553161207454865,
+      "learning_rate": 1.9993484042424588e-05,
+      "loss": 0.3873,
+      "step": 717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4812560176886251,
+      "learning_rate": 1.9993416701863736e-05,
+      "loss": 0.2393,
+      "step": 718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3210916333620273,
+      "learning_rate": 1.9993349015231297e-05,
+      "loss": 0.8372,
+      "step": 719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5269059279825339,
+      "learning_rate": 1.999328098252961e-05,
+      "loss": 0.2579,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6285141235409536,
+      "learning_rate": 1.9993212603761032e-05,
+      "loss": 0.2367,
+      "step": 721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.564077425147305,
+      "learning_rate": 1.9993143878927933e-05,
+      "loss": 0.413,
+      "step": 722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5407667855885592,
+      "learning_rate": 1.999307480803269e-05,
+      "loss": 0.6998,
+      "step": 723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44099260765567994,
+      "learning_rate": 1.9993005391077694e-05,
+      "loss": 0.2509,
+      "step": 724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6939970930509446,
+      "learning_rate": 1.999293562806535e-05,
+      "loss": 0.4579,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5590999653786924,
+      "learning_rate": 1.999286551899808e-05,
+      "loss": 0.2503,
+      "step": 726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4594339329962618,
+      "learning_rate": 1.9992795063878304e-05,
+      "loss": 0.2918,
+      "step": 727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6897597721436182,
+      "learning_rate": 1.9992724262708466e-05,
+      "loss": 0.5618,
+      "step": 728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6934232532251842,
+      "learning_rate": 1.9992653115491015e-05,
+      "loss": 0.4563,
+      "step": 729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.457017970343513,
+      "learning_rate": 1.9992581622228418e-05,
+      "loss": 0.3187,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.331623318724993,
+      "learning_rate": 1.9992509782923148e-05,
+      "loss": 0.5542,
+      "step": 731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5444148013267605,
+      "learning_rate": 1.9992437597577692e-05,
+      "loss": 0.3955,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4182804082754334,
+      "learning_rate": 1.9992365066194554e-05,
+      "loss": 0.3364,
+      "step": 733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4630481339095031,
+      "learning_rate": 1.9992292188776244e-05,
+      "loss": 0.2866,
+      "step": 734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.576845422602978,
+      "learning_rate": 1.9992218965325283e-05,
+      "loss": 0.6871,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8118373704300686,
+      "learning_rate": 1.999214539584421e-05,
+      "loss": 0.369,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5701701832074956,
+      "learning_rate": 1.9992071480335574e-05,
+      "loss": 0.3343,
+      "step": 737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.592621937478156,
+      "learning_rate": 1.9991997218801925e-05,
+      "loss": 0.4622,
+      "step": 738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5980367247552338,
+      "learning_rate": 1.999192261124585e-05,
+      "loss": 0.3063,
+      "step": 739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40888204345849005,
+      "learning_rate": 1.9991847657669922e-05,
+      "loss": 0.2738,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6537608743254626,
+      "learning_rate": 1.9991772358076738e-05,
+      "loss": 0.3047,
+      "step": 741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5674094793046328,
+      "learning_rate": 1.999169671246891e-05,
+      "loss": 0.3443,
+      "step": 742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6513678869487316,
+      "learning_rate": 1.999162072084905e-05,
+      "loss": 0.4706,
+      "step": 743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7395606121601905,
+      "learning_rate": 1.9991544383219796e-05,
+      "loss": 0.4724,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4922112233983982,
+      "learning_rate": 1.999146769958379e-05,
+      "loss": 0.312,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4493305840997312,
+      "learning_rate": 1.999139066994369e-05,
+      "loss": 0.2616,
+      "step": 746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.197209300931008,
+      "learning_rate": 1.9991313294302155e-05,
+      "loss": 0.4788,
+      "step": 747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4994103566217086,
+      "learning_rate": 1.9991235572661874e-05,
+      "loss": 0.3094,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6447067429291433,
+      "learning_rate": 1.9991157505025534e-05,
+      "loss": 0.9033,
+      "step": 749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5606319354875383,
+      "learning_rate": 1.999107909139584e-05,
+      "loss": 0.3895,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5734727183073857,
+      "learning_rate": 1.9991000331775506e-05,
+      "loss": 0.3423,
+      "step": 751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.569467318382504,
+      "learning_rate": 1.999092122616726e-05,
+      "loss": 0.2506,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5087244938226237,
+      "learning_rate": 1.9990841774573843e-05,
+      "loss": 0.331,
+      "step": 753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34419700613207677,
+      "learning_rate": 1.9990761976998e-05,
+      "loss": 0.0829,
+      "step": 754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5802236112684132,
+      "learning_rate": 1.9990681833442503e-05,
+      "loss": 0.3881,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9103632693548557,
+      "learning_rate": 1.9990601343910126e-05,
+      "loss": 0.6007,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45808385012447494,
+      "learning_rate": 1.999052050840365e-05,
+      "loss": 0.2621,
+      "step": 757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4230696341008969,
+      "learning_rate": 1.9990439326925882e-05,
+      "loss": 0.2886,
+      "step": 758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.78737798009791,
+      "learning_rate": 1.9990357799479626e-05,
+      "loss": 0.342,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5573592556824364,
+      "learning_rate": 1.9990275926067713e-05,
+      "loss": 0.2124,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6546682929758058,
+      "learning_rate": 1.9990193706692972e-05,
+      "loss": 0.3749,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.533380714290272,
+      "learning_rate": 1.9990111141358252e-05,
+      "loss": 0.6221,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5379176222315797,
+      "learning_rate": 1.9990028230066413e-05,
+      "loss": 0.2695,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4715225267416928,
+      "learning_rate": 1.998994497282033e-05,
+      "loss": 0.252,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4811990499682349,
+      "learning_rate": 1.9989861369622877e-05,
+      "loss": 0.2896,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.886555411853447,
+      "learning_rate": 1.9989777420476954e-05,
+      "loss": 0.4155,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9790462773253676,
+      "learning_rate": 1.998969312538547e-05,
+      "loss": 0.3598,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1859233395253255,
+      "learning_rate": 1.9989608484351343e-05,
+      "loss": 0.676,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5154902003353229,
+      "learning_rate": 1.9989523497377505e-05,
+      "loss": 0.3098,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6483670299556568,
+      "learning_rate": 1.9989438164466896e-05,
+      "loss": 0.3439,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39362482228028733,
+      "learning_rate": 1.9989352485622472e-05,
+      "loss": 0.1467,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.74006668147056,
+      "learning_rate": 1.9989266460847207e-05,
+      "loss": 0.4542,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8813615902844071,
+      "learning_rate": 1.998918009014407e-05,
+      "loss": 0.3356,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.841448242464731,
+      "learning_rate": 1.9989093373516053e-05,
+      "loss": 0.7304,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0736502504903374,
+      "learning_rate": 1.9989006310966162e-05,
+      "loss": 0.4557,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5013227982869481,
+      "learning_rate": 1.9988918902497417e-05,
+      "loss": 0.2518,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5610738984393089,
+      "learning_rate": 1.998883114811284e-05,
+      "loss": 0.3094,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8664247677141819,
+      "learning_rate": 1.9988743047815465e-05,
+      "loss": 0.4851,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1921957110147678,
+      "learning_rate": 1.9988654601608354e-05,
+      "loss": 0.4159,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8570412985712648,
+      "learning_rate": 1.998856580949456e-05,
+      "loss": 0.6539,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5304787388263474,
+      "learning_rate": 1.998847667147716e-05,
+      "loss": 0.3169,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6695906793491527,
+      "learning_rate": 1.9988387187559245e-05,
+      "loss": 0.3891,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4991033760441603,
+      "learning_rate": 1.9988297357743906e-05,
+      "loss": 0.1626,
+      "step": 783
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6294509961056738,
+      "learning_rate": 1.9988207182034264e-05,
+      "loss": 0.4898,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7317674592716401,
+      "learning_rate": 1.9988116660433435e-05,
+      "loss": 0.3378,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1238616873997904,
+      "learning_rate": 1.9988025792944558e-05,
+      "loss": 0.3753,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7119062202083866,
+      "learning_rate": 1.9987934579570776e-05,
+      "loss": 0.3211,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7562474683264186,
+      "learning_rate": 1.9987843020315248e-05,
+      "loss": 0.3854,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5043603296184589,
+      "learning_rate": 1.9987751115181147e-05,
+      "loss": 0.3109,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0146693352319747,
+      "learning_rate": 1.9987658864171653e-05,
+      "loss": 0.5063,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42337892738407523,
+      "learning_rate": 1.9987566267289963e-05,
+      "loss": 0.2544,
+      "step": 791
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2475416323106774,
+      "learning_rate": 1.9987473324539285e-05,
+      "loss": 0.7866,
+      "step": 792
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7025050670031587,
+      "learning_rate": 1.998738003592283e-05,
+      "loss": 0.3059,
+      "step": 793
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5384616703375238,
+      "learning_rate": 1.9987286401443838e-05,
+      "loss": 0.3314,
+      "step": 794
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0796972648685363,
+      "learning_rate": 1.9987192421105546e-05,
+      "loss": 0.5532,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34313687983856905,
+      "learning_rate": 1.998709809491121e-05,
+      "loss": 0.1962,
+      "step": 796
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3985139635769342,
+      "learning_rate": 1.9987003422864094e-05,
+      "loss": 0.2577,
+      "step": 797
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6243134259537804,
+      "learning_rate": 1.998690840496748e-05,
+      "loss": 0.8693,
+      "step": 798
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7373614038171744,
+      "learning_rate": 1.9986813041224662e-05,
+      "loss": 0.2976,
+      "step": 799
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6195132227874984,
+      "learning_rate": 1.9986717331638935e-05,
+      "loss": 0.3861,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5595292667728948,
+      "learning_rate": 1.9986621276213616e-05,
+      "loss": 0.3883,
+      "step": 801
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5434227894915622,
+      "learning_rate": 1.998652487495203e-05,
+      "loss": 0.2684,
+      "step": 802
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4591481879778902,
+      "learning_rate": 1.998642812785752e-05,
+      "loss": 0.2694,
+      "step": 803
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6614380862805616,
+      "learning_rate": 1.998633103493343e-05,
+      "loss": 0.8139,
+      "step": 804
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4670262324947827,
+      "learning_rate": 1.998623359618313e-05,
+      "loss": 0.3484,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.564486768316231,
+      "learning_rate": 1.9986135811609983e-05,
+      "loss": 0.2648,
+      "step": 806
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.053932627292119,
+      "learning_rate": 1.998603768121739e-05,
+      "loss": 0.5622,
+      "step": 807
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4672012049877789,
+      "learning_rate": 1.9985939205008734e-05,
+      "loss": 0.3168,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4293014248611022,
+      "learning_rate": 1.998584038298744e-05,
+      "loss": 0.2512,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6436970530457229,
+      "learning_rate": 1.998574121515692e-05,
+      "loss": 0.3514,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0177245118126792,
+      "learning_rate": 1.998564170152061e-05,
+      "loss": 0.5856,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45703131510777767,
+      "learning_rate": 1.9985541842081957e-05,
+      "loss": 0.3059,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48741084543212293,
+      "learning_rate": 1.9985441636844424e-05,
+      "loss": 0.413,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4244781136880331,
+      "learning_rate": 1.998534108581147e-05,
+      "loss": 0.6553,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46494993418916686,
+      "learning_rate": 1.998524018898659e-05,
+      "loss": 0.2524,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5302682621937632,
+      "learning_rate": 1.9985138946373266e-05,
+      "loss": 0.3199,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43980037601885463,
+      "learning_rate": 1.9985037357975013e-05,
+      "loss": 0.3068,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.504431850220822,
+      "learning_rate": 1.9984935423795345e-05,
+      "loss": 0.3338,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8249104939952835,
+      "learning_rate": 1.998483314383779e-05,
+      "loss": 0.4644,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5695000365147458,
+      "learning_rate": 1.9984730518105897e-05,
+      "loss": 0.3879,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4473843111395704,
+      "learning_rate": 1.9984627546603214e-05,
+      "loss": 0.3239,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3833543869741045,
+      "learning_rate": 1.9984524229333307e-05,
+      "loss": 0.1974,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4942323075738801,
+      "learning_rate": 1.9984420566299756e-05,
+      "loss": 0.347,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5670416194035222,
+      "learning_rate": 1.998431655750615e-05,
+      "loss": 0.3483,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5051832663125587,
+      "learning_rate": 1.998421220295609e-05,
+      "loss": 0.3617,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6905171954763243,
+      "learning_rate": 1.9984107502653193e-05,
+      "loss": 0.7626,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4875542594639041,
+      "learning_rate": 1.9984002456601082e-05,
+      "loss": 0.2914,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6393314261217024,
+      "learning_rate": 1.9983897064803396e-05,
+      "loss": 0.4494,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5626873081314722,
+      "learning_rate": 1.9983791327263782e-05,
+      "loss": 0.4125,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4740622156552829,
+      "learning_rate": 1.9983685243985905e-05,
+      "loss": 0.3176,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45140349812898456,
+      "learning_rate": 1.9983578814973437e-05,
+      "loss": 0.2087,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.582391697868439,
+      "learning_rate": 1.9983472040230063e-05,
+      "loss": 0.3153,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.504699200239562,
+      "learning_rate": 1.998336491975948e-05,
+      "loss": 0.3052,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6280795918613231,
+      "learning_rate": 1.9983257453565402e-05,
+      "loss": 0.8253,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.538834623854235,
+      "learning_rate": 1.9983149641651546e-05,
+      "loss": 0.3547,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.661409036567593,
+      "learning_rate": 1.998304148402165e-05,
+      "loss": 0.3122,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48427368476034816,
+      "learning_rate": 1.9982932980679455e-05,
+      "loss": 0.2972,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8611669536442395,
+      "learning_rate": 1.998282413162872e-05,
+      "loss": 0.5054,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5038129864771659,
+      "learning_rate": 1.9982714936873215e-05,
+      "loss": 0.3187,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6294109039118858,
+      "learning_rate": 1.998260539641672e-05,
+      "loss": 0.4348,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.582952653039707,
+      "learning_rate": 1.998249551026303e-05,
+      "loss": 0.4443,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5631283283567661,
+      "learning_rate": 1.998238527841595e-05,
+      "loss": 0.2381,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40488497045596294,
+      "learning_rate": 1.9982274700879295e-05,
+      "loss": 0.1833,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6296325473623197,
+      "learning_rate": 1.9982163777656902e-05,
+      "loss": 0.3783,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5638688872776851,
+      "learning_rate": 1.9982052508752605e-05,
+      "loss": 0.2327,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9837522407153001,
+      "learning_rate": 1.998194089417025e-05,
+      "loss": 0.5804,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8229817506353978,
+      "learning_rate": 1.9981828933913722e-05,
+      "loss": 0.5947,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5743285905374217,
+      "learning_rate": 1.9981716627986882e-05,
+      "loss": 0.2498,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.440183500729955,
+      "learning_rate": 1.9981603976393625e-05,
+      "loss": 0.2566,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.769652766589286,
+      "learning_rate": 1.9981490979137853e-05,
+      "loss": 0.8859,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5989281680159326,
+      "learning_rate": 1.9981377636223477e-05,
+      "loss": 0.2521,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0097503020108294,
+      "learning_rate": 1.998126394765442e-05,
+      "loss": 0.4525,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7673720570657228,
+      "learning_rate": 1.9981149913434626e-05,
+      "loss": 0.4724,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4981975468603891,
+      "learning_rate": 1.9981035533568035e-05,
+      "loss": 0.3007,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.340649236428933,
+      "learning_rate": 1.998092080805862e-05,
+      "loss": 0.1132,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.652425187067407,
+      "learning_rate": 1.9980805736910337e-05,
+      "loss": 0.4076,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5959681924395248,
+      "learning_rate": 1.9980690320127188e-05,
+      "loss": 0.3472,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4708566945755925,
+      "learning_rate": 1.998057455771316e-05,
+      "loss": 0.4343,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6946441906172074,
+      "learning_rate": 1.9980458449672263e-05,
+      "loss": 0.7589,
+      "step": 859
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.546655269731215,
+      "learning_rate": 1.998034199600852e-05,
+      "loss": 0.3238,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4237690098340433,
+      "learning_rate": 1.9980225196725964e-05,
+      "loss": 0.2079,
+      "step": 861
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1400927331869855,
+      "learning_rate": 1.998010805182864e-05,
+      "loss": 0.5551,
+      "step": 862
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5594124097189067,
+      "learning_rate": 1.9979990561320597e-05,
+      "loss": 0.3234,
+      "step": 863
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8241251910035258,
+      "learning_rate": 1.9979872725205915e-05,
+      "loss": 0.3545,
+      "step": 864
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6342272397960347,
+      "learning_rate": 1.997975454348867e-05,
+      "loss": 0.7562,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5442992828218223,
+      "learning_rate": 1.9979636016172952e-05,
+      "loss": 0.3479,
+      "step": 866
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0426475164986881,
+      "learning_rate": 1.9979517143262867e-05,
+      "loss": 0.2534,
+      "step": 867
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8241038141038709,
+      "learning_rate": 1.9979397924762537e-05,
+      "loss": 0.3189,
+      "step": 868
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8868676701867368,
+      "learning_rate": 1.9979278360676082e-05,
+      "loss": 0.3811,
+      "step": 869
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8775413046921993,
+      "learning_rate": 1.9979158451007648e-05,
+      "loss": 0.4496,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.841188215520176,
+      "learning_rate": 1.9979038195761386e-05,
+      "loss": 0.5085,
+      "step": 871
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46194012023312764,
+      "learning_rate": 1.997891759494146e-05,
+      "loss": 0.2981,
+      "step": 872
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7829547374288088,
+      "learning_rate": 1.9978796648552045e-05,
+      "loss": 0.415,
+      "step": 873
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7663806220882737,
+      "learning_rate": 1.9978675356597334e-05,
+      "loss": 0.2796,
+      "step": 874
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6505647168387442,
+      "learning_rate": 1.9978553719081523e-05,
+      "loss": 0.3577,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6693024697212153,
+      "learning_rate": 1.997843173600883e-05,
+      "loss": 0.3511,
+      "step": 876
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4101677187766488,
+      "learning_rate": 1.997830940738347e-05,
+      "loss": 0.5603,
+      "step": 877
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8793601862533259,
+      "learning_rate": 1.9978186733209686e-05,
+      "loss": 0.3397,
+      "step": 878
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.587777319443702,
+      "learning_rate": 1.997806371349172e-05,
+      "loss": 0.2368,
+      "step": 879
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5639544073283919,
+      "learning_rate": 1.9977940348233845e-05,
+      "loss": 0.4186,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7457185631167834,
+      "learning_rate": 1.997781663744032e-05,
+      "loss": 0.2677,
+      "step": 881
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6447444340438591,
+      "learning_rate": 1.9977692581115436e-05,
+      "loss": 0.3951,
+      "step": 882
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6502688235687761,
+      "learning_rate": 1.9977568179263484e-05,
+      "loss": 0.8296,
+      "step": 883
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4514393605711949,
+      "learning_rate": 1.9977443431888778e-05,
+      "loss": 0.2656,
+      "step": 884
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0553305711117273,
+      "learning_rate": 1.9977318338995632e-05,
+      "loss": 0.4572,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8407200121591809,
+      "learning_rate": 1.9977192900588385e-05,
+      "loss": 0.5764,
+      "step": 886
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47881812578845484,
+      "learning_rate": 1.9977067116671374e-05,
+      "loss": 0.1567,
+      "step": 887
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4873118349825849,
+      "learning_rate": 1.9976940987248956e-05,
+      "loss": 0.3673,
+      "step": 888
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4079005001201457,
+      "learning_rate": 1.9976814512325503e-05,
+      "loss": 0.8054,
+      "step": 889
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5093928927192002,
+      "learning_rate": 1.9976687691905394e-05,
+      "loss": 0.2287,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7518705232066628,
+      "learning_rate": 1.9976560525993015e-05,
+      "loss": 0.4441,
+      "step": 891
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6356126660781017,
+      "learning_rate": 1.9976433014592776e-05,
+      "loss": 0.4113,
+      "step": 892
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.52789164129456,
+      "learning_rate": 1.9976305157709092e-05,
+      "loss": 0.2405,
+      "step": 893
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6451124335109072,
+      "learning_rate": 1.9976176955346392e-05,
+      "loss": 0.3365,
+      "step": 894
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4228472921247635,
+      "learning_rate": 1.9976048407509107e-05,
+      "loss": 0.3027,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5781676443680558,
+      "learning_rate": 1.99759195142017e-05,
+      "loss": 0.3938,
+      "step": 896
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5641365163135668,
+      "learning_rate": 1.9975790275428625e-05,
+      "loss": 0.3254,
+      "step": 897
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5717025160344602,
+      "learning_rate": 1.9975660691194365e-05,
+      "loss": 0.4194,
+      "step": 898
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0582116646632875,
+      "learning_rate": 1.99755307615034e-05,
+      "loss": 0.5222,
+      "step": 899
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4722208730822841,
+      "learning_rate": 1.997540048636024e-05,
+      "loss": 0.2897,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6038363034287668,
+      "learning_rate": 1.997526986576938e-05,
+      "loss": 0.3378,
+      "step": 901
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5519184370509896,
+      "learning_rate": 1.9975138899735366e-05,
+      "loss": 0.3994,
+      "step": 902
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4720499415776914,
+      "learning_rate": 1.9975007588262715e-05,
+      "loss": 0.3905,
+      "step": 903
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5569224560379596,
+      "learning_rate": 1.9974875931355977e-05,
+      "loss": 0.3047,
+      "step": 904
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7713492034663683,
+      "learning_rate": 1.9974743929019717e-05,
+      "loss": 0.3841,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5967554524752199,
+      "learning_rate": 1.99746115812585e-05,
+      "loss": 0.3936,
+      "step": 906
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4020352157362454,
+      "learning_rate": 1.997447888807692e-05,
+      "loss": 0.1904,
+      "step": 907
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4793739647988955,
+      "learning_rate": 1.997434584947956e-05,
+      "loss": 0.3606,
+      "step": 908
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5324891885968183,
+      "learning_rate": 1.9974212465471037e-05,
+      "loss": 0.3638,
+      "step": 909
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.711921743120288,
+      "learning_rate": 1.9974078736055963e-05,
+      "loss": 0.4732,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4754960806126236,
+      "learning_rate": 1.997394466123897e-05,
+      "loss": 0.3441,
+      "step": 911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4828296923573561,
+      "learning_rate": 1.99738102410247e-05,
+      "loss": 0.3559,
+      "step": 912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4723973954273907,
+      "learning_rate": 1.9973675475417814e-05,
+      "loss": 0.1828,
+      "step": 913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9490921519164291,
+      "learning_rate": 1.9973540364422973e-05,
+      "loss": 0.4723,
+      "step": 914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5314162635369581,
+      "learning_rate": 1.997340490804486e-05,
+      "loss": 0.3647,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48861096170235063,
+      "learning_rate": 1.9973269106288163e-05,
+      "loss": 0.387,
+      "step": 916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.635391914689005,
+      "learning_rate": 1.997313295915759e-05,
+      "loss": 0.3044,
+      "step": 917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46713708953393285,
+      "learning_rate": 1.9972996466657846e-05,
+      "loss": 0.3306,
+      "step": 918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.478474730302629,
+      "learning_rate": 1.9972859628793663e-05,
+      "loss": 0.3477,
+      "step": 919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8346975964677359,
+      "learning_rate": 1.9972722445569782e-05,
+      "loss": 0.3668,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44496567184720826,
+      "learning_rate": 1.997258491699095e-05,
+      "loss": 0.2714,
+      "step": 921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.387562812852257,
+      "learning_rate": 1.9972447043061933e-05,
+      "loss": 0.7089,
+      "step": 922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4650130501105177,
+      "learning_rate": 1.9972308823787504e-05,
+      "loss": 0.2542,
+      "step": 923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.49044841015158935,
+      "learning_rate": 1.9972170259172444e-05,
+      "loss": 0.3412,
+      "step": 924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9751351988159547,
+      "learning_rate": 1.9972031349221563e-05,
+      "loss": 0.585,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4007795809187631,
+      "learning_rate": 1.9971892093939663e-05,
+      "loss": 0.2822,
+      "step": 926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4016861009078226,
+      "learning_rate": 1.9971752493331568e-05,
+      "loss": 0.2736,
+      "step": 927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5568432011062416,
+      "learning_rate": 1.9971612547402116e-05,
+      "loss": 0.3634,
+      "step": 928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8775709744925506,
+      "learning_rate": 1.9971472256156147e-05,
+      "loss": 0.507,
+      "step": 929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5513163113977712,
+      "learning_rate": 1.997133161959852e-05,
+      "loss": 0.2609,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5854383825688924,
+      "learning_rate": 1.9971190637734113e-05,
+      "loss": 0.3781,
+      "step": 931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5414854941618819,
+      "learning_rate": 1.99710493105678e-05,
+      "loss": 0.3819,
+      "step": 932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36193619268358984,
+      "learning_rate": 1.9970907638104483e-05,
+      "loss": 0.1874,
+      "step": 933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0396706344808258,
+      "learning_rate": 1.9970765620349058e-05,
+      "loss": 0.6366,
+      "step": 934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5805743491039917,
+      "learning_rate": 1.997062325730645e-05,
+      "loss": 0.36,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4610257475867486,
+      "learning_rate": 1.997048054898159e-05,
+      "loss": 0.2515,
+      "step": 936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3399968664903208,
+      "learning_rate": 1.997033749537941e-05,
+      "loss": 0.5433,
+      "step": 937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5558017309777695,
+      "learning_rate": 1.9970194096504877e-05,
+      "loss": 0.2575,
+      "step": 938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47350587576542036,
+      "learning_rate": 1.9970050352362952e-05,
+      "loss": 0.2149,
+      "step": 939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8210880489033746,
+      "learning_rate": 1.996990626295861e-05,
+      "loss": 0.4412,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5633944668244224,
+      "learning_rate": 1.9969761828296843e-05,
+      "loss": 0.8621,
+      "step": 941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5279281352946498,
+      "learning_rate": 1.9969617048382653e-05,
+      "loss": 0.3227,
+      "step": 942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5300481438173281,
+      "learning_rate": 1.996947192322105e-05,
+      "loss": 0.3458,
+      "step": 943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7143815231199013,
+      "learning_rate": 1.9969326452817068e-05,
+      "loss": 0.4565,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5568644549729681,
+      "learning_rate": 1.9969180637175737e-05,
+      "loss": 0.3141,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4280076221992522,
+      "learning_rate": 1.9969034476302108e-05,
+      "loss": 0.0969,
+      "step": 946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5319356976602545,
+      "learning_rate": 1.996888797020125e-05,
+      "loss": 0.3553,
+      "step": 947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5872685383034628,
+      "learning_rate": 1.9968741118878224e-05,
+      "loss": 0.3226,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0190650467630764,
+      "learning_rate": 1.9968593922338125e-05,
+      "loss": 0.4937,
+      "step": 949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6087067671856272,
+      "learning_rate": 1.9968446380586045e-05,
+      "loss": 0.3468,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5114568446529008,
+      "learning_rate": 1.9968298493627096e-05,
+      "loss": 0.2784,
+      "step": 951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40876696190949857,
+      "learning_rate": 1.99681502614664e-05,
+      "loss": 0.2605,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1545115092345783,
+      "learning_rate": 1.9968001684109086e-05,
+      "loss": 0.6098,
+      "step": 953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48476631935801145,
+      "learning_rate": 1.9967852761560304e-05,
+      "loss": 0.3156,
+      "step": 954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4489912788177858,
+      "learning_rate": 1.996770349382521e-05,
+      "loss": 0.3845,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8817666801720644,
+      "learning_rate": 1.9967553880908973e-05,
+      "loss": 0.3831,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5005095482186253,
+      "learning_rate": 1.996740392281677e-05,
+      "loss": 0.3326,
+      "step": 957
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41208570683327056,
+      "learning_rate": 1.9967253619553805e-05,
+      "loss": 0.1964,
+      "step": 958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9145023410100701,
+      "learning_rate": 1.996710297112527e-05,
+      "loss": 0.3778,
+      "step": 959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5094709013843017,
+      "learning_rate": 1.9966951977536387e-05,
+      "loss": 0.3345,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1238884116633348,
+      "learning_rate": 1.996680063879239e-05,
+      "loss": 0.6331,
+      "step": 961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6022319418093043,
+      "learning_rate": 1.9966648954898515e-05,
+      "loss": 0.3348,
+      "step": 962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5012059527869338,
+      "learning_rate": 1.9966496925860014e-05,
+      "loss": 0.2811,
+      "step": 963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34219376584073213,
+      "learning_rate": 1.996634455168215e-05,
+      "loss": 0.2451,
+      "step": 964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2382160129106405,
+      "learning_rate": 1.9966191832370208e-05,
+      "loss": 0.4533,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5785626838524301,
+      "learning_rate": 1.9966038767929468e-05,
+      "loss": 0.3016,
+      "step": 966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5612011179574358,
+      "learning_rate": 1.9965885358365234e-05,
+      "loss": 0.3872,
+      "step": 967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4325677209096839,
+      "learning_rate": 1.996573160368282e-05,
+      "loss": 0.6171,
+      "step": 968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4714099776950298,
+      "learning_rate": 1.996557750388755e-05,
+      "loss": 0.2267,
+      "step": 969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45563509639747796,
+      "learning_rate": 1.996542305898476e-05,
+      "loss": 0.2878,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5373298863023417,
+      "learning_rate": 1.9965268268979794e-05,
+      "loss": 0.357,
+      "step": 971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.588654271043847,
+      "learning_rate": 1.996511313387802e-05,
+      "loss": 0.2835,
+      "step": 972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4807502575156875,
+      "learning_rate": 1.9964957653684804e-05,
+      "loss": 0.8575,
+      "step": 973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4442004066106364,
+      "learning_rate": 1.9964801828405536e-05,
+      "loss": 0.6195,
+      "step": 974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4943797589189964,
+      "learning_rate": 1.9964645658045607e-05,
+      "loss": 0.2719,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39577055533039807,
+      "learning_rate": 1.9964489142610426e-05,
+      "loss": 0.1685,
+      "step": 976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1935907452878245,
+      "learning_rate": 1.996433228210542e-05,
+      "loss": 0.5387,
+      "step": 977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5879709174870033,
+      "learning_rate": 1.996417507653601e-05,
+      "loss": 0.2531,
+      "step": 978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5808913193155929,
+      "learning_rate": 1.9964017525907646e-05,
+      "loss": 0.365,
+      "step": 979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8732982117118,
+      "learning_rate": 1.9963859630225786e-05,
+      "loss": 0.7228,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5626726887011435,
+      "learning_rate": 1.9963701389495896e-05,
+      "loss": 0.3331,
+      "step": 981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.77938332796842,
+      "learning_rate": 1.9963542803723452e-05,
+      "loss": 0.4066,
+      "step": 982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42358202474178924,
+      "learning_rate": 1.996338387291395e-05,
+      "loss": 0.2655,
+      "step": 983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7147571092797101,
+      "learning_rate": 1.9963224597072896e-05,
+      "loss": 0.3894,
+      "step": 984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6968938961710522,
+      "learning_rate": 1.99630649762058e-05,
+      "loss": 0.3112,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6853819380607388,
+      "learning_rate": 1.996290501031819e-05,
+      "loss": 0.4494,
+      "step": 986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4657733992528829,
+      "learning_rate": 1.996274469941561e-05,
+      "loss": 0.317,
+      "step": 987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9112005971743489,
+      "learning_rate": 1.9962584043503616e-05,
+      "loss": 0.4292,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49945616956948385,
+      "learning_rate": 1.9962423042587756e-05,
+      "loss": 0.2444,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5901638109218509,
+      "learning_rate": 1.996226169667362e-05,
+      "loss": 0.3631,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5898119279938763,
+      "learning_rate": 1.9962100005766783e-05,
+      "loss": 0.3696,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37546046574179853,
+      "learning_rate": 1.9961937969872858e-05,
+      "loss": 0.0755,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.538046268884345,
+      "learning_rate": 1.996177558899745e-05,
+      "loss": 0.3463,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7582586586626521,
+      "learning_rate": 1.9961612863146175e-05,
+      "loss": 0.4278,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4647624246924399,
+      "learning_rate": 1.9961449792324677e-05,
+      "loss": 0.2752,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46627636277037593,
+      "learning_rate": 1.9961286376538607e-05,
+      "loss": 0.2918,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4470087570646943,
+      "learning_rate": 1.996112261579361e-05,
+      "loss": 0.8131,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40123913886301504,
+      "learning_rate": 1.9960958510095373e-05,
+      "loss": 0.248,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48846965331054437,
+      "learning_rate": 1.9960794059449564e-05,
+      "loss": 0.3432,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6562528678683638,
+      "learning_rate": 1.996062926386189e-05,
+      "loss": 0.4336,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9824380086974319,
+      "learning_rate": 1.996046412333805e-05,
+      "loss": 0.472,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.70434564261008,
+      "learning_rate": 1.996029863788377e-05,
+      "loss": 0.335,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47245421383507813,
+      "learning_rate": 1.9960132807504772e-05,
+      "loss": 0.3284,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39877422949668473,
+      "learning_rate": 1.9959966632206804e-05,
+      "loss": 0.2873,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5732104169024447,
+      "learning_rate": 1.995980011199562e-05,
+      "loss": 0.3251,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5576190638879068,
+      "learning_rate": 1.9959633246876987e-05,
+      "loss": 0.3643,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48385803317161563,
+      "learning_rate": 1.995946603685668e-05,
+      "loss": 0.3651,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46973803118307117,
+      "learning_rate": 1.99592984819405e-05,
+      "loss": 0.2235,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31711744184779644,
+      "learning_rate": 1.9959130582134234e-05,
+      "loss": 0.2126,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6060670327723078,
+      "learning_rate": 1.995896233744371e-05,
+      "loss": 0.4179,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39326789263335804,
+      "learning_rate": 1.9958793747874744e-05,
+      "loss": 0.2994,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6352828886575169,
+      "learning_rate": 1.995862481343318e-05,
+      "loss": 0.4093,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.117282848504116,
+      "learning_rate": 1.9958455534124867e-05,
+      "loss": 0.6806,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38513507477952136,
+      "learning_rate": 1.9958285909955668e-05,
+      "loss": 0.2518,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49782104906396,
+      "learning_rate": 1.9958115940931454e-05,
+      "loss": 0.3526,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5610251919492635,
+      "learning_rate": 1.9957945627058115e-05,
+      "loss": 0.3765,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.349097678917595,
+      "learning_rate": 1.995777496834155e-05,
+      "loss": 0.2702,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6490771901337339,
+      "learning_rate": 1.9957603964787662e-05,
+      "loss": 0.2902,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5038053547197864,
+      "learning_rate": 1.9957432616402377e-05,
+      "loss": 0.4018,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0330461815808312,
+      "learning_rate": 1.995726092319163e-05,
+      "loss": 0.4524,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4900679690379944,
+      "learning_rate": 1.9957088885161366e-05,
+      "loss": 0.3125,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46117236213253127,
+      "learning_rate": 1.9956916502317537e-05,
+      "loss": 0.395,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4574938839171714,
+      "learning_rate": 1.9956743774666124e-05,
+      "loss": 0.2827,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36527000688924915,
+      "learning_rate": 1.99565707022131e-05,
+      "loss": 0.2012,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.771745956925933,
+      "learning_rate": 1.995639728496446e-05,
+      "loss": 0.859,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5436455796015267,
+      "learning_rate": 1.9956223522926212e-05,
+      "loss": 0.3379,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5224236889719983,
+      "learning_rate": 1.995604941610437e-05,
+      "loss": 0.3158,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8151837330528591,
+      "learning_rate": 1.9955874964504964e-05,
+      "loss": 0.5744,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31072521368728423,
+      "learning_rate": 1.995570016813404e-05,
+      "loss": 0.158,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6735818235853589,
+      "learning_rate": 1.995552502699764e-05,
+      "loss": 0.3841,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6394142288022795,
+      "learning_rate": 1.9955349541101844e-05,
+      "loss": 0.3293,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7060434285177977,
+      "learning_rate": 1.995517371045272e-05,
+      "loss": 0.4121,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7891476065439303,
+      "learning_rate": 1.9954997535056354e-05,
+      "loss": 0.361,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6378419743565197,
+      "learning_rate": 1.9954821014918857e-05,
+      "loss": 0.3563,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48004431946828835,
+      "learning_rate": 1.995464415004633e-05,
+      "loss": 0.2107,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3995032608655963,
+      "learning_rate": 1.9954466940444913e-05,
+      "loss": 0.2438,
+      "step": 1036
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1159782193087138,
+      "learning_rate": 1.9954289386120728e-05,
+      "loss": 0.5159,
+      "step": 1037
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5618332940945832,
+      "learning_rate": 1.995411148707993e-05,
+      "loss": 0.3351,
+      "step": 1038
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5215455455748942,
+      "learning_rate": 1.995393324332868e-05,
+      "loss": 0.3607,
+      "step": 1039
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8208653345275203,
+      "learning_rate": 1.9953754654873148e-05,
+      "loss": 0.5031,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3994022765006455,
+      "learning_rate": 1.995357572171952e-05,
+      "loss": 0.1645,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49644502143567504,
+      "learning_rate": 1.9953396443873996e-05,
+      "loss": 0.322,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7034835367330016,
+      "learning_rate": 1.995321682134278e-05,
+      "loss": 0.433,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7908560910972711,
+      "learning_rate": 1.995303685413209e-05,
+      "loss": 0.4042,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49210532169863536,
+      "learning_rate": 1.9952856542248168e-05,
+      "loss": 0.3592,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.646689735534446,
+      "learning_rate": 1.995267588569725e-05,
+      "loss": 0.4115,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4048654510347636,
+      "learning_rate": 1.9952494884485593e-05,
+      "loss": 0.2289,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4411817060201455,
+      "learning_rate": 1.9952313538619467e-05,
+      "loss": 0.2335,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2620590338316189,
+      "learning_rate": 1.995213184810515e-05,
+      "loss": 0.7683,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5329777387061266,
+      "learning_rate": 1.9951949812948933e-05,
+      "loss": 0.3357,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5525441866753129,
+      "learning_rate": 1.9951767433157126e-05,
+      "loss": 0.3914,
+      "step": 1051
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0809939876830499,
+      "learning_rate": 1.9951584708736038e-05,
+      "loss": 0.7235,
+      "step": 1052
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3968852627552618,
+      "learning_rate": 1.9951401639692e-05,
+      "loss": 0.1826,
+      "step": 1053
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4945020554203922,
+      "learning_rate": 1.9951218226031354e-05,
+      "loss": 0.3181,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42878235628024847,
+      "learning_rate": 1.9951034467760446e-05,
+      "loss": 0.3214,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.904246221547212,
+      "learning_rate": 1.995085036488564e-05,
+      "loss": 0.5193,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4788553603759133,
+      "learning_rate": 1.9950665917413318e-05,
+      "loss": 0.2582,
+      "step": 1057
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.497482325015419,
+      "learning_rate": 1.995048112534986e-05,
+      "loss": 0.3624,
+      "step": 1058
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1997009804190655,
+      "learning_rate": 1.995029598870167e-05,
+      "loss": 0.6325,
+      "step": 1059
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3760169516809768,
+      "learning_rate": 1.995011050747516e-05,
+      "loss": 0.1641,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0218299290087096,
+      "learning_rate": 1.994992468167675e-05,
+      "loss": 0.5896,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5325934080764194,
+      "learning_rate": 1.9949738511312872e-05,
+      "loss": 0.4082,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5189018963919375,
+      "learning_rate": 1.994955199638998e-05,
+      "loss": 0.2595,
+      "step": 1063
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8985392734797513,
+      "learning_rate": 1.994936513691453e-05,
+      "loss": 0.8639,
+      "step": 1064
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.616301320543746,
+      "learning_rate": 1.9949177932892997e-05,
+      "loss": 0.4081,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43255047491592685,
+      "learning_rate": 1.9948990384331853e-05,
+      "loss": 0.2952,
+      "step": 1066
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4972852098105318,
+      "learning_rate": 1.9948802491237608e-05,
+      "loss": 0.2134,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7883153357673944,
+      "learning_rate": 1.994861425361675e-05,
+      "loss": 0.525,
+      "step": 1068
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6785269587723624,
+      "learning_rate": 1.9948425671475816e-05,
+      "loss": 0.3279,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5178554482526371,
+      "learning_rate": 1.9948236744821327e-05,
+      "loss": 0.2993,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9167086626332699,
+      "learning_rate": 1.994804747365983e-05,
+      "loss": 0.4278,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5356860464606096,
+      "learning_rate": 1.994785785799787e-05,
+      "loss": 0.3144,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4967397434179642,
+      "learning_rate": 1.9947667897842027e-05,
+      "loss": 0.2778,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6967124996922901,
+      "learning_rate": 1.994747759319887e-05,
+      "loss": 0.3812,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48475894899954397,
+      "learning_rate": 1.994728694407499e-05,
+      "loss": 0.2961,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9083581496874299,
+      "learning_rate": 1.9947095950476992e-05,
+      "loss": 0.5082,
+      "step": 1076
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7444906820012983,
+      "learning_rate": 1.994690461241149e-05,
+      "loss": 0.3969,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46270242209941526,
+      "learning_rate": 1.994671292988511e-05,
+      "loss": 0.3147,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7264460965731782,
+      "learning_rate": 1.9946520902904485e-05,
+      "loss": 0.5275,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43704367289000573,
+      "learning_rate": 1.994632853147627e-05,
+      "loss": 0.1736,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5335354016333684,
+      "learning_rate": 1.9946135815607128e-05,
+      "loss": 0.3071,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4954050549137195,
+      "learning_rate": 1.9945942755303727e-05,
+      "loss": 0.2834,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0935397207678366,
+      "learning_rate": 1.994574935057276e-05,
+      "loss": 0.4421,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.438437072971696,
+      "learning_rate": 1.994555560142092e-05,
+      "loss": 0.3371,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8748799090587484,
+      "learning_rate": 1.9945361507854914e-05,
+      "loss": 0.5721,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49367308799806964,
+      "learning_rate": 1.9945167069881468e-05,
+      "loss": 0.3301,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5471671685840018,
+      "learning_rate": 1.9944972287507316e-05,
+      "loss": 0.2885,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44722860542267706,
+      "learning_rate": 1.99447771607392e-05,
+      "loss": 0.2388,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4399451723523636,
+      "learning_rate": 1.9944581689583878e-05,
+      "loss": 0.3063,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49743000266769444,
+      "learning_rate": 1.994438587404812e-05,
+      "loss": 0.3283,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2104431749854316,
+      "learning_rate": 1.994418971413871e-05,
+      "loss": 0.5652,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2263233603048742,
+      "learning_rate": 1.994399320986243e-05,
+      "loss": 0.7256,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5740931856541661,
+      "learning_rate": 1.99437963612261e-05,
+      "loss": 0.2438,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39197406731229295,
+      "learning_rate": 1.9943599168236526e-05,
+      "loss": 0.2557,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.245863493813506,
+      "learning_rate": 1.9943401630900543e-05,
+      "loss": 0.5469,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5865781893361174,
+      "learning_rate": 1.9943203749224986e-05,
+      "loss": 0.3007,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3374424837824597,
+      "learning_rate": 1.9943005523216713e-05,
+      "loss": 0.581,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5545592888985182,
+      "learning_rate": 1.9942806952882587e-05,
+      "loss": 0.3663,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4807505633477306,
+      "learning_rate": 1.994260803822948e-05,
+      "loss": 0.2451,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3769247168053469,
+      "learning_rate": 1.994240877926429e-05,
+      "loss": 0.2144,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7380917848688296,
+      "learning_rate": 1.994220917599391e-05,
+      "loss": 0.4029,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5737493430661074,
+      "learning_rate": 1.994200922842525e-05,
+      "loss": 0.3119,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.063385576846166,
+      "learning_rate": 1.994180893656524e-05,
+      "loss": 0.8398,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3920732373420577,
+      "learning_rate": 1.9941608300420815e-05,
+      "loss": 0.7834,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7518563438767537,
+      "learning_rate": 1.9941407319998918e-05,
+      "loss": 0.3335,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45728680330538174,
+      "learning_rate": 1.9941205995306517e-05,
+      "loss": 0.2271,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7047347617250996,
+      "learning_rate": 1.994100432635058e-05,
+      "loss": 0.5542,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7164373543084449,
+      "learning_rate": 1.9940802313138092e-05,
+      "loss": 0.3288,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4610040084120822,
+      "learning_rate": 1.994059995567604e-05,
+      "loss": 0.2943,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6608778183845305,
+      "learning_rate": 1.9940397253971447e-05,
+      "loss": 0.3997,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6010430641256433,
+      "learning_rate": 1.9940194208031322e-05,
+      "loss": 0.3541,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5892069559400017,
+      "learning_rate": 1.9939990817862696e-05,
+      "loss": 0.3829,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39423776153888224,
+      "learning_rate": 1.9939787083472616e-05,
+      "loss": 0.3134,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3783860241639249,
+      "learning_rate": 1.993958300486814e-05,
+      "loss": 0.2465,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5857598729183383,
+      "learning_rate": 1.9939378582056332e-05,
+      "loss": 0.3123,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.451790292271993,
+      "learning_rate": 1.993917381504427e-05,
+      "loss": 0.8534,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43312088945702426,
+      "learning_rate": 1.9938968703839045e-05,
+      "loss": 0.3252,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5774867354453822,
+      "learning_rate": 1.9938763248447762e-05,
+      "loss": 0.3247,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5606158673409164,
+      "learning_rate": 1.9938557448877536e-05,
+      "loss": 0.3491,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3791222671011608,
+      "learning_rate": 1.9938351305135492e-05,
+      "loss": 0.235,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9555671938828716,
+      "learning_rate": 1.993814481722877e-05,
+      "loss": 0.8035,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5317081120210266,
+      "learning_rate": 1.9937937985164518e-05,
+      "loss": 0.3285,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7753229601610528,
+      "learning_rate": 1.9937730808949905e-05,
+      "loss": 0.3667,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5828111764073042,
+      "learning_rate": 1.99375232885921e-05,
+      "loss": 0.4137,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39500188850271334,
+      "learning_rate": 1.9937315424098288e-05,
+      "loss": 0.291,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41533959678148713,
+      "learning_rate": 1.9937107215475673e-05,
+      "loss": 0.2114,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5307863364980268,
+      "learning_rate": 1.9936898662731463e-05,
+      "loss": 0.3774,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1478798383437205,
+      "learning_rate": 1.9936689765872878e-05,
+      "loss": 0.5261,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46423088674225926,
+      "learning_rate": 1.9936480524907154e-05,
+      "loss": 0.3142,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4766471909599253,
+      "learning_rate": 1.9936270939841536e-05,
+      "loss": 0.3808,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.155229588534361,
+      "learning_rate": 1.9936061010683285e-05,
+      "loss": 0.5931,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3543521865584727,
+      "learning_rate": 1.9935850737439667e-05,
+      "loss": 0.134,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4931860315991214,
+      "learning_rate": 1.9935640120117965e-05,
+      "loss": 0.3337,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.390135619937858,
+      "learning_rate": 1.9935429158725475e-05,
+      "loss": 0.8683,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5975249328681871,
+      "learning_rate": 1.9935217853269497e-05,
+      "loss": 0.3176,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.625255175254,
+      "learning_rate": 1.9935006203757354e-05,
+      "loss": 0.4517,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5236283609437692,
+      "learning_rate": 1.9934794210196374e-05,
+      "loss": 0.3806,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.25258242792276325,
+      "learning_rate": 1.9934581872593893e-05,
+      "loss": 0.1451,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6981887311269585,
+      "learning_rate": 1.9934369190957275e-05,
+      "loss": 0.4007,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8433075775942684,
+      "learning_rate": 1.9934156165293878e-05,
+      "loss": 0.6068,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4589248024137913,
+      "learning_rate": 1.9933942795611075e-05,
+      "loss": 0.2664,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5854364839569616,
+      "learning_rate": 1.9933729081916266e-05,
+      "loss": 0.4127,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1202946340701216,
+      "learning_rate": 1.9933515024216844e-05,
+      "loss": 0.6557,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32203550195598013,
+      "learning_rate": 1.9933300622520225e-05,
+      "loss": 0.1755,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47969552743697114,
+      "learning_rate": 1.9933085876833833e-05,
+      "loss": 0.2807,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0268504627365178,
+      "learning_rate": 1.99328707871651e-05,
+      "loss": 0.6002,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6890581144690542,
+      "learning_rate": 1.9932655353521483e-05,
+      "loss": 0.4666,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5003936995109405,
+      "learning_rate": 1.9932439575910436e-05,
+      "loss": 0.2906,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5008695250916106,
+      "learning_rate": 1.9932223454339435e-05,
+      "loss": 0.3429,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43940406343985966,
+      "learning_rate": 1.993200698881596e-05,
+      "loss": 0.2448,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37846887106776983,
+      "learning_rate": 1.9931790179347514e-05,
+      "loss": 0.2156,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9187364850397228,
+      "learning_rate": 1.99315730259416e-05,
+      "loss": 0.6074,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4227557946650834,
+      "learning_rate": 1.9931355528605738e-05,
+      "loss": 0.3318,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6158280113417645,
+      "learning_rate": 1.993113768734746e-05,
+      "loss": 0.3654,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1925768363660976,
+      "learning_rate": 1.9930919502174312e-05,
+      "loss": 0.4372,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2978244213918772,
+      "learning_rate": 1.993070097309385e-05,
+      "loss": 0.1655,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4938521213731747,
+      "learning_rate": 1.9930482100113642e-05,
+      "loss": 0.3306,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.026122384333359,
+      "learning_rate": 1.9930262883241265e-05,
+      "loss": 0.4183,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9371568067117475,
+      "learning_rate": 1.9930043322484306e-05,
+      "loss": 0.552,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5631135620050378,
+      "learning_rate": 1.992982341785038e-05,
+      "loss": 0.3326,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5588638465425702,
+      "learning_rate": 1.9929603169347095e-05,
+      "loss": 0.3178,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.096666482734466,
+      "learning_rate": 1.9929382576982076e-05,
+      "loss": 0.4737,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3927720395948921,
+      "learning_rate": 1.9929161640762968e-05,
+      "loss": 0.281,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7510863349671362,
+      "learning_rate": 1.992894036069742e-05,
+      "loss": 0.5086,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6070472615776253,
+      "learning_rate": 1.992871873679309e-05,
+      "loss": 0.3358,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4624380279404961,
+      "learning_rate": 1.9928496769057662e-05,
+      "loss": 0.2936,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0190631314339442,
+      "learning_rate": 1.9928274457498818e-05,
+      "loss": 0.7872,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4957965334493548,
+      "learning_rate": 1.9928051802124252e-05,
+      "loss": 0.3326,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47560370552216386,
+      "learning_rate": 1.9927828802941683e-05,
+      "loss": 0.3471,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5550644991870759,
+      "learning_rate": 1.9927605459958825e-05,
+      "loss": 0.3929,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4000080669193282,
+      "learning_rate": 1.992738177318342e-05,
+      "loss": 0.1703,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6353595781264442,
+      "learning_rate": 1.992715774262321e-05,
+      "loss": 0.3254,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5679976742869749,
+      "learning_rate": 1.992693336828596e-05,
+      "loss": 0.3704,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5135060035524516,
+      "learning_rate": 1.9926708650179426e-05,
+      "loss": 0.2752,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5082286635920554,
+      "learning_rate": 1.9926483588311402e-05,
+      "loss": 0.3722,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7864428636362328,
+      "learning_rate": 1.9926258182689677e-05,
+      "loss": 0.5978,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4977399864736148,
+      "learning_rate": 1.992603243332206e-05,
+      "loss": 0.2923,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3823089546041058,
+      "learning_rate": 1.9925806340216365e-05,
+      "loss": 0.2497,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5603124203509675,
+      "learning_rate": 1.9925579903380425e-05,
+      "loss": 0.3242,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4985678408576438,
+      "learning_rate": 1.9925353122822077e-05,
+      "loss": 0.4399,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37733182664093273,
+      "learning_rate": 1.992512599854918e-05,
+      "loss": 0.269,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1088873935769934,
+      "learning_rate": 1.9924898530569594e-05,
+      "loss": 0.6713,
+      "step": 1182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5461847080526667,
+      "learning_rate": 1.99246707188912e-05,
+      "loss": 0.3439,
+      "step": 1183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3354047490782291,
+      "learning_rate": 1.9924442563521885e-05,
+      "loss": 0.1924,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5354214457370254,
+      "learning_rate": 1.992421406446955e-05,
+      "loss": 0.3413,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9941015836479209,
+      "learning_rate": 1.9923985221742112e-05,
+      "loss": 0.5636,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4010314605169855,
+      "learning_rate": 1.992375603534749e-05,
+      "loss": 0.2394,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.578704678511741,
+      "learning_rate": 1.9923526505293623e-05,
+      "loss": 0.844,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5145164289755612,
+      "learning_rate": 1.9923296631588462e-05,
+      "loss": 0.3961,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42736151507104164,
+      "learning_rate": 1.9923066414239965e-05,
+      "loss": 0.2347,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.39926855460341876,
+      "learning_rate": 1.9922835853256103e-05,
+      "loss": 0.2363,
+      "step": 1191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6873331025968474,
+      "learning_rate": 1.9922604948644865e-05,
+      "loss": 0.4264,
+      "step": 1192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5172055478516355,
+      "learning_rate": 1.992237370041424e-05,
+      "loss": 0.3225,
+      "step": 1193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5461826946326311,
+      "learning_rate": 1.9922142108572245e-05,
+      "loss": 0.3776,
+      "step": 1194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8171086797305573,
+      "learning_rate": 1.9921910173126894e-05,
+      "loss": 0.545,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.545195871554347,
+      "learning_rate": 1.9921677894086217e-05,
+      "loss": 0.3055,
+      "step": 1196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.296707240427443,
+      "learning_rate": 1.9921445271458263e-05,
+      "loss": 0.2576,
+      "step": 1197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0695270171865263,
+      "learning_rate": 1.992121230525109e-05,
+      "loss": 0.5985,
+      "step": 1198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5373484107110249,
+      "learning_rate": 1.992097899547276e-05,
+      "loss": 0.3029,
+      "step": 1199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.521812372352844,
+      "learning_rate": 1.992074534213135e-05,
+      "loss": 0.3079,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3838297456653332,
+      "learning_rate": 1.9920511345234956e-05,
+      "loss": 0.7985,
+      "step": 1201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48375303096121197,
+      "learning_rate": 1.9920277004791682e-05,
+      "loss": 0.3086,
+      "step": 1202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.552106262015009,
+      "learning_rate": 1.992004232080964e-05,
+      "loss": 0.3058,
+      "step": 1203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5531874559226082,
+      "learning_rate": 1.9919807293296963e-05,
+      "loss": 0.3231,
+      "step": 1204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5455190994044464,
+      "learning_rate": 1.9919571922261784e-05,
+      "loss": 0.3043,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2944476032556527,
+      "learning_rate": 1.9919336207712258e-05,
+      "loss": 0.7407,
+      "step": 1206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9246456693780484,
+      "learning_rate": 1.991910014965654e-05,
+      "loss": 0.5139,
+      "step": 1207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5754654948616781,
+      "learning_rate": 1.9918863748102818e-05,
+      "loss": 0.2991,
+      "step": 1208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5043664025461552,
+      "learning_rate": 1.9918627003059266e-05,
+      "loss": 0.3907,
+      "step": 1209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3706019964753908,
+      "learning_rate": 1.9918389914534086e-05,
+      "loss": 0.2143,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5300152129471692,
+      "learning_rate": 1.9918152482535494e-05,
+      "loss": 0.3016,
+      "step": 1211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5248817577305702,
+      "learning_rate": 1.9917914707071703e-05,
+      "loss": 0.4038,
+      "step": 1212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.617387402551657,
+      "learning_rate": 1.991767658815096e-05,
+      "loss": 0.3454,
+      "step": 1213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4336728307512205,
+      "learning_rate": 1.9917438125781497e-05,
+      "loss": 0.2924,
+      "step": 1214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.72894687779715,
+      "learning_rate": 1.991719931997158e-05,
+      "loss": 0.5171,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3394246258776419,
+      "learning_rate": 1.9916960170729475e-05,
+      "loss": 0.2009,
+      "step": 1216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44296100608782896,
+      "learning_rate": 1.9916720678063467e-05,
+      "loss": 0.3119,
+      "step": 1217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2136487371261702,
+      "learning_rate": 1.9916480841981853e-05,
+      "loss": 0.6634,
+      "step": 1218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9793411528882585,
+      "learning_rate": 1.991624066249293e-05,
+      "loss": 0.6157,
+      "step": 1219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3788574100278588,
+      "learning_rate": 1.9916000139605013e-05,
+      "loss": 0.2773,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5509266148859141,
+      "learning_rate": 1.9915759273326447e-05,
+      "loss": 0.4402,
+      "step": 1221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35834376031323334,
+      "learning_rate": 1.9915518063665556e-05,
+      "loss": 0.1633,
+      "step": 1222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4119437399996039,
+      "learning_rate": 1.991527651063071e-05,
+      "loss": 0.2417,
+      "step": 1223
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6280447549338211,
+      "learning_rate": 1.9915034614230256e-05,
+      "loss": 0.4287,
+      "step": 1224
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5800209981404687,
+      "learning_rate": 1.9914792374472584e-05,
+      "loss": 0.4533,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43318599109249784,
+      "learning_rate": 1.991454979136608e-05,
+      "loss": 0.262,
+      "step": 1226
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8602780715330212,
+      "learning_rate": 1.991430686491914e-05,
+      "loss": 0.5644,
+      "step": 1227
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37876834389922714,
+      "learning_rate": 1.9914063595140184e-05,
+      "loss": 0.3045,
+      "step": 1228
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41026526676860525,
+      "learning_rate": 1.9913819982037627e-05,
+      "loss": 0.2101,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5216085347059072,
+      "learning_rate": 1.991357602561991e-05,
+      "loss": 0.2894,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9685343917475587,
+      "learning_rate": 1.9913331725895485e-05,
+      "loss": 0.5665,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49616526614313705,
+      "learning_rate": 1.9913087082872808e-05,
+      "loss": 0.3046,
+      "step": 1232
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40009900721312874,
+      "learning_rate": 1.9912842096560348e-05,
+      "loss": 0.2986,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6011999434725922,
+      "learning_rate": 1.9912596766966598e-05,
+      "loss": 0.3332,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3754282911061259,
+      "learning_rate": 1.9912351094100043e-05,
+      "loss": 0.2408,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4790995862974184,
+      "learning_rate": 1.99121050779692e-05,
+      "loss": 0.313,
+      "step": 1236
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9519799944694786,
+      "learning_rate": 1.9911858718582583e-05,
+      "loss": 0.5728,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48524535965068866,
+      "learning_rate": 1.9911612015948726e-05,
+      "loss": 0.3642,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6463784806450834,
+      "learning_rate": 1.9911364970076167e-05,
+      "loss": 0.2903,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5487653396668991,
+      "learning_rate": 1.9911117580973468e-05,
+      "loss": 0.3872,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28116441246936236,
+      "learning_rate": 1.9910869848649192e-05,
+      "loss": 0.1866,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7251262440666021,
+      "learning_rate": 1.991062177311192e-05,
+      "loss": 0.4662,
+      "step": 1242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4629712657903426,
+      "learning_rate": 1.9910373354370245e-05,
+      "loss": 0.504,
+      "step": 1243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4424282521343179,
+      "learning_rate": 1.9910124592432763e-05,
+      "loss": 0.3082,
+      "step": 1244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7088073521939233,
+      "learning_rate": 1.9909875487308096e-05,
+      "loss": 0.4297,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45330422783598706,
+      "learning_rate": 1.9909626039004862e-05,
+      "loss": 0.1286,
+      "step": 1246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6475879211850505,
+      "learning_rate": 1.990937624753171e-05,
+      "loss": 0.3348,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5626112796455328,
+      "learning_rate": 1.990912611289728e-05,
+      "loss": 0.3295,
+      "step": 1248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.169279671276457,
+      "learning_rate": 1.9908875635110244e-05,
+      "loss": 0.5194,
+      "step": 1249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7257332887197935,
+      "learning_rate": 1.990862481417927e-05,
+      "loss": 0.4001,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6408089696432202,
+      "learning_rate": 1.990837365011304e-05,
+      "loss": 0.3905,
+      "step": 1251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.628381425763534,
+      "learning_rate": 1.9908122142920262e-05,
+      "loss": 0.2872,
+      "step": 1252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.393935167434499,
+      "learning_rate": 1.9907870292609633e-05,
+      "loss": 0.1497,
+      "step": 1253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3577178025888306,
+      "learning_rate": 1.990761809918989e-05,
+      "loss": 0.4176,
+      "step": 1254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0334997177234264,
+      "learning_rate": 1.9907365562669753e-05,
+      "loss": 0.6454,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4657792094873402,
+      "learning_rate": 1.9907112683057974e-05,
+      "loss": 0.2523,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6638005020345322,
+      "learning_rate": 1.9906859460363307e-05,
+      "loss": 0.4164,
+      "step": 1257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2400648252392563,
+      "learning_rate": 1.9906605894594525e-05,
+      "loss": 0.6687,
+      "step": 1258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5219896184442447,
+      "learning_rate": 1.990635198576041e-05,
+      "loss": 0.2149,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7292828300759986,
+      "learning_rate": 1.9906097733869746e-05,
+      "loss": 0.3419,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.141928613905644,
+      "learning_rate": 1.990584313893135e-05,
+      "loss": 0.5934,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.352641454123304,
+      "learning_rate": 1.9905588200954025e-05,
+      "loss": 0.1808,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.743654610148312,
+      "learning_rate": 1.990533291994661e-05,
+      "loss": 0.4418,
+      "step": 1263
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5783868326461529,
+      "learning_rate": 1.9905077295917944e-05,
+      "loss": 0.3728,
+      "step": 1264
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6927545700509946,
+      "learning_rate": 1.9904821328876873e-05,
+      "loss": 0.2807,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9029660267620941,
+      "learning_rate": 1.9904565018832267e-05,
+      "loss": 0.4061,
+      "step": 1266
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48832856958609416,
+      "learning_rate": 1.9904308365792998e-05,
+      "loss": 0.3221,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6071012810126546,
+      "learning_rate": 1.9904051369767958e-05,
+      "loss": 0.3727,
+      "step": 1268
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4333485493904293,
+      "learning_rate": 1.9903794030766047e-05,
+      "loss": 0.216,
+      "step": 1269
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8550028384302811,
+      "learning_rate": 1.9903536348796172e-05,
+      "loss": 0.5802,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5444160978702124,
+      "learning_rate": 1.9903278323867262e-05,
+      "loss": 0.502,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49056768853210037,
+      "learning_rate": 1.9903019955988246e-05,
+      "loss": 0.2712,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1139660490114864,
+      "learning_rate": 1.9902761245168078e-05,
+      "loss": 0.659,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.36944374947304287,
+      "learning_rate": 1.990250219141571e-05,
+      "loss": 0.1944,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4744817910147028,
+      "learning_rate": 1.990224279474012e-05,
+      "loss": 0.2431,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5063465239101601,
+      "learning_rate": 1.990198305515029e-05,
+      "loss": 0.3569,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8148200431071357,
+      "learning_rate": 1.9901722972655207e-05,
+      "loss": 0.4337,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4741481971958385,
+      "learning_rate": 1.990146254726389e-05,
+      "loss": 0.3101,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5962509814169538,
+      "learning_rate": 1.9901201778985344e-05,
+      "loss": 0.4426,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38573360540688933,
+      "learning_rate": 1.9900940667828606e-05,
+      "loss": 0.259,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40658105084541624,
+      "learning_rate": 1.9900679213802724e-05,
+      "loss": 0.2474,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7175034272163154,
+      "learning_rate": 1.9900417416916742e-05,
+      "loss": 0.4537,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5669166964944146,
+      "learning_rate": 1.9900155277179734e-05,
+      "loss": 0.3968,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7815627889065767,
+      "learning_rate": 1.989989279460077e-05,
+      "loss": 0.313,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9383698484651365,
+      "learning_rate": 1.989962996918895e-05,
+      "loss": 0.5734,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4976846772814327,
+      "learning_rate": 1.9899366800953367e-05,
+      "loss": 0.2426,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38644317778329124,
+      "learning_rate": 1.9899103289903137e-05,
+      "loss": 0.2545,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.545966418703407,
+      "learning_rate": 1.9898839436047384e-05,
+      "loss": 0.3516,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8579484617086817,
+      "learning_rate": 1.989857523939525e-05,
+      "loss": 0.4972,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.455372961518016,
+      "learning_rate": 1.9898310699955884e-05,
+      "loss": 0.3282,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6335961270391565,
+      "learning_rate": 1.989804581773844e-05,
+      "loss": 0.3331,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.50902091626044,
+      "learning_rate": 1.989778059275209e-05,
+      "loss": 0.3618,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30150229710949894,
+      "learning_rate": 1.989751502500603e-05,
+      "loss": 0.2083,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9675402387132627,
+      "learning_rate": 1.9897249114509453e-05,
+      "loss": 0.533,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5243771162575476,
+      "learning_rate": 1.9896982861271558e-05,
+      "loss": 0.3046,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5198603235923646,
+      "learning_rate": 1.9896716265301577e-05,
+      "loss": 0.3362,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1509175904091664,
+      "learning_rate": 1.9896449326608734e-05,
+      "loss": 0.8126,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.23527713331733763,
+      "learning_rate": 1.9896182045202278e-05,
+      "loss": 0.1337,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5617916958589548,
+      "learning_rate": 1.989591442109146e-05,
+      "loss": 0.3786,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5253914577139904,
+      "learning_rate": 1.9895646454285558e-05,
+      "loss": 0.4054,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3003271411328984,
+      "learning_rate": 1.989537814479384e-05,
+      "loss": 0.1277,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4684808438512099,
+      "learning_rate": 1.9895109492625604e-05,
+      "loss": 0.2995,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5034967279795296,
+      "learning_rate": 1.9894840497790154e-05,
+      "loss": 0.3855,
+      "step": 1303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6015096789873883,
+      "learning_rate": 1.98945711602968e-05,
+      "loss": 0.4516,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43202696608905033,
+      "learning_rate": 1.9894301480154873e-05,
+      "loss": 0.2953,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39747979018767193,
+      "learning_rate": 1.989403145737371e-05,
+      "loss": 0.2653,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5408971085490359,
+      "learning_rate": 1.989376109196266e-05,
+      "loss": 0.3496,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4058095462488843,
+      "learning_rate": 1.9893490383931095e-05,
+      "loss": 0.2257,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6788930294125746,
+      "learning_rate": 1.989321933328838e-05,
+      "loss": 0.8471,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7626511562337264,
+      "learning_rate": 1.9892947940043906e-05,
+      "loss": 0.5442,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5049748058118498,
+      "learning_rate": 1.989267620420707e-05,
+      "loss": 0.2766,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7666799027992355,
+      "learning_rate": 1.9892404125787283e-05,
+      "loss": 0.438,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3823035156163941,
+      "learning_rate": 1.9892131704793968e-05,
+      "loss": 0.2047,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47813218026484006,
+      "learning_rate": 1.9891858941236554e-05,
+      "loss": 0.2231,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.573911958015997,
+      "learning_rate": 1.989158583512449e-05,
+      "loss": 0.377,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7540145368248665,
+      "learning_rate": 1.989131238646724e-05,
+      "loss": 0.4964,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.534381050112668,
+      "learning_rate": 1.989103859527426e-05,
+      "loss": 0.3059,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4482636946041032,
+      "learning_rate": 1.9890764461555044e-05,
+      "loss": 0.2821,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43344662744253243,
+      "learning_rate": 1.9890489985319077e-05,
+      "loss": 0.268,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5202040997266357,
+      "learning_rate": 1.989021516657587e-05,
+      "loss": 0.2804,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2201274898081262,
+      "learning_rate": 1.9889940005334932e-05,
+      "loss": 0.5302,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8568685374679005,
+      "learning_rate": 1.98896645016058e-05,
+      "loss": 0.5544,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45238693312134165,
+      "learning_rate": 1.9889388655398015e-05,
+      "loss": 0.3015,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45007359982134754,
+      "learning_rate": 1.9889112466721122e-05,
+      "loss": 0.3122,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3702787792276908,
+      "learning_rate": 1.9888835935584686e-05,
+      "loss": 0.2158,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44701134918728447,
+      "learning_rate": 1.9888559061998294e-05,
+      "loss": 0.2895,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5825053647444507,
+      "learning_rate": 1.9888281845971522e-05,
+      "loss": 0.3207,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9820095173697092,
+      "learning_rate": 1.988800428751398e-05,
+      "loss": 0.6004,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46305327622259507,
+      "learning_rate": 1.988772638663527e-05,
+      "loss": 0.3056,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3268004204954384,
+      "learning_rate": 1.9887448143345022e-05,
+      "loss": 0.6153,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3606881840853597,
+      "learning_rate": 1.988716955765287e-05,
+      "loss": 0.2031,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45448857925172526,
+      "learning_rate": 1.988689062956846e-05,
+      "loss": 0.2976,
+      "step": 1332
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8464056685167116,
+      "learning_rate": 1.9886611359101455e-05,
+      "loss": 0.4598,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8824103472910778,
+      "learning_rate": 1.9886331746261523e-05,
+      "loss": 0.4065,
+      "step": 1334
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4968644088356621,
+      "learning_rate": 1.988605179105835e-05,
+      "loss": 0.3155,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5104609846598314,
+      "learning_rate": 1.9885771493501625e-05,
+      "loss": 0.3783,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28315552118650017,
+      "learning_rate": 1.9885490853601058e-05,
+      "loss": 0.1439,
+      "step": 1337
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.441602883979966,
+      "learning_rate": 1.988520987136637e-05,
+      "loss": 0.3015,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4708608501945979,
+      "learning_rate": 1.9884928546807286e-05,
+      "loss": 0.3942,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7510346237626402,
+      "learning_rate": 1.9884646879933555e-05,
+      "loss": 0.5117,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48347633854447297,
+      "learning_rate": 1.9884364870754925e-05,
+      "loss": 0.2751,
+      "step": 1341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5360278255553699,
+      "learning_rate": 1.988408251928117e-05,
+      "loss": 0.6755,
+      "step": 1342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38517016065839627,
+      "learning_rate": 1.9883799825522056e-05,
+      "loss": 0.2931,
+      "step": 1343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.36798370244603723,
+      "learning_rate": 1.988351678948738e-05,
+      "loss": 0.1964,
+      "step": 1344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8098718422245624,
+      "learning_rate": 1.9883233411186947e-05,
+      "loss": 0.5179,
+      "step": 1345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1491988511215931,
+      "learning_rate": 1.9882949690630563e-05,
+      "loss": 0.634,
+      "step": 1346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38455018173049227,
+      "learning_rate": 1.9882665627828054e-05,
+      "loss": 0.2628,
+      "step": 1347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.134503438141174,
+      "learning_rate": 1.988238122278926e-05,
+      "loss": 0.6759,
+      "step": 1348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45620155579332167,
+      "learning_rate": 1.9882096475524032e-05,
+      "loss": 0.3142,
+      "step": 1349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6149984382187618,
+      "learning_rate": 1.988181138604223e-05,
+      "loss": 0.2522,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.534257884253937,
+      "learning_rate": 1.988152595435372e-05,
+      "loss": 0.3617,
+      "step": 1351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6039783722918978,
+      "learning_rate": 1.9881240180468394e-05,
+      "loss": 0.3739,
+      "step": 1352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5317008736940727,
+      "learning_rate": 1.988095406439614e-05,
+      "loss": 0.1945,
+      "step": 1353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7544906573407119,
+      "learning_rate": 1.9880667606146878e-05,
+      "loss": 0.4314,
+      "step": 1354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.461252463212955,
+      "learning_rate": 1.988038080573052e-05,
+      "loss": 0.3904,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6937852377504348,
+      "learning_rate": 1.9880093663157e-05,
+      "loss": 0.3911,
+      "step": 1356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5038983872790718,
+      "learning_rate": 1.9879806178436258e-05,
+      "loss": 0.355,
+      "step": 1357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6390378111434681,
+      "learning_rate": 1.9879518351578254e-05,
+      "loss": 0.2838,
+      "step": 1358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38546418075325506,
+      "learning_rate": 1.9879230182592958e-05,
+      "loss": 0.3129,
+      "step": 1359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49492171454772504,
+      "learning_rate": 1.9878941671490342e-05,
+      "loss": 0.2142,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8425537889343386,
+      "learning_rate": 1.9878652818280402e-05,
+      "loss": 0.5642,
+      "step": 1361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4003643201454913,
+      "learning_rate": 1.9878363622973137e-05,
+      "loss": 0.3595,
+      "step": 1362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43574709207441126,
+      "learning_rate": 1.987807408557857e-05,
+      "loss": 0.2916,
+      "step": 1363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.517603021457836,
+      "learning_rate": 1.987778420610672e-05,
+      "loss": 0.3182,
+      "step": 1364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40541367749597956,
+      "learning_rate": 1.9877493984567623e-05,
+      "loss": 0.2377,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.595146514308601,
+      "learning_rate": 1.9877203420971338e-05,
+      "loss": 0.3073,
+      "step": 1366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3996742522387982,
+      "learning_rate": 1.9876912515327925e-05,
+      "loss": 0.3877,
+      "step": 1367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6372451758270241,
+      "learning_rate": 1.9876621267647452e-05,
+      "loss": 0.4087,
+      "step": 1368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5014611845347008,
+      "learning_rate": 1.9876329677940015e-05,
+      "loss": 0.3651,
+      "step": 1369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4638886549372142,
+      "learning_rate": 1.9876037746215703e-05,
+      "loss": 0.3335,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5265298020277424,
+      "learning_rate": 1.9875745472484627e-05,
+      "loss": 0.2185,
+      "step": 1371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38274955304444164,
+      "learning_rate": 1.987545285675691e-05,
+      "loss": 0.261,
+      "step": 1372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7594647538424038,
+      "learning_rate": 1.9875159899042685e-05,
+      "loss": 0.4655,
+      "step": 1373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.653896701468227,
+      "learning_rate": 1.98748665993521e-05,
+      "loss": 0.3104,
+      "step": 1374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39040508633732796,
+      "learning_rate": 1.987457295769531e-05,
+      "loss": 0.3225,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0234810635667073,
+      "learning_rate": 1.9874278974082482e-05,
+      "loss": 0.5349,
+      "step": 1376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34396246269087427,
+      "learning_rate": 1.9873984648523796e-05,
+      "loss": 0.1595,
+      "step": 1377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5609481420866197,
+      "learning_rate": 1.9873689981029445e-05,
+      "loss": 0.3817,
+      "step": 1378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5179111274026793,
+      "learning_rate": 1.9873394971609636e-05,
+      "loss": 0.3453,
+      "step": 1379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46525695462885736,
+      "learning_rate": 1.9873099620274585e-05,
+      "loss": 0.327,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6728477544118068,
+      "learning_rate": 1.987280392703452e-05,
+      "loss": 0.433,
+      "step": 1381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6129052674102026,
+      "learning_rate": 1.987250789189968e-05,
+      "loss": 0.437,
+      "step": 1382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33943298098310193,
+      "learning_rate": 1.987221151488031e-05,
+      "loss": 0.215,
+      "step": 1383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4695780101175386,
+      "learning_rate": 1.9871914795986683e-05,
+      "loss": 0.2768,
+      "step": 1384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9439529862803306,
+      "learning_rate": 1.987161773522907e-05,
+      "loss": 0.5373,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4313605518517955,
+      "learning_rate": 1.9871320332617762e-05,
+      "loss": 0.2314,
+      "step": 1386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4968256366599863,
+      "learning_rate": 1.9871022588163057e-05,
+      "loss": 0.3409,
+      "step": 1387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4379547696004578,
+      "learning_rate": 1.987072450187526e-05,
+      "loss": 0.8263,
+      "step": 1388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3114063006115645,
+      "learning_rate": 1.98704260737647e-05,
+      "loss": 0.117,
+      "step": 1389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5394045691022546,
+      "learning_rate": 1.9870127303841708e-05,
+      "loss": 0.393,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6326035735789947,
+      "learning_rate": 1.9869828192116634e-05,
+      "loss": 0.3016,
+      "step": 1391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.73994607381943,
+      "learning_rate": 1.986952873859983e-05,
+      "loss": 0.4724,
+      "step": 1392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43056343619915194,
+      "learning_rate": 1.9869228943301677e-05,
+      "loss": 0.2817,
+      "step": 1393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6526090883250378,
+      "learning_rate": 1.9868928806232545e-05,
+      "loss": 0.4252,
+      "step": 1394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5255751811418452,
+      "learning_rate": 1.9868628327402833e-05,
+      "loss": 0.3748,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3704563536878913,
+      "learning_rate": 1.9868327506822948e-05,
+      "loss": 0.3383,
+      "step": 1396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4162983542173049,
+      "learning_rate": 1.9868026344503307e-05,
+      "loss": 0.1772,
+      "step": 1397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5122113190467039,
+      "learning_rate": 1.9867724840454336e-05,
+      "loss": 0.3366,
+      "step": 1398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46115541335438065,
+      "learning_rate": 1.986742299468648e-05,
+      "loss": 0.2477,
+      "step": 1399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5151180809976266,
+      "learning_rate": 1.9867120807210188e-05,
+      "loss": 0.7334,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.158542189747334,
+      "learning_rate": 1.9866818278035926e-05,
+      "loss": 0.4279,
+      "step": 1401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6994009052744022,
+      "learning_rate": 1.9866515407174174e-05,
+      "loss": 0.3324,
+      "step": 1402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0880865146791605,
+      "learning_rate": 1.9866212194635414e-05,
+      "loss": 0.2424,
+      "step": 1403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9025391903348476,
+      "learning_rate": 1.986590864043015e-05,
+      "loss": 0.4065,
+      "step": 1404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2725994136331205,
+      "learning_rate": 1.98656047445689e-05,
+      "loss": 0.3837,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1275735184080333,
+      "learning_rate": 1.9865300507062177e-05,
+      "loss": 0.3384,
+      "step": 1406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6833289165964801,
+      "learning_rate": 1.986499592792052e-05,
+      "loss": 0.501,
+      "step": 1407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7700107274879846,
+      "learning_rate": 1.9864691007154486e-05,
+      "loss": 0.3731,
+      "step": 1408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4943906235284774,
+      "learning_rate": 1.9864385744774618e-05,
+      "loss": 0.1505,
+      "step": 1409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7558856050651058,
+      "learning_rate": 1.9864080140791503e-05,
+      "loss": 0.3196,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9313219061687138,
+      "learning_rate": 1.9863774195215714e-05,
+      "loss": 0.359,
+      "step": 1411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.319474351714657,
+      "learning_rate": 1.9863467908057846e-05,
+      "loss": 0.4746,
+      "step": 1412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7970602847741426,
+      "learning_rate": 1.986316127932851e-05,
+      "loss": 0.4742,
+      "step": 1413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.602170573188123,
+      "learning_rate": 1.9862854309038324e-05,
+      "loss": 0.3773,
+      "step": 1414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9078365075980418,
+      "learning_rate": 1.9862546997197917e-05,
+      "loss": 0.2455,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6767700019137007,
+      "learning_rate": 1.9862239343817932e-05,
+      "loss": 0.2878,
+      "step": 1416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0559677425033875,
+      "learning_rate": 1.9861931348909024e-05,
+      "loss": 0.3493,
+      "step": 1417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0023429161123776,
+      "learning_rate": 1.9861623012481853e-05,
+      "loss": 0.4367,
+      "step": 1418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9368686272084427,
+      "learning_rate": 1.9861314334547105e-05,
+      "loss": 0.3789,
+      "step": 1419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9117183827974491,
+      "learning_rate": 1.9861005315115466e-05,
+      "loss": 0.3442,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3748896935498363,
+      "learning_rate": 1.9860695954197635e-05,
+      "loss": 0.3071,
+      "step": 1421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6790630927567458,
+      "learning_rate": 1.9860386251804327e-05,
+      "loss": 0.2388,
+      "step": 1422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8602801769926192,
+      "learning_rate": 1.9860076207946268e-05,
+      "loss": 0.352,
+      "step": 1423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4437574717264712,
+      "learning_rate": 1.9859765822634194e-05,
+      "loss": 0.5806,
+      "step": 1424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8911016493029239,
+      "learning_rate": 1.9859455095878853e-05,
+      "loss": 0.4489,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8138852682470921,
+      "learning_rate": 1.9859144027691006e-05,
+      "loss": 0.3231,
+      "step": 1426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3279680757064087,
+      "learning_rate": 1.9858832618081427e-05,
+      "loss": 0.2975,
+      "step": 1427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8824804769850708,
+      "learning_rate": 1.9858520867060897e-05,
+      "loss": 0.1917,
+      "step": 1428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7736102829235555,
+      "learning_rate": 1.9858208774640213e-05,
+      "loss": 0.3659,
+      "step": 1429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.003301610323338,
+      "learning_rate": 1.9857896340830182e-05,
+      "loss": 0.4005,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6699557262792868,
+      "learning_rate": 1.9857583565641627e-05,
+      "loss": 0.6112,
+      "step": 1431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7784077780986679,
+      "learning_rate": 1.9857270449085378e-05,
+      "loss": 0.2576,
+      "step": 1432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7558353459824572,
+      "learning_rate": 1.9856956991172272e-05,
+      "loss": 0.7202,
+      "step": 1433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8726908894911565,
+      "learning_rate": 1.9856643191913173e-05,
+      "loss": 0.3263,
+      "step": 1434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6394845806469756,
+      "learning_rate": 1.9856329051318942e-05,
+      "loss": 0.2556,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.364157046676915,
+      "learning_rate": 1.9856014569400463e-05,
+      "loss": 0.5657,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6999329889210596,
+      "learning_rate": 1.985569974616862e-05,
+      "loss": 0.3046,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8304526967619219,
+      "learning_rate": 1.9855384581634322e-05,
+      "loss": 0.2348,
+      "step": 1438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6726255986180282,
+      "learning_rate": 1.985506907580848e-05,
+      "loss": 0.4075,
+      "step": 1439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6573471554110082,
+      "learning_rate": 1.9854753228702016e-05,
+      "loss": 0.3843,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5886305659040076,
+      "learning_rate": 1.9854437040325872e-05,
+      "loss": 0.234,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7112981770432031,
+      "learning_rate": 1.9854120510691e-05,
+      "loss": 0.4008,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6036621461010034,
+      "learning_rate": 1.9853803639808357e-05,
+      "loss": 0.2906,
+      "step": 1443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5678057149957291,
+      "learning_rate": 1.9853486427688918e-05,
+      "loss": 0.3099,
+      "step": 1444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5707284738168328,
+      "learning_rate": 1.9853168874343665e-05,
+      "loss": 0.3151,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6543721185475424,
+      "learning_rate": 1.98528509797836e-05,
+      "loss": 0.4543,
+      "step": 1446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5152518450197626,
+      "learning_rate": 1.985253274401973e-05,
+      "loss": 0.3259,
+      "step": 1447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7630087118038243,
+      "learning_rate": 1.985221416706307e-05,
+      "loss": 0.5161,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6039997332315664,
+      "learning_rate": 1.9851895248924662e-05,
+      "loss": 0.2395,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5091314902898457,
+      "learning_rate": 1.9851575989615545e-05,
+      "loss": 0.3048,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4961746423041666,
+      "learning_rate": 1.9851256389146774e-05,
+      "loss": 0.089,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8407751460559815,
+      "learning_rate": 1.985093644752942e-05,
+      "loss": 0.5311,
+      "step": 1452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5369866218744758,
+      "learning_rate": 1.9850616164774556e-05,
+      "loss": 0.3195,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6396346126247083,
+      "learning_rate": 1.985029554089328e-05,
+      "loss": 0.3142,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6021130089990159,
+      "learning_rate": 1.9849974575896695e-05,
+      "loss": 0.2219,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8167722928601668,
+      "learning_rate": 1.984965326979591e-05,
+      "loss": 0.3154,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6078781381764322,
+      "learning_rate": 1.984933162260206e-05,
+      "loss": 0.4259,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5111068434766219,
+      "learning_rate": 1.9849009634326275e-05,
+      "loss": 0.3663,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.491679245588098,
+      "learning_rate": 1.984868730497971e-05,
+      "loss": 0.3022,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6657778643771437,
+      "learning_rate": 1.9848364634573533e-05,
+      "loss": 0.461,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4499622319663323,
+      "learning_rate": 1.984804162311891e-05,
+      "loss": 0.1674,
+      "step": 1461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43775680064572114,
+      "learning_rate": 1.9847718270627022e-05,
+      "loss": 0.2935,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9724835720183442,
+      "learning_rate": 1.9847394577109083e-05,
+      "loss": 0.6095,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6832214021917391,
+      "learning_rate": 1.984707054257629e-05,
+      "loss": 0.4568,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.553754646433227,
+      "learning_rate": 1.9846746167039864e-05,
+      "loss": 0.2975,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4793911200155322,
+      "learning_rate": 1.9846421450511045e-05,
+      "loss": 0.3857,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.553399104439865,
+      "learning_rate": 1.9846096393001074e-05,
+      "loss": 0.1008,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4279760960702405,
+      "learning_rate": 1.984577099452121e-05,
+      "loss": 0.2942,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0356861473365662,
+      "learning_rate": 1.984544525508272e-05,
+      "loss": 0.5771,
+      "step": 1469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4577243781894562,
+      "learning_rate": 1.9845119174696882e-05,
+      "loss": 0.4015,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42424685772485127,
+      "learning_rate": 1.984479275337499e-05,
+      "loss": 0.2457,
+      "step": 1471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4301253843484638,
+      "learning_rate": 1.984446599112835e-05,
+      "loss": 0.838,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3919599300947234,
+      "learning_rate": 1.9844138887968273e-05,
+      "loss": 0.304,
+      "step": 1473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3932132520222729,
+      "learning_rate": 1.9843811443906093e-05,
+      "loss": 0.1543,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6901366164674826,
+      "learning_rate": 1.9843483658953148e-05,
+      "loss": 0.4529,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7955472019255666,
+      "learning_rate": 1.9843155533120782e-05,
+      "loss": 0.5426,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47244127096043176,
+      "learning_rate": 1.9842827066420366e-05,
+      "loss": 0.2129,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4441680857479455,
+      "learning_rate": 1.9842498258863274e-05,
+      "loss": 0.362,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5322606042879695,
+      "learning_rate": 1.9842169110460885e-05,
+      "loss": 0.295,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.404128094447175,
+      "learning_rate": 1.9841839621224606e-05,
+      "loss": 0.2315,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5781529263016978,
+      "learning_rate": 1.9841509791165847e-05,
+      "loss": 0.3367,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6389770846973686,
+      "learning_rate": 1.9841179620296022e-05,
+      "loss": 0.4023,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5240438297258404,
+      "learning_rate": 1.9840849108626574e-05,
+      "loss": 0.3152,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6238085399437978,
+      "learning_rate": 1.984051825616894e-05,
+      "loss": 0.34,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5489870116923643,
+      "learning_rate": 1.9840187062934583e-05,
+      "loss": 0.3452,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3663527210417489,
+      "learning_rate": 1.9839855528934972e-05,
+      "loss": 0.2432,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6521194126080185,
+      "learning_rate": 1.983952365418159e-05,
+      "loss": 0.3821,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48007821958494157,
+      "learning_rate": 1.9839191438685922e-05,
+      "loss": 0.3019,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5161949916231795,
+      "learning_rate": 1.9838858882459483e-05,
+      "loss": 0.3056,
+      "step": 1489
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5390200772732368,
+      "learning_rate": 1.9838525985513783e-05,
+      "loss": 0.2989,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8864658034838391,
+      "learning_rate": 1.9838192747860345e-05,
+      "loss": 0.6144,
+      "step": 1491
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5180367210317303,
+      "learning_rate": 1.9837859169510723e-05,
+      "loss": 0.4018,
+      "step": 1492
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4120010395143323,
+      "learning_rate": 1.9837525250476454e-05,
+      "loss": 0.2938,
+      "step": 1493
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40632841819630106,
+      "learning_rate": 1.9837190990769115e-05,
+      "loss": 0.2472,
+      "step": 1494
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7107951744200354,
+      "learning_rate": 1.9836856390400273e-05,
+      "loss": 0.3808,
+      "step": 1495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5256133147742968,
+      "learning_rate": 1.9836521449381515e-05,
+      "loss": 0.3597,
+      "step": 1496
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43642958130027093,
+      "learning_rate": 1.9836186167724443e-05,
+      "loss": 0.3331,
+      "step": 1497
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6079517801611382,
+      "learning_rate": 1.983585054544067e-05,
+      "loss": 0.4103,
+      "step": 1498
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46354758872951146,
+      "learning_rate": 1.9835514582541812e-05,
+      "loss": 0.3661,
+      "step": 1499
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2972644570786185,
+      "learning_rate": 1.983517827903951e-05,
+      "loss": 0.1053,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3795868350126114,
+      "learning_rate": 1.9834841634945402e-05,
+      "loss": 0.2982,
+      "step": 1501
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5592247863268277,
+      "learning_rate": 1.9834504650271157e-05,
+      "loss": 0.3494,
+      "step": 1502
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6832464565355145,
+      "learning_rate": 1.9834167325028436e-05,
+      "loss": 0.4326,
+      "step": 1503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45112271937913245,
+      "learning_rate": 1.9833829659228923e-05,
+      "loss": 0.3454,
+      "step": 1504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6484508195836393,
+      "learning_rate": 1.983349165288431e-05,
+      "loss": 0.3971,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3254094644109379,
+      "learning_rate": 1.983315330600631e-05,
+      "loss": 0.2295,
+      "step": 1506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45949992766591424,
+      "learning_rate": 1.983281461860663e-05,
+      "loss": 0.2378,
+      "step": 1507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6729297248346193,
+      "learning_rate": 1.9832475590697e-05,
+      "loss": 0.4185,
+      "step": 1508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5529243330352177,
+      "learning_rate": 1.9832136222289168e-05,
+      "loss": 0.3886,
+      "step": 1509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6263171618684181,
+      "learning_rate": 1.983179651339488e-05,
+      "loss": 0.3395,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4995660700715559,
+      "learning_rate": 1.9831456464025897e-05,
+      "loss": 0.3677,
+      "step": 1511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32612760199585944,
+      "learning_rate": 1.9831116074194006e-05,
+      "loss": 0.1966,
+      "step": 1512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.507043580890754,
+      "learning_rate": 1.9830775343910984e-05,
+      "loss": 0.2792,
+      "step": 1513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5548136660748703,
+      "learning_rate": 1.9830434273188636e-05,
+      "loss": 0.3315,
+      "step": 1514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1903386244763454,
+      "learning_rate": 1.9830092862038773e-05,
+      "loss": 0.5327,
+      "step": 1515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.35362906803195593,
+      "learning_rate": 1.9829751110473215e-05,
+      "loss": 0.0803,
+      "step": 1516
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38747733634272685,
+      "learning_rate": 1.98294090185038e-05,
+      "loss": 0.3157,
+      "step": 1517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3734505271613158,
+      "learning_rate": 1.9829066586142375e-05,
+      "loss": 0.2002,
+      "step": 1518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9309793973511313,
+      "learning_rate": 1.982872381340079e-05,
+      "loss": 0.5253,
+      "step": 1519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8832068734492745,
+      "learning_rate": 1.982838070029093e-05,
+      "loss": 0.3373,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5060812038856458,
+      "learning_rate": 1.9828037246824664e-05,
+      "loss": 0.3777,
+      "step": 1521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7545887148384214,
+      "learning_rate": 1.9827693453013892e-05,
+      "loss": 0.5325,
+      "step": 1522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3855633161515275,
+      "learning_rate": 1.982734931887052e-05,
+      "loss": 0.2315,
+      "step": 1523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3295751571087598,
+      "learning_rate": 1.982700484440646e-05,
+      "loss": 0.2219,
+      "step": 1524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4242518325493269,
+      "learning_rate": 1.982666002963365e-05,
+      "loss": 0.3222,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4857458074762703,
+      "learning_rate": 1.982631487456402e-05,
+      "loss": 0.3091,
+      "step": 1526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8730410681906142,
+      "learning_rate": 1.9825969379209533e-05,
+      "loss": 0.5321,
+      "step": 1527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8632641640956932,
+      "learning_rate": 1.9825623543582145e-05,
+      "loss": 0.4488,
+      "step": 1528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38414737870140997,
+      "learning_rate": 1.982527736769384e-05,
+      "loss": 0.2726,
+      "step": 1529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4745151571052424,
+      "learning_rate": 1.9824930851556604e-05,
+      "loss": 0.3034,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9141713621897183,
+      "learning_rate": 1.982458399518243e-05,
+      "loss": 0.5761,
+      "step": 1531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4592062472149273,
+      "learning_rate": 1.9824236798583338e-05,
+      "loss": 0.3236,
+      "step": 1532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32912152990907256,
+      "learning_rate": 1.9823889261771346e-05,
+      "loss": 0.2727,
+      "step": 1533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9264614208288615,
+      "learning_rate": 1.9823541384758492e-05,
+      "loss": 0.4646,
+      "step": 1534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43380933167627445,
+      "learning_rate": 1.982319316755682e-05,
+      "loss": 0.3267,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3434369284109887,
+      "learning_rate": 1.9822844610178394e-05,
+      "loss": 0.1541,
+      "step": 1536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42805356727710603,
+      "learning_rate": 1.982249571263528e-05,
+      "loss": 0.35,
+      "step": 1537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46768921780410705,
+      "learning_rate": 1.9822146474939563e-05,
+      "loss": 0.3597,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3205738141307784,
+      "learning_rate": 1.9821796897103334e-05,
+      "loss": 0.171,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4918143379538728,
+      "learning_rate": 1.98214469791387e-05,
+      "loss": 0.3352,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38168005214386574,
+      "learning_rate": 1.9821096721057787e-05,
+      "loss": 0.3347,
+      "step": 1541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9996924570187796,
+      "learning_rate": 1.982074612287271e-05,
+      "loss": 0.4523,
+      "step": 1542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.532892984371382,
+      "learning_rate": 1.982039518459562e-05,
+      "loss": 0.4597,
+      "step": 1543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4556488663844575,
+      "learning_rate": 1.9820043906238667e-05,
+      "loss": 0.3112,
+      "step": 1544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5333818960745033,
+      "learning_rate": 1.9819692287814014e-05,
+      "loss": 0.3715,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3205794074373295,
+      "learning_rate": 1.981934032933384e-05,
+      "loss": 0.1031,
+      "step": 1546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5161864800464732,
+      "learning_rate": 1.981898803081033e-05,
+      "loss": 0.3052,
+      "step": 1547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7795833825785495,
+      "learning_rate": 1.981863539225569e-05,
+      "loss": 0.5345,
+      "step": 1548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4615357704091355,
+      "learning_rate": 1.9818282413682127e-05,
+      "loss": 0.3508,
+      "step": 1549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3946497104947286,
+      "learning_rate": 1.981792909510187e-05,
+      "loss": 0.2904,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3454321623744058,
+      "learning_rate": 1.9817575436527147e-05,
+      "loss": 0.2079,
+      "step": 1551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4495745099321259,
+      "learning_rate": 1.981722143797021e-05,
+      "loss": 0.2907,
+      "step": 1552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45876952116799197,
+      "learning_rate": 1.9816867099443314e-05,
+      "loss": 0.3153,
+      "step": 1553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1485942385955623,
+      "learning_rate": 1.9816512420958734e-05,
+      "loss": 0.5895,
+      "step": 1554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9054974881219597,
+      "learning_rate": 1.9816157402528753e-05,
+      "loss": 0.5669,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4243340339417857,
+      "learning_rate": 1.9815802044165663e-05,
+      "loss": 0.2367,
+      "step": 1556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5642073908469096,
+      "learning_rate": 1.981544634588177e-05,
+      "loss": 0.3691,
+      "step": 1557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46111573851136795,
+      "learning_rate": 1.9815090307689392e-05,
+      "loss": 0.2229,
+      "step": 1558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42590791206403383,
+      "learning_rate": 1.9814733929600857e-05,
+      "loss": 0.2272,
+      "step": 1559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9421334937751409,
+      "learning_rate": 1.981437721162851e-05,
+      "loss": 0.5379,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.561819349052626,
+      "learning_rate": 1.98140201537847e-05,
+      "loss": 0.3477,
+      "step": 1561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4240617607652223,
+      "learning_rate": 1.9813662756081794e-05,
+      "loss": 0.2214,
+      "step": 1562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0596622839127672,
+      "learning_rate": 1.9813305018532172e-05,
+      "loss": 0.7146,
+      "step": 1563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3903654621157037,
+      "learning_rate": 1.981294694114822e-05,
+      "loss": 0.2457,
+      "step": 1564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4339515383696826,
+      "learning_rate": 1.9812588523942334e-05,
+      "loss": 0.2166,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.650174074607824,
+      "learning_rate": 1.981222976692693e-05,
+      "loss": 0.5498,
+      "step": 1566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2077256521150712,
+      "learning_rate": 1.981187067011443e-05,
+      "loss": 0.6577,
+      "step": 1567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6294600660337998,
+      "learning_rate": 1.9811511233517275e-05,
+      "loss": 0.313,
+      "step": 1568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.560575934466919,
+      "learning_rate": 1.9811151457147904e-05,
+      "loss": 0.29,
+      "step": 1569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3676919839173703,
+      "learning_rate": 1.981079134101878e-05,
+      "loss": 0.214,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5136664835618114,
+      "learning_rate": 1.9810430885142377e-05,
+      "loss": 0.312,
+      "step": 1571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4254358022411515,
+      "learning_rate": 1.981007008953117e-05,
+      "loss": 0.4594,
+      "step": 1572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7586444608485484,
+      "learning_rate": 1.9809708954197658e-05,
+      "loss": 0.3544,
+      "step": 1573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5827592667394669,
+      "learning_rate": 1.980934747915435e-05,
+      "loss": 0.2916,
+      "step": 1574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6997799572006387,
+      "learning_rate": 1.9808985664413757e-05,
+      "loss": 0.4934,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34869623686568857,
+      "learning_rate": 1.9808623509988415e-05,
+      "loss": 0.2492,
+      "step": 1576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7051869049693643,
+      "learning_rate": 1.980826101589086e-05,
+      "loss": 0.3112,
+      "step": 1577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.375367808092143,
+      "learning_rate": 1.980789818213365e-05,
+      "loss": 0.4905,
+      "step": 1578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6868012547392646,
+      "learning_rate": 1.9807535008729347e-05,
+      "loss": 0.8162,
+      "step": 1579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5252690640932575,
+      "learning_rate": 1.980717149569053e-05,
+      "loss": 0.3133,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.467302492898827,
+      "learning_rate": 1.9806807643029786e-05,
+      "loss": 0.3452,
+      "step": 1581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0265829484598235,
+      "learning_rate": 1.9806443450759715e-05,
+      "loss": 0.3143,
+      "step": 1582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5886112657839901,
+      "learning_rate": 1.9806078918892925e-05,
+      "loss": 0.34,
+      "step": 1583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5920413529763409,
+      "learning_rate": 1.9805714047442045e-05,
+      "loss": 0.3122,
+      "step": 1584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6764117396792537,
+      "learning_rate": 1.9805348836419712e-05,
+      "loss": 0.3036,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5594852265265499,
+      "learning_rate": 1.9804983285838567e-05,
+      "loss": 0.3569,
+      "step": 1586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6844695489408091,
+      "learning_rate": 1.9804617395711275e-05,
+      "loss": 0.4128,
+      "step": 1587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6307190369431338,
+      "learning_rate": 1.9804251166050505e-05,
+      "loss": 0.3458,
+      "step": 1588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41952195036620993,
+      "learning_rate": 1.9803884596868937e-05,
+      "loss": 0.3168,
+      "step": 1589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3453562469345302,
+      "learning_rate": 1.9803517688179264e-05,
+      "loss": 0.2204,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0262728387774869,
+      "learning_rate": 1.9803150439994202e-05,
+      "loss": 0.5397,
+      "step": 1591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5369925676511108,
+      "learning_rate": 1.9802782852326456e-05,
+      "loss": 0.3018,
+      "step": 1592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7179121095737457,
+      "learning_rate": 1.9802414925188766e-05,
+      "loss": 0.3874,
+      "step": 1593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7147170188690286,
+      "learning_rate": 1.9802046658593867e-05,
+      "loss": 0.5446,
+      "step": 1594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4854192777883944,
+      "learning_rate": 1.9801678052554512e-05,
+      "loss": 0.2607,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3895561134200103,
+      "learning_rate": 1.9801309107083465e-05,
+      "loss": 0.2836,
+      "step": 1596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3859364875104156,
+      "learning_rate": 1.9800939822193512e-05,
+      "loss": 0.2896,
+      "step": 1597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4179295750829518,
+      "learning_rate": 1.980057019789743e-05,
+      "loss": 0.2141,
+      "step": 1598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5931917930493955,
+      "learning_rate": 1.9800200234208022e-05,
+      "loss": 0.3975,
+      "step": 1599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5242128316638529,
+      "learning_rate": 1.9799829931138107e-05,
+      "loss": 0.4053,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4126511092046684,
+      "learning_rate": 1.9799459288700498e-05,
+      "loss": 0.2093,
+      "step": 1601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3701533373340219,
+      "learning_rate": 1.9799088306908035e-05,
+      "loss": 0.2427,
+      "step": 1602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1308744292919486,
+      "learning_rate": 1.9798716985773567e-05,
+      "loss": 0.6998,
+      "step": 1603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3758134428810142,
+      "learning_rate": 1.9798345325309952e-05,
+      "loss": 0.2527,
+      "step": 1604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5785694558525246,
+      "learning_rate": 1.9797973325530058e-05,
+      "loss": 0.4437,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7455283306651675,
+      "learning_rate": 1.9797600986446773e-05,
+      "loss": 0.5741,
+      "step": 1606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49548617080975527,
+      "learning_rate": 1.9797228308072982e-05,
+      "loss": 0.3097,
+      "step": 1607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.399945434741605,
+      "learning_rate": 1.97968552904216e-05,
+      "loss": 0.2833,
+      "step": 1608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3260983798124931,
+      "learning_rate": 1.9796481933505535e-05,
+      "loss": 0.1926,
+      "step": 1609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5046829217920206,
+      "learning_rate": 1.9796108237337724e-05,
+      "loss": 0.3538,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.588275165004359,
+      "learning_rate": 1.979573420193111e-05,
+      "loss": 0.3805,
+      "step": 1611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45169272979663194,
+      "learning_rate": 1.9795359827298643e-05,
+      "loss": 0.3902,
+      "step": 1612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45282344751971243,
+      "learning_rate": 1.979498511345328e-05,
+      "loss": 0.3399,
+      "step": 1613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3358407517147635,
+      "learning_rate": 1.9794610060408007e-05,
+      "loss": 0.1913,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.468574580020193,
+      "learning_rate": 1.979423466817581e-05,
+      "loss": 0.3122,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49239486490812656,
+      "learning_rate": 1.9793858936769683e-05,
+      "loss": 0.3336,
+      "step": 1616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4719865931595738,
+      "learning_rate": 1.9793482866202645e-05,
+      "loss": 0.3269,
+      "step": 1617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8462916216082507,
+      "learning_rate": 1.9793106456487717e-05,
+      "loss": 0.5964,
+      "step": 1618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7204368957434376,
+      "learning_rate": 1.9792729707637935e-05,
+      "loss": 0.3242,
+      "step": 1619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36900163439375716,
+      "learning_rate": 1.979235261966634e-05,
+      "loss": 0.323,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3250448460505923,
+      "learning_rate": 1.9791975192586e-05,
+      "loss": 0.1878,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7114004359000305,
+      "learning_rate": 1.9791597426409973e-05,
+      "loss": 0.4203,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48779469367220873,
+      "learning_rate": 1.9791219321151356e-05,
+      "loss": 0.4049,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33688741860596855,
+      "learning_rate": 1.979084087682323e-05,
+      "loss": 0.2401,
+      "step": 1624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6457978863500597,
+      "learning_rate": 1.9790462093438707e-05,
+      "loss": 0.3773,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43997114105386825,
+      "learning_rate": 1.9790082971010903e-05,
+      "loss": 0.3352,
+      "step": 1626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8578347622033227,
+      "learning_rate": 1.9789703509552947e-05,
+      "loss": 0.4459,
+      "step": 1627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3320817142902381,
+      "learning_rate": 1.978932370907798e-05,
+      "loss": 0.3051,
+      "step": 1628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46309576333499364,
+      "learning_rate": 1.978894356959915e-05,
+      "loss": 0.4221,
+      "step": 1629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23451118651819808,
+      "learning_rate": 1.978856309112963e-05,
+      "loss": 0.1448,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4679183307690478,
+      "learning_rate": 1.978818227368259e-05,
+      "loss": 0.2926,
+      "step": 1631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4429604060875303,
+      "learning_rate": 1.9787801117271213e-05,
+      "loss": 0.3402,
+      "step": 1632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8594224955091179,
+      "learning_rate": 1.978741962190871e-05,
+      "loss": 0.6281,
+      "step": 1633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6065069575331038,
+      "learning_rate": 1.9787037787608287e-05,
+      "loss": 0.3928,
+      "step": 1634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42824444304880876,
+      "learning_rate": 1.9786655614383163e-05,
+      "loss": 0.3295,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30050373194288177,
+      "learning_rate": 1.978627310224658e-05,
+      "loss": 0.2354,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5963019421892746,
+      "learning_rate": 1.9785890251211777e-05,
+      "loss": 0.1868,
+      "step": 1637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.400518818725669,
+      "learning_rate": 1.9785507061292017e-05,
+      "loss": 0.2862,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9336374686602577,
+      "learning_rate": 1.978512353250057e-05,
+      "loss": 0.5577,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.35757051249823835,
+      "learning_rate": 1.978473966485071e-05,
+      "loss": 0.2641,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4439429484546262,
+      "learning_rate": 1.978435545835574e-05,
+      "loss": 0.3602,
+      "step": 1641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3921207587665929,
+      "learning_rate": 1.978397091302896e-05,
+      "loss": 0.1743,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7618586521311463,
+      "learning_rate": 1.9783586028883688e-05,
+      "loss": 0.4513,
+      "step": 1643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3751913533863891,
+      "learning_rate": 1.9783200805933252e-05,
+      "loss": 0.2761,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7805939473702508,
+      "learning_rate": 1.9782815244190997e-05,
+      "loss": 0.5491,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6195714209412733,
+      "learning_rate": 1.9782429343670267e-05,
+      "loss": 0.5169,
+      "step": 1646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4918147444122574,
+      "learning_rate": 1.978204310438443e-05,
+      "loss": 0.2578,
+      "step": 1647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3435562889326771,
+      "learning_rate": 1.9781656526346863e-05,
+      "loss": 0.2275,
+      "step": 1648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7237826939442035,
+      "learning_rate": 1.9781269609570945e-05,
+      "loss": 0.3185,
+      "step": 1649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5178707759432076,
+      "learning_rate": 1.978088235407009e-05,
+      "loss": 0.2934,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7923454047135651,
+      "learning_rate": 1.978049475985769e-05,
+      "loss": 0.4562,
+      "step": 1651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5428509403531181,
+      "learning_rate": 1.9780106826947184e-05,
+      "loss": 0.3805,
+      "step": 1652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45702446485133047,
+      "learning_rate": 1.9779718555351997e-05,
+      "loss": 0.2591,
+      "step": 1653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36296431359883,
+      "learning_rate": 1.9779329945085578e-05,
+      "loss": 0.2195,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6704714680728022,
+      "learning_rate": 1.9778940996161382e-05,
+      "loss": 0.4751,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43102537372482086,
+      "learning_rate": 1.9778551708592883e-05,
+      "loss": 0.2816,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7987627103272095,
+      "learning_rate": 1.9778162082393554e-05,
+      "loss": 0.4378,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4586157829032407,
+      "learning_rate": 1.9777772117576893e-05,
+      "loss": 0.8086,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5097364905331209,
+      "learning_rate": 1.977738181415641e-05,
+      "loss": 0.2658,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5945672810230304,
+      "learning_rate": 1.977699117214561e-05,
+      "loss": 0.2941,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3399558203328304,
+      "learning_rate": 1.9776600191558025e-05,
+      "loss": 0.2018,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6445674144621936,
+      "learning_rate": 1.9776208872407202e-05,
+      "loss": 0.3421,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7888347334580998,
+      "learning_rate": 1.9775817214706682e-05,
+      "loss": 0.3533,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7188794356920819,
+      "learning_rate": 1.977542521847003e-05,
+      "loss": 0.3961,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5233002822851152,
+      "learning_rate": 1.9775032883710826e-05,
+      "loss": 0.3212,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.22069561834964382,
+      "learning_rate": 1.9774640210442654e-05,
+      "loss": 0.1167,
+      "step": 1666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4538299309152852,
+      "learning_rate": 1.977424719867911e-05,
+      "loss": 0.3369,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5837962589537384,
+      "learning_rate": 1.9773853848433806e-05,
+      "loss": 0.3332,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2822390523629652,
+      "learning_rate": 1.9773460159720365e-05,
+      "loss": 0.499,
+      "step": 1669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6098981703977231,
+      "learning_rate": 1.977306613255242e-05,
+      "loss": 0.3464,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5010711098035411,
+      "learning_rate": 1.977267176694361e-05,
+      "loss": 0.3058,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4940495865152775,
+      "learning_rate": 1.97722770629076e-05,
+      "loss": 0.3481,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6839400527952039,
+      "learning_rate": 1.9771882020458055e-05,
+      "loss": 0.3003,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4901359588398224,
+      "learning_rate": 1.9771486639608657e-05,
+      "loss": 0.3097,
+      "step": 1674
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38335500215663265,
+      "learning_rate": 1.9771090920373096e-05,
+      "loss": 0.2784,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0457668872874353,
+      "learning_rate": 1.9770694862765077e-05,
+      "loss": 0.6059,
+      "step": 1676
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4529475843166634,
+      "learning_rate": 1.9770298466798316e-05,
+      "loss": 0.2996,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9647389731875299,
+      "learning_rate": 1.9769901732486542e-05,
+      "loss": 0.545,
+      "step": 1678
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4138833322271508,
+      "learning_rate": 1.9769504659843486e-05,
+      "loss": 0.3303,
+      "step": 1679
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37714688186515205,
+      "learning_rate": 1.9769107248882904e-05,
+      "loss": 0.2794,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.539646531653099,
+      "learning_rate": 1.976870949961856e-05,
+      "loss": 0.2455,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9706385236557056,
+      "learning_rate": 1.9768311412064224e-05,
+      "loss": 0.8027,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4506680655535206,
+      "learning_rate": 1.9767912986233685e-05,
+      "loss": 0.2213,
+      "step": 1683
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4818504309643764,
+      "learning_rate": 1.976751422214074e-05,
+      "loss": 0.3729,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6867809896350154,
+      "learning_rate": 1.9767115119799197e-05,
+      "loss": 0.5068,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4081668174851612,
+      "learning_rate": 1.9766715679222875e-05,
+      "loss": 0.2312,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3395615177013707,
+      "learning_rate": 1.9766315900425613e-05,
+      "loss": 0.2423,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7966060680798426,
+      "learning_rate": 1.976591578342125e-05,
+      "loss": 0.8855,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.380188460807513,
+      "learning_rate": 1.9765515328223644e-05,
+      "loss": 0.2262,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7703079045710871,
+      "learning_rate": 1.9765114534846662e-05,
+      "loss": 0.5256,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5143745240081559,
+      "learning_rate": 1.9764713403304183e-05,
+      "loss": 0.3954,
+      "step": 1691
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46836814781278047,
+      "learning_rate": 1.97643119336101e-05,
+      "loss": 0.2407,
+      "step": 1692
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3690528313572445,
+      "learning_rate": 1.976391012577831e-05,
+      "loss": 0.1872,
+      "step": 1693
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2780455908200286,
+      "learning_rate": 1.9763507979822737e-05,
+      "loss": 0.7456,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4214444867559315,
+      "learning_rate": 1.97631054957573e-05,
+      "loss": 0.2931,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5189345658634478,
+      "learning_rate": 1.9762702673595943e-05,
+      "loss": 0.3177,
+      "step": 1696
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.880757909211329,
+      "learning_rate": 1.9762299513352604e-05,
+      "loss": 0.5314,
+      "step": 1697
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4845554317665074,
+      "learning_rate": 1.976189601504126e-05,
+      "loss": 0.2937,
+      "step": 1698
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30040774511408,
+      "learning_rate": 1.9761492178675876e-05,
+      "loss": 0.1939,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1642293277701257,
+      "learning_rate": 1.9761088004270435e-05,
+      "loss": 0.7888,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43107411175540156,
+      "learning_rate": 1.976068349183894e-05,
+      "loss": 0.2908,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7767979561549465,
+      "learning_rate": 1.9760278641395395e-05,
+      "loss": 0.447,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.562402325785457,
+      "learning_rate": 1.9759873452953816e-05,
+      "loss": 0.3703,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4995224833626151,
+      "learning_rate": 1.9759467926528242e-05,
+      "loss": 0.3042,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27285547576376873,
+      "learning_rate": 1.975906206213271e-05,
+      "loss": 0.1633,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9079324937335543,
+      "learning_rate": 1.9758655859781282e-05,
+      "loss": 0.6932,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40677426359666474,
+      "learning_rate": 1.975824931948802e-05,
+      "loss": 0.2871,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45580786137028356,
+      "learning_rate": 1.975784244126701e-05,
+      "loss": 0.3893,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9194512800758327,
+      "learning_rate": 1.9757435225132325e-05,
+      "loss": 0.3548,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44115743053683315,
+      "learning_rate": 1.9757027671098083e-05,
+      "loss": 0.3031,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2825713405110046,
+      "learning_rate": 1.9756619779178393e-05,
+      "loss": 0.2427,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1849133538579537,
+      "learning_rate": 1.9756211549387378e-05,
+      "loss": 0.4944,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36690700948942095,
+      "learning_rate": 1.975580298173918e-05,
+      "loss": 0.3057,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7245805282332856,
+      "learning_rate": 1.975539407624794e-05,
+      "loss": 0.5556,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4612175857613643,
+      "learning_rate": 1.9754984832927825e-05,
+      "loss": 0.3085,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44090499684018897,
+      "learning_rate": 1.9754575251793006e-05,
+      "loss": 0.3089,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5288877897785201,
+      "learning_rate": 1.975416533285766e-05,
+      "loss": 0.3071,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3106501640407626,
+      "learning_rate": 1.9753755076135994e-05,
+      "loss": 0.587,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3528842608221892,
+      "learning_rate": 1.9753344481642205e-05,
+      "loss": 0.2396,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6296848852224964,
+      "learning_rate": 1.975293354939052e-05,
+      "loss": 0.4465,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5365509567706639,
+      "learning_rate": 1.975252227939516e-05,
+      "loss": 0.2626,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3821923286062335,
+      "learning_rate": 1.9752110671670375e-05,
+      "loss": 0.1844,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40328934838267844,
+      "learning_rate": 1.975169872623042e-05,
+      "loss": 0.3276,
+      "step": 1723
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9932472840383428,
+      "learning_rate": 1.9751286443089555e-05,
+      "loss": 0.5172,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39449755648213397,
+      "learning_rate": 1.975087382226206e-05,
+      "loss": 0.2579,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34791684251849897,
+      "learning_rate": 1.9750460863762225e-05,
+      "loss": 0.2135,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4824628574743985,
+      "learning_rate": 1.9750047567604348e-05,
+      "loss": 0.3734,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45556148548025616,
+      "learning_rate": 1.9749633933802743e-05,
+      "loss": 0.22,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6722527683545421,
+      "learning_rate": 1.9749219962371736e-05,
+      "loss": 0.4172,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3213518952540688,
+      "learning_rate": 1.9748805653325657e-05,
+      "loss": 0.5264,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43129412639939235,
+      "learning_rate": 1.974839100667886e-05,
+      "loss": 0.3006,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5744145659725003,
+      "learning_rate": 1.97479760224457e-05,
+      "loss": 0.3189,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28489126228750195,
+      "learning_rate": 1.9747560700640552e-05,
+      "loss": 0.1805,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4691600450533372,
+      "learning_rate": 1.97471450412778e-05,
+      "loss": 0.2848,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48444296804977377,
+      "learning_rate": 1.9746729044371826e-05,
+      "loss": 0.2677,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6659386268209997,
+      "learning_rate": 1.9746312709937047e-05,
+      "loss": 0.5739,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4936719526225396,
+      "learning_rate": 1.974589603798788e-05,
+      "loss": 0.3041,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34597533311036405,
+      "learning_rate": 1.974547902853875e-05,
+      "loss": 0.2043,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38018595031990254,
+      "learning_rate": 1.9745061681604104e-05,
+      "loss": 0.3154,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8001416151657222,
+      "learning_rate": 1.974464399719839e-05,
+      "loss": 0.461,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5315768234902671,
+      "learning_rate": 1.974422597533607e-05,
+      "loss": 0.3053,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5469926220469847,
+      "learning_rate": 1.9743807616031624e-05,
+      "loss": 0.3805,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7185184120264385,
+      "learning_rate": 1.974338891929954e-05,
+      "loss": 0.3472,
+      "step": 1743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5697328626613448,
+      "learning_rate": 1.9742969885154318e-05,
+      "loss": 0.3864,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2519361023605971,
+      "learning_rate": 1.974255051361047e-05,
+      "loss": 0.076,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5995345198599,
+      "learning_rate": 1.974213080468251e-05,
+      "loss": 0.3818,
+      "step": 1746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40605771267747187,
+      "learning_rate": 1.9741710758384985e-05,
+      "loss": 0.3281,
+      "step": 1747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.852458852226233,
+      "learning_rate": 1.9741290374732434e-05,
+      "loss": 0.4804,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6862694898087204,
+      "learning_rate": 1.9740869653739413e-05,
+      "loss": 0.4156,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5487997218358596,
+      "learning_rate": 1.97404485954205e-05,
+      "loss": 0.3663,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.303352834458015,
+      "learning_rate": 1.974002719979027e-05,
+      "loss": 0.1821,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5481110458066554,
+      "learning_rate": 1.973960546686331e-05,
+      "loss": 0.3175,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6217773627623907,
+      "learning_rate": 1.973918339665424e-05,
+      "loss": 0.4053,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4157060592887071,
+      "learning_rate": 1.9738760989177665e-05,
+      "loss": 0.3342,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48771030560903766,
+      "learning_rate": 1.9738338244448214e-05,
+      "loss": 0.3222,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5874946088172913,
+      "learning_rate": 1.9737915162480527e-05,
+      "loss": 0.4358,
+      "step": 1756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37947933594438027,
+      "learning_rate": 1.973749174328926e-05,
+      "loss": 0.2179,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4789152218360017,
+      "learning_rate": 1.9737067986889072e-05,
+      "loss": 0.3133,
+      "step": 1758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45954057561314693,
+      "learning_rate": 1.973664389329464e-05,
+      "loss": 0.3215,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0575391871829922,
+      "learning_rate": 1.9736219462520645e-05,
+      "loss": 0.6064,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5316241369605608,
+      "learning_rate": 1.973579469458179e-05,
+      "loss": 0.2216,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5200064785449022,
+      "learning_rate": 1.9735369589492786e-05,
+      "loss": 0.3547,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3661703504092863,
+      "learning_rate": 1.973494414726835e-05,
+      "loss": 0.2464,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6885337064692816,
+      "learning_rate": 1.9734518367923216e-05,
+      "loss": 0.3472,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44627825575802316,
+      "learning_rate": 1.973409225147213e-05,
+      "loss": 0.3789,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5653737927716992,
+      "learning_rate": 1.973366579792985e-05,
+      "loss": 0.3915,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47244377328827575,
+      "learning_rate": 1.9733239007311137e-05,
+      "loss": 0.2816,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5118989988502627,
+      "learning_rate": 1.973281187963078e-05,
+      "loss": 0.3149,
+      "step": 1768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8961397366295146,
+      "learning_rate": 1.9732384414903562e-05,
+      "loss": 0.5801,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40915889685555634,
+      "learning_rate": 1.9731956613144297e-05,
+      "loss": 0.3385,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36489091879616176,
+      "learning_rate": 1.9731528474367787e-05,
+      "loss": 0.1888,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4653678958356583,
+      "learning_rate": 1.9731099998588865e-05,
+      "loss": 0.3244,
+      "step": 1772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7889078900037176,
+      "learning_rate": 1.973067118582237e-05,
+      "loss": 0.4141,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48457358345367246,
+      "learning_rate": 1.973024203608315e-05,
+      "loss": 0.2676,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5849351746372535,
+      "learning_rate": 1.9729812549386066e-05,
+      "loss": 0.3811,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6100840543617714,
+      "learning_rate": 1.9729382725745997e-05,
+      "loss": 0.3586,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32169239000575955,
+      "learning_rate": 1.9728952565177817e-05,
+      "loss": 0.1859,
+      "step": 1777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4926537571689009,
+      "learning_rate": 1.972852206769643e-05,
+      "loss": 0.3794,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4956203678206292,
+      "learning_rate": 1.972809123331674e-05,
+      "loss": 0.8704,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.432781530735122,
+      "learning_rate": 1.972766006205367e-05,
+      "loss": 0.2187,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8836133604422183,
+      "learning_rate": 1.9727228553922152e-05,
+      "loss": 0.5072,
+      "step": 1781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6000939456264764,
+      "learning_rate": 1.9726796708937125e-05,
+      "loss": 0.3967,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36428401062594407,
+      "learning_rate": 1.972636452711355e-05,
+      "loss": 0.2402,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35233509057631474,
+      "learning_rate": 1.9725932008466383e-05,
+      "loss": 0.1392,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7541938557788923,
+      "learning_rate": 1.9725499153010613e-05,
+      "loss": 0.4919,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4648704692571199,
+      "learning_rate": 1.9725065960761225e-05,
+      "loss": 0.3193,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7643519431832295,
+      "learning_rate": 1.9724632431733223e-05,
+      "loss": 0.3782,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4383417741200626,
+      "learning_rate": 1.9724198565941616e-05,
+      "loss": 0.2711,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5752027956523929,
+      "learning_rate": 1.972376436340143e-05,
+      "loss": 0.3578,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3491638947718094,
+      "learning_rate": 1.9723329824127703e-05,
+      "loss": 0.2435,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9210790563154199,
+      "learning_rate": 1.9722894948135485e-05,
+      "loss": 0.5047,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4745370515957307,
+      "learning_rate": 1.972245973543983e-05,
+      "loss": 0.3136,
+      "step": 1792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8457754702291259,
+      "learning_rate": 1.9722024186055812e-05,
+      "loss": 0.4412,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5232986705641869,
+      "learning_rate": 1.9721588299998516e-05,
+      "loss": 0.3358,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43026069963986585,
+      "learning_rate": 1.9721152077283038e-05,
+      "loss": 0.2955,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3170651313977208,
+      "learning_rate": 1.9720715517924476e-05,
+      "loss": 0.2067,
+      "step": 1796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.055213009657008,
+      "learning_rate": 1.9720278621937954e-05,
+      "loss": 0.4355,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39566676103877046,
+      "learning_rate": 1.9719841389338605e-05,
+      "loss": 0.2776,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45559232573659136,
+      "learning_rate": 1.9719403820141563e-05,
+      "loss": 0.4138,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9887763688604866,
+      "learning_rate": 1.971896591436199e-05,
+      "loss": 0.2925,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3638885992702867,
+      "learning_rate": 1.971852767201504e-05,
+      "loss": 0.2546,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6125066940881474,
+      "learning_rate": 1.97180890931159e-05,
+      "loss": 0.4316,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.645523555658483,
+      "learning_rate": 1.971765017767975e-05,
+      "loss": 0.3526,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48292600807911545,
+      "learning_rate": 1.9717210925721792e-05,
+      "loss": 0.3628,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.510701994227429,
+      "learning_rate": 1.9716771337257235e-05,
+      "loss": 0.2944,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.512257770671276,
+      "learning_rate": 1.9716331412301304e-05,
+      "loss": 0.3456,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3723557207434571,
+      "learning_rate": 1.971589115086924e-05,
+      "loss": 0.2188,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4776975335292782,
+      "learning_rate": 1.971545055297628e-05,
+      "loss": 0.3271,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4671253964539537,
+      "learning_rate": 1.971500961863768e-05,
+      "loss": 0.4177,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4226489406088548,
+      "learning_rate": 1.9714568347868722e-05,
+      "loss": 0.2196,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36354839297692376,
+      "learning_rate": 1.9714126740684677e-05,
+      "loss": 0.3186,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4618017892562303,
+      "learning_rate": 1.9713684797100843e-05,
+      "loss": 0.7304,
+      "step": 1812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35343600195296193,
+      "learning_rate": 1.9713242517132518e-05,
+      "loss": 0.2205,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4373763331330048,
+      "learning_rate": 1.9712799900795026e-05,
+      "loss": 0.3891,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7587560756410427,
+      "learning_rate": 1.971235694810369e-05,
+      "loss": 0.5362,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38486193172658234,
+      "learning_rate": 1.971191365907385e-05,
+      "loss": 0.2663,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3515214902888066,
+      "learning_rate": 1.9711470033720856e-05,
+      "loss": 0.168,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48100670484562935,
+      "learning_rate": 1.9711026072060077e-05,
+      "loss": 0.3562,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39630337446911107,
+      "learning_rate": 1.971058177410688e-05,
+      "loss": 0.2998,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8004872736078943,
+      "learning_rate": 1.9710137139876653e-05,
+      "loss": 0.4614,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.49674968425439975,
+      "learning_rate": 1.9709692169384794e-05,
+      "loss": 0.4202,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3863426434060569,
+      "learning_rate": 1.9709246862646712e-05,
+      "loss": 0.2811,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.22159405934145307,
+      "learning_rate": 1.9708801219677832e-05,
+      "loss": 0.0938,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.116757279703663,
+      "learning_rate": 1.970835524049358e-05,
+      "loss": 0.6516,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4811222290577782,
+      "learning_rate": 1.9707908925109405e-05,
+      "loss": 0.3042,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4106501961078986,
+      "learning_rate": 1.9707462273540762e-05,
+      "loss": 0.3154,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8876173475980776,
+      "learning_rate": 1.9707015285803117e-05,
+      "loss": 0.5606,
+      "step": 1827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4763770328566096,
+      "learning_rate": 1.9706567961911952e-05,
+      "loss": 0.3034,
+      "step": 1828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2731838409066813,
+      "learning_rate": 1.9706120301882755e-05,
+      "loss": 0.1925,
+      "step": 1829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6050226311696162,
+      "learning_rate": 1.9705672305731027e-05,
+      "loss": 0.3795,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5254253951284472,
+      "learning_rate": 1.9705223973472288e-05,
+      "loss": 0.3244,
+      "step": 1831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7309290009909968,
+      "learning_rate": 1.9704775305122057e-05,
+      "loss": 0.4473,
+      "step": 1832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5142034091004006,
+      "learning_rate": 1.9704326300695874e-05,
+      "loss": 0.3584,
+      "step": 1833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41208408396675505,
+      "learning_rate": 1.9703876960209292e-05,
+      "loss": 0.2997,
+      "step": 1834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4215025386501301,
+      "learning_rate": 1.970342728367787e-05,
+      "loss": 0.207,
+      "step": 1835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1215430532038166,
+      "learning_rate": 1.9702977271117172e-05,
+      "loss": 0.3586,
+      "step": 1836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40290356090073365,
+      "learning_rate": 1.9702526922542788e-05,
+      "loss": 0.3035,
+      "step": 1837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4444886784774347,
+      "learning_rate": 1.970207623797032e-05,
+      "loss": 0.367,
+      "step": 1838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7960973226307128,
+      "learning_rate": 1.9701625217415366e-05,
+      "loss": 0.4945,
+      "step": 1839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40419674822028395,
+      "learning_rate": 1.9701173860893547e-05,
+      "loss": 0.2819,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3928574273644804,
+      "learning_rate": 1.9700722168420493e-05,
+      "loss": 0.1876,
+      "step": 1841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6832258064420349,
+      "learning_rate": 1.9700270140011852e-05,
+      "loss": 0.3169,
+      "step": 1842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5857771512420725,
+      "learning_rate": 1.969981777568327e-05,
+      "loss": 0.3352,
+      "step": 1843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4742736573868894,
+      "learning_rate": 1.969936507545042e-05,
+      "loss": 0.3776,
+      "step": 1844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.448882200382126,
+      "learning_rate": 1.969891203932897e-05,
+      "loss": 0.3702,
+      "step": 1845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3997430046247253,
+      "learning_rate": 1.9698458667334616e-05,
+      "loss": 0.2028,
+      "step": 1846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42185324820247727,
+      "learning_rate": 1.969800495948305e-05,
+      "loss": 0.2546,
+      "step": 1847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.581614305117135,
+      "learning_rate": 1.9697550915789992e-05,
+      "loss": 0.3695,
+      "step": 1848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46363413685206445,
+      "learning_rate": 1.9697096536271166e-05,
+      "loss": 0.2585,
+      "step": 1849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42718596651710894,
+      "learning_rate": 1.96966418209423e-05,
+      "loss": 0.3225,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4101683380916672,
+      "learning_rate": 1.9696186769819146e-05,
+      "loss": 0.8429,
+      "step": 1851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.406650262457706,
+      "learning_rate": 1.9695731382917463e-05,
+      "loss": 0.2515,
+      "step": 1852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2883048378866706,
+      "learning_rate": 1.9695275660253017e-05,
+      "loss": 0.236,
+      "step": 1853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5570129462774673,
+      "learning_rate": 1.9694819601841593e-05,
+      "loss": 0.4075,
+      "step": 1854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6608487821497272,
+      "learning_rate": 1.969436320769898e-05,
+      "loss": 0.324,
+      "step": 1855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.695684745170112,
+      "learning_rate": 1.969390647784099e-05,
+      "loss": 0.4257,
+      "step": 1856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42950676535452104,
+      "learning_rate": 1.9693449412283435e-05,
+      "loss": 0.3572,
+      "step": 1857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.422128890021055,
+      "learning_rate": 1.9692992011042143e-05,
+      "loss": 0.2884,
+      "step": 1858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4976931054505805,
+      "learning_rate": 1.969253427413295e-05,
+      "loss": 0.3108,
+      "step": 1859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42506504553438845,
+      "learning_rate": 1.969207620157172e-05,
+      "loss": 0.2968,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44845845302893456,
+      "learning_rate": 1.9691617793374304e-05,
+      "loss": 0.2973,
+      "step": 1861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.463892625466207,
+      "learning_rate": 1.969115904955658e-05,
+      "loss": 0.2725,
+      "step": 1862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2968732447910472,
+      "learning_rate": 1.9690699970134436e-05,
+      "loss": 0.7922,
+      "step": 1863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7591209376006144,
+      "learning_rate": 1.9690240555123767e-05,
+      "loss": 0.4119,
+      "step": 1864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37942803845107786,
+      "learning_rate": 1.9689780804540487e-05,
+      "loss": 0.2586,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5185336534105881,
+      "learning_rate": 1.968932071840051e-05,
+      "loss": 0.4203,
+      "step": 1866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41377909230155135,
+      "learning_rate": 1.968886029671978e-05,
+      "loss": 0.2385,
+      "step": 1867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3525014148154284,
+      "learning_rate": 1.968839953951423e-05,
+      "loss": 0.2215,
+      "step": 1868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4768967276948403,
+      "learning_rate": 1.968793844679982e-05,
+      "loss": 0.3546,
+      "step": 1869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5977294660884478,
+      "learning_rate": 1.9687477018592517e-05,
+      "loss": 0.3633,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4148338563966621,
+      "learning_rate": 1.9687015254908305e-05,
+      "loss": 0.375,
+      "step": 1871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5783402829003099,
+      "learning_rate": 1.968655315576317e-05,
+      "loss": 0.4202,
+      "step": 1872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3409097705300037,
+      "learning_rate": 1.9686090721173118e-05,
+      "loss": 0.2646,
+      "step": 1873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3731720064413754,
+      "learning_rate": 1.9685627951154154e-05,
+      "loss": 0.259,
+      "step": 1874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.056227121717924,
+      "learning_rate": 1.9685164845722317e-05,
+      "loss": 0.6084,
+      "step": 1875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47496672745326424,
+      "learning_rate": 1.9684701404893635e-05,
+      "loss": 0.331,
+      "step": 1876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6892217331540745,
+      "learning_rate": 1.9684237628684162e-05,
+      "loss": 0.4158,
+      "step": 1877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5381495786374161,
+      "learning_rate": 1.9683773517109954e-05,
+      "loss": 0.3442,
+      "step": 1878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4442211799150528,
+      "learning_rate": 1.9683309070187088e-05,
+      "loss": 0.2373,
+      "step": 1879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4516614938674143,
+      "learning_rate": 1.9682844287931643e-05,
+      "loss": 0.2681,
+      "step": 1880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46034638626235896,
+      "learning_rate": 1.9682379170359717e-05,
+      "loss": 0.379,
+      "step": 1881
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7113109005722336,
+      "learning_rate": 1.9681913717487418e-05,
+      "loss": 0.2923,
+      "step": 1882
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46398582128192567,
+      "learning_rate": 1.9681447929330864e-05,
+      "loss": 0.343,
+      "step": 1883
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8005165026287019,
+      "learning_rate": 1.9680981805906182e-05,
+      "loss": 0.5295,
+      "step": 1884
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5318756913666476,
+      "learning_rate": 1.968051534722952e-05,
+      "loss": 0.222,
+      "step": 1885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36430519799466327,
+      "learning_rate": 1.968004855331702e-05,
+      "loss": 0.2523,
+      "step": 1886
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.379727729095134,
+      "learning_rate": 1.9679581424184862e-05,
+      "loss": 0.803,
+      "step": 1887
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6338746679719071,
+      "learning_rate": 1.9679113959849213e-05,
+      "loss": 0.3052,
+      "step": 1888
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.436401101542336,
+      "learning_rate": 1.9678646160326268e-05,
+      "loss": 0.3473,
+      "step": 1889
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9931177098972608,
+      "learning_rate": 1.9678178025632217e-05,
+      "loss": 0.5409,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4048200278519992,
+      "learning_rate": 1.9677709555783278e-05,
+      "loss": 0.1736,
+      "step": 1891
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5199305893202164,
+      "learning_rate": 1.9677240750795677e-05,
+      "loss": 0.3528,
+      "step": 1892
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45682355339831787,
+      "learning_rate": 1.967677161068564e-05,
+      "loss": 0.3718,
+      "step": 1893
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6801158394462933,
+      "learning_rate": 1.9676302135469424e-05,
+      "loss": 0.3543,
+      "step": 1894
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4675467552847163,
+      "learning_rate": 1.9675832325163277e-05,
+      "loss": 0.2873,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.21108403861959,
+      "learning_rate": 1.9675362179783472e-05,
+      "loss": 0.6525,
+      "step": 1896
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4128238832269528,
+      "learning_rate": 1.967489169934629e-05,
+      "loss": 0.3157,
+      "step": 1897
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4572471040151339,
+      "learning_rate": 1.9674420883868032e-05,
+      "loss": 0.2685,
+      "step": 1898
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5102926952782346,
+      "learning_rate": 1.9673949733364987e-05,
+      "loss": 0.3365,
+      "step": 1899
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.801037699421989,
+      "learning_rate": 1.9673478247853482e-05,
+      "loss": 0.5427,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40519304528153915,
+      "learning_rate": 1.967300642734984e-05,
+      "loss": 0.2522,
+      "step": 1901
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43817917000824186,
+      "learning_rate": 1.96725342718704e-05,
+      "loss": 0.2323,
+      "step": 1902
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0546181162648953,
+      "learning_rate": 1.967206178143152e-05,
+      "loss": 0.6465,
+      "step": 1903
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40293228795698843,
+      "learning_rate": 1.967158895604955e-05,
+      "loss": 0.2422,
+      "step": 1904
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4840118868245216,
+      "learning_rate": 1.967111579574087e-05,
+      "loss": 0.3647,
+      "step": 1905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.856390248834586,
+      "learning_rate": 1.967064230052187e-05,
+      "loss": 0.3756,
+      "step": 1906
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3372824122208772,
+      "learning_rate": 1.9670168470408942e-05,
+      "loss": 0.2382,
+      "step": 1907
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0445851645817175,
+      "learning_rate": 1.9669694305418498e-05,
+      "loss": 0.3723,
+      "step": 1908
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5236158385845334,
+      "learning_rate": 1.9669219805566954e-05,
+      "loss": 0.3219,
+      "step": 1909
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.489704534862564,
+      "learning_rate": 1.966874497087074e-05,
+      "loss": 0.3341,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8448033082087018,
+      "learning_rate": 1.9668269801346305e-05,
+      "loss": 0.4509,
+      "step": 1911
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.440510759931443,
+      "learning_rate": 1.966779429701011e-05,
+      "loss": 0.3606,
+      "step": 1912
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49307606252366426,
+      "learning_rate": 1.9667318457878604e-05,
+      "loss": 0.3268,
+      "step": 1913
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4100642263944072,
+      "learning_rate": 1.966684228396828e-05,
+      "loss": 0.2066,
+      "step": 1914
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9405707203282181,
+      "learning_rate": 1.9666365775295622e-05,
+      "loss": 0.4985,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46413787240405263,
+      "learning_rate": 1.966588893187714e-05,
+      "loss": 0.2986,
+      "step": 1916
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.532938541056413,
+      "learning_rate": 1.966541175372933e-05,
+      "loss": 0.3551,
+      "step": 1917
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8985917488823684,
+      "learning_rate": 1.966493424086873e-05,
+      "loss": 0.6217,
+      "step": 1918
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33445042184657847,
+      "learning_rate": 1.9664456393311876e-05,
+      "loss": 0.2323,
+      "step": 1919
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5402494374756727,
+      "learning_rate": 1.966397821107531e-05,
+      "loss": 0.2774,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45417449458805104,
+      "learning_rate": 1.9663499694175596e-05,
+      "loss": 0.2966,
+      "step": 1921
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41078506159188094,
+      "learning_rate": 1.96630208426293e-05,
+      "loss": 0.3234,
+      "step": 1922
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.880978732867112,
+      "learning_rate": 1.966254165645301e-05,
+      "loss": 0.5256,
+      "step": 1923
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.397876864335666,
+      "learning_rate": 1.9662062135663316e-05,
+      "loss": 0.2994,
+      "step": 1924
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36149216851821453,
+      "learning_rate": 1.9661582280276828e-05,
+      "loss": 0.2885,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32020921506739036,
+      "learning_rate": 1.9661102090310157e-05,
+      "loss": 0.1698,
+      "step": 1926
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0554723084064312,
+      "learning_rate": 1.9660621565779943e-05,
+      "loss": 0.4549,
+      "step": 1927
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39709968758603115,
+      "learning_rate": 1.966014070670281e-05,
+      "loss": 0.2861,
+      "step": 1928
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4627118390227328,
+      "learning_rate": 1.965965951309543e-05,
+      "loss": 0.37,
+      "step": 1929
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5652502961754505,
+      "learning_rate": 1.965917798497445e-05,
+      "loss": 0.3301,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5014863993079642,
+      "learning_rate": 1.9658696122356556e-05,
+      "loss": 0.3489,
+      "step": 1931
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3653922709051474,
+      "learning_rate": 1.965821392525843e-05,
+      "loss": 0.2062,
+      "step": 1932
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5667706777693249,
+      "learning_rate": 1.9657731393696768e-05,
+      "loss": 0.3548,
+      "step": 1933
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36616733693133574,
+      "learning_rate": 1.9657248527688285e-05,
+      "loss": 0.2127,
+      "step": 1934
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0432472752070607,
+      "learning_rate": 1.9656765327249697e-05,
+      "loss": 0.5552,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5719621038132259,
+      "learning_rate": 1.9656281792397745e-05,
+      "loss": 0.4127,
+      "step": 1936
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37457017917808055,
+      "learning_rate": 1.965579792314917e-05,
+      "loss": 0.2565,
+      "step": 1937
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30681311507200504,
+      "learning_rate": 1.9655313719520726e-05,
+      "loss": 0.2221,
+      "step": 1938
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8535758857629228,
+      "learning_rate": 1.9654829181529186e-05,
+      "loss": 0.5047,
+      "step": 1939
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5719271930360567,
+      "learning_rate": 1.965434430919132e-05,
+      "loss": 0.286,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5563675709499672,
+      "learning_rate": 1.9653859102523936e-05,
+      "loss": 0.364,
+      "step": 1941
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3137731119716958,
+      "learning_rate": 1.965337356154382e-05,
+      "loss": 0.8021,
+      "step": 1942
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38001594373431286,
+      "learning_rate": 1.9652887686267795e-05,
+      "loss": 0.2393,
+      "step": 1943
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2948874349792371,
+      "learning_rate": 1.965240147671268e-05,
+      "loss": 0.1511,
+      "step": 1944
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6219444278818353,
+      "learning_rate": 1.965191493289532e-05,
+      "loss": 0.4022,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48882053571073614,
+      "learning_rate": 1.9651428054832562e-05,
+      "loss": 0.3009,
+      "step": 1946
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0667319837664195,
+      "learning_rate": 1.9650940842541265e-05,
+      "loss": 0.4638,
+      "step": 1947
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4139384433063878,
+      "learning_rate": 1.9650453296038302e-05,
+      "loss": 0.3544,
+      "step": 1948
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5010220701794562,
+      "learning_rate": 1.9649965415340553e-05,
+      "loss": 0.3336,
+      "step": 1949
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28320481878320053,
+      "learning_rate": 1.964947720046492e-05,
+      "loss": 0.1546,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2928573922057744,
+      "learning_rate": 1.9648988651428308e-05,
+      "loss": 0.5374,
+      "step": 1951
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48585857871631216,
+      "learning_rate": 1.964849976824763e-05,
+      "loss": 0.2929,
+      "step": 1952
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4246659305786616,
+      "learning_rate": 1.964801055093982e-05,
+      "loss": 0.3102,
+      "step": 1953
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.487500029193504,
+      "learning_rate": 1.964752099952182e-05,
+      "loss": 0.8572,
+      "step": 1954
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4105755914914621,
+      "learning_rate": 1.9647031114010585e-05,
+      "loss": 0.3072,
+      "step": 1955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3978007522631092,
+      "learning_rate": 1.9646540894423074e-05,
+      "loss": 0.2497,
+      "step": 1956
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6591002883094217,
+      "learning_rate": 1.964605034077627e-05,
+      "loss": 0.5882,
+      "step": 1957
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34811743917000226,
+      "learning_rate": 1.9645559453087158e-05,
+      "loss": 0.235,
+      "step": 1958
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6295042608541986,
+      "learning_rate": 1.9645068231372733e-05,
+      "loss": 0.426,
+      "step": 1959
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3943625001062919,
+      "learning_rate": 1.9644576675650012e-05,
+      "loss": 0.2832,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4604908836391139,
+      "learning_rate": 1.9644084785936014e-05,
+      "loss": 0.3313,
+      "step": 1961
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0978664581234787,
+      "learning_rate": 1.9643592562247776e-05,
+      "loss": 0.5663,
+      "step": 1962
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5053998885488958,
+      "learning_rate": 1.964310000460234e-05,
+      "loss": 0.2447,
+      "step": 1963
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4433806067056563,
+      "learning_rate": 1.964260711301677e-05,
+      "loss": 0.2844,
+      "step": 1964
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5933816056127612,
+      "learning_rate": 1.9642113887508127e-05,
+      "loss": 0.4007,
+      "step": 1965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9395519950185571,
+      "learning_rate": 1.9641620328093496e-05,
+      "loss": 0.5341,
+      "step": 1966
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6664142487597251,
+      "learning_rate": 1.964112643478997e-05,
+      "loss": 0.3412,
+      "step": 1967
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5458830028835255,
+      "learning_rate": 1.9640632207614647e-05,
+      "loss": 0.3502,
+      "step": 1968
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4650976263027919,
+      "learning_rate": 1.9640137646584646e-05,
+      "loss": 0.3399,
+      "step": 1969
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2908370498165278,
+      "learning_rate": 1.963964275171709e-05,
+      "loss": 0.1562,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7007220925956545,
+      "learning_rate": 1.9639147523029125e-05,
+      "loss": 0.4742,
+      "step": 1971
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5483497042322073,
+      "learning_rate": 1.963865196053789e-05,
+      "loss": 0.3736,
+      "step": 1972
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42343628698039426,
+      "learning_rate": 1.9638156064260555e-05,
+      "loss": 0.1913,
+      "step": 1973
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4601079222814512,
+      "learning_rate": 1.9637659834214294e-05,
+      "loss": 0.4321,
+      "step": 1974
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3014578591026609,
+      "learning_rate": 1.9637163270416283e-05,
+      "loss": 0.7311,
+      "step": 1975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31662376400744924,
+      "learning_rate": 1.9636666372883722e-05,
+      "loss": 0.2132,
+      "step": 1976
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.423523284773505,
+      "learning_rate": 1.963616914163382e-05,
+      "loss": 0.2908,
+      "step": 1977
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.229874934088239,
+      "learning_rate": 1.9635671576683798e-05,
+      "loss": 0.7617,
+      "step": 1978
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4814629313067314,
+      "learning_rate": 1.9635173678050878e-05,
+      "loss": 0.3127,
+      "step": 1979
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5001839315639627,
+      "learning_rate": 1.963467544575231e-05,
+      "loss": 0.3492,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5230200200146033,
+      "learning_rate": 1.9634176879805347e-05,
+      "loss": 0.3975,
+      "step": 1981
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31796381358669257,
+      "learning_rate": 1.9633677980227254e-05,
+      "loss": 0.2032,
+      "step": 1982
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.562655816237142,
+      "learning_rate": 1.96331787470353e-05,
+      "loss": 0.3209,
+      "step": 1983
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3877665388095057,
+      "learning_rate": 1.9632679180246787e-05,
+      "loss": 0.3566,
+      "step": 1984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5555037398364667,
+      "learning_rate": 1.9632179279879006e-05,
+      "loss": 0.4005,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4124910862080048,
+      "learning_rate": 1.963167904594927e-05,
+      "loss": 0.3273,
+      "step": 1986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6000884096195452,
+      "learning_rate": 1.9631178478474905e-05,
+      "loss": 0.3107,
+      "step": 1987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4283814322598163,
+      "learning_rate": 1.9630677577473242e-05,
+      "loss": 0.3005,
+      "step": 1988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4370673442250718,
+      "learning_rate": 1.963017634296163e-05,
+      "loss": 0.2721,
+      "step": 1989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4521900634802621,
+      "learning_rate": 1.9629674774957425e-05,
+      "loss": 0.3517,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5563225794926104,
+      "learning_rate": 1.9629172873477995e-05,
+      "loss": 0.3998,
+      "step": 1991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4409425742430659,
+      "learning_rate": 1.9628670638540722e-05,
+      "loss": 0.2862,
+      "step": 1992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3474166819860802,
+      "learning_rate": 1.9628168070163e-05,
+      "loss": 0.6459,
+      "step": 1993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6413772554148557,
+      "learning_rate": 1.9627665168362234e-05,
+      "loss": 0.4017,
+      "step": 1994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4979080455824859,
+      "learning_rate": 1.9627161933155833e-05,
+      "loss": 0.3773,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47734635309643114,
+      "learning_rate": 1.962665836456123e-05,
+      "loss": 0.3279,
+      "step": 1996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4702257531545852,
+      "learning_rate": 1.9626154462595863e-05,
+      "loss": 0.2722,
+      "step": 1997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36757418615763193,
+      "learning_rate": 1.9625650227277182e-05,
+      "loss": 0.2572,
+      "step": 1998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0698780573684648,
+      "learning_rate": 1.9625145658622644e-05,
+      "loss": 0.3931,
+      "step": 1999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4586305864242774,
+      "learning_rate": 1.962464075664973e-05,
+      "loss": 0.2985,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4799034083760339,
+      "learning_rate": 1.9624135521375914e-05,
+      "loss": 0.345,
+      "step": 2001
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8087456132963655,
+      "learning_rate": 1.9623629952818705e-05,
+      "loss": 0.4573,
+      "step": 2002
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.273169211369182,
+      "learning_rate": 1.9623124050995603e-05,
+      "loss": 0.1335,
+      "step": 2003
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44311381479955264,
+      "learning_rate": 1.9622617815924125e-05,
+      "loss": 0.3134,
+      "step": 2004
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.874639231040764,
+      "learning_rate": 1.962211124762181e-05,
+      "loss": 0.5349,
+      "step": 2005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.910928440837772,
+      "learning_rate": 1.9621604346106197e-05,
+      "loss": 0.5488,
+      "step": 2006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.730448021870774,
+      "learning_rate": 1.9621097111394837e-05,
+      "loss": 0.334,
+      "step": 2007
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8847582506626019,
+      "learning_rate": 1.9620589543505297e-05,
+      "loss": 0.4115,
+      "step": 2008
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26085846379198124,
+      "learning_rate": 1.9620081642455155e-05,
+      "loss": 0.109,
+      "step": 2009
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37726397681575125,
+      "learning_rate": 1.9619573408262004e-05,
+      "loss": 0.2366,
+      "step": 2010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.216495069611042,
+      "learning_rate": 1.9619064840943432e-05,
+      "loss": 0.8289,
+      "step": 2011
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4878648634362295,
+      "learning_rate": 1.9618555940517062e-05,
+      "loss": 0.2654,
+      "step": 2012
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.093933887287843,
+      "learning_rate": 1.9618046707000515e-05,
+      "loss": 0.4438,
+      "step": 2013
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.040222786715266,
+      "learning_rate": 1.9617537140411423e-05,
+      "loss": 0.6686,
+      "step": 2014
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4288877331493517,
+      "learning_rate": 1.961702724076743e-05,
+      "loss": 0.2278,
+      "step": 2015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.609857521709017,
+      "learning_rate": 1.96165170080862e-05,
+      "loss": 0.3412,
+      "step": 2016
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3849639020833395,
+      "learning_rate": 1.9616006442385403e-05,
+      "loss": 0.5256,
+      "step": 2017
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0769342424347998,
+      "learning_rate": 1.9615495543682712e-05,
+      "loss": 0.3283,
+      "step": 2018
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.679860697691885,
+      "learning_rate": 1.9614984311995825e-05,
+      "loss": 0.4145,
+      "step": 2019
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7572448145592617,
+      "learning_rate": 1.9614472747342445e-05,
+      "loss": 0.4164,
+      "step": 2020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0369229362139345,
+      "learning_rate": 1.9613960849740284e-05,
+      "loss": 0.4782,
+      "step": 2021
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43789872014124714,
+      "learning_rate": 1.9613448619207077e-05,
+      "loss": 0.172,
+      "step": 2022
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8217791474880007,
+      "learning_rate": 1.9612936055760557e-05,
+      "loss": 0.4245,
+      "step": 2023
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7758649033699128,
+      "learning_rate": 1.9612423159418474e-05,
+      "loss": 0.4332,
+      "step": 2024
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5558536412840944,
+      "learning_rate": 1.9611909930198588e-05,
+      "loss": 0.305,
+      "step": 2025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.763245662074697,
+      "learning_rate": 1.9611396368118675e-05,
+      "loss": 0.68,
+      "step": 2026
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6710808942620035,
+      "learning_rate": 1.961088247319652e-05,
+      "loss": 0.4107,
+      "step": 2027
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36096294284561037,
+      "learning_rate": 1.961036824544992e-05,
+      "loss": 0.2201,
+      "step": 2028
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5810650256593232,
+      "learning_rate": 1.960985368489668e-05,
+      "loss": 0.2909,
+      "step": 2029
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5690149812729766,
+      "learning_rate": 1.9609338791554623e-05,
+      "loss": 0.5314,
+      "step": 2030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6393362197055744,
+      "learning_rate": 1.960882356544157e-05,
+      "loss": 0.2609,
+      "step": 2031
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6976100461836577,
+      "learning_rate": 1.9608308006575373e-05,
+      "loss": 0.4226,
+      "step": 2032
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.479240310964179,
+      "learning_rate": 1.9607792114973884e-05,
+      "loss": 0.5768,
+      "step": 2033
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3668310724900021,
+      "learning_rate": 1.9607275890654967e-05,
+      "loss": 0.2626,
+      "step": 2034
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.857454535850804,
+      "learning_rate": 1.9606759333636498e-05,
+      "loss": 0.2436,
+      "step": 2035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9569480593524808,
+      "learning_rate": 1.9606242443936368e-05,
+      "loss": 0.4232,
+      "step": 2036
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5653570617669198,
+      "learning_rate": 1.9605725221572475e-05,
+      "loss": 0.3075,
+      "step": 2037
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2376985969066299,
+      "learning_rate": 1.960520766656273e-05,
+      "loss": 0.594,
+      "step": 2038
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6424647493461252,
+      "learning_rate": 1.9604689778925056e-05,
+      "loss": 0.3995,
+      "step": 2039
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4235494564972714,
+      "learning_rate": 1.960417155867739e-05,
+      "loss": 0.3047,
+      "step": 2040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5912471531591597,
+      "learning_rate": 1.960365300583767e-05,
+      "loss": 0.2393,
+      "step": 2041
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.029688638479728,
+      "learning_rate": 1.960313412042387e-05,
+      "loss": 0.5656,
+      "step": 2042
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4732960711478903,
+      "learning_rate": 1.960261490245394e-05,
+      "loss": 0.3001,
+      "step": 2043
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5220022647134588,
+      "learning_rate": 1.9602095351945872e-05,
+      "loss": 0.3535,
+      "step": 2044
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5883266373730363,
+      "learning_rate": 1.9601575468917654e-05,
+      "loss": 0.4899,
+      "step": 2045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3387809210650048,
+      "learning_rate": 1.9601055253387292e-05,
+      "loss": 0.2403,
+      "step": 2046
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5623556957444193,
+      "learning_rate": 1.9600534705372795e-05,
+      "loss": 0.3546,
+      "step": 2047
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5773352363021086,
+      "learning_rate": 1.9600013824892198e-05,
+      "loss": 0.3738,
+      "step": 2048
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4232267894991952,
+      "learning_rate": 1.9599492611963533e-05,
+      "loss": 0.29,
+      "step": 2049
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5190140255118185,
+      "learning_rate": 1.9598971066604854e-05,
+      "loss": 0.818,
+      "step": 2050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6714832479028133,
+      "learning_rate": 1.9598449188834218e-05,
+      "loss": 0.314,
+      "step": 2051
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44430442128339687,
+      "learning_rate": 1.95979269786697e-05,
+      "loss": 0.2978,
+      "step": 2052
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9537546657131378,
+      "learning_rate": 1.959740443612938e-05,
+      "loss": 0.5501,
+      "step": 2053
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2242161983506829,
+      "learning_rate": 1.9596881561231363e-05,
+      "loss": 0.1592,
+      "step": 2054
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48220806957042844,
+      "learning_rate": 1.9596358353993747e-05,
+      "loss": 0.3105,
+      "step": 2055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6976369230595237,
+      "learning_rate": 1.959583481443465e-05,
+      "loss": 0.4052,
+      "step": 2056
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8327429933334323,
+      "learning_rate": 1.9595310942572212e-05,
+      "loss": 0.768,
+      "step": 2057
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44394301554498045,
+      "learning_rate": 1.9594786738424566e-05,
+      "loss": 0.2232,
+      "step": 2058
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7342598439395362,
+      "learning_rate": 1.959426220200987e-05,
+      "loss": 0.4006,
+      "step": 2059
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.427795716449992,
+      "learning_rate": 1.9593737333346286e-05,
+      "loss": 0.2239,
+      "step": 2060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4343683489018406,
+      "learning_rate": 1.9593212132451992e-05,
+      "loss": 0.2375,
+      "step": 2061
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6866265621347991,
+      "learning_rate": 1.959268659934517e-05,
+      "loss": 0.7592,
+      "step": 2062
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.46897540943464006,
+      "learning_rate": 1.9592160734044027e-05,
+      "loss": 0.3492,
+      "step": 2063
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.46889413154603504,
+      "learning_rate": 1.9591634536566766e-05,
+      "loss": 0.271,
+      "step": 2064
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0019504066133773,
+      "learning_rate": 1.9591108006931618e-05,
+      "loss": 0.5227,
+      "step": 2065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4826247658886829,
+      "learning_rate": 1.9590581145156812e-05,
+      "loss": 0.2342,
+      "step": 2066
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.46222434250869515,
+      "learning_rate": 1.959005395126059e-05,
+      "loss": 0.2521,
+      "step": 2067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5843926885025996,
+      "learning_rate": 1.9589526425261213e-05,
+      "loss": 0.4092,
+      "step": 2068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8389100016932158,
+      "learning_rate": 1.958899856717695e-05,
+      "loss": 0.5575,
+      "step": 2069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4238709325497757,
+      "learning_rate": 1.958847037702608e-05,
+      "loss": 0.334,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4507205040580907,
+      "learning_rate": 1.9587941854826892e-05,
+      "loss": 0.3119,
+      "step": 2071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47985631218897673,
+      "learning_rate": 1.9587413000597687e-05,
+      "loss": 0.3043,
+      "step": 2072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7661228750951127,
+      "learning_rate": 1.9586883814356785e-05,
+      "loss": 0.4541,
+      "step": 2073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3672906869250385,
+      "learning_rate": 1.958635429612251e-05,
+      "loss": 0.2207,
+      "step": 2074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4746221052006005,
+      "learning_rate": 1.9585824445913194e-05,
+      "loss": 0.3761,
+      "step": 2075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4915279326174938,
+      "learning_rate": 1.9585294263747192e-05,
+      "loss": 0.3161,
+      "step": 2076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49898984543053004,
+      "learning_rate": 1.9584763749642862e-05,
+      "loss": 0.3595,
+      "step": 2077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5341824181711823,
+      "learning_rate": 1.9584232903618576e-05,
+      "loss": 0.3403,
+      "step": 2078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38017236195959025,
+      "learning_rate": 1.9583701725692716e-05,
+      "loss": 0.3137,
+      "step": 2079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32745929971134763,
+      "learning_rate": 1.9583170215883677e-05,
+      "loss": 0.2298,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8625293763295988,
+      "learning_rate": 1.9582638374209864e-05,
+      "loss": 0.5644,
+      "step": 2081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4296796221626104,
+      "learning_rate": 1.9582106200689698e-05,
+      "loss": 0.3539,
+      "step": 2082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43436138803461416,
+      "learning_rate": 1.9581573695341607e-05,
+      "loss": 0.3427,
+      "step": 2083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9284517598696714,
+      "learning_rate": 1.9581040858184028e-05,
+      "loss": 0.3858,
+      "step": 2084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38693079051285023,
+      "learning_rate": 1.958050768923542e-05,
+      "loss": 0.3074,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4582866470000151,
+      "learning_rate": 1.957997418851424e-05,
+      "loss": 0.2932,
+      "step": 2086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7347478591787949,
+      "learning_rate": 1.9579440356038966e-05,
+      "loss": 0.3689,
+      "step": 2087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3424729743541756,
+      "learning_rate": 1.9578906191828086e-05,
+      "loss": 0.2313,
+      "step": 2088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6525889969280737,
+      "learning_rate": 1.9578371695900097e-05,
+      "loss": 0.4339,
+      "step": 2089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49133362050474455,
+      "learning_rate": 1.9577836868273504e-05,
+      "loss": 0.3199,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.46103788217869224,
+      "learning_rate": 1.9577301708966837e-05,
+      "loss": 0.3133,
+      "step": 2091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3968124504129748,
+      "learning_rate": 1.9576766217998622e-05,
+      "loss": 0.2731,
+      "step": 2092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5598574134391687,
+      "learning_rate": 1.9576230395387403e-05,
+      "loss": 0.3231,
+      "step": 2093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.507072225088975,
+      "learning_rate": 1.9575694241151737e-05,
+      "loss": 0.3042,
+      "step": 2094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4229706741547905,
+      "learning_rate": 1.9575157755310193e-05,
+      "loss": 0.3297,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2787555528563692,
+      "learning_rate": 1.957462093788135e-05,
+      "loss": 0.677,
+      "step": 2096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40537782087377044,
+      "learning_rate": 1.957408378888379e-05,
+      "loss": 0.245,
+      "step": 2097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6392618964501624,
+      "learning_rate": 1.957354630833612e-05,
+      "loss": 0.4378,
+      "step": 2098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44083595616310706,
+      "learning_rate": 1.957300849625696e-05,
+      "loss": 0.3578,
+      "step": 2099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.277591819404161,
+      "learning_rate": 1.9572470352664923e-05,
+      "loss": 0.1437,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6445522078777529,
+      "learning_rate": 1.957193187757865e-05,
+      "loss": 0.4184,
+      "step": 2101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.06003956945609,
+      "learning_rate": 1.9571393071016785e-05,
+      "loss": 0.7141,
+      "step": 2102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3327901938604386,
+      "learning_rate": 1.9570853932997993e-05,
+      "loss": 0.2564,
+      "step": 2103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5349142642984711,
+      "learning_rate": 1.957031446354094e-05,
+      "loss": 0.4241,
+      "step": 2104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7450362254049,
+      "learning_rate": 1.9569774662664306e-05,
+      "loss": 0.5647,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2616609861296739,
+      "learning_rate": 1.9569234530386792e-05,
+      "loss": 0.1297,
+      "step": 2106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4211284765636636,
+      "learning_rate": 1.9568694066727095e-05,
+      "loss": 0.311,
+      "step": 2107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4618973142603389,
+      "learning_rate": 1.9568153271703932e-05,
+      "loss": 0.8132,
+      "step": 2108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.738279974487151,
+      "learning_rate": 1.9567612145336036e-05,
+      "loss": 0.4929,
+      "step": 2109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44960921087855577,
+      "learning_rate": 1.9567070687642142e-05,
+      "loss": 0.317,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4821347287488108,
+      "learning_rate": 1.9566528898641e-05,
+      "loss": 0.3665,
+      "step": 2111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32569543755682334,
+      "learning_rate": 1.9565986778351376e-05,
+      "loss": 0.1666,
+      "step": 2112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48347428505913215,
+      "learning_rate": 1.9565444326792038e-05,
+      "loss": 0.2783,
+      "step": 2113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5037634230328805,
+      "learning_rate": 1.9564901543981776e-05,
+      "loss": 0.7917,
+      "step": 2114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3697719835295979,
+      "learning_rate": 1.9564358429939386e-05,
+      "loss": 0.3053,
+      "step": 2115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5447348890954887,
+      "learning_rate": 1.9563814984683674e-05,
+      "loss": 0.3506,
+      "step": 2116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.304263404947526,
+      "learning_rate": 1.9563271208233462e-05,
+      "loss": 0.8144,
+      "step": 2117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3159438724391187,
+      "learning_rate": 1.9562727100607577e-05,
+      "loss": 0.1621,
+      "step": 2118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41587057997986904,
+      "learning_rate": 1.956218266182486e-05,
+      "loss": 0.2664,
+      "step": 2119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0253769563866983,
+      "learning_rate": 1.9561637891904176e-05,
+      "loss": 0.568,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6434762446373379,
+      "learning_rate": 1.9561092790864376e-05,
+      "loss": 0.4318,
+      "step": 2121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5164150783625712,
+      "learning_rate": 1.956054735872435e-05,
+      "loss": 0.3499,
+      "step": 2122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.437135421408849,
+      "learning_rate": 1.9560001595502978e-05,
+      "loss": 0.3035,
+      "step": 2123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3042150377766891,
+      "learning_rate": 1.955945550121916e-05,
+      "loss": 0.1825,
+      "step": 2124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.506806034858422,
+      "learning_rate": 1.9558909075891812e-05,
+      "loss": 0.3272,
+      "step": 2125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47732056135622875,
+      "learning_rate": 1.955836231953985e-05,
+      "loss": 0.3269,
+      "step": 2126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.968804033210661,
+      "learning_rate": 1.9557815232182216e-05,
+      "loss": 0.4565,
+      "step": 2127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3925570608975105,
+      "learning_rate": 1.9557267813837848e-05,
+      "loss": 0.3436,
+      "step": 2128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9226390952793319,
+      "learning_rate": 1.955672006452571e-05,
+      "loss": 0.4685,
+      "step": 2129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4344185349778874,
+      "learning_rate": 1.9556171984264765e-05,
+      "loss": 0.2778,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3822100632725823,
+      "learning_rate": 1.9555623573073994e-05,
+      "loss": 0.2504,
+      "step": 2131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.166974059789117,
+      "learning_rate": 1.955507483097239e-05,
+      "loss": 0.548,
+      "step": 2132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7183519526044224,
+      "learning_rate": 1.9554525757978958e-05,
+      "loss": 0.4007,
+      "step": 2133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40625089003812276,
+      "learning_rate": 1.955397635411271e-05,
+      "loss": 0.3013,
+      "step": 2134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6427276061652427,
+      "learning_rate": 1.955342661939267e-05,
+      "loss": 0.4044,
+      "step": 2135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7481016203756692,
+      "learning_rate": 1.9552876553837878e-05,
+      "loss": 0.2465,
+      "step": 2136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4078942359314576,
+      "learning_rate": 1.955232615746738e-05,
+      "loss": 0.2344,
+      "step": 2137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9841112838622158,
+      "learning_rate": 1.9551775430300238e-05,
+      "loss": 0.4383,
+      "step": 2138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5783608548387257,
+      "learning_rate": 1.9551224372355523e-05,
+      "loss": 0.2947,
+      "step": 2139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.513156088021875,
+      "learning_rate": 1.955067298365232e-05,
+      "loss": 0.3523,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4460222044906592,
+      "learning_rate": 1.9550121264209724e-05,
+      "loss": 0.8275,
+      "step": 2141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47796263241095566,
+      "learning_rate": 1.9549569214046837e-05,
+      "loss": 0.261,
+      "step": 2142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5068551991890565,
+      "learning_rate": 1.954901683318278e-05,
+      "loss": 0.2742,
+      "step": 2143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7854309469545647,
+      "learning_rate": 1.9548464121636678e-05,
+      "loss": 0.39,
+      "step": 2144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9562992698117687,
+      "learning_rate": 1.954791107942768e-05,
+      "loss": 0.4997,
+      "step": 2145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49884170414180384,
+      "learning_rate": 1.9547357706574926e-05,
+      "loss": 0.2523,
+      "step": 2146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43289131925381763,
+      "learning_rate": 1.9546804003097588e-05,
+      "loss": 0.3589,
+      "step": 2147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1386748667131321,
+      "learning_rate": 1.9546249969014836e-05,
+      "loss": 0.673,
+      "step": 2148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.478383355031457,
+      "learning_rate": 1.954569560434586e-05,
+      "loss": 0.283,
+      "step": 2149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5539994701491447,
+      "learning_rate": 1.9545140909109854e-05,
+      "loss": 0.3762,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35583342252654265,
+      "learning_rate": 1.954458588332603e-05,
+      "loss": 0.168,
+      "step": 2151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47482346170901524,
+      "learning_rate": 1.9544030527013603e-05,
+      "loss": 0.236,
+      "step": 2152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.488695547220567,
+      "learning_rate": 1.9543474840191817e-05,
+      "loss": 0.7431,
+      "step": 2153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4557534247583365,
+      "learning_rate": 1.9542918822879902e-05,
+      "loss": 0.3779,
+      "step": 2154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33288832250828193,
+      "learning_rate": 1.954236247509712e-05,
+      "loss": 0.2592,
+      "step": 2155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4796173843723215,
+      "learning_rate": 1.954180579686274e-05,
+      "loss": 0.3337,
+      "step": 2156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4089067855760762,
+      "learning_rate": 1.954124878819603e-05,
+      "loss": 0.2759,
+      "step": 2157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4690219866747868,
+      "learning_rate": 1.9540691449116286e-05,
+      "loss": 0.2872,
+      "step": 2158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.445945068109367,
+      "learning_rate": 1.9540133779642806e-05,
+      "loss": 0.3255,
+      "step": 2159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.620237870223999,
+      "learning_rate": 1.9539575779794906e-05,
+      "loss": 0.4513,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4622798158558348,
+      "learning_rate": 1.9539017449591905e-05,
+      "loss": 0.3245,
+      "step": 2161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40301614161542826,
+      "learning_rate": 1.9538458789053143e-05,
+      "loss": 0.3361,
+      "step": 2162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4533281560660749,
+      "learning_rate": 1.9537899798197963e-05,
+      "loss": 0.2963,
+      "step": 2163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4209930222368525,
+      "learning_rate": 1.953734047704572e-05,
+      "loss": 0.281,
+      "step": 2164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36565295658400504,
+      "learning_rate": 1.9536780825615788e-05,
+      "loss": 0.2151,
+      "step": 2165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4422205941031057,
+      "learning_rate": 1.9536220843927544e-05,
+      "loss": 0.3669,
+      "step": 2166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4385665259665185,
+      "learning_rate": 1.953566053200039e-05,
+      "loss": 0.3155,
+      "step": 2167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6424769906790054,
+      "learning_rate": 1.953509988985371e-05,
+      "loss": 0.4169,
+      "step": 2168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.194460975030824,
+      "learning_rate": 1.953453891750694e-05,
+      "loss": 0.7403,
+      "step": 2169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3717051558823366,
+      "learning_rate": 1.9533977614979493e-05,
+      "loss": 0.2938,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3201587539393018,
+      "learning_rate": 1.9533415982290813e-05,
+      "loss": 0.2192,
+      "step": 2171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7368872184262916,
+      "learning_rate": 1.9532854019460346e-05,
+      "loss": 0.4229,
+      "step": 2172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44181930693740384,
+      "learning_rate": 1.9532291726507557e-05,
+      "loss": 0.2848,
+      "step": 2173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4664201908407062,
+      "learning_rate": 1.9531729103451912e-05,
+      "loss": 0.35,
+      "step": 2174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.454446688689338,
+      "learning_rate": 1.9531166150312902e-05,
+      "loss": 0.605,
+      "step": 2175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46776768368543914,
+      "learning_rate": 1.9530602867110015e-05,
+      "loss": 0.2971,
+      "step": 2176
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3477231628548521,
+      "learning_rate": 1.953003925386276e-05,
+      "loss": 0.2378,
+      "step": 2177
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5094617271180392,
+      "learning_rate": 1.9529475310590656e-05,
+      "loss": 0.3366,
+      "step": 2178
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46931079454798147,
+      "learning_rate": 1.9528911037313233e-05,
+      "loss": 0.3177,
+      "step": 2179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.178746644251515,
+      "learning_rate": 1.952834643405003e-05,
+      "loss": 0.6243,
+      "step": 2180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9884192839837473,
+      "learning_rate": 1.95277815008206e-05,
+      "loss": 0.5839,
+      "step": 2181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4751766716695899,
+      "learning_rate": 1.9527216237644508e-05,
+      "loss": 0.2844,
+      "step": 2182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.587711364919638,
+      "learning_rate": 1.9526650644541326e-05,
+      "loss": 0.3622,
+      "step": 2183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8459515983868039,
+      "learning_rate": 1.952608472153064e-05,
+      "loss": 0.3906,
+      "step": 2184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4486221294280846,
+      "learning_rate": 1.952551846863205e-05,
+      "loss": 0.2203,
+      "step": 2185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6553354179919787,
+      "learning_rate": 1.9524951885865165e-05,
+      "loss": 0.3764,
+      "step": 2186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.830584238414141,
+      "learning_rate": 1.952438497324961e-05,
+      "loss": 0.7863,
+      "step": 2187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5252954020431466,
+      "learning_rate": 1.9523817730805008e-05,
+      "loss": 0.2538,
+      "step": 2188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8584255441697742,
+      "learning_rate": 1.952325015855101e-05,
+      "loss": 0.5336,
+      "step": 2189
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3471580515229295,
+      "learning_rate": 1.9522682256507268e-05,
+      "loss": 0.2511,
+      "step": 2190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4088705601951199,
+      "learning_rate": 1.952211402469345e-05,
+      "loss": 0.2464,
+      "step": 2191
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2705970535049596,
+      "learning_rate": 1.952154546312923e-05,
+      "loss": 0.6611,
+      "step": 2192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6454425353434776,
+      "learning_rate": 1.9520976571834304e-05,
+      "loss": 0.4209,
+      "step": 2193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44634869722763765,
+      "learning_rate": 1.9520407350828364e-05,
+      "loss": 0.2355,
+      "step": 2194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7939220131831405,
+      "learning_rate": 1.9519837800131127e-05,
+      "loss": 0.4377,
+      "step": 2195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3147220383048209,
+      "learning_rate": 1.9519267919762318e-05,
+      "loss": 0.1702,
+      "step": 2196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.703105533799731,
+      "learning_rate": 1.951869770974167e-05,
+      "loss": 0.3809,
+      "step": 2197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47335849847144024,
+      "learning_rate": 1.951812717008893e-05,
+      "loss": 0.2881,
+      "step": 2198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1606156210211465,
+      "learning_rate": 1.951755630082386e-05,
+      "loss": 0.7156,
+      "step": 2199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45947436250339385,
+      "learning_rate": 1.9516985101966218e-05,
+      "loss": 0.3269,
+      "step": 2200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43358484416349463,
+      "learning_rate": 1.9516413573535794e-05,
+      "loss": 0.3523,
+      "step": 2201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5472429483088781,
+      "learning_rate": 1.9515841715552376e-05,
+      "loss": 0.3946,
+      "step": 2202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34276882389222835,
+      "learning_rate": 1.951526952803577e-05,
+      "loss": 0.1918,
+      "step": 2203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5604731992436495,
+      "learning_rate": 1.951469701100579e-05,
+      "loss": 0.3188,
+      "step": 2204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5485943753095354,
+      "learning_rate": 1.951412416448226e-05,
+      "loss": 0.4226,
+      "step": 2205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38949595158304373,
+      "learning_rate": 1.951355098848502e-05,
+      "loss": 0.3067,
+      "step": 2206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5935212725749861,
+      "learning_rate": 1.9512977483033916e-05,
+      "loss": 0.4256,
+      "step": 2207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3030094098212264,
+      "learning_rate": 1.9512403648148813e-05,
+      "loss": 0.1428,
+      "step": 2208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3956227850830546,
+      "learning_rate": 1.951182948384958e-05,
+      "loss": 0.2861,
+      "step": 2209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4697965880003851,
+      "learning_rate": 1.9511254990156103e-05,
+      "loss": 0.3211,
+      "step": 2210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9172394638687885,
+      "learning_rate": 1.9510680167088275e-05,
+      "loss": 0.4232,
+      "step": 2211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43535146422603604,
+      "learning_rate": 1.9510105014665998e-05,
+      "loss": 0.3297,
+      "step": 2212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5773882269858571,
+      "learning_rate": 1.9509529532909196e-05,
+      "loss": 0.4388,
+      "step": 2213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3137998718551292,
+      "learning_rate": 1.9508953721837795e-05,
+      "loss": 0.2522,
+      "step": 2214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35939192870248576,
+      "learning_rate": 1.9508377581471735e-05,
+      "loss": 0.1858,
+      "step": 2215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5116306737278765,
+      "learning_rate": 1.9507801111830967e-05,
+      "loss": 0.3576,
+      "step": 2216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5731809416440271,
+      "learning_rate": 1.950722431293546e-05,
+      "loss": 0.3353,
+      "step": 2217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47304259161573736,
+      "learning_rate": 1.950664718480518e-05,
+      "loss": 0.3236,
+      "step": 2218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6276525740396246,
+      "learning_rate": 1.9506069727460116e-05,
+      "loss": 0.4444,
+      "step": 2219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.090832100391312,
+      "learning_rate": 1.9505491940920268e-05,
+      "loss": 0.6109,
+      "step": 2220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2481407623113843,
+      "learning_rate": 1.9504913825205643e-05,
+      "loss": 0.1352,
+      "step": 2221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4448099721555093,
+      "learning_rate": 1.950433538033626e-05,
+      "loss": 0.3269,
+      "step": 2222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.866009184797271,
+      "learning_rate": 1.950375660633215e-05,
+      "loss": 0.55,
+      "step": 2223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4384851116001498,
+      "learning_rate": 1.950317750321336e-05,
+      "loss": 0.2842,
+      "step": 2224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6313666828316659,
+      "learning_rate": 1.950259807099994e-05,
+      "loss": 0.3842,
+      "step": 2225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4331107293530284,
+      "learning_rate": 1.950201830971196e-05,
+      "loss": 0.3443,
+      "step": 2226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23533665688629704,
+      "learning_rate": 1.9501438219369492e-05,
+      "loss": 0.1392,
+      "step": 2227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6139970636009742,
+      "learning_rate": 1.9500857799992628e-05,
+      "loss": 0.4041,
+      "step": 2228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48155996698312703,
+      "learning_rate": 1.9500277051601465e-05,
+      "loss": 0.3663,
+      "step": 2229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4152605894643503,
+      "learning_rate": 1.949969597421612e-05,
+      "loss": 0.2154,
+      "step": 2230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6361750555191724,
+      "learning_rate": 1.9499114567856708e-05,
+      "loss": 0.4438,
+      "step": 2231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2016891543780488,
+      "learning_rate": 1.9498532832543372e-05,
+      "loss": 0.6434,
+      "step": 2232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39561891747852557,
+      "learning_rate": 1.9497950768296246e-05,
+      "loss": 0.2549,
+      "step": 2233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31869872802169985,
+      "learning_rate": 1.9497368375135497e-05,
+      "loss": 0.2302,
+      "step": 2234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8584671892353136,
+      "learning_rate": 1.949678565308129e-05,
+      "loss": 0.4976,
+      "step": 2235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8802500200900603,
+      "learning_rate": 1.9496202602153805e-05,
+      "loss": 0.4712,
+      "step": 2236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35855578518631354,
+      "learning_rate": 1.949561922237323e-05,
+      "loss": 0.2521,
+      "step": 2237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5343010087481445,
+      "learning_rate": 1.9495035513759772e-05,
+      "loss": 0.386,
+      "step": 2238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3763359873365377,
+      "learning_rate": 1.9494451476333637e-05,
+      "loss": 0.2417,
+      "step": 2239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7266720426527591,
+      "learning_rate": 1.949386711011506e-05,
+      "loss": 0.3591,
+      "step": 2240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3434275838027053,
+      "learning_rate": 1.9493282415124274e-05,
+      "loss": 0.279,
+      "step": 2241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.793014395205871,
+      "learning_rate": 1.9492697391381523e-05,
+      "loss": 0.3409,
+      "step": 2242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4304945592429074,
+      "learning_rate": 1.949211203890707e-05,
+      "loss": 0.2789,
+      "step": 2243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8892252947519116,
+      "learning_rate": 1.949152635772119e-05,
+      "loss": 0.617,
+      "step": 2244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.362436878968347,
+      "learning_rate": 1.9490940347844156e-05,
+      "loss": 0.3207,
+      "step": 2245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44349164878713104,
+      "learning_rate": 1.9490354009296268e-05,
+      "loss": 0.3616,
+      "step": 2246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3274107448446957,
+      "learning_rate": 1.9489767342097824e-05,
+      "loss": 0.2196,
+      "step": 2247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8783237744052049,
+      "learning_rate": 1.948918034626915e-05,
+      "loss": 0.5678,
+      "step": 2248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43709607755935653,
+      "learning_rate": 1.948859302183057e-05,
+      "loss": 0.329,
+      "step": 2249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4099301199962942,
+      "learning_rate": 1.9488005368802415e-05,
+      "loss": 0.2837,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5420335066249002,
+      "learning_rate": 1.948741738720505e-05,
+      "loss": 0.4656,
+      "step": 2251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3731959010848403,
+      "learning_rate": 1.9486829077058823e-05,
+      "loss": 0.3333,
+      "step": 2252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2855765128398837,
+      "learning_rate": 1.9486240438384114e-05,
+      "loss": 0.2129,
+      "step": 2253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8666950180014855,
+      "learning_rate": 1.9485651471201306e-05,
+      "loss": 0.4876,
+      "step": 2254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46097536285247287,
+      "learning_rate": 1.94850621755308e-05,
+      "loss": 0.3323,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7361756550595048,
+      "learning_rate": 1.9484472551392993e-05,
+      "loss": 0.4123,
+      "step": 2256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3658556105623814,
+      "learning_rate": 1.9483882598808315e-05,
+      "loss": 0.3458,
+      "step": 2257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44969033758873034,
+      "learning_rate": 1.948329231779719e-05,
+      "loss": 0.3461,
+      "step": 2258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5560341222882007,
+      "learning_rate": 1.9482701708380056e-05,
+      "loss": 0.2802,
+      "step": 2259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4553291061872389,
+      "learning_rate": 1.948211077057737e-05,
+      "loss": 0.2485,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4033237221641209,
+      "learning_rate": 1.9481519504409596e-05,
+      "loss": 0.2866,
+      "step": 2261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5367463676843832,
+      "learning_rate": 1.948092790989721e-05,
+      "loss": 0.3968,
+      "step": 2262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5626236268127532,
+      "learning_rate": 1.94803359870607e-05,
+      "loss": 0.3514,
+      "step": 2263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4338608687064876,
+      "learning_rate": 1.947974373592056e-05,
+      "loss": 0.3034,
+      "step": 2264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4131007253237356,
+      "learning_rate": 1.9479151156497303e-05,
+      "loss": 0.3456,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7775496063551903,
+      "learning_rate": 1.9478558248811448e-05,
+      "loss": 0.3321,
+      "step": 2266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35258832797577216,
+      "learning_rate": 1.947796501288353e-05,
+      "loss": 0.2487,
+      "step": 2267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6701083819158502,
+      "learning_rate": 1.947737144873409e-05,
+      "loss": 0.3394,
+      "step": 2268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43898287408668396,
+      "learning_rate": 1.9476777556383685e-05,
+      "loss": 0.3219,
+      "step": 2269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3596482929485019,
+      "learning_rate": 1.947618333585288e-05,
+      "loss": 0.2842,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1019436057037184,
+      "learning_rate": 1.947558878716225e-05,
+      "loss": 0.7142,
+      "step": 2271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5411126947410136,
+      "learning_rate": 1.9474993910332394e-05,
+      "loss": 0.3461,
+      "step": 2272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40898556290402976,
+      "learning_rate": 1.9474398705383904e-05,
+      "loss": 0.2542,
+      "step": 2273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6471286583545248,
+      "learning_rate": 1.9473803172337396e-05,
+      "loss": 0.3206,
+      "step": 2274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8779107686075143,
+      "learning_rate": 1.947320731121349e-05,
+      "loss": 0.5398,
+      "step": 2275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3532457572292047,
+      "learning_rate": 1.947261112203282e-05,
+      "loss": 0.2389,
+      "step": 2276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4493598183529648,
+      "learning_rate": 1.947201460481604e-05,
+      "loss": 0.3389,
+      "step": 2277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2898301269699424,
+      "learning_rate": 1.9471417759583796e-05,
+      "loss": 0.7665,
+      "step": 2278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.465301976628202,
+      "learning_rate": 1.9470820586356763e-05,
+      "loss": 0.2486,
+      "step": 2279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8359470240166336,
+      "learning_rate": 1.9470223085155622e-05,
+      "loss": 0.3916,
+      "step": 2280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36417049272575086,
+      "learning_rate": 1.946962525600106e-05,
+      "loss": 0.2749,
+      "step": 2281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39093612328850025,
+      "learning_rate": 1.9469027098913787e-05,
+      "loss": 0.2384,
+      "step": 2282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4785923875213445,
+      "learning_rate": 1.946842861391451e-05,
+      "loss": 0.876,
+      "step": 2283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49453376260267884,
+      "learning_rate": 1.9467829801023957e-05,
+      "loss": 0.3659,
+      "step": 2284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5000531898361844,
+      "learning_rate": 1.9467230660262864e-05,
+      "loss": 0.3059,
+      "step": 2285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5643620845088193,
+      "learning_rate": 1.9466631191651984e-05,
+      "loss": 0.305,
+      "step": 2286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4382366231139346,
+      "learning_rate": 1.9466031395212073e-05,
+      "loss": 0.3067,
+      "step": 2287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40747576339356095,
+      "learning_rate": 1.9465431270963898e-05,
+      "loss": 0.2737,
+      "step": 2288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4389132877484424,
+      "learning_rate": 1.9464830818928247e-05,
+      "loss": 0.2912,
+      "step": 2289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8213346014038311,
+      "learning_rate": 1.946423003912591e-05,
+      "loss": 0.648,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3995884995275768,
+      "learning_rate": 1.94636289315777e-05,
+      "loss": 0.3271,
+      "step": 2291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8771463518079066,
+      "learning_rate": 1.946302749630442e-05,
+      "loss": 0.4139,
+      "step": 2292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3162617732642643,
+      "learning_rate": 1.9462425733326906e-05,
+      "loss": 0.2405,
+      "step": 2293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40209065551847967,
+      "learning_rate": 1.9461823642666e-05,
+      "loss": 0.2874,
+      "step": 2294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9720138998527125,
+      "learning_rate": 1.9461221224342544e-05,
+      "loss": 0.5947,
+      "step": 2295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46874994093569994,
+      "learning_rate": 1.9460618478377406e-05,
+      "loss": 0.342,
+      "step": 2296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4541039308298201,
+      "learning_rate": 1.9460015404791456e-05,
+      "loss": 0.3137,
+      "step": 2297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9474289796905376,
+      "learning_rate": 1.9459412003605577e-05,
+      "loss": 0.595,
+      "step": 2298
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3041258639698369,
+      "learning_rate": 1.945880827484067e-05,
+      "loss": 0.1296,
+      "step": 2299
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44826921009046905,
+      "learning_rate": 1.9458204218517638e-05,
+      "loss": 0.2813,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4353294044673283,
+      "learning_rate": 1.94575998346574e-05,
+      "loss": 0.3258,
+      "step": 2301
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9102026658374495,
+      "learning_rate": 1.945699512328089e-05,
+      "loss": 0.4023,
+      "step": 2302
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4436564510925583,
+      "learning_rate": 1.9456390084409044e-05,
+      "loss": 0.3012,
+      "step": 2303
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8056583922169369,
+      "learning_rate": 1.9455784718062813e-05,
+      "loss": 0.5757,
+      "step": 2304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2843505676702236,
+      "learning_rate": 1.9455179024263166e-05,
+      "loss": 0.1952,
+      "step": 2305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42052209498560444,
+      "learning_rate": 1.9454573003031078e-05,
+      "loss": 0.2669,
+      "step": 2306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9245941501277933,
+      "learning_rate": 1.945396665438753e-05,
+      "loss": 0.6041,
+      "step": 2307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5978445007136682,
+      "learning_rate": 1.9453359978353524e-05,
+      "loss": 0.4349,
+      "step": 2308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32742532663412943,
+      "learning_rate": 1.945275297495007e-05,
+      "loss": 0.2599,
+      "step": 2309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9431148146992495,
+      "learning_rate": 1.9452145644198185e-05,
+      "loss": 0.6554,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3051067841482634,
+      "learning_rate": 1.9451537986118904e-05,
+      "loss": 0.1965,
+      "step": 2311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4158374417954412,
+      "learning_rate": 1.945093000073327e-05,
+      "loss": 0.2121,
+      "step": 2312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4787539021841934,
+      "learning_rate": 1.9450321688062336e-05,
+      "loss": 0.3388,
+      "step": 2313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1867093489703773,
+      "learning_rate": 1.944971304812717e-05,
+      "loss": 0.5359,
+      "step": 2314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31635180255683376,
+      "learning_rate": 1.9449104080948842e-05,
+      "loss": 0.2358,
+      "step": 2315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9261529169569618,
+      "learning_rate": 1.9448494786548448e-05,
+      "loss": 0.8326,
+      "step": 2316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34949286273506625,
+      "learning_rate": 1.944788516494709e-05,
+      "loss": 0.2497,
+      "step": 2317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4379583841036261,
+      "learning_rate": 1.944727521616587e-05,
+      "loss": 0.2183,
+      "step": 2318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8861251508067872,
+      "learning_rate": 1.9446664940225917e-05,
+      "loss": 0.4202,
+      "step": 2319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5479306060921134,
+      "learning_rate": 1.9446054337148364e-05,
+      "loss": 0.3957,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8068260676833023,
+      "learning_rate": 1.9445443406954357e-05,
+      "loss": 0.4384,
+      "step": 2321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4521357323513071,
+      "learning_rate": 1.9444832149665048e-05,
+      "loss": 0.285,
+      "step": 2322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3373624792377293,
+      "learning_rate": 1.944422056530161e-05,
+      "loss": 0.1889,
+      "step": 2323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5131559776901216,
+      "learning_rate": 1.944360865388522e-05,
+      "loss": 0.3261,
+      "step": 2324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4789087105248987,
+      "learning_rate": 1.9442996415437066e-05,
+      "loss": 0.3175,
+      "step": 2325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0279162829610848,
+      "learning_rate": 1.9442383849978354e-05,
+      "loss": 0.6445,
+      "step": 2326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5037670907171229,
+      "learning_rate": 1.9441770957530295e-05,
+      "loss": 0.3263,
+      "step": 2327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4267882635388313,
+      "learning_rate": 1.9441157738114114e-05,
+      "loss": 0.2697,
+      "step": 2328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3900503798338192,
+      "learning_rate": 1.9440544191751046e-05,
+      "loss": 0.2927,
+      "step": 2329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3245214667402623,
+      "learning_rate": 1.943993031846234e-05,
+      "loss": 0.2375,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7409088978080296,
+      "learning_rate": 1.9439316118269248e-05,
+      "loss": 0.3682,
+      "step": 2331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5008460271734648,
+      "learning_rate": 1.943870159119305e-05,
+      "loss": 0.3801,
+      "step": 2332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41324478033839196,
+      "learning_rate": 1.9438086737255022e-05,
+      "loss": 0.2974,
+      "step": 2333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0615661938603904,
+      "learning_rate": 1.9437471556476454e-05,
+      "loss": 0.6605,
+      "step": 2334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7578715878324432,
+      "learning_rate": 1.9436856048878653e-05,
+      "loss": 0.4057,
+      "step": 2335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4578545746760707,
+      "learning_rate": 1.943624021448293e-05,
+      "loss": 0.2886,
+      "step": 2336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46684580870173853,
+      "learning_rate": 1.9435624053310617e-05,
+      "loss": 0.335,
+      "step": 2337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2697068340642194,
+      "learning_rate": 1.943500756538305e-05,
+      "loss": 0.1427,
+      "step": 2338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.620415357649466,
+      "learning_rate": 1.943439075072157e-05,
+      "loss": 0.2752,
+      "step": 2339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4502906967970935,
+      "learning_rate": 1.9433773609347553e-05,
+      "loss": 0.3357,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5019698091806195,
+      "learning_rate": 1.9433156141282356e-05,
+      "loss": 0.3581,
+      "step": 2341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6175844702796243,
+      "learning_rate": 1.943253834654737e-05,
+      "loss": 0.4041,
+      "step": 2342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.365708234558843,
+      "learning_rate": 1.9431920225163984e-05,
+      "loss": 0.2691,
+      "step": 2343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45935395249114086,
+      "learning_rate": 1.9431301777153607e-05,
+      "loss": 0.2848,
+      "step": 2344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44585991482447607,
+      "learning_rate": 1.9430683002537657e-05,
+      "loss": 0.327,
+      "step": 2345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3728807303254334,
+      "learning_rate": 1.9430063901337562e-05,
+      "loss": 0.3105,
+      "step": 2346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7637717215205987,
+      "learning_rate": 1.9429444473574753e-05,
+      "loss": 0.5145,
+      "step": 2347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4234374308721751,
+      "learning_rate": 1.942882471927069e-05,
+      "loss": 0.2724,
+      "step": 2348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.569665903816151,
+      "learning_rate": 1.9428204638446834e-05,
+      "loss": 0.412,
+      "step": 2349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4707793907074113,
+      "learning_rate": 1.9427584231124656e-05,
+      "loss": 0.3092,
+      "step": 2350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2978071148204672,
+      "learning_rate": 1.942696349732564e-05,
+      "loss": 0.1615,
+      "step": 2351
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4934235765164325,
+      "learning_rate": 1.9426342437071287e-05,
+      "loss": 0.3496,
+      "step": 2352
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5296265448238974,
+      "learning_rate": 1.94257210503831e-05,
+      "loss": 0.4341,
+      "step": 2353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4039824220929225,
+      "learning_rate": 1.9425099337282596e-05,
+      "loss": 0.2677,
+      "step": 2354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46538885633625465,
+      "learning_rate": 1.942447729779131e-05,
+      "loss": 0.3583,
+      "step": 2355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5022552896111364,
+      "learning_rate": 1.9423854931930778e-05,
+      "loss": 0.3459,
+      "step": 2356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2656556336693924,
+      "learning_rate": 1.9423232239722557e-05,
+      "loss": 0.1004,
+      "step": 2357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3751636957128666,
+      "learning_rate": 1.9422609221188208e-05,
+      "loss": 0.2678,
+      "step": 2358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7635549619993852,
+      "learning_rate": 1.9421985876349307e-05,
+      "loss": 0.5,
+      "step": 2359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4246343907147413,
+      "learning_rate": 1.942136220522744e-05,
+      "loss": 0.3152,
+      "step": 2360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40637458685864913,
+      "learning_rate": 1.9420738207844202e-05,
+      "loss": 0.2499,
+      "step": 2361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2699260617905255,
+      "learning_rate": 1.9420113884221207e-05,
+      "loss": 0.7759,
+      "step": 2362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2895707578621004,
+      "learning_rate": 1.9419489234380077e-05,
+      "loss": 0.1907,
+      "step": 2363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42114540707427844,
+      "learning_rate": 1.9418864258342433e-05,
+      "loss": 0.3008,
+      "step": 2364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.823988721160775,
+      "learning_rate": 1.941823895612993e-05,
+      "loss": 0.5296,
+      "step": 2365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5848088756752329,
+      "learning_rate": 1.9417613327764214e-05,
+      "loss": 0.3968,
+      "step": 2366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4707688843312788,
+      "learning_rate": 1.9416987373266957e-05,
+      "loss": 0.2714,
+      "step": 2367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4300757765992352,
+      "learning_rate": 1.941636109265983e-05,
+      "loss": 0.3517,
+      "step": 2368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3547634392299474,
+      "learning_rate": 1.941573448596452e-05,
+      "loss": 0.2276,
+      "step": 2369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4239024769417483,
+      "learning_rate": 1.9415107553202736e-05,
+      "loss": 0.3535,
+      "step": 2370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5859513694469434,
+      "learning_rate": 1.9414480294396178e-05,
+      "loss": 0.3019,
+      "step": 2371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41543686750719744,
+      "learning_rate": 1.9413852709566574e-05,
+      "loss": 0.3323,
+      "step": 2372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4297104608283397,
+      "learning_rate": 1.9413224798735655e-05,
+      "loss": 0.3159,
+      "step": 2373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9676280782400583,
+      "learning_rate": 1.9412596561925164e-05,
+      "loss": 0.5264,
+      "step": 2374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7162830264380677,
+      "learning_rate": 1.9411967999156866e-05,
+      "loss": 0.5078,
+      "step": 2375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3800369126423308,
+      "learning_rate": 1.9411339110452512e-05,
+      "loss": 0.3029,
+      "step": 2376
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3623660175814287,
+      "learning_rate": 1.9410709895833895e-05,
+      "loss": 0.2252,
+      "step": 2377
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7698720451832731,
+      "learning_rate": 1.9410080355322797e-05,
+      "loss": 0.4136,
+      "step": 2378
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5038883234863918,
+      "learning_rate": 1.9409450488941018e-05,
+      "loss": 0.3308,
+      "step": 2379
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.523722918460601,
+      "learning_rate": 1.9408820296710377e-05,
+      "loss": 0.3026,
+      "step": 2380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6180731282065387,
+      "learning_rate": 1.9408189778652694e-05,
+      "loss": 0.473,
+      "step": 2381
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4264018524998352,
+      "learning_rate": 1.94075589347898e-05,
+      "loss": 0.3437,
+      "step": 2382
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34952776740085484,
+      "learning_rate": 1.940692776514355e-05,
+      "loss": 0.1839,
+      "step": 2383
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3725548545644662,
+      "learning_rate": 1.9406296269735792e-05,
+      "loss": 0.2575,
+      "step": 2384
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5378553612475551,
+      "learning_rate": 1.94056644485884e-05,
+      "loss": 0.3269,
+      "step": 2385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.312515301867269,
+      "learning_rate": 1.940503230172325e-05,
+      "loss": 0.8482,
+      "step": 2386
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6711137585190222,
+      "learning_rate": 1.940439982916224e-05,
+      "loss": 0.3149,
+      "step": 2387
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39804796254758124,
+      "learning_rate": 1.9403767030927265e-05,
+      "loss": 0.3114,
+      "step": 2388
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49038512112971977,
+      "learning_rate": 1.9403133907040245e-05,
+      "loss": 0.3422,
+      "step": 2389
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2498696865928933,
+      "learning_rate": 1.94025004575231e-05,
+      "loss": 0.1121,
+      "step": 2390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4388817507166522,
+      "learning_rate": 1.940186668239777e-05,
+      "loss": 0.3111,
+      "step": 2391
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5706651361437006,
+      "learning_rate": 1.94012325816862e-05,
+      "loss": 0.3399,
+      "step": 2392
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1738840800470578,
+      "learning_rate": 1.9400598155410352e-05,
+      "loss": 0.4457,
+      "step": 2393
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4022071782421249,
+      "learning_rate": 1.939996340359219e-05,
+      "loss": 0.2773,
+      "step": 2394
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3049991474196785,
+      "learning_rate": 1.9399328326253702e-05,
+      "loss": 0.2383,
+      "step": 2395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5410903011939309,
+      "learning_rate": 1.9398692923416877e-05,
+      "loss": 0.3912,
+      "step": 2396
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38073227562966827,
+      "learning_rate": 1.9398057195103725e-05,
+      "loss": 0.2435,
+      "step": 2397
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3803216316707985,
+      "learning_rate": 1.9397421141336252e-05,
+      "loss": 0.6396,
+      "step": 2398
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5643971893815023,
+      "learning_rate": 1.9396784762136488e-05,
+      "loss": 0.3818,
+      "step": 2399
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31425978785959,
+      "learning_rate": 1.9396148057526473e-05,
+      "loss": 0.2359,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.311778884841662,
+      "learning_rate": 1.9395511027528257e-05,
+      "loss": 0.653,
+      "step": 2401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3134910613406851,
+      "learning_rate": 1.9394873672163896e-05,
+      "loss": 0.1867,
+      "step": 2402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45026996141169107,
+      "learning_rate": 1.9394235991455464e-05,
+      "loss": 0.2147,
+      "step": 2403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6240035351919085,
+      "learning_rate": 1.9393597985425044e-05,
+      "loss": 0.3687,
+      "step": 2404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0016231929095378,
+      "learning_rate": 1.939295965409473e-05,
+      "loss": 0.4781,
+      "step": 2405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39450446781921794,
+      "learning_rate": 1.9392320997486624e-05,
+      "loss": 0.234,
+      "step": 2406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6254131653942641,
+      "learning_rate": 1.939168201562285e-05,
+      "loss": 0.3663,
+      "step": 2407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3195029509674968,
+      "learning_rate": 1.939104270852553e-05,
+      "loss": 0.2335,
+      "step": 2408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46483875879314784,
+      "learning_rate": 1.9390403076216805e-05,
+      "loss": 0.3006,
+      "step": 2409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1589225599424164,
+      "learning_rate": 1.9389763118718824e-05,
+      "loss": 0.4651,
+      "step": 2410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6169719370131748,
+      "learning_rate": 1.938912283605375e-05,
+      "loss": 0.3751,
+      "step": 2411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3554253129647085,
+      "learning_rate": 1.9388482228243758e-05,
+      "loss": 0.3024,
+      "step": 2412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4486534251987144,
+      "learning_rate": 1.938784129531103e-05,
+      "loss": 0.2002,
+      "step": 2413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47231607456050406,
+      "learning_rate": 1.938720003727776e-05,
+      "loss": 0.2697,
+      "step": 2414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4328718171018966,
+      "learning_rate": 1.9386558454166158e-05,
+      "loss": 0.3028,
+      "step": 2415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9168616970784886,
+      "learning_rate": 1.938591654599844e-05,
+      "loss": 0.324,
+      "step": 2416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0797620646008737,
+      "learning_rate": 1.9385274312796834e-05,
+      "loss": 0.6565,
+      "step": 2417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.400563368589376,
+      "learning_rate": 1.9384631754583586e-05,
+      "loss": 0.2876,
+      "step": 2418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7285786800582772,
+      "learning_rate": 1.938398887138094e-05,
+      "loss": 0.2817,
+      "step": 2419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2902928252182725,
+      "learning_rate": 1.9383345663211162e-05,
+      "loss": 0.2576,
+      "step": 2420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3972358838624784,
+      "learning_rate": 1.938270213009653e-05,
+      "loss": 0.2703,
+      "step": 2421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8769939771723178,
+      "learning_rate": 1.9382058272059326e-05,
+      "loss": 0.5173,
+      "step": 2422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5224940687326848,
+      "learning_rate": 1.9381414089121848e-05,
+      "loss": 0.3176,
+      "step": 2423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4746503525476009,
+      "learning_rate": 1.9380769581306404e-05,
+      "loss": 0.2982,
+      "step": 2424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5547203830668347,
+      "learning_rate": 1.9380124748635312e-05,
+      "loss": 0.4058,
+      "step": 2425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8275991864380573,
+      "learning_rate": 1.9379479591130903e-05,
+      "loss": 0.197,
+      "step": 2426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.395529200272553,
+      "learning_rate": 1.937883410881552e-05,
+      "loss": 0.2868,
+      "step": 2427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5053460012049817,
+      "learning_rate": 1.937818830171151e-05,
+      "loss": 0.3728,
+      "step": 2428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0284565975396147,
+      "learning_rate": 1.937754216984125e-05,
+      "loss": 0.5913,
+      "step": 2429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41865327079166226,
+      "learning_rate": 1.9376895713227106e-05,
+      "loss": 0.2927,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5272358314902269,
+      "learning_rate": 1.9376248931891463e-05,
+      "loss": 0.3642,
+      "step": 2431
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7547226733060595,
+      "learning_rate": 1.9375601825856724e-05,
+      "loss": 0.4149,
+      "step": 2432
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37814727594751524,
+      "learning_rate": 1.93749543951453e-05,
+      "loss": 0.2974,
+      "step": 2433
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4377703315425029,
+      "learning_rate": 1.9374306639779606e-05,
+      "loss": 0.2508,
+      "step": 2434
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46933076479616304,
+      "learning_rate": 1.9373658559782075e-05,
+      "loss": 0.3411,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38251871677416266,
+      "learning_rate": 1.9373010155175155e-05,
+      "loss": 0.2416,
+      "step": 2436
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4471766073972423,
+      "learning_rate": 1.9372361425981293e-05,
+      "loss": 0.7063,
+      "step": 2437
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7185286092475877,
+      "learning_rate": 1.937171237222296e-05,
+      "loss": 0.4922,
+      "step": 2438
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3228268085409524,
+      "learning_rate": 1.9371062993922627e-05,
+      "loss": 0.238,
+      "step": 2439
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3947789797377527,
+      "learning_rate": 1.937041329110279e-05,
+      "loss": 0.2659,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4828332809183167,
+      "learning_rate": 1.9369763263785942e-05,
+      "loss": 0.2961,
+      "step": 2441
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42032450024642715,
+      "learning_rate": 1.936911291199459e-05,
+      "loss": 0.1853,
+      "step": 2442
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42092663555747223,
+      "learning_rate": 1.9368462235751267e-05,
+      "loss": 0.3511,
+      "step": 2443
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8694030752192309,
+      "learning_rate": 1.93678112350785e-05,
+      "loss": 0.5347,
+      "step": 2444
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3919234530205258,
+      "learning_rate": 1.936715990999883e-05,
+      "loss": 0.1909,
+      "step": 2445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.43121382968926086,
+      "learning_rate": 1.9366508260534816e-05,
+      "loss": 0.2989,
+      "step": 2446
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41099307028572735,
+      "learning_rate": 1.936585628670902e-05,
+      "loss": 0.3394,
+      "step": 2447
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3230654577529209,
+      "learning_rate": 1.936520398854403e-05,
+      "loss": 0.2249,
+      "step": 2448
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5695656142368685,
+      "learning_rate": 1.9364551366062426e-05,
+      "loss": 0.3446,
+      "step": 2449
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7429372957735979,
+      "learning_rate": 1.9363898419286812e-05,
+      "loss": 0.5617,
+      "step": 2450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5023685924135727,
+      "learning_rate": 1.9363245148239796e-05,
+      "loss": 0.3299,
+      "step": 2451
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45158885365379514,
+      "learning_rate": 1.9362591552944004e-05,
+      "loss": 0.2682,
+      "step": 2452
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4618495334104154,
+      "learning_rate": 1.9361937633422066e-05,
+      "loss": 0.3204,
+      "step": 2453
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28793626450866544,
+      "learning_rate": 1.9361283389696637e-05,
+      "loss": 0.2239,
+      "step": 2454
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4599655334941602,
+      "learning_rate": 1.9360628821790365e-05,
+      "loss": 0.3066,
+      "step": 2455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1546352730802452,
+      "learning_rate": 1.9359973929725915e-05,
+      "loss": 0.5382,
+      "step": 2456
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41592337368392673,
+      "learning_rate": 1.9359318713525974e-05,
+      "loss": 0.303,
+      "step": 2457
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5380319613336388,
+      "learning_rate": 1.9358663173213227e-05,
+      "loss": 0.8324,
+      "step": 2458
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2911517817560607,
+      "learning_rate": 1.9358007308810377e-05,
+      "loss": 0.23,
+      "step": 2459
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32365945054881184,
+      "learning_rate": 1.9357351120340137e-05,
+      "loss": 0.2067,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5107713270649263,
+      "learning_rate": 1.935669460782523e-05,
+      "loss": 0.3869,
+      "step": 2461
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5950877636646806,
+      "learning_rate": 1.935603777128839e-05,
+      "loss": 0.3651,
+      "step": 2462
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4329840077090914,
+      "learning_rate": 1.9355380610752364e-05,
+      "loss": 0.2897,
+      "step": 2463
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5263592276254527,
+      "learning_rate": 1.9354723126239913e-05,
+      "loss": 0.368,
+      "step": 2464
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8971518502199158,
+      "learning_rate": 1.93540653177738e-05,
+      "loss": 0.5344,
+      "step": 2465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3520567138494093,
+      "learning_rate": 1.9353407185376805e-05,
+      "loss": 0.2685,
+      "step": 2466
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4554545967398719,
+      "learning_rate": 1.9352748729071727e-05,
+      "loss": 0.3259,
+      "step": 2467
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4117401158899818,
+      "learning_rate": 1.9352089948881358e-05,
+      "loss": 0.2224,
+      "step": 2468
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4564482028613299,
+      "learning_rate": 1.935143084482852e-05,
+      "loss": 0.2908,
+      "step": 2469
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6895418863474747,
+      "learning_rate": 1.935077141693603e-05,
+      "loss": 0.447,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5274251237370065,
+      "learning_rate": 1.935011166522673e-05,
+      "loss": 0.3838,
+      "step": 2471
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4138804782363925,
+      "learning_rate": 1.9349451589723465e-05,
+      "loss": 0.2777,
+      "step": 2472
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5436885501881874,
+      "learning_rate": 1.9348791190449092e-05,
+      "loss": 0.4258,
+      "step": 2473
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3883095750427234,
+      "learning_rate": 1.9348130467426486e-05,
+      "loss": 0.1774,
+      "step": 2474
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3857708543855049,
+      "learning_rate": 1.934746942067852e-05,
+      "loss": 0.2547,
+      "step": 2475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5736300598247335,
+      "learning_rate": 1.9346808050228093e-05,
+      "loss": 0.4197,
+      "step": 2476
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7074214232035989,
+      "learning_rate": 1.9346146356098103e-05,
+      "loss": 0.5278,
+      "step": 2477
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.442403841592901,
+      "learning_rate": 1.9345484338311467e-05,
+      "loss": 0.2977,
+      "step": 2478
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4645088493036005,
+      "learning_rate": 1.9344821996891106e-05,
+      "loss": 0.3155,
+      "step": 2479
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4166778230104967,
+      "learning_rate": 1.9344159331859965e-05,
+      "loss": 0.1801,
+      "step": 2480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4899829826259577,
+      "learning_rate": 1.9343496343240994e-05,
+      "loss": 0.1517,
+      "step": 2481
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5196827332828534,
+      "learning_rate": 1.9342833031057138e-05,
+      "loss": 0.3627,
+      "step": 2482
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46032261971842325,
+      "learning_rate": 1.934216939533138e-05,
+      "loss": 0.367,
+      "step": 2483
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8366442088496777,
+      "learning_rate": 1.9341505436086695e-05,
+      "loss": 0.441,
+      "step": 2484
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48325853141372255,
+      "learning_rate": 1.934084115334608e-05,
+      "loss": 0.263,
+      "step": 2485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38690498341366925,
+      "learning_rate": 1.9340176547132536e-05,
+      "loss": 0.2036,
+      "step": 2486
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41703692608099363,
+      "learning_rate": 1.9339511617469082e-05,
+      "loss": 0.3184,
+      "step": 2487
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47705445064380225,
+      "learning_rate": 1.9338846364378742e-05,
+      "loss": 0.2599,
+      "step": 2488
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.186205154115369,
+      "learning_rate": 1.9338180787884554e-05,
+      "loss": 0.5164,
+      "step": 2489
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5865526165668318,
+      "learning_rate": 1.9337514888009566e-05,
+      "loss": 0.4188,
+      "step": 2490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3801379438763642,
+      "learning_rate": 1.933684866477684e-05,
+      "loss": 0.264,
+      "step": 2491
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2829332968836524,
+      "learning_rate": 1.9336182118209442e-05,
+      "loss": 0.192,
+      "step": 2492
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5463688788581329,
+      "learning_rate": 1.9335515248330462e-05,
+      "loss": 0.3456,
+      "step": 2493
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.719184449716863,
+      "learning_rate": 1.933484805516299e-05,
+      "loss": 0.3116,
+      "step": 2494
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8530292958738765,
+      "learning_rate": 1.9334180538730133e-05,
+      "loss": 0.3951,
+      "step": 2495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5021582635497452,
+      "learning_rate": 1.9333512699055004e-05,
+      "loss": 0.7942,
+      "step": 2496
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37553192886357234,
+      "learning_rate": 1.933284453616073e-05,
+      "loss": 0.3011,
+      "step": 2497
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27112912159156816,
+      "learning_rate": 1.9332176050070454e-05,
+      "loss": 0.2007,
+      "step": 2498
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.763728328487966,
+      "learning_rate": 1.9331507240807322e-05,
+      "loss": 0.4586,
+      "step": 2499
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6138837661580483,
+      "learning_rate": 1.9330838108394497e-05,
+      "loss": 0.327,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.051207841064682,
+      "learning_rate": 1.933016865285515e-05,
+      "loss": 0.438,
+      "step": 2501
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5530078844473687,
+      "learning_rate": 1.932949887421246e-05,
+      "loss": 0.3198,
+      "step": 2502
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4045931786615891,
+      "learning_rate": 1.9328828772489623e-05,
+      "loss": 0.293,
+      "step": 2503
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3240936130174602,
+      "learning_rate": 1.932815834770985e-05,
+      "loss": 0.1162,
+      "step": 2504
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37395828600806214,
+      "learning_rate": 1.9327487599896355e-05,
+      "loss": 0.2758,
+      "step": 2505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4175861001760937,
+      "learning_rate": 1.9326816529072364e-05,
+      "loss": 0.3197,
+      "step": 2506
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7256771592836744,
+      "learning_rate": 1.932614513526112e-05,
+      "loss": 0.3772,
+      "step": 2507
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6586431832270313,
+      "learning_rate": 1.932547341848587e-05,
+      "loss": 0.4013,
+      "step": 2508
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5348147230679424,
+      "learning_rate": 1.932480137876988e-05,
+      "loss": 0.3208,
+      "step": 2509
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39809772127850734,
+      "learning_rate": 1.9324129016136414e-05,
+      "loss": 0.265,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29378693207938744,
+      "learning_rate": 1.9323456330608762e-05,
+      "loss": 0.2083,
+      "step": 2511
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41811342753632974,
+      "learning_rate": 1.932278332221022e-05,
+      "loss": 0.3463,
+      "step": 2512
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7113376176543778,
+      "learning_rate": 1.932210999096409e-05,
+      "loss": 0.5525,
+      "step": 2513
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4728687902245757,
+      "learning_rate": 1.9321436336893693e-05,
+      "loss": 0.2821,
+      "step": 2514
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3785533771206013,
+      "learning_rate": 1.9320762360022354e-05,
+      "loss": 0.2939,
+      "step": 2515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4423604779271077,
+      "learning_rate": 1.932008806037342e-05,
+      "loss": 0.2852,
+      "step": 2516
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34758961266810223,
+      "learning_rate": 1.9319413437970236e-05,
+      "loss": 0.2658,
+      "step": 2517
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40220948422232616,
+      "learning_rate": 1.9318738492836165e-05,
+      "loss": 0.2915,
+      "step": 2518
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515572698759552,
+      "learning_rate": 1.931806322499458e-05,
+      "loss": 0.4523,
+      "step": 2519
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5901676572616871,
+      "learning_rate": 1.9317387634468867e-05,
+      "loss": 0.2373,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42464998973281115,
+      "learning_rate": 1.9316711721282423e-05,
+      "loss": 0.3158,
+      "step": 2521
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5342049535459824,
+      "learning_rate": 1.931603548545865e-05,
+      "loss": 0.3712,
+      "step": 2522
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3737853946547916,
+      "learning_rate": 1.9315358927020965e-05,
+      "loss": 0.3404,
+      "step": 2523
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28710083849023404,
+      "learning_rate": 1.931468204599281e-05,
+      "loss": 0.2055,
+      "step": 2524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.221810801453039,
+      "learning_rate": 1.931400484239761e-05,
+      "loss": 0.8419,
+      "step": 2525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4907234180429648,
+      "learning_rate": 1.931332731625882e-05,
+      "loss": 0.3614,
+      "step": 2526
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37691010597331226,
+      "learning_rate": 1.931264946759991e-05,
+      "loss": 0.2088,
+      "step": 2527
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7873124730767418,
+      "learning_rate": 1.931197129644435e-05,
+      "loss": 0.5543,
+      "step": 2528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5526920490936016,
+      "learning_rate": 1.9311292802815626e-05,
+      "loss": 0.4587,
+      "step": 2529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28601007260420575,
+      "learning_rate": 1.931061398673723e-05,
+      "loss": 0.1858,
+      "step": 2530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4887404163838093,
+      "learning_rate": 1.930993484823267e-05,
+      "loss": 0.3604,
+      "step": 2531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5338039650083546,
+      "learning_rate": 1.9309255387325468e-05,
+      "loss": 0.3452,
+      "step": 2532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.34880351685054517,
+      "learning_rate": 1.9308575604039156e-05,
+      "loss": 0.2113,
+      "step": 2533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50223678765854,
+      "learning_rate": 1.9307895498397268e-05,
+      "loss": 0.3496,
+      "step": 2534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0383332882780774,
+      "learning_rate": 1.9307215070423354e-05,
+      "loss": 0.5223,
+      "step": 2535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40725782340222744,
+      "learning_rate": 1.930653432014099e-05,
+      "loss": 0.3043,
+      "step": 2536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0892283717210414,
+      "learning_rate": 1.930585324757374e-05,
+      "loss": 0.5579,
+      "step": 2537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3285256564934091,
+      "learning_rate": 1.9305171852745192e-05,
+      "loss": 0.2542,
+      "step": 2538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3941296085844982,
+      "learning_rate": 1.9304490135678944e-05,
+      "loss": 0.2776,
+      "step": 2539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8550640205170302,
+      "learning_rate": 1.9303808096398602e-05,
+      "loss": 0.4352,
+      "step": 2540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7635534341325637,
+      "learning_rate": 1.9303125734927784e-05,
+      "loss": 0.5426,
+      "step": 2541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3402636225143411,
+      "learning_rate": 1.9302443051290126e-05,
+      "loss": 0.2776,
+      "step": 2542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6774987084654007,
+      "learning_rate": 1.930176004550926e-05,
+      "loss": 0.3408,
+      "step": 2543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33787908931664956,
+      "learning_rate": 1.930107671760885e-05,
+      "loss": 0.2045,
+      "step": 2544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47972956125767385,
+      "learning_rate": 1.9300393067612545e-05,
+      "loss": 0.2875,
+      "step": 2545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4447447023685629,
+      "learning_rate": 1.9299709095544035e-05,
+      "loss": 0.3332,
+      "step": 2546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9395633232005106,
+      "learning_rate": 1.9299024801426994e-05,
+      "loss": 0.4455,
+      "step": 2547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4374552033580974,
+      "learning_rate": 1.9298340185285127e-05,
+      "loss": 0.2683,
+      "step": 2548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6582182316770087,
+      "learning_rate": 1.929765524714214e-05,
+      "loss": 0.4531,
+      "step": 2549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3174736595811298,
+      "learning_rate": 1.929696998702175e-05,
+      "loss": 0.1964,
+      "step": 2550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40811572548117436,
+      "learning_rate": 1.929628440494769e-05,
+      "loss": 0.2882,
+      "step": 2551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8862066628895517,
+      "learning_rate": 1.92955985009437e-05,
+      "loss": 0.4916,
+      "step": 2552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0030933658275236,
+      "learning_rate": 1.9294912275033533e-05,
+      "loss": 0.3916,
+      "step": 2553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4245887044072022,
+      "learning_rate": 1.9294225727240958e-05,
+      "loss": 0.2804,
+      "step": 2554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5604093210603397,
+      "learning_rate": 1.9293538857589737e-05,
+      "loss": 0.4478,
+      "step": 2555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.22753638244231286,
+      "learning_rate": 1.9292851666103673e-05,
+      "loss": 0.1424,
+      "step": 2556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.393215750130867,
+      "learning_rate": 1.9292164152806553e-05,
+      "loss": 0.321,
+      "step": 2557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42867114089154623,
+      "learning_rate": 1.9291476317722185e-05,
+      "loss": 0.3651,
+      "step": 2558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0719463340093578,
+      "learning_rate": 1.9290788160874395e-05,
+      "loss": 0.5735,
+      "step": 2559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42211388783214765,
+      "learning_rate": 1.9290099682287012e-05,
+      "loss": 0.2557,
+      "step": 2560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.346065145365312,
+      "learning_rate": 1.928941088198387e-05,
+      "loss": 0.877,
+      "step": 2561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4098509316745687,
+      "learning_rate": 1.9288721759988834e-05,
+      "loss": 0.372,
+      "step": 2562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42484258362884286,
+      "learning_rate": 1.9288032316325762e-05,
+      "loss": 0.2418,
+      "step": 2563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35294191462631713,
+      "learning_rate": 1.928734255101853e-05,
+      "loss": 0.2576,
+      "step": 2564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5300039653477053,
+      "learning_rate": 1.9286652464091024e-05,
+      "loss": 0.4365,
+      "step": 2565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36024518354461466,
+      "learning_rate": 1.9285962055567144e-05,
+      "loss": 0.2375,
+      "step": 2566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.676422785771166,
+      "learning_rate": 1.9285271325470794e-05,
+      "loss": 0.4694,
+      "step": 2567
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8126513636201301,
+      "learning_rate": 1.92845802738259e-05,
+      "loss": 0.5042,
+      "step": 2568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4114429506728939,
+      "learning_rate": 1.928388890065639e-05,
+      "loss": 0.2532,
+      "step": 2569
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3213342473563277,
+      "learning_rate": 1.928319720598621e-05,
+      "loss": 0.27,
+      "step": 2570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40903951950701906,
+      "learning_rate": 1.9282505189839305e-05,
+      "loss": 0.2594,
+      "step": 2571
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44270353797659523,
+      "learning_rate": 1.9281812852239647e-05,
+      "loss": 0.3207,
+      "step": 2572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43969504549066846,
+      "learning_rate": 1.928112019321121e-05,
+      "loss": 0.297,
+      "step": 2573
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6514459661898435,
+      "learning_rate": 1.9280427212777976e-05,
+      "loss": 0.4354,
+      "step": 2574
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5278709615714426,
+      "learning_rate": 1.927973391096395e-05,
+      "loss": 0.334,
+      "step": 2575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3225729087069408,
+      "learning_rate": 1.9279040287793138e-05,
+      "loss": 0.1843,
+      "step": 2576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38192438331171524,
+      "learning_rate": 1.9278346343289555e-05,
+      "loss": 0.285,
+      "step": 2577
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4238320027450154,
+      "learning_rate": 1.9277652077477242e-05,
+      "loss": 0.3222,
+      "step": 2578
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5692015716239939,
+      "learning_rate": 1.9276957490380236e-05,
+      "loss": 0.3179,
+      "step": 2579
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6989023862803406,
+      "learning_rate": 1.927626258202259e-05,
+      "loss": 0.4818,
+      "step": 2580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47851777806709617,
+      "learning_rate": 1.927556735242837e-05,
+      "loss": 0.3246,
+      "step": 2581
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37801509190886534,
+      "learning_rate": 1.9274871801621652e-05,
+      "loss": 0.2653,
+      "step": 2582
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31902918198499913,
+      "learning_rate": 1.927417592962652e-05,
+      "loss": 0.2124,
+      "step": 2583
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5711538165034585,
+      "learning_rate": 1.9273479736467077e-05,
+      "loss": 0.3241,
+      "step": 2584
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48424088195332443,
+      "learning_rate": 1.9272783222167424e-05,
+      "loss": 0.3557,
+      "step": 2585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5128003089466692,
+      "learning_rate": 1.9272086386751693e-05,
+      "loss": 0.3375,
+      "step": 2586
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5824582826960971,
+      "learning_rate": 1.927138923024401e-05,
+      "loss": 0.3402,
+      "step": 2587
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3477545206134362,
+      "learning_rate": 1.927069175266851e-05,
+      "loss": 0.2719,
+      "step": 2588
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35480404863309206,
+      "learning_rate": 1.926999395404936e-05,
+      "loss": 0.1912,
+      "step": 2589
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4088392855070457,
+      "learning_rate": 1.9269295834410715e-05,
+      "loss": 0.3091,
+      "step": 2590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4128220609130996,
+      "learning_rate": 1.9268597393776753e-05,
+      "loss": 0.3596,
+      "step": 2591
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6523103375473858,
+      "learning_rate": 1.9267898632171663e-05,
+      "loss": 0.4379,
+      "step": 2592
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42876866445114575,
+      "learning_rate": 1.9267199549619643e-05,
+      "loss": 0.2967,
+      "step": 2593
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4201511676991754,
+      "learning_rate": 1.92665001461449e-05,
+      "loss": 0.3146,
+      "step": 2594
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23291446601377253,
+      "learning_rate": 1.9265800421771655e-05,
+      "loss": 0.1366,
+      "step": 2595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43171545122840194,
+      "learning_rate": 1.926510037652414e-05,
+      "loss": 0.3457,
+      "step": 2596
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5129067503078448,
+      "learning_rate": 1.9264400010426598e-05,
+      "loss": 0.4012,
+      "step": 2597
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41831673800447344,
+      "learning_rate": 1.926369932350328e-05,
+      "loss": 0.3836,
+      "step": 2598
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30934776416133264,
+      "learning_rate": 1.9262998315778453e-05,
+      "loss": 0.1444,
+      "step": 2599
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.46632659562332673,
+      "learning_rate": 1.9262296987276395e-05,
+      "loss": 0.348,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27229649864445243,
+      "learning_rate": 1.9261595338021388e-05,
+      "loss": 0.2331,
+      "step": 2601
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40189885147596294,
+      "learning_rate": 1.9260893368037735e-05,
+      "loss": 0.1689,
+      "step": 2602
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41329355608342555,
+      "learning_rate": 1.9260191077349742e-05,
+      "loss": 0.3672,
+      "step": 2603
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6931020159640395,
+      "learning_rate": 1.925948846598173e-05,
+      "loss": 0.5568,
+      "step": 2604
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3821818839566917,
+      "learning_rate": 1.925878553395803e-05,
+      "loss": 0.2077,
+      "step": 2605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36378068853839796,
+      "learning_rate": 1.9258082281302988e-05,
+      "loss": 0.3239,
+      "step": 2606
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3489085591030622,
+      "learning_rate": 1.925737870804095e-05,
+      "loss": 0.1991,
+      "step": 2607
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4365373846944763,
+      "learning_rate": 1.925667481419629e-05,
+      "loss": 0.3204,
+      "step": 2608
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37061176255326356,
+      "learning_rate": 1.9255970599793376e-05,
+      "loss": 0.3358,
+      "step": 2609
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5448112276722097,
+      "learning_rate": 1.92552660648566e-05,
+      "loss": 0.4135,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7625349028762688,
+      "learning_rate": 1.925456120941036e-05,
+      "loss": 0.4468,
+      "step": 2611
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3891806726343858,
+      "learning_rate": 1.925385603347906e-05,
+      "loss": 0.2244,
+      "step": 2612
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3134500837415368,
+      "learning_rate": 1.9253150537087123e-05,
+      "loss": 0.2471,
+      "step": 2613
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6358650633518437,
+      "learning_rate": 1.9252444720258982e-05,
+      "loss": 0.3955,
+      "step": 2614
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4146219135777843,
+      "learning_rate": 1.925173858301908e-05,
+      "loss": 0.3018,
+      "step": 2615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5250021533099688,
+      "learning_rate": 1.9251032125391867e-05,
+      "loss": 0.3568,
+      "step": 2616
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3748774238993723,
+      "learning_rate": 1.925032534740181e-05,
+      "loss": 0.3043,
+      "step": 2617
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4535619682866479,
+      "learning_rate": 1.9249618249073384e-05,
+      "loss": 0.2748,
+      "step": 2618
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7356712302689734,
+      "learning_rate": 1.9248910830431073e-05,
+      "loss": 0.5015,
+      "step": 2619
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4561318815496416,
+      "learning_rate": 1.924820309149938e-05,
+      "loss": 0.3636,
+      "step": 2620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40170842706925475,
+      "learning_rate": 1.9247495032302812e-05,
+      "loss": 0.317,
+      "step": 2621
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3116553045717942,
+      "learning_rate": 1.924678665286589e-05,
+      "loss": 0.2193,
+      "step": 2622
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0170395298052468,
+      "learning_rate": 1.924607795321314e-05,
+      "loss": 0.4508,
+      "step": 2623
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4078089046880881,
+      "learning_rate": 1.924536893336911e-05,
+      "loss": 0.3254,
+      "step": 2624
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43203476595198353,
+      "learning_rate": 1.924465959335835e-05,
+      "loss": 0.3156,
+      "step": 2625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8382291099953593,
+      "learning_rate": 1.924394993320543e-05,
+      "loss": 0.4348,
+      "step": 2626
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4875964972050568,
+      "learning_rate": 1.9243239952934918e-05,
+      "loss": 0.3129,
+      "step": 2627
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3377201613499536,
+      "learning_rate": 1.9242529652571405e-05,
+      "loss": 0.1133,
+      "step": 2628
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4149281039334817,
+      "learning_rate": 1.9241819032139487e-05,
+      "loss": 0.3227,
+      "step": 2629
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48560499229576615,
+      "learning_rate": 1.9241108091663774e-05,
+      "loss": 0.3227,
+      "step": 2630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3978598633144965,
+      "learning_rate": 1.9240396831168883e-05,
+      "loss": 0.4226,
+      "step": 2631
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5892322349385228,
+      "learning_rate": 1.923968525067945e-05,
+      "loss": 0.3879,
+      "step": 2632
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5121453834736848,
+      "learning_rate": 1.923897335022011e-05,
+      "loss": 0.3623,
+      "step": 2633
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4052392757884597,
+      "learning_rate": 1.9238261129815526e-05,
+      "loss": 0.2655,
+      "step": 2634
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3905272412860965,
+      "learning_rate": 1.9237548589490355e-05,
+      "loss": 0.14,
+      "step": 2635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7394376527945423,
+      "learning_rate": 1.923683572926927e-05,
+      "loss": 0.3439,
+      "step": 2636
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8275157121617628,
+      "learning_rate": 1.9236122549176963e-05,
+      "loss": 0.3755,
+      "step": 2637
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9844646238383457,
+      "learning_rate": 1.923540904923813e-05,
+      "loss": 0.3717,
+      "step": 2638
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3459704137199816,
+      "learning_rate": 1.9234695229477475e-05,
+      "loss": 0.3101,
+      "step": 2639
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31297296064647323,
+      "learning_rate": 1.9233981089919727e-05,
+      "loss": 0.1909,
+      "step": 2640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6016717619270547,
+      "learning_rate": 1.9233266630589607e-05,
+      "loss": 0.2894,
+      "step": 2641
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5144862160701495,
+      "learning_rate": 1.923255185151186e-05,
+      "loss": 0.3252,
+      "step": 2642
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0734016508045525,
+      "learning_rate": 1.923183675271124e-05,
+      "loss": 0.4787,
+      "step": 2643
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6162045563854098,
+      "learning_rate": 1.9231121334212513e-05,
+      "loss": 0.2236,
+      "step": 2644
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45479225577934984,
+      "learning_rate": 1.9230405596040448e-05,
+      "loss": 0.2862,
+      "step": 2645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5405861989390848,
+      "learning_rate": 1.922968953821984e-05,
+      "loss": 0.8193,
+      "step": 2646
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3822968065664251,
+      "learning_rate": 1.9228973160775474e-05,
+      "loss": 0.2538,
+      "step": 2647
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4312183020139809,
+      "learning_rate": 1.9228256463732165e-05,
+      "loss": 0.2563,
+      "step": 2648
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.614927999518781,
+      "learning_rate": 1.9227539447114732e-05,
+      "loss": 0.3677,
+      "step": 2649
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1379887071076775,
+      "learning_rate": 1.9226822110948005e-05,
+      "loss": 0.6745,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37894151707122936,
+      "learning_rate": 1.9226104455256827e-05,
+      "loss": 0.2201,
+      "step": 2651
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4781041246729594,
+      "learning_rate": 1.9225386480066046e-05,
+      "loss": 0.3187,
+      "step": 2652
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4630776461667929,
+      "learning_rate": 1.9224668185400528e-05,
+      "loss": 0.3508,
+      "step": 2653
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38185126415206494,
+      "learning_rate": 1.922394957128515e-05,
+      "loss": 0.2651,
+      "step": 2654
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4365198237742083,
+      "learning_rate": 1.9223230637744792e-05,
+      "loss": 0.2976,
+      "step": 2655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6072299775806103,
+      "learning_rate": 1.9222511384804355e-05,
+      "loss": 0.8611,
+      "step": 2656
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35629856336310844,
+      "learning_rate": 1.9221791812488746e-05,
+      "loss": 0.2504,
+      "step": 2657
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9970693080774518,
+      "learning_rate": 1.9221071920822882e-05,
+      "loss": 0.6513,
+      "step": 2658
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7995130307427668,
+      "learning_rate": 1.922035170983169e-05,
+      "loss": 0.5226,
+      "step": 2659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3808491244278673,
+      "learning_rate": 1.9219631179540125e-05,
+      "loss": 0.3165,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3197238539222895,
+      "learning_rate": 1.921891032997312e-05,
+      "loss": 0.1838,
+      "step": 2661
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.503646870246312,
+      "learning_rate": 1.921818916115565e-05,
+      "loss": 0.794,
+      "step": 2662
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39543891144707394,
+      "learning_rate": 1.9217467673112685e-05,
+      "loss": 0.2965,
+      "step": 2663
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9269117583869534,
+      "learning_rate": 1.921674586586921e-05,
+      "loss": 0.4454,
+      "step": 2664
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4249499526884838,
+      "learning_rate": 1.9216023739450222e-05,
+      "loss": 0.3679,
+      "step": 2665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4217638497070136,
+      "learning_rate": 1.921530129388073e-05,
+      "loss": 0.2539,
+      "step": 2666
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.277953569574096,
+      "learning_rate": 1.921457852918575e-05,
+      "loss": 0.1021,
+      "step": 2667
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44369267599406603,
+      "learning_rate": 1.921385544539031e-05,
+      "loss": 0.3468,
+      "step": 2668
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4842457706418174,
+      "learning_rate": 1.9213132042519453e-05,
+      "loss": 0.2911,
+      "step": 2669
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.836503549357935,
+      "learning_rate": 1.921240832059823e-05,
+      "loss": 0.376,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8464103679581624,
+      "learning_rate": 1.9211684279651703e-05,
+      "loss": 0.5637,
+      "step": 2671
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4795410800772796,
+      "learning_rate": 1.9210959919704946e-05,
+      "loss": 0.2707,
+      "step": 2672
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3464321091289899,
+      "learning_rate": 1.9210235240783044e-05,
+      "loss": 0.2455,
+      "step": 2673
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.184053359160218,
+      "learning_rate": 1.920951024291109e-05,
+      "loss": 0.4233,
+      "step": 2674
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4197906790697942,
+      "learning_rate": 1.9208784926114194e-05,
+      "loss": 0.3018,
+      "step": 2675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9864362176828683,
+      "learning_rate": 1.9208059290417468e-05,
+      "loss": 0.5272,
+      "step": 2676
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42562445797911436,
+      "learning_rate": 1.9207333335846048e-05,
+      "loss": 0.311,
+      "step": 2677
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4449285311849742,
+      "learning_rate": 1.920660706242507e-05,
+      "loss": 0.2907,
+      "step": 2678
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4022883840813125,
+      "learning_rate": 1.9205880470179682e-05,
+      "loss": 0.2319,
+      "step": 2679
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45241222820165394,
+      "learning_rate": 1.9205153559135056e-05,
+      "loss": 0.2982,
+      "step": 2680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3837648287989111,
+      "learning_rate": 1.9204426329316354e-05,
+      "loss": 0.28,
+      "step": 2681
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1774690784929287,
+      "learning_rate": 1.9203698780748765e-05,
+      "loss": 0.5684,
+      "step": 2682
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9706697136795795,
+      "learning_rate": 1.920297091345748e-05,
+      "loss": 0.4298,
+      "step": 2683
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44832358661289956,
+      "learning_rate": 1.9202242727467713e-05,
+      "loss": 0.2841,
+      "step": 2684
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40119273159907204,
+      "learning_rate": 1.9201514222804672e-05,
+      "loss": 0.2904,
+      "step": 2685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5340909714198232,
+      "learning_rate": 1.9200785399493592e-05,
+      "loss": 0.3694,
+      "step": 2686
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6551279965817546,
+      "learning_rate": 1.9200056257559706e-05,
+      "loss": 0.2458,
+      "step": 2687
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2765480970724306,
+      "learning_rate": 1.9199326797028268e-05,
+      "loss": 0.4481,
+      "step": 2688
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4896591885875324,
+      "learning_rate": 1.9198597017924543e-05,
+      "loss": 0.3536,
+      "step": 2689
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3569245432977872,
+      "learning_rate": 1.9197866920273794e-05,
+      "loss": 0.2064,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33094841706483585,
+      "learning_rate": 1.919713650410131e-05,
+      "loss": 0.2783,
+      "step": 2691
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.46907427883634445,
+      "learning_rate": 1.9196405769432385e-05,
+      "loss": 0.3683,
+      "step": 2692
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.540396516824427,
+      "learning_rate": 1.9195674716292326e-05,
+      "loss": 0.2609,
+      "step": 2693
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7820904049432057,
+      "learning_rate": 1.9194943344706444e-05,
+      "loss": 0.423,
+      "step": 2694
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4341198750989717,
+      "learning_rate": 1.919421165470007e-05,
+      "loss": 0.7056,
+      "step": 2695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41243600637594274,
+      "learning_rate": 1.919347964629854e-05,
+      "loss": 0.217,
+      "step": 2696
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29327859156718405,
+      "learning_rate": 1.919274731952721e-05,
+      "loss": 0.2332,
+      "step": 2697
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9189693213979191,
+      "learning_rate": 1.919201467441143e-05,
+      "loss": 0.629,
+      "step": 2698
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46368652568716096,
+      "learning_rate": 1.919128171097658e-05,
+      "loss": 0.3495,
+      "step": 2699
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.594946940268688,
+      "learning_rate": 1.919054842924804e-05,
+      "loss": 0.341,
+      "step": 2700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4819190307628047,
+      "learning_rate": 1.9189814829251204e-05,
+      "loss": 0.3548,
+      "step": 2701
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48033771181806284,
+      "learning_rate": 1.9189080911011474e-05,
+      "loss": 0.3313,
+      "step": 2702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38210522382790274,
+      "learning_rate": 1.9188346674554267e-05,
+      "loss": 0.1974,
+      "step": 2703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36886884191893365,
+      "learning_rate": 1.918761211990501e-05,
+      "loss": 0.2936,
+      "step": 2704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47789844802984893,
+      "learning_rate": 1.918687724708914e-05,
+      "loss": 0.3213,
+      "step": 2705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5360226368299564,
+      "learning_rate": 1.918614205613211e-05,
+      "loss": 0.3334,
+      "step": 2706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.878108287963956,
+      "learning_rate": 1.9185406547059367e-05,
+      "loss": 0.5888,
+      "step": 2707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48964078753235296,
+      "learning_rate": 1.91846707198964e-05,
+      "loss": 0.3056,
+      "step": 2708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4418702371752559,
+      "learning_rate": 1.9183934574668674e-05,
+      "loss": 0.3089,
+      "step": 2709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4469509833893016,
+      "learning_rate": 1.918319811140169e-05,
+      "loss": 0.3066,
+      "step": 2710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5447582634426448,
+      "learning_rate": 1.9182461330120952e-05,
+      "loss": 0.3778,
+      "step": 2711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38486688340331743,
+      "learning_rate": 1.918172423085197e-05,
+      "loss": 0.3022,
+      "step": 2712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3383846109182444,
+      "learning_rate": 1.9180986813620276e-05,
+      "loss": 0.2254,
+      "step": 2713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5701648234097344,
+      "learning_rate": 1.9180249078451406e-05,
+      "loss": 0.3499,
+      "step": 2714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4083474638888638,
+      "learning_rate": 1.9179511025370902e-05,
+      "loss": 0.3619,
+      "step": 2715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.41177540040311006,
+      "learning_rate": 1.9178772654404323e-05,
+      "loss": 0.3057,
+      "step": 2716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3629090355785191,
+      "learning_rate": 1.9178033965577243e-05,
+      "loss": 0.2618,
+      "step": 2717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5828011824693402,
+      "learning_rate": 1.9177294958915246e-05,
+      "loss": 0.4277,
+      "step": 2718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44338515108899634,
+      "learning_rate": 1.9176555634443912e-05,
+      "loss": 0.1719,
+      "step": 2719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3385222291686445,
+      "learning_rate": 1.9175815992188856e-05,
+      "loss": 0.2838,
+      "step": 2720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5432606412549738,
+      "learning_rate": 1.9175076032175685e-05,
+      "loss": 0.3732,
+      "step": 2721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5378990298129169,
+      "learning_rate": 1.9174335754430026e-05,
+      "loss": 0.4871,
+      "step": 2722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.544793784552123,
+      "learning_rate": 1.9173595158977515e-05,
+      "loss": 0.2266,
+      "step": 2723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34300168100501205,
+      "learning_rate": 1.9172854245843796e-05,
+      "loss": 0.3132,
+      "step": 2724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35438645648218475,
+      "learning_rate": 1.917211301505453e-05,
+      "loss": 0.2455,
+      "step": 2725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37041691780405817,
+      "learning_rate": 1.9171371466635385e-05,
+      "loss": 0.1947,
+      "step": 2726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.41229526848301834,
+      "learning_rate": 1.9170629600612044e-05,
+      "loss": 0.3676,
+      "step": 2727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3968209388434555,
+      "learning_rate": 1.916988741701019e-05,
+      "loss": 0.3847,
+      "step": 2728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45513126229457,
+      "learning_rate": 1.9169144915855532e-05,
+      "loss": 0.1219,
+      "step": 2729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4630940115074412,
+      "learning_rate": 1.9168402097173774e-05,
+      "loss": 0.3857,
+      "step": 2730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31886087385947326,
+      "learning_rate": 1.916765896099065e-05,
+      "loss": 0.1737,
+      "step": 2731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38274972925488704,
+      "learning_rate": 1.916691550733189e-05,
+      "loss": 0.2351,
+      "step": 2732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3942346069361589,
+      "learning_rate": 1.9166171736223244e-05,
+      "loss": 0.3809,
+      "step": 2733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7779362864741345,
+      "learning_rate": 1.9165427647690457e-05,
+      "loss": 0.5735,
+      "step": 2734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7486116425794241,
+      "learning_rate": 1.916468324175931e-05,
+      "loss": 0.5144,
+      "step": 2735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4035043380202906,
+      "learning_rate": 1.9163938518455577e-05,
+      "loss": 0.2557,
+      "step": 2736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2841072985025231,
+      "learning_rate": 1.9163193477805042e-05,
+      "loss": 0.2075,
+      "step": 2737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4792294242182115,
+      "learning_rate": 1.9162448119833515e-05,
+      "loss": 0.4132,
+      "step": 2738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.440562055294282,
+      "learning_rate": 1.9161702444566803e-05,
+      "loss": 0.3153,
+      "step": 2739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43331696516541285,
+      "learning_rate": 1.9160956452030728e-05,
+      "loss": 0.3573,
+      "step": 2740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7305226944821447,
+      "learning_rate": 1.9160210142251127e-05,
+      "loss": 0.4819,
+      "step": 2741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39263699843117555,
+      "learning_rate": 1.9159463515253842e-05,
+      "loss": 0.2561,
+      "step": 2742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28498177963286886,
+      "learning_rate": 1.9158716571064728e-05,
+      "loss": 0.2061,
+      "step": 2743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.509629190991603,
+      "learning_rate": 1.9157969309709656e-05,
+      "loss": 0.4165,
+      "step": 2744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.343301559511347,
+      "learning_rate": 1.9157221731214498e-05,
+      "loss": 0.2715,
+      "step": 2745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9309225184678585,
+      "learning_rate": 1.9156473835605146e-05,
+      "loss": 0.6432,
+      "step": 2746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8504928341821835,
+      "learning_rate": 1.9155725622907496e-05,
+      "loss": 0.5317,
+      "step": 2747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3641216283918297,
+      "learning_rate": 1.9154977093147467e-05,
+      "loss": 0.2938,
+      "step": 2748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36303716960804644,
+      "learning_rate": 1.915422824635097e-05,
+      "loss": 0.148,
+      "step": 2749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6009779568927565,
+      "learning_rate": 1.9153479082543945e-05,
+      "loss": 0.4414,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3125761277927421,
+      "learning_rate": 1.9152729601752334e-05,
+      "loss": 0.2688,
+      "step": 2751
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4369004478121061,
+      "learning_rate": 1.9151979804002086e-05,
+      "loss": 0.2762,
+      "step": 2752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8335443101192821,
+      "learning_rate": 1.9151229689319177e-05,
+      "loss": 0.5237,
+      "step": 2753
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.41561726711084734,
+      "learning_rate": 1.9150479257729576e-05,
+      "loss": 0.3092,
+      "step": 2754
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9114026808149253,
+      "learning_rate": 1.9149728509259268e-05,
+      "loss": 0.3958,
+      "step": 2755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37242842792677777,
+      "learning_rate": 1.9148977443934257e-05,
+      "loss": 0.313,
+      "step": 2756
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45442449929965356,
+      "learning_rate": 1.914822606178055e-05,
+      "loss": 0.328,
+      "step": 2757
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29522579252993075,
+      "learning_rate": 1.914747436282417e-05,
+      "loss": 0.1609,
+      "step": 2758
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5405467495766618,
+      "learning_rate": 1.9146722347091145e-05,
+      "loss": 0.385,
+      "step": 2759
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3952204533027063,
+      "learning_rate": 1.9145970014607517e-05,
+      "loss": 0.297,
+      "step": 2760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4252943635395237,
+      "learning_rate": 1.914521736539934e-05,
+      "loss": 0.5132,
+      "step": 2761
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42261706849570274,
+      "learning_rate": 1.9144464399492682e-05,
+      "loss": 0.2417,
+      "step": 2762
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3112984251374848,
+      "learning_rate": 1.9143711116913614e-05,
+      "loss": 0.2201,
+      "step": 2763
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5503114825951091,
+      "learning_rate": 1.9142957517688226e-05,
+      "loss": 0.3508,
+      "step": 2764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.290525038080281,
+      "learning_rate": 1.9142203601842607e-05,
+      "loss": 0.626,
+      "step": 2765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4340454110617708,
+      "learning_rate": 1.9141449369402873e-05,
+      "loss": 0.305,
+      "step": 2766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0459985522737103,
+      "learning_rate": 1.914069482039514e-05,
+      "loss": 0.5296,
+      "step": 2767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4969749645117052,
+      "learning_rate": 1.913993995484554e-05,
+      "loss": 0.2863,
+      "step": 2768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40985606509878225,
+      "learning_rate": 1.913918477278021e-05,
+      "loss": 0.2643,
+      "step": 2769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6034420793845431,
+      "learning_rate": 1.9138429274225306e-05,
+      "loss": 0.3003,
+      "step": 2770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9476623773086797,
+      "learning_rate": 1.913767345920699e-05,
+      "loss": 0.2913,
+      "step": 2771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7160270392478728,
+      "learning_rate": 1.9136917327751433e-05,
+      "loss": 0.3226,
+      "step": 2772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8619252386421343,
+      "learning_rate": 1.913616087988482e-05,
+      "loss": 0.4995,
+      "step": 2773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1457060818764169,
+      "learning_rate": 1.9135404115633354e-05,
+      "loss": 0.6252,
+      "step": 2774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4366886822707938,
+      "learning_rate": 1.9134647035023233e-05,
+      "loss": 0.2177,
+      "step": 2775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37860818327377915,
+      "learning_rate": 1.913388963808068e-05,
+      "loss": 0.2423,
+      "step": 2776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9684838256494043,
+      "learning_rate": 1.9133131924831917e-05,
+      "loss": 0.4987,
+      "step": 2777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8367814228740066,
+      "learning_rate": 1.9132373895303193e-05,
+      "loss": 0.2736,
+      "step": 2778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2098389822749138,
+      "learning_rate": 1.9131615549520752e-05,
+      "loss": 0.4306,
+      "step": 2779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8197725081398024,
+      "learning_rate": 1.913085688751086e-05,
+      "loss": 0.3871,
+      "step": 2780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49208863169300127,
+      "learning_rate": 1.913009790929978e-05,
+      "loss": 0.2306,
+      "step": 2781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3526381294259932,
+      "learning_rate": 1.9129338614913808e-05,
+      "loss": 0.2522,
+      "step": 2782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4756058297463948,
+      "learning_rate": 1.912857900437923e-05,
+      "loss": 0.326,
+      "step": 2783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45837754491678995,
+      "learning_rate": 1.9127819077722353e-05,
+      "loss": 0.3238,
+      "step": 2784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4324033883404848,
+      "learning_rate": 1.9127058834969494e-05,
+      "loss": 0.4619,
+      "step": 2785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0480337876957362,
+      "learning_rate": 1.9126298276146982e-05,
+      "loss": 0.5592,
+      "step": 2786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.41564243767044307,
+      "learning_rate": 1.912553740128115e-05,
+      "loss": 0.2959,
+      "step": 2787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2892471983222418,
+      "learning_rate": 1.9124776210398354e-05,
+      "loss": 0.1853,
+      "step": 2788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7875199409762493,
+      "learning_rate": 1.9124014703524946e-05,
+      "loss": 0.5366,
+      "step": 2789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4263502403630036,
+      "learning_rate": 1.9123252880687303e-05,
+      "loss": 0.3377,
+      "step": 2790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5112805395238533,
+      "learning_rate": 1.9122490741911806e-05,
+      "loss": 0.2663,
+      "step": 2791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.525359587134657,
+      "learning_rate": 1.9121728287224844e-05,
+      "loss": 0.4025,
+      "step": 2792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4425050759597999,
+      "learning_rate": 1.9120965516652828e-05,
+      "loss": 0.2785,
+      "step": 2793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2788256255576857,
+      "learning_rate": 1.912020243022217e-05,
+      "loss": 0.1893,
+      "step": 2794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47566417935386807,
+      "learning_rate": 1.911943902795929e-05,
+      "loss": 0.3708,
+      "step": 2795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47443133106229063,
+      "learning_rate": 1.9118675309890628e-05,
+      "loss": 0.3175,
+      "step": 2796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5872484241415183,
+      "learning_rate": 1.911791127604263e-05,
+      "loss": 0.4337,
+      "step": 2797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9705192208886395,
+      "learning_rate": 1.9117146926441757e-05,
+      "loss": 0.3478,
+      "step": 2798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3903267981435643,
+      "learning_rate": 1.9116382261114484e-05,
+      "loss": 0.2852,
+      "step": 2799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3265402016275641,
+      "learning_rate": 1.911561728008728e-05,
+      "loss": 0.2788,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7350955291555213,
+      "learning_rate": 1.9114851983386646e-05,
+      "loss": 0.4241,
+      "step": 2801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4414743371523098,
+      "learning_rate": 1.9114086371039078e-05,
+      "loss": 0.3136,
+      "step": 2802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4236622091499315,
+      "learning_rate": 1.911332044307109e-05,
+      "loss": 0.3239,
+      "step": 2803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5945493542557609,
+      "learning_rate": 1.9112554199509207e-05,
+      "loss": 0.3062,
+      "step": 2804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4080833519256838,
+      "learning_rate": 1.911178764037996e-05,
+      "loss": 0.2928,
+      "step": 2805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.803133176210868,
+      "learning_rate": 1.9111020765709905e-05,
+      "loss": 0.5323,
+      "step": 2806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4639683487323455,
+      "learning_rate": 1.9110253575525593e-05,
+      "loss": 0.3419,
+      "step": 2807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48221297823220266,
+      "learning_rate": 1.910948606985359e-05,
+      "loss": 0.3114,
+      "step": 2808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4188949921186524,
+      "learning_rate": 1.9108718248720472e-05,
+      "loss": 0.2776,
+      "step": 2809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4180698257150617,
+      "learning_rate": 1.9107950112152838e-05,
+      "loss": 0.2458,
+      "step": 2810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33142474909538355,
+      "learning_rate": 1.910718166017728e-05,
+      "loss": 0.2246,
+      "step": 2811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3976551718650196,
+      "learning_rate": 1.910641289282041e-05,
+      "loss": 0.3871,
+      "step": 2812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.754813347285477,
+      "learning_rate": 1.910564381010886e-05,
+      "loss": 0.5863,
+      "step": 2813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4022736855798634,
+      "learning_rate": 1.9104874412069253e-05,
+      "loss": 0.2171,
+      "step": 2814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32538048947625997,
+      "learning_rate": 1.9104104698728235e-05,
+      "loss": 0.2844,
+      "step": 2815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4742319547411235,
+      "learning_rate": 1.9103334670112468e-05,
+      "loss": 0.3209,
+      "step": 2816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3539327872330429,
+      "learning_rate": 1.9102564326248608e-05,
+      "loss": 0.1927,
+      "step": 2817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5182586066974086,
+      "learning_rate": 1.910179366716334e-05,
+      "loss": 0.4189,
+      "step": 2818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46291557914978526,
+      "learning_rate": 1.9101022692883348e-05,
+      "loss": 0.3427,
+      "step": 2819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3223453767011489,
+      "learning_rate": 1.910025140343533e-05,
+      "loss": 0.2285,
+      "step": 2820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4671981056244132,
+      "learning_rate": 1.9099479798845997e-05,
+      "loss": 0.2945,
+      "step": 2821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48019689418661415,
+      "learning_rate": 1.9098707879142072e-05,
+      "loss": 0.3136,
+      "step": 2822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4030849015440731,
+      "learning_rate": 1.9097935644350284e-05,
+      "loss": 0.2743,
+      "step": 2823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5656401007116433,
+      "learning_rate": 1.9097163094497374e-05,
+      "loss": 0.349,
+      "step": 2824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6376831416506765,
+      "learning_rate": 1.9096390229610095e-05,
+      "loss": 0.557,
+      "step": 2825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4153453663473684,
+      "learning_rate": 1.9095617049715217e-05,
+      "loss": 0.3185,
+      "step": 2826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34493890499970015,
+      "learning_rate": 1.9094843554839513e-05,
+      "loss": 0.2722,
+      "step": 2827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3024656467389478,
+      "learning_rate": 1.9094069745009766e-05,
+      "loss": 0.2226,
+      "step": 2828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7295724146360978,
+      "learning_rate": 1.9093295620252776e-05,
+      "loss": 0.4354,
+      "step": 2829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3640806422459361,
+      "learning_rate": 1.9092521180595347e-05,
+      "loss": 0.2594,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3788662395591776,
+      "learning_rate": 1.9091746426064303e-05,
+      "loss": 0.3483,
+      "step": 2831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5866543462727462,
+      "learning_rate": 1.9090971356686473e-05,
+      "loss": 0.3439,
+      "step": 2832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2907543494961357,
+      "learning_rate": 1.909019597248869e-05,
+      "loss": 0.2154,
+      "step": 2833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.440487120983383,
+      "learning_rate": 1.9089420273497813e-05,
+      "loss": 0.2879,
+      "step": 2834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3480069803830963,
+      "learning_rate": 1.9088644259740708e-05,
+      "loss": 0.3217,
+      "step": 2835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5148676360714839,
+      "learning_rate": 1.9087867931244238e-05,
+      "loss": 0.364,
+      "step": 2836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9445440875281329,
+      "learning_rate": 1.9087091288035293e-05,
+      "loss": 0.4219,
+      "step": 2837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6342925351854132,
+      "learning_rate": 1.908631433014077e-05,
+      "loss": 0.4129,
+      "step": 2838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32986836933112085,
+      "learning_rate": 1.9085537057587568e-05,
+      "loss": 0.3206,
+      "step": 2839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5347354807326525,
+      "learning_rate": 1.9084759470402612e-05,
+      "loss": 0.247,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4500849622065172,
+      "learning_rate": 1.9083981568612828e-05,
+      "loss": 0.3415,
+      "step": 2841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4047582580502225,
+      "learning_rate": 1.9083203352245148e-05,
+      "loss": 0.255,
+      "step": 2842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4042265707783651,
+      "learning_rate": 1.9082424821326532e-05,
+      "loss": 0.3031,
+      "step": 2843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4400449057120318,
+      "learning_rate": 1.9081645975883928e-05,
+      "loss": 0.325,
+      "step": 2844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5280907990199353,
+      "learning_rate": 1.908086681594432e-05,
+      "loss": 0.3971,
+      "step": 2845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4402381574442051,
+      "learning_rate": 1.908008734153468e-05,
+      "loss": 0.3617,
+      "step": 2846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34365348876661916,
+      "learning_rate": 1.9079307552682013e-05,
+      "loss": 0.2332,
+      "step": 2847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3630634485960283,
+      "learning_rate": 1.907852744941331e-05,
+      "loss": 0.2653,
+      "step": 2848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9110965112820536,
+      "learning_rate": 1.9077747031755594e-05,
+      "loss": 0.6379,
+      "step": 2849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4468249874232147,
+      "learning_rate": 1.9076966299735887e-05,
+      "loss": 0.1208,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3606662643476382,
+      "learning_rate": 1.9076185253381227e-05,
+      "loss": 0.2985,
+      "step": 2851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6445782349707652,
+      "learning_rate": 1.9075403892718664e-05,
+      "loss": 0.533,
+      "step": 2852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3657364361197772,
+      "learning_rate": 1.9074622217775253e-05,
+      "loss": 0.1565,
+      "step": 2853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.352593659165976,
+      "learning_rate": 1.9073840228578068e-05,
+      "loss": 0.2859,
+      "step": 2854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4152158108112304,
+      "learning_rate": 1.9073057925154184e-05,
+      "loss": 0.3108,
+      "step": 2855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5767800434899424,
+      "learning_rate": 1.9072275307530692e-05,
+      "loss": 0.2169,
+      "step": 2856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.406189229853736,
+      "learning_rate": 1.9071492375734698e-05,
+      "loss": 0.3657,
+      "step": 2857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7434134856812202,
+      "learning_rate": 1.9070709129793313e-05,
+      "loss": 0.536,
+      "step": 2858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30162267512857593,
+      "learning_rate": 1.906992556973366e-05,
+      "loss": 0.1912,
+      "step": 2859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.360668147936046,
+      "learning_rate": 1.906914169558288e-05,
+      "loss": 0.2589,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8302107796578109,
+      "learning_rate": 1.9068357507368108e-05,
+      "loss": 0.6726,
+      "step": 2861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43014673007102716,
+      "learning_rate": 1.9067573005116506e-05,
+      "loss": 0.3113,
+      "step": 2862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.351210921130384,
+      "learning_rate": 1.9066788188855237e-05,
+      "loss": 0.2871,
+      "step": 2863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8047851678299188,
+      "learning_rate": 1.906600305861149e-05,
+      "loss": 0.4959,
+      "step": 2864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47323207583686905,
+      "learning_rate": 1.906521761441244e-05,
+      "loss": 0.2408,
+      "step": 2865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3269727310027329,
+      "learning_rate": 1.90644318562853e-05,
+      "loss": 0.1871,
+      "step": 2866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4722268465786524,
+      "learning_rate": 1.9063645784257274e-05,
+      "loss": 0.3434,
+      "step": 2867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8141264243435448,
+      "learning_rate": 1.906285939835558e-05,
+      "loss": 0.5588,
+      "step": 2868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3784784877528985,
+      "learning_rate": 1.9062072698607457e-05,
+      "loss": 0.2705,
+      "step": 2869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6574080394539428,
+      "learning_rate": 1.9061285685040148e-05,
+      "loss": 0.3812,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4694759377072518,
+      "learning_rate": 1.9060498357680905e-05,
+      "loss": 0.3247,
+      "step": 2871
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3590925137545595,
+      "learning_rate": 1.905971071655699e-05,
+      "loss": 0.243,
+      "step": 2872
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.43092525195270276,
+      "learning_rate": 1.9058922761695684e-05,
+      "loss": 0.1675,
+      "step": 2873
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.509367055075115,
+      "learning_rate": 1.9058134493124275e-05,
+      "loss": 0.3637,
+      "step": 2874
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3456206104797771,
+      "learning_rate": 1.9057345910870054e-05,
+      "loss": 0.3076,
+      "step": 2875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0697084543942041,
+      "learning_rate": 1.905655701496034e-05,
+      "loss": 0.3674,
+      "step": 2876
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40182912681545313,
+      "learning_rate": 1.9055767805422438e-05,
+      "loss": 0.2836,
+      "step": 2877
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.473138542860798,
+      "learning_rate": 1.905497828228369e-05,
+      "loss": 0.3122,
+      "step": 2878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3813765291907576,
+      "learning_rate": 1.9054188445571435e-05,
+      "loss": 0.2404,
+      "step": 2879
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7796531744569375,
+      "learning_rate": 1.905339829531302e-05,
+      "loss": 0.448,
+      "step": 2880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42795114627019815,
+      "learning_rate": 1.9052607831535812e-05,
+      "loss": 0.2856,
+      "step": 2881
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4012186112644199,
+      "learning_rate": 1.9051817054267184e-05,
+      "loss": 0.3072,
+      "step": 2882
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.748819417416899,
+      "learning_rate": 1.9051025963534526e-05,
+      "loss": 0.4127,
+      "step": 2883
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41468273696256386,
+      "learning_rate": 1.9050234559365223e-05,
+      "loss": 0.3054,
+      "step": 2884
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5002904393428474,
+      "learning_rate": 1.904944284178669e-05,
+      "loss": 0.3061,
+      "step": 2885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4156710785103392,
+      "learning_rate": 1.9048650810826333e-05,
+      "loss": 0.3083,
+      "step": 2886
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40806478413793273,
+      "learning_rate": 1.9047858466511594e-05,
+      "loss": 0.3049,
+      "step": 2887
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4128113081838106,
+      "learning_rate": 1.9047065808869902e-05,
+      "loss": 0.2899,
+      "step": 2888
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3442654178253383,
+      "learning_rate": 1.9046272837928713e-05,
+      "loss": 0.1134,
+      "step": 2889
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3907830303341359,
+      "learning_rate": 1.9045479553715482e-05,
+      "loss": 0.2846,
+      "step": 2890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5828855253121485,
+      "learning_rate": 1.9044685956257686e-05,
+      "loss": 0.4289,
+      "step": 2891
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.514484579636209,
+      "learning_rate": 1.9043892045582804e-05,
+      "loss": 0.3811,
+      "step": 2892
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3266743076958912,
+      "learning_rate": 1.9043097821718327e-05,
+      "loss": 0.2815,
+      "step": 2893
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5032500743515141,
+      "learning_rate": 1.9042303284691762e-05,
+      "loss": 0.3833,
+      "step": 2894
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4345386460910853,
+      "learning_rate": 1.9041508434530622e-05,
+      "loss": 0.2607,
+      "step": 2895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41878196971218673,
+      "learning_rate": 1.9040713271262438e-05,
+      "loss": 0.2993,
+      "step": 2896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8099840083560745,
+      "learning_rate": 1.9039917794914736e-05,
+      "loss": 0.5032,
+      "step": 2897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42788121734475587,
+      "learning_rate": 1.9039122005515074e-05,
+      "loss": 0.3714,
+      "step": 2898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3654518107930367,
+      "learning_rate": 1.9038325903091003e-05,
+      "loss": 0.218,
+      "step": 2899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37284915313610173,
+      "learning_rate": 1.90375294876701e-05,
+      "loss": 0.2112,
+      "step": 2900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.25242807424076,
+      "learning_rate": 1.9036732759279935e-05,
+      "loss": 0.6335,
+      "step": 2901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3850326090104052,
+      "learning_rate": 1.9035935717948102e-05,
+      "loss": 0.2102,
+      "step": 2902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6094638643634038,
+      "learning_rate": 1.9035138363702206e-05,
+      "loss": 0.392,
+      "step": 2903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9291079961641789,
+      "learning_rate": 1.9034340696569858e-05,
+      "loss": 0.635,
+      "step": 2904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6338307726280604,
+      "learning_rate": 1.9033542716578677e-05,
+      "loss": 0.248,
+      "step": 2905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3460124838063629,
+      "learning_rate": 1.90327444237563e-05,
+      "loss": 0.2289,
+      "step": 2906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.568956704643555,
+      "learning_rate": 1.9031945818130373e-05,
+      "loss": 0.7279,
+      "step": 2907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4521370738515677,
+      "learning_rate": 1.9031146899728555e-05,
+      "loss": 0.2614,
+      "step": 2908
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9225487734393911,
+      "learning_rate": 1.9030347668578506e-05,
+      "loss": 0.5389,
+      "step": 2909
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5007109970727628,
+      "learning_rate": 1.90295481247079e-05,
+      "loss": 0.3537,
+      "step": 2910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3697140213680634,
+      "learning_rate": 1.902874826814444e-05,
+      "loss": 0.2837,
+      "step": 2911
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25441038379523395,
+      "learning_rate": 1.902794809891581e-05,
+      "loss": 0.1146,
+      "step": 2912
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9960926038338246,
+      "learning_rate": 1.9027147617049727e-05,
+      "loss": 0.5207,
+      "step": 2913
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49115998471316646,
+      "learning_rate": 1.9026346822573906e-05,
+      "loss": 0.2785,
+      "step": 2914
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6090023738796487,
+      "learning_rate": 1.902554571551609e-05,
+      "loss": 0.3221,
+      "step": 2915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1039710139860757,
+      "learning_rate": 1.902474429590401e-05,
+      "loss": 0.735,
+      "step": 2916
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49261792202989546,
+      "learning_rate": 1.9023942563765422e-05,
+      "loss": 0.2849,
+      "step": 2917
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28069746124257844,
+      "learning_rate": 1.9023140519128093e-05,
+      "loss": 0.1992,
+      "step": 2918
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3144734644202736,
+      "learning_rate": 1.9022338162019794e-05,
+      "loss": 0.5199,
+      "step": 2919
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5300530829580868,
+      "learning_rate": 1.9021535492468313e-05,
+      "loss": 0.3645,
+      "step": 2920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5599830716860857,
+      "learning_rate": 1.9020732510501445e-05,
+      "loss": 0.3213,
+      "step": 2921
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4318688829311013,
+      "learning_rate": 1.9019929216147002e-05,
+      "loss": 0.3321,
+      "step": 2922
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40200132881097284,
+      "learning_rate": 1.9019125609432793e-05,
+      "loss": 0.2933,
+      "step": 2923
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35823983826684375,
+      "learning_rate": 1.9018321690386656e-05,
+      "loss": 0.2411,
+      "step": 2924
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2091907994532067,
+      "learning_rate": 1.9017517459036426e-05,
+      "loss": 0.4403,
+      "step": 2925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44885633440024797,
+      "learning_rate": 1.9016712915409953e-05,
+      "loss": 0.2987,
+      "step": 2926
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6097786376505031,
+      "learning_rate": 1.90159080595351e-05,
+      "loss": 0.3783,
+      "step": 2927
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4166969282231294,
+      "learning_rate": 1.901510289143974e-05,
+      "loss": 0.5974,
+      "step": 2928
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39087853057358346,
+      "learning_rate": 1.901429741115175e-05,
+      "loss": 0.2789,
+      "step": 2929
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4707325846249934,
+      "learning_rate": 1.901349161869903e-05,
+      "loss": 0.3342,
+      "step": 2930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5044461165624786,
+      "learning_rate": 1.9012685514109487e-05,
+      "loss": 0.2705,
+      "step": 2931
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.446221372568942,
+      "learning_rate": 1.9011879097411028e-05,
+      "loss": 0.2971,
+      "step": 2932
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7000113910533794,
+      "learning_rate": 1.9011072368631586e-05,
+      "loss": 0.4128,
+      "step": 2933
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6515818280851625,
+      "learning_rate": 1.9010265327799092e-05,
+      "loss": 0.3242,
+      "step": 2934
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4527641519593625,
+      "learning_rate": 1.90094579749415e-05,
+      "loss": 0.271,
+      "step": 2935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34571287724653443,
+      "learning_rate": 1.9008650310086768e-05,
+      "loss": 0.3034,
+      "step": 2936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3554877938716127,
+      "learning_rate": 1.900784233326286e-05,
+      "loss": 0.3289,
+      "step": 2937
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4412511425860835,
+      "learning_rate": 1.9007034044497757e-05,
+      "loss": 0.2159,
+      "step": 2938
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7452586967068775,
+      "learning_rate": 1.9006225443819456e-05,
+      "loss": 0.4174,
+      "step": 2939
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.353204527219416,
+      "learning_rate": 1.900541653125595e-05,
+      "loss": 0.8481,
+      "step": 2940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4632062818338658,
+      "learning_rate": 1.9004607306835263e-05,
+      "loss": 0.218,
+      "step": 2941
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5341120108017874,
+      "learning_rate": 1.900379777058541e-05,
+      "loss": 0.3449,
+      "step": 2942
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9366471498287889,
+      "learning_rate": 1.9002987922534427e-05,
+      "loss": 0.5459,
+      "step": 2943
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27741904876939394,
+      "learning_rate": 1.900217776271036e-05,
+      "loss": 0.1587,
+      "step": 2944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3941750294590425,
+      "learning_rate": 1.9001367291141264e-05,
+      "loss": 0.6409,
+      "step": 2945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.48903414699985137,
+      "learning_rate": 1.9000556507855204e-05,
+      "loss": 0.353,
+      "step": 2946
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4693605982996768,
+      "learning_rate": 1.8999745412880264e-05,
+      "loss": 0.3303,
+      "step": 2947
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5832080658668408,
+      "learning_rate": 1.8998934006244522e-05,
+      "loss": 0.3351,
+      "step": 2948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5745640754829965,
+      "learning_rate": 1.8998122287976085e-05,
+      "loss": 0.4208,
+      "step": 2949
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3367288088164626,
+      "learning_rate": 1.899731025810306e-05,
+      "loss": 0.2534,
+      "step": 2950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38374292249387476,
+      "learning_rate": 1.8996497916653565e-05,
+      "loss": 0.1868,
+      "step": 2951
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4710186990718137,
+      "learning_rate": 1.899568526365574e-05,
+      "loss": 0.8454,
+      "step": 2952
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6019236351354754,
+      "learning_rate": 1.8994872299137715e-05,
+      "loss": 0.462,
+      "step": 2953
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3551118807958904,
+      "learning_rate": 1.8994059023127655e-05,
+      "loss": 0.2787,
+      "step": 2954
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4704132793170269,
+      "learning_rate": 1.899324543565371e-05,
+      "loss": 0.3093,
+      "step": 2955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37749498137552984,
+      "learning_rate": 1.899243153674407e-05,
+      "loss": 0.1934,
+      "step": 2956
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4254951358246995,
+      "learning_rate": 1.8991617326426907e-05,
+      "loss": 0.2868,
+      "step": 2957
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6753680434296535,
+      "learning_rate": 1.8990802804730424e-05,
+      "loss": 0.4295,
+      "step": 2958
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7358847969565454,
+      "learning_rate": 1.8989987971682828e-05,
+      "loss": 0.3888,
+      "step": 2959
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40847603026118584,
+      "learning_rate": 1.8989172827312337e-05,
+      "loss": 0.3701,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5456246498843511,
+      "learning_rate": 1.8988357371647173e-05,
+      "loss": 0.2735,
+      "step": 2961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31106052444760784,
+      "learning_rate": 1.8987541604715584e-05,
+      "loss": 0.2288,
+      "step": 2962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36068213668776067,
+      "learning_rate": 1.898672552654581e-05,
+      "loss": 0.2461,
+      "step": 2963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9711275231098498,
+      "learning_rate": 1.8985909137166122e-05,
+      "loss": 0.4776,
+      "step": 2964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3911974616653507,
+      "learning_rate": 1.8985092436604783e-05,
+      "loss": 0.329,
+      "step": 2965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4923858502324332,
+      "learning_rate": 1.8984275424890085e-05,
+      "loss": 0.3115,
+      "step": 2966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5945187050292371,
+      "learning_rate": 1.8983458102050313e-05,
+      "loss": 0.4092,
+      "step": 2967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24507124970481212,
+      "learning_rate": 1.8982640468113774e-05,
+      "loss": 0.197,
+      "step": 2968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5955934559945066,
+      "learning_rate": 1.898182252310878e-05,
+      "loss": 0.4072,
+      "step": 2969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.429200057143055,
+      "learning_rate": 1.8981004267063658e-05,
+      "loss": 0.3242,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5983194080034205,
+      "learning_rate": 1.8980185700006744e-05,
+      "loss": 0.4249,
+      "step": 2971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42932383596148743,
+      "learning_rate": 1.8979366821966386e-05,
+      "loss": 0.3293,
+      "step": 2972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3954252758600323,
+      "learning_rate": 1.8978547632970943e-05,
+      "loss": 0.3225,
+      "step": 2973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30943424037499556,
+      "learning_rate": 1.897772813304878e-05,
+      "loss": 0.0774,
+      "step": 2974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2955268319884107,
+      "learning_rate": 1.8976908322228277e-05,
+      "loss": 0.2291,
+      "step": 2975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8219325687989213,
+      "learning_rate": 1.897608820053783e-05,
+      "loss": 0.5307,
+      "step": 2976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3811432509694952,
+      "learning_rate": 1.8975267768005828e-05,
+      "loss": 0.2864,
+      "step": 2977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4285338176452864,
+      "learning_rate": 1.897444702466069e-05,
+      "loss": 0.3331,
+      "step": 2978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9754019324573616,
+      "learning_rate": 1.897362597053084e-05,
+      "loss": 0.5513,
+      "step": 2979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28062601130507825,
+      "learning_rate": 1.897280460564471e-05,
+      "loss": 0.1539,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5016667633376,
+      "learning_rate": 1.897198293003074e-05,
+      "loss": 0.3324,
+      "step": 2981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5883370722958858,
+      "learning_rate": 1.8971160943717387e-05,
+      "loss": 0.3618,
+      "step": 2982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7596055078716403,
+      "learning_rate": 1.8970338646733112e-05,
+      "loss": 0.3371,
+      "step": 2983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34230411328604804,
+      "learning_rate": 1.8969516039106402e-05,
+      "loss": 0.2392,
+      "step": 2984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42128346482721546,
+      "learning_rate": 1.8968693120865734e-05,
+      "loss": 0.3327,
+      "step": 2985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6798764575206757,
+      "learning_rate": 1.896786989203961e-05,
+      "loss": 0.3974,
+      "step": 2986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3784241264767709,
+      "learning_rate": 1.896704635265654e-05,
+      "loss": 0.2136,
+      "step": 2987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.276853121304933,
+      "learning_rate": 1.8966222502745034e-05,
+      "loss": 0.5163,
+      "step": 2988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44923140900964686,
+      "learning_rate": 1.8965398342333632e-05,
+      "loss": 0.3497,
+      "step": 2989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2639236438544276,
+      "learning_rate": 1.896457387145087e-05,
+      "loss": 0.1798,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1097809048080438,
+      "learning_rate": 1.8963749090125302e-05,
+      "loss": 0.6302,
+      "step": 2991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.234947244764597,
+      "learning_rate": 1.8962923998385487e-05,
+      "loss": 0.6959,
+      "step": 2992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4410351618966857,
+      "learning_rate": 1.896209859626e-05,
+      "loss": 0.258,
+      "step": 2993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9236199272405345,
+      "learning_rate": 1.8961272883777424e-05,
+      "loss": 0.4043,
+      "step": 2994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8780417499501578,
+      "learning_rate": 1.8960446860966353e-05,
+      "loss": 0.5626,
+      "step": 2995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2345376241037861,
+      "learning_rate": 1.895962052785539e-05,
+      "loss": 0.1591,
+      "step": 2996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4550241241419985,
+      "learning_rate": 1.895879388447316e-05,
+      "loss": 0.347,
+      "step": 2997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0166192508391756,
+      "learning_rate": 1.8957966930848278e-05,
+      "loss": 0.4882,
+      "step": 2998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4244725281572897,
+      "learning_rate": 1.8957139667009388e-05,
+      "loss": 0.3237,
+      "step": 2999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9454011346453592,
+      "learning_rate": 1.8956312092985135e-05,
+      "loss": 0.453,
+      "step": 3000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4287534901474559,
+      "learning_rate": 1.895548420880418e-05,
+      "loss": 0.3222,
+      "step": 3001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3976559492680062,
+      "learning_rate": 1.8954656014495193e-05,
+      "loss": 0.2943,
+      "step": 3002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.256110448009788,
+      "learning_rate": 1.8953827510086855e-05,
+      "loss": 0.128,
+      "step": 3003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6654967054817726,
+      "learning_rate": 1.8952998695607848e-05,
+      "loss": 0.4478,
+      "step": 3004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4407137936298601,
+      "learning_rate": 1.895216957108689e-05,
+      "loss": 0.2961,
+      "step": 3005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4160984346255178,
+      "learning_rate": 1.8951340136552677e-05,
+      "loss": 0.3153,
+      "step": 3006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.078401739989723,
+      "learning_rate": 1.8950510392033945e-05,
+      "loss": 0.669,
+      "step": 3007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.33084119777448523,
+      "learning_rate": 1.8949680337559422e-05,
+      "loss": 0.2457,
+      "step": 3008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3027583624026009,
+      "learning_rate": 1.8948849973157855e-05,
+      "loss": 0.2139,
+      "step": 3009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9818615820212763,
+      "learning_rate": 1.8948019298858e-05,
+      "loss": 0.4861,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4073751640836383,
+      "learning_rate": 1.8947188314688614e-05,
+      "loss": 0.3296,
+      "step": 3011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7477288026893142,
+      "learning_rate": 1.8946357020678484e-05,
+      "loss": 0.5517,
+      "step": 3012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38402672958445916,
+      "learning_rate": 1.89455254168564e-05,
+      "loss": 0.2941,
+      "step": 3013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.378465393496418,
+      "learning_rate": 1.8944693503251154e-05,
+      "loss": 0.2712,
+      "step": 3014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38536621633444945,
+      "learning_rate": 1.8943861279891555e-05,
+      "loss": 0.19,
+      "step": 3015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.112414636701315,
+      "learning_rate": 1.8943028746806423e-05,
+      "loss": 0.3941,
+      "step": 3016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4202034683893572,
+      "learning_rate": 1.8942195904024593e-05,
+      "loss": 0.315,
+      "step": 3017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5026074471147242,
+      "learning_rate": 1.89413627515749e-05,
+      "loss": 0.8241,
+      "step": 3018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6070421657926213,
+      "learning_rate": 1.89405292894862e-05,
+      "loss": 0.3541,
+      "step": 3019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41453425610717626,
+      "learning_rate": 1.8939695517787355e-05,
+      "loss": 0.3023,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29993170844835293,
+      "learning_rate": 1.893886143650724e-05,
+      "loss": 0.2491,
+      "step": 3021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6600485195325879,
+      "learning_rate": 1.893802704567474e-05,
+      "loss": 0.3919,
+      "step": 3022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4849563665909965,
+      "learning_rate": 1.8937192345318745e-05,
+      "loss": 0.2887,
+      "step": 3023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.121191174841304,
+      "learning_rate": 1.8936357335468164e-05,
+      "loss": 0.7787,
+      "step": 3024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4336597655448138,
+      "learning_rate": 1.8935522016151914e-05,
+      "loss": 0.3233,
+      "step": 3025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30783811468715805,
+      "learning_rate": 1.8934686387398916e-05,
+      "loss": 0.2047,
+      "step": 3026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31539444070821426,
+      "learning_rate": 1.8933850449238118e-05,
+      "loss": 0.2582,
+      "step": 3027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7055654305578103,
+      "learning_rate": 1.893301420169846e-05,
+      "loss": 0.4926,
+      "step": 3028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3686311554656913,
+      "learning_rate": 1.893217764480891e-05,
+      "loss": 0.2407,
+      "step": 3029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1909355193895523,
+      "learning_rate": 1.8931340778598427e-05,
+      "loss": 0.6177,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4663985038033136,
+      "learning_rate": 1.8930503603095996e-05,
+      "loss": 0.8403,
+      "step": 3031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3190431419463848,
+      "learning_rate": 1.892966611833061e-05,
+      "loss": 0.1833,
+      "step": 3032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4511305241542145,
+      "learning_rate": 1.892882832433127e-05,
+      "loss": 0.3602,
+      "step": 3033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7068053529043494,
+      "learning_rate": 1.8927990221126992e-05,
+      "loss": 0.3597,
+      "step": 3034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42401792713410896,
+      "learning_rate": 1.8927151808746794e-05,
+      "loss": 0.284,
+      "step": 3035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0660654062947625,
+      "learning_rate": 1.8926313087219715e-05,
+      "loss": 0.4866,
+      "step": 3036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4213915786757873,
+      "learning_rate": 1.8925474056574797e-05,
+      "loss": 0.3367,
+      "step": 3037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5059010362430902,
+      "learning_rate": 1.8924634716841095e-05,
+      "loss": 0.3145,
+      "step": 3038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6904293276334135,
+      "learning_rate": 1.8923795068047676e-05,
+      "loss": 0.3638,
+      "step": 3039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32126069141038616,
+      "learning_rate": 1.892295511022362e-05,
+      "loss": 0.2206,
+      "step": 3040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49072173398664753,
+      "learning_rate": 1.8922114843398008e-05,
+      "loss": 0.2853,
+      "step": 3041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6178054617716159,
+      "learning_rate": 1.8921274267599948e-05,
+      "loss": 0.3293,
+      "step": 3042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3910184013031428,
+      "learning_rate": 1.8920433382858543e-05,
+      "loss": 0.6896,
+      "step": 3043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.43153226746326395,
+      "learning_rate": 1.8919592189202907e-05,
+      "loss": 0.3513,
+      "step": 3044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4102488838420799,
+      "learning_rate": 1.8918750686662182e-05,
+      "loss": 0.3131,
+      "step": 3045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47947338773237685,
+      "learning_rate": 1.8917908875265507e-05,
+      "loss": 0.3053,
+      "step": 3046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3460541287559192,
+      "learning_rate": 1.8917066755042028e-05,
+      "loss": 0.2311,
+      "step": 3047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5716077686683415,
+      "learning_rate": 1.891622432602091e-05,
+      "loss": 0.4332,
+      "step": 3048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5029071813624045,
+      "learning_rate": 1.8915381588231327e-05,
+      "loss": 0.2939,
+      "step": 3049
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44046657183956867,
+      "learning_rate": 1.8914538541702466e-05,
+      "loss": 0.2922,
+      "step": 3050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7995531866883153,
+      "learning_rate": 1.8913695186463517e-05,
+      "loss": 0.4911,
+      "step": 3051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36615694629870316,
+      "learning_rate": 1.8912851522543687e-05,
+      "loss": 0.2274,
+      "step": 3052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32321900527474984,
+      "learning_rate": 1.891200754997219e-05,
+      "loss": 0.2136,
+      "step": 3053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2596204295464923,
+      "learning_rate": 1.8911163268778257e-05,
+      "loss": 0.8114,
+      "step": 3054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7144167478169268,
+      "learning_rate": 1.891031867899112e-05,
+      "loss": 0.4086,
+      "step": 3055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6253671899765448,
+      "learning_rate": 1.8909473780640037e-05,
+      "loss": 0.3703,
+      "step": 3056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45134821280279186,
+      "learning_rate": 1.8908628573754254e-05,
+      "loss": 0.3021,
+      "step": 3057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22005853180138574,
+      "learning_rate": 1.890778305836305e-05,
+      "loss": 0.1269,
+      "step": 3058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4524272668639035,
+      "learning_rate": 1.89069372344957e-05,
+      "loss": 0.264,
+      "step": 3059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6209278718230915,
+      "learning_rate": 1.8906091102181495e-05,
+      "loss": 0.4388,
+      "step": 3060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4937884114799369,
+      "learning_rate": 1.890524466144974e-05,
+      "loss": 0.3877,
+      "step": 3061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41369254008640416,
+      "learning_rate": 1.8904397912329745e-05,
+      "loss": 0.2448,
+      "step": 3062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5973610783320861,
+      "learning_rate": 1.8903550854850834e-05,
+      "loss": 0.3712,
+      "step": 3063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.656951303682737,
+      "learning_rate": 1.890270348904234e-05,
+      "loss": 0.4271,
+      "step": 3064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2452544715442981,
+      "learning_rate": 1.8901855814933607e-05,
+      "loss": 0.1632,
+      "step": 3065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5677559684846096,
+      "learning_rate": 1.890100783255399e-05,
+      "loss": 0.4578,
+      "step": 3066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5891915481483365,
+      "learning_rate": 1.890015954193285e-05,
+      "loss": 0.4746,
+      "step": 3067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35663869657398584,
+      "learning_rate": 1.8899310943099573e-05,
+      "loss": 0.2468,
+      "step": 3068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5247041604668781,
+      "learning_rate": 1.8898462036083537e-05,
+      "loss": 0.3401,
+      "step": 3069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0380567894925528,
+      "learning_rate": 1.8897612820914147e-05,
+      "loss": 0.5755,
+      "step": 3070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2670561595530922,
+      "learning_rate": 1.8896763297620805e-05,
+      "loss": 0.1841,
+      "step": 3071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9455444266024311,
+      "learning_rate": 1.8895913466232937e-05,
+      "loss": 0.6592,
+      "step": 3072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41053313749568826,
+      "learning_rate": 1.8895063326779965e-05,
+      "loss": 0.3663,
+      "step": 3073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36267065242506263,
+      "learning_rate": 1.8894212879291332e-05,
+      "loss": 0.2047,
+      "step": 3074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4786448101222816,
+      "learning_rate": 1.8893362123796488e-05,
+      "loss": 0.2759,
+      "step": 3075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4093384984549816,
+      "learning_rate": 1.88925110603249e-05,
+      "loss": 0.3397,
+      "step": 3076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.845745579057613,
+      "learning_rate": 1.8891659688906033e-05,
+      "loss": 0.4151,
+      "step": 3077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47292178771932475,
+      "learning_rate": 1.8890808009569376e-05,
+      "loss": 0.3051,
+      "step": 3078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.49182209863150383,
+      "learning_rate": 1.8889956022344414e-05,
+      "loss": 0.3479,
+      "step": 3079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46806648165181836,
+      "learning_rate": 1.8889103727260666e-05,
+      "loss": 0.2757,
+      "step": 3080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3120414790673563,
+      "learning_rate": 1.888825112434763e-05,
+      "loss": 0.2244,
+      "step": 3081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2679215891920346,
+      "learning_rate": 1.8887398213634848e-05,
+      "loss": 0.7125,
+      "step": 3082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6403408615039261,
+      "learning_rate": 1.8886544995151844e-05,
+      "loss": 0.38,
+      "step": 3083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33151253151833543,
+      "learning_rate": 1.8885691468928166e-05,
+      "loss": 0.2932,
+      "step": 3084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7277961331311545,
+      "learning_rate": 1.8884837634993377e-05,
+      "loss": 0.5574,
+      "step": 3085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.340274101058813,
+      "learning_rate": 1.8883983493377045e-05,
+      "loss": 0.1544,
+      "step": 3086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4429892097106967,
+      "learning_rate": 1.8883129044108744e-05,
+      "loss": 0.3517,
+      "step": 3087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3928554684234481,
+      "learning_rate": 1.8882274287218067e-05,
+      "loss": 0.2741,
+      "step": 3088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4743146801940653,
+      "learning_rate": 1.8881419222734615e-05,
+      "loss": 0.3134,
+      "step": 3089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5239268770428587,
+      "learning_rate": 1.8880563850687995e-05,
+      "loss": 0.4167,
+      "step": 3090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8884943070873123,
+      "learning_rate": 1.8879708171107828e-05,
+      "loss": 0.4566,
+      "step": 3091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5371132744319287,
+      "learning_rate": 1.8878852184023754e-05,
+      "loss": 0.3002,
+      "step": 3092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32469681817681595,
+      "learning_rate": 1.887799588946541e-05,
+      "loss": 0.2159,
+      "step": 3093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2594353839536387,
+      "learning_rate": 1.8877139287462446e-05,
+      "loss": 0.3695,
+      "step": 3094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6393985107547493,
+      "learning_rate": 1.8876282378044535e-05,
+      "loss": 0.4849,
+      "step": 3095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3707643080188264,
+      "learning_rate": 1.8875425161241345e-05,
+      "loss": 0.3116,
+      "step": 3096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5094016079937749,
+      "learning_rate": 1.887456763708256e-05,
+      "loss": 0.3134,
+      "step": 3097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2347285327171995,
+      "learning_rate": 1.8873709805597884e-05,
+      "loss": 0.1598,
+      "step": 3098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40752913954631803,
+      "learning_rate": 1.8872851666817017e-05,
+      "loss": 0.3338,
+      "step": 3099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5350281103112107,
+      "learning_rate": 1.887199322076968e-05,
+      "loss": 0.3887,
+      "step": 3100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6124219414870147,
+      "learning_rate": 1.8871134467485597e-05,
+      "loss": 0.2902,
+      "step": 3101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5018307054344358,
+      "learning_rate": 1.8870275406994513e-05,
+      "loss": 0.3797,
+      "step": 3102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1961520940052548,
+      "learning_rate": 1.886941603932617e-05,
+      "loss": 0.7538,
+      "step": 3103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32363681224514806,
+      "learning_rate": 1.886855636451033e-05,
+      "loss": 0.2232,
+      "step": 3104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4023577478846728,
+      "learning_rate": 1.8867696382576767e-05,
+      "loss": 0.2839,
+      "step": 3105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.437139177965491,
+      "learning_rate": 1.886683609355526e-05,
+      "loss": 0.3293,
+      "step": 3106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6121045236750057,
+      "learning_rate": 1.8865975497475596e-05,
+      "loss": 0.333,
+      "step": 3107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4427691977419533,
+      "learning_rate": 1.8865114594367585e-05,
+      "loss": 0.3435,
+      "step": 3108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6110743595027336,
+      "learning_rate": 1.8864253384261036e-05,
+      "loss": 0.4338,
+      "step": 3109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.580897320895251,
+      "learning_rate": 1.8863391867185774e-05,
+      "loss": 0.3527,
+      "step": 3110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29260609606365173,
+      "learning_rate": 1.8862530043171633e-05,
+      "loss": 0.1696,
+      "step": 3111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4524691334794982,
+      "learning_rate": 1.8861667912248456e-05,
+      "loss": 0.3644,
+      "step": 3112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8502053958943329,
+      "learning_rate": 1.8860805474446103e-05,
+      "loss": 0.5187,
+      "step": 3113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34026675743767276,
+      "learning_rate": 1.8859942729794433e-05,
+      "loss": 0.2207,
+      "step": 3114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5803398218349507,
+      "learning_rate": 1.885907967832333e-05,
+      "loss": 0.4215,
+      "step": 3115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5413184538995245,
+      "learning_rate": 1.885821632006268e-05,
+      "loss": 0.385,
+      "step": 3116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.267538412442085,
+      "learning_rate": 1.8857352655042378e-05,
+      "loss": 0.1605,
+      "step": 3117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9481877105032955,
+      "learning_rate": 1.885648868329234e-05,
+      "loss": 0.5323,
+      "step": 3118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7094437605080949,
+      "learning_rate": 1.8855624404842472e-05,
+      "loss": 0.4595,
+      "step": 3119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3795999302636725,
+      "learning_rate": 1.8854759819722713e-05,
+      "loss": 0.2596,
+      "step": 3120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3263785752350479,
+      "learning_rate": 1.8853894927963004e-05,
+      "loss": 0.8709,
+      "step": 3121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6358746810174966,
+      "learning_rate": 1.8853029729593296e-05,
+      "loss": 0.4608,
+      "step": 3122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3726386152293299,
+      "learning_rate": 1.8852164224643546e-05,
+      "loss": 0.2343,
+      "step": 3123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5836516260644941,
+      "learning_rate": 1.885129841314373e-05,
+      "loss": 0.3314,
+      "step": 3124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.49963353752080175,
+      "learning_rate": 1.8850432295123832e-05,
+      "loss": 0.2861,
+      "step": 3125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4523725849030945,
+      "learning_rate": 1.8849565870613844e-05,
+      "loss": 0.3031,
+      "step": 3126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47633458502229503,
+      "learning_rate": 1.8848699139643768e-05,
+      "loss": 0.3484,
+      "step": 3127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4572869976239571,
+      "learning_rate": 1.8847832102243626e-05,
+      "loss": 0.3506,
+      "step": 3128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4738886849479047,
+      "learning_rate": 1.8846964758443434e-05,
+      "loss": 0.3156,
+      "step": 3129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3158973478786652,
+      "learning_rate": 1.8846097108273234e-05,
+      "loss": 0.2454,
+      "step": 3130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6088936520152631,
+      "learning_rate": 1.8845229151763072e-05,
+      "loss": 0.3999,
+      "step": 3131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.341198313608363,
+      "learning_rate": 1.884436088894301e-05,
+      "loss": 0.2854,
+      "step": 3132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8632109606054379,
+      "learning_rate": 1.8843492319843105e-05,
+      "loss": 0.5362,
+      "step": 3133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8698131777727494,
+      "learning_rate": 1.884262344449344e-05,
+      "loss": 0.5899,
+      "step": 3134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40345140665949575,
+      "learning_rate": 1.8841754262924106e-05,
+      "loss": 0.2877,
+      "step": 3135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4323619276330218,
+      "learning_rate": 1.8840884775165204e-05,
+      "loss": 0.3597,
+      "step": 3136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23494410269272176,
+      "learning_rate": 1.8840014981246843e-05,
+      "loss": 0.121,
+      "step": 3137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39915844033516507,
+      "learning_rate": 1.8839144881199144e-05,
+      "loss": 0.2998,
+      "step": 3138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3686412428770773,
+      "learning_rate": 1.8838274475052233e-05,
+      "loss": 0.8266,
+      "step": 3139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3427600739038595,
+      "learning_rate": 1.883740376283626e-05,
+      "loss": 0.2858,
+      "step": 3140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4077783868641141,
+      "learning_rate": 1.8836532744581377e-05,
+      "loss": 0.3271,
+      "step": 3141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7926122519324589,
+      "learning_rate": 1.8835661420317745e-05,
+      "loss": 0.4978,
+      "step": 3142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25284561606242717,
+      "learning_rate": 1.8834789790075536e-05,
+      "loss": 0.1756,
+      "step": 3143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38142146209876004,
+      "learning_rate": 1.8833917853884935e-05,
+      "loss": 0.2853,
+      "step": 3144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.088916732216867,
+      "learning_rate": 1.8833045611776143e-05,
+      "loss": 0.801,
+      "step": 3145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.618587140267945,
+      "learning_rate": 1.8832173063779357e-05,
+      "loss": 0.3862,
+      "step": 3146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4643314856293626,
+      "learning_rate": 1.8831300209924797e-05,
+      "loss": 0.3202,
+      "step": 3147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40377489613397566,
+      "learning_rate": 1.8830427050242693e-05,
+      "loss": 0.3142,
+      "step": 3148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2564840205627374,
+      "learning_rate": 1.8829553584763278e-05,
+      "loss": 0.1487,
+      "step": 3149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3232933389811095,
+      "learning_rate": 1.8828679813516806e-05,
+      "loss": 0.2236,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5490085608060375,
+      "learning_rate": 1.8827805736533528e-05,
+      "loss": 0.4398,
+      "step": 3151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5491388527656633,
+      "learning_rate": 1.8826931353843717e-05,
+      "loss": 0.4112,
+      "step": 3152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39943577631082894,
+      "learning_rate": 1.8826056665477654e-05,
+      "loss": 0.2803,
+      "step": 3153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.773866438726851,
+      "learning_rate": 1.8825181671465628e-05,
+      "loss": 0.5776,
+      "step": 3154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3216913090997973,
+      "learning_rate": 1.882430637183794e-05,
+      "loss": 0.2737,
+      "step": 3155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.290138086584041,
+      "learning_rate": 1.8823430766624905e-05,
+      "loss": 0.1886,
+      "step": 3156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4431178972907859,
+      "learning_rate": 1.8822554855856838e-05,
+      "loss": 0.8918,
+      "step": 3157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5797258534093347,
+      "learning_rate": 1.8821678639564075e-05,
+      "loss": 0.4781,
+      "step": 3158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3837048306336846,
+      "learning_rate": 1.8820802117776963e-05,
+      "loss": 0.2224,
+      "step": 3159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42205950077117843,
+      "learning_rate": 1.8819925290525854e-05,
+      "loss": 0.3351,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3076485977032022,
+      "learning_rate": 1.8819048157841105e-05,
+      "loss": 0.2032,
+      "step": 3161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5041804621749775,
+      "learning_rate": 1.8818170719753104e-05,
+      "loss": 0.3487,
+      "step": 3162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38071028970387893,
+      "learning_rate": 1.8817292976292227e-05,
+      "loss": 0.33,
+      "step": 3163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4558400988040032,
+      "learning_rate": 1.8816414927488877e-05,
+      "loss": 0.3912,
+      "step": 3164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45130705580419683,
+      "learning_rate": 1.8815536573373453e-05,
+      "loss": 0.2836,
+      "step": 3165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8510761019378265,
+      "learning_rate": 1.8814657913976377e-05,
+      "loss": 0.302,
+      "step": 3166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.429963063044294,
+      "learning_rate": 1.881377894932808e-05,
+      "loss": 0.3323,
+      "step": 3167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29400249907992115,
+      "learning_rate": 1.8812899679458993e-05,
+      "loss": 0.23,
+      "step": 3168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.464996279016933,
+      "learning_rate": 1.8812020104399572e-05,
+      "loss": 0.3656,
+      "step": 3169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46758247447598833,
+      "learning_rate": 1.8811140224180273e-05,
+      "loss": 0.3112,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41433320208215574,
+      "learning_rate": 1.8810260038831564e-05,
+      "loss": 0.291,
+      "step": 3171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.452144073174074,
+      "learning_rate": 1.8809379548383932e-05,
+      "loss": 0.2709,
+      "step": 3172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.178725602164265,
+      "learning_rate": 1.8808498752867863e-05,
+      "loss": 0.7155,
+      "step": 3173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34409535391847335,
+      "learning_rate": 1.880761765231386e-05,
+      "loss": 0.2906,
+      "step": 3174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7192857758620904,
+      "learning_rate": 1.8806736246752443e-05,
+      "loss": 0.4855,
+      "step": 3175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3361678418408267,
+      "learning_rate": 1.8805854536214122e-05,
+      "loss": 0.2721,
+      "step": 3176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37055400507860653,
+      "learning_rate": 1.8804972520729443e-05,
+      "loss": 0.262,
+      "step": 3177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9664731315219579,
+      "learning_rate": 1.8804090200328938e-05,
+      "loss": 0.5504,
+      "step": 3178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42261767384952514,
+      "learning_rate": 1.880320757504317e-05,
+      "loss": 0.2837,
+      "step": 3179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.614889244744422,
+      "learning_rate": 1.8802324644902704e-05,
+      "loss": 0.3954,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5047054055116674,
+      "learning_rate": 1.880144140993811e-05,
+      "loss": 0.365,
+      "step": 3181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2809330896067468,
+      "learning_rate": 1.880055787017998e-05,
+      "loss": 0.1917,
+      "step": 3182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4424624416740233,
+      "learning_rate": 1.8799674025658913e-05,
+      "loss": 0.3117,
+      "step": 3183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4259910851470908,
+      "learning_rate": 1.879878987640551e-05,
+      "loss": 0.3169,
+      "step": 3184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5477900879594761,
+      "learning_rate": 1.879790542245039e-05,
+      "loss": 0.4208,
+      "step": 3185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5038440134908405,
+      "learning_rate": 1.8797020663824187e-05,
+      "loss": 0.3235,
+      "step": 3186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4738666383574769,
+      "learning_rate": 1.8796135600557534e-05,
+      "loss": 0.3355,
+      "step": 3187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4251583580601876,
+      "learning_rate": 1.8795250232681085e-05,
+      "loss": 0.9205,
+      "step": 3188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.253214970145997,
+      "learning_rate": 1.8794364560225496e-05,
+      "loss": 0.1462,
+      "step": 3189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6313789827965512,
+      "learning_rate": 1.8793478583221448e-05,
+      "loss": 0.3916,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4616837157366747,
+      "learning_rate": 1.879259230169961e-05,
+      "loss": 0.3263,
+      "step": 3191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3453392906380078,
+      "learning_rate": 1.8791705715690675e-05,
+      "loss": 0.2672,
+      "step": 3192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7657077369940369,
+      "learning_rate": 1.8790818825225355e-05,
+      "loss": 0.5951,
+      "step": 3193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4873348104716959,
+      "learning_rate": 1.8789931630334353e-05,
+      "loss": 0.3264,
+      "step": 3194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3471370066679422,
+      "learning_rate": 1.8789044131048397e-05,
+      "loss": 0.2468,
+      "step": 3195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5258610482459171,
+      "learning_rate": 1.8788156327398225e-05,
+      "loss": 0.2788,
+      "step": 3196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5999552621196282,
+      "learning_rate": 1.8787268219414572e-05,
+      "loss": 0.4748,
+      "step": 3197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5376375761332872,
+      "learning_rate": 1.87863798071282e-05,
+      "loss": 0.3582,
+      "step": 3198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42167059924717126,
+      "learning_rate": 1.8785491090569876e-05,
+      "loss": 0.2779,
+      "step": 3199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.048591073850127,
+      "learning_rate": 1.878460206977037e-05,
+      "loss": 0.7897,
+      "step": 3200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3659218143881057,
+      "learning_rate": 1.8783712744760475e-05,
+      "loss": 0.1953,
+      "step": 3201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4075025882352195,
+      "learning_rate": 1.878282311557098e-05,
+      "loss": 0.2572,
+      "step": 3202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.43005347413976514,
+      "learning_rate": 1.8781933182232702e-05,
+      "loss": 0.3816,
+      "step": 3203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5294171894124521,
+      "learning_rate": 1.8781042944776457e-05,
+      "loss": 0.3735,
+      "step": 3204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4682006514675936,
+      "learning_rate": 1.8780152403233073e-05,
+      "loss": 0.29,
+      "step": 3205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4391156031332297,
+      "learning_rate": 1.8779261557633385e-05,
+      "loss": 0.696,
+      "step": 3206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34352071114583893,
+      "learning_rate": 1.8778370408008247e-05,
+      "loss": 0.256,
+      "step": 3207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31215051226110274,
+      "learning_rate": 1.877747895438852e-05,
+      "loss": 0.2139,
+      "step": 3208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7384179652760221,
+      "learning_rate": 1.8776587196805077e-05,
+      "loss": 0.4823,
+      "step": 3209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6360730312478228,
+      "learning_rate": 1.8775695135288794e-05,
+      "loss": 0.4391,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46584518363056193,
+      "learning_rate": 1.8774802769870564e-05,
+      "loss": 0.3308,
+      "step": 3211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4706672925631658,
+      "learning_rate": 1.8773910100581294e-05,
+      "loss": 0.3272,
+      "step": 3212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2877101741955395,
+      "learning_rate": 1.8773017127451893e-05,
+      "loss": 0.2086,
+      "step": 3213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4410049272946392,
+      "learning_rate": 1.877212385051329e-05,
+      "loss": 0.2814,
+      "step": 3214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4041381893357965,
+      "learning_rate": 1.8771230269796412e-05,
+      "loss": 0.3087,
+      "step": 3215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7639969423662106,
+      "learning_rate": 1.877033638533221e-05,
+      "loss": 0.4319,
+      "step": 3216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4278602034149037,
+      "learning_rate": 1.876944219715163e-05,
+      "loss": 0.3137,
+      "step": 3217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48961249607176893,
+      "learning_rate": 1.876854770528565e-05,
+      "loss": 0.3156,
+      "step": 3218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44431744410002555,
+      "learning_rate": 1.8767652909765236e-05,
+      "loss": 0.3015,
+      "step": 3219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33106297024120673,
+      "learning_rate": 1.8766757810621383e-05,
+      "loss": 0.3407,
+      "step": 3220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28688345276164595,
+      "learning_rate": 1.876586240788508e-05,
+      "loss": 0.1918,
+      "step": 3221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5071273372638695,
+      "learning_rate": 1.876496670158734e-05,
+      "loss": 0.3566,
+      "step": 3222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4110348765229278,
+      "learning_rate": 1.876407069175918e-05,
+      "loss": 0.316,
+      "step": 3223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9657130710337607,
+      "learning_rate": 1.876317437843163e-05,
+      "loss": 0.6621,
+      "step": 3224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4092046974694814,
+      "learning_rate": 1.8762277761635725e-05,
+      "loss": 0.2433,
+      "step": 3225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39059696176709385,
+      "learning_rate": 1.876138084140252e-05,
+      "loss": 0.3564,
+      "step": 3226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3427885145541495,
+      "learning_rate": 1.876048361776307e-05,
+      "loss": 0.2631,
+      "step": 3227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2848821101035255,
+      "learning_rate": 1.8759586090748454e-05,
+      "loss": 0.1501,
+      "step": 3228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4979226409712117,
+      "learning_rate": 1.8758688260389747e-05,
+      "loss": 0.3348,
+      "step": 3229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7883755482032335,
+      "learning_rate": 1.8757790126718044e-05,
+      "loss": 0.849,
+      "step": 3230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3038487339560472,
+      "learning_rate": 1.8756891689764444e-05,
+      "loss": 0.2654,
+      "step": 3231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.426868655627405,
+      "learning_rate": 1.8755992949560058e-05,
+      "loss": 0.3428,
+      "step": 3232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34227594083645335,
+      "learning_rate": 1.8755093906136016e-05,
+      "loss": 0.2369,
+      "step": 3233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4159884317750057,
+      "learning_rate": 1.8754194559523447e-05,
+      "loss": 0.2368,
+      "step": 3234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4452450173245431,
+      "learning_rate": 1.87532949097535e-05,
+      "loss": 0.3102,
+      "step": 3235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8178285752035969,
+      "learning_rate": 1.8752394956857322e-05,
+      "loss": 0.5632,
+      "step": 3236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6512942111877557,
+      "learning_rate": 1.8751494700866088e-05,
+      "loss": 0.491,
+      "step": 3237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36820713984463505,
+      "learning_rate": 1.8750594141810964e-05,
+      "loss": 0.205,
+      "step": 3238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32327909485014955,
+      "learning_rate": 1.8749693279723146e-05,
+      "loss": 0.241,
+      "step": 3239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3048291367110827,
+      "learning_rate": 1.874879211463382e-05,
+      "loss": 0.6865,
+      "step": 3240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38934075541803975,
+      "learning_rate": 1.8747890646574204e-05,
+      "loss": 0.2433,
+      "step": 3241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8793842949876021,
+      "learning_rate": 1.874698887557551e-05,
+      "loss": 0.5371,
+      "step": 3242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4606832588802145,
+      "learning_rate": 1.8746086801668964e-05,
+      "loss": 0.3555,
+      "step": 3243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34513483530045763,
+      "learning_rate": 1.8745184424885815e-05,
+      "loss": 0.2315,
+      "step": 3244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6050070424752414,
+      "learning_rate": 1.87442817452573e-05,
+      "loss": 0.2856,
+      "step": 3245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30067116685715717,
+      "learning_rate": 1.8743378762814685e-05,
+      "loss": 0.2406,
+      "step": 3246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3879704127020492,
+      "learning_rate": 1.874247547758924e-05,
+      "loss": 0.2495,
+      "step": 3247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0069201260432241,
+      "learning_rate": 1.8741571889612248e-05,
+      "loss": 0.5666,
+      "step": 3248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6375161367998676,
+      "learning_rate": 1.8740667998914996e-05,
+      "loss": 0.4336,
+      "step": 3249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42327093852418024,
+      "learning_rate": 1.8739763805528782e-05,
+      "loss": 0.306,
+      "step": 3250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41315398403240317,
+      "learning_rate": 1.8738859309484926e-05,
+      "loss": 0.2775,
+      "step": 3251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4775464807987565,
+      "learning_rate": 1.8737954510814752e-05,
+      "loss": 0.306,
+      "step": 3252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3845598831291664,
+      "learning_rate": 1.8737049409549586e-05,
+      "loss": 0.2899,
+      "step": 3253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7332887356293013,
+      "learning_rate": 1.8736144005720775e-05,
+      "loss": 0.3581,
+      "step": 3254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5740626229995096,
+      "learning_rate": 1.8735238299359672e-05,
+      "loss": 0.3243,
+      "step": 3255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37707225799385136,
+      "learning_rate": 1.8734332290497642e-05,
+      "loss": 0.285,
+      "step": 3256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2926649412762437,
+      "learning_rate": 1.8733425979166063e-05,
+      "loss": 0.1384,
+      "step": 3257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4234930784962659,
+      "learning_rate": 1.8732519365396314e-05,
+      "loss": 0.3293,
+      "step": 3258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44694857769056656,
+      "learning_rate": 1.87316124492198e-05,
+      "loss": 0.3063,
+      "step": 3259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.881998655230814,
+      "learning_rate": 1.8730705230667916e-05,
+      "loss": 0.4143,
+      "step": 3260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47126328441633636,
+      "learning_rate": 1.8729797709772088e-05,
+      "loss": 0.3326,
+      "step": 3261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3424134567721488,
+      "learning_rate": 1.872888988656374e-05,
+      "loss": 0.309,
+      "step": 3262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5270321114224565,
+      "learning_rate": 1.8727981761074315e-05,
+      "loss": 0.375,
+      "step": 3263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6391706970436589,
+      "learning_rate": 1.872707333333525e-05,
+      "loss": 0.3429,
+      "step": 3264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4914387054979702,
+      "learning_rate": 1.8726164603378016e-05,
+      "loss": 0.3219,
+      "step": 3265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6902772708493562,
+      "learning_rate": 1.8725255571234075e-05,
+      "loss": 0.4276,
+      "step": 3266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.252273761014582,
+      "learning_rate": 1.872434623693491e-05,
+      "loss": 0.1688,
+      "step": 3267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43328993906220703,
+      "learning_rate": 1.8723436600512007e-05,
+      "loss": 0.3001,
+      "step": 3268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5775971207484492,
+      "learning_rate": 1.8722526661996872e-05,
+      "loss": 0.6708,
+      "step": 3269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4552562263007585,
+      "learning_rate": 1.8721616421421017e-05,
+      "loss": 0.3101,
+      "step": 3270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3964430436328616,
+      "learning_rate": 1.8720705878815953e-05,
+      "loss": 0.2978,
+      "step": 3271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9240033860694061,
+      "learning_rate": 1.8719795034213226e-05,
+      "loss": 0.5123,
+      "step": 3272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35059453995859596,
+      "learning_rate": 1.871888388764437e-05,
+      "loss": 0.1863,
+      "step": 3273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32757580915834644,
+      "learning_rate": 1.8717972439140938e-05,
+      "loss": 0.23,
+      "step": 3274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5944268003456217,
+      "learning_rate": 1.8717060688734495e-05,
+      "loss": 0.41,
+      "step": 3275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1405676959172855,
+      "learning_rate": 1.871614863645662e-05,
+      "loss": 0.5975,
+      "step": 3276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4561238387683592,
+      "learning_rate": 1.871523628233889e-05,
+      "loss": 0.267,
+      "step": 3277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5709846215706098,
+      "learning_rate": 1.87143236264129e-05,
+      "loss": 0.4121,
+      "step": 3278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32475528845468604,
+      "learning_rate": 1.871341066871026e-05,
+      "loss": 0.2167,
+      "step": 3279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35223528951570476,
+      "learning_rate": 1.8712497409262582e-05,
+      "loss": 0.2124,
+      "step": 3280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.107902808950662,
+      "learning_rate": 1.8711583848101492e-05,
+      "loss": 0.6202,
+      "step": 3281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5448763076692719,
+      "learning_rate": 1.871066998525863e-05,
+      "loss": 0.3702,
+      "step": 3282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43760624896546474,
+      "learning_rate": 1.870975582076564e-05,
+      "loss": 0.2446,
+      "step": 3283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49297898663282697,
+      "learning_rate": 1.8708841354654184e-05,
+      "loss": 0.4166,
+      "step": 3284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3323663731490816,
+      "learning_rate": 1.870792658695592e-05,
+      "loss": 0.2341,
+      "step": 3285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49578418381735967,
+      "learning_rate": 1.870701151770254e-05,
+      "loss": 0.2838,
+      "step": 3286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4174917171345786,
+      "learning_rate": 1.870609614692572e-05,
+      "loss": 0.2761,
+      "step": 3287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.750839879220727,
+      "learning_rate": 1.8705180474657166e-05,
+      "loss": 0.4965,
+      "step": 3288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5129470258434858,
+      "learning_rate": 1.8704264500928588e-05,
+      "loss": 0.2986,
+      "step": 3289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42347571147632856,
+      "learning_rate": 1.87033482257717e-05,
+      "loss": 0.2931,
+      "step": 3290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.481272218463836,
+      "learning_rate": 1.8702431649218245e-05,
+      "loss": 0.2205,
+      "step": 3291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30216937260105364,
+      "learning_rate": 1.870151477129995e-05,
+      "loss": 0.2286,
+      "step": 3292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5601484432369667,
+      "learning_rate": 1.8700597592048576e-05,
+      "loss": 0.3279,
+      "step": 3293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4983673447612351,
+      "learning_rate": 1.869968011149588e-05,
+      "loss": 0.3449,
+      "step": 3294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48122935402031364,
+      "learning_rate": 1.8698762329673636e-05,
+      "loss": 0.3335,
+      "step": 3295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5877751356413182,
+      "learning_rate": 1.869784424661363e-05,
+      "loss": 0.3133,
+      "step": 3296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3422421447813925,
+      "learning_rate": 1.8696925862347647e-05,
+      "loss": 0.2158,
+      "step": 3297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3768728050060615,
+      "learning_rate": 1.8696007176907494e-05,
+      "loss": 0.2928,
+      "step": 3298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6163884429509207,
+      "learning_rate": 1.869508819032499e-05,
+      "loss": 0.3752,
+      "step": 3299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8648706566024917,
+      "learning_rate": 1.8694168902631957e-05,
+      "loss": 0.4133,
+      "step": 3300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4157955496291116,
+      "learning_rate": 1.8693249313860225e-05,
+      "loss": 0.3008,
+      "step": 3301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3660210547536385,
+      "learning_rate": 1.8692329424041648e-05,
+      "loss": 0.3064,
+      "step": 3302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.267248482819533,
+      "learning_rate": 1.8691409233208072e-05,
+      "loss": 0.1172,
+      "step": 3303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6430966014648096,
+      "learning_rate": 1.869048874139137e-05,
+      "loss": 0.3144,
+      "step": 3304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.666856224468795,
+      "learning_rate": 1.8689567948623417e-05,
+      "loss": 0.3695,
+      "step": 3305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5142155690233231,
+      "learning_rate": 1.86886468549361e-05,
+      "loss": 0.3119,
+      "step": 3306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4921167365321923,
+      "learning_rate": 1.8687725460361315e-05,
+      "loss": 0.3192,
+      "step": 3307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6095335014762302,
+      "learning_rate": 1.868680376493097e-05,
+      "loss": 0.4577,
+      "step": 3308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3542443554161993,
+      "learning_rate": 1.8685881768676983e-05,
+      "loss": 0.1199,
+      "step": 3309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3551171004629968,
+      "learning_rate": 1.868495947163129e-05,
+      "loss": 0.2994,
+      "step": 3310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4032175463163124,
+      "learning_rate": 1.8684036873825817e-05,
+      "loss": 0.3191,
+      "step": 3311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9000679546075138,
+      "learning_rate": 1.8683113975292522e-05,
+      "loss": 0.5656,
+      "step": 3312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4240286543723701,
+      "learning_rate": 1.8682190776063368e-05,
+      "loss": 0.2259,
+      "step": 3313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4201800746491544,
+      "learning_rate": 1.8681267276170315e-05,
+      "loss": 0.3116,
+      "step": 3314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8506988193417204,
+      "learning_rate": 1.8680343475645354e-05,
+      "loss": 0.5609,
+      "step": 3315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3969631858789526,
+      "learning_rate": 1.8679419374520467e-05,
+      "loss": 0.2653,
+      "step": 3316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6065235075604374,
+      "learning_rate": 1.8678494972827665e-05,
+      "loss": 0.419,
+      "step": 3317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32114184153525965,
+      "learning_rate": 1.8677570270598956e-05,
+      "loss": 0.2346,
+      "step": 3318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40387175409091025,
+      "learning_rate": 1.8676645267866356e-05,
+      "loss": 0.2204,
+      "step": 3319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6801443170904025,
+      "learning_rate": 1.867571996466191e-05,
+      "loss": 0.4311,
+      "step": 3320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5159272006258127,
+      "learning_rate": 1.867479436101765e-05,
+      "loss": 0.3767,
+      "step": 3321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4089571870928136,
+      "learning_rate": 1.8673868456965635e-05,
+      "loss": 0.2527,
+      "step": 3322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48397213976412556,
+      "learning_rate": 1.867294225253793e-05,
+      "loss": 0.3643,
+      "step": 3323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3446079467709578,
+      "learning_rate": 1.8672015747766606e-05,
+      "loss": 0.205,
+      "step": 3324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8503150555575159,
+      "learning_rate": 1.8671088942683752e-05,
+      "loss": 0.5678,
+      "step": 3325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39654527040197246,
+      "learning_rate": 1.8670161837321457e-05,
+      "loss": 0.2591,
+      "step": 3326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0220278362714632,
+      "learning_rate": 1.8669234431711833e-05,
+      "loss": 0.5651,
+      "step": 3327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5649493509801116,
+      "learning_rate": 1.866830672588699e-05,
+      "loss": 0.3735,
+      "step": 3328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3635380013575951,
+      "learning_rate": 1.866737871987906e-05,
+      "loss": 0.2953,
+      "step": 3329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3433034425476781,
+      "learning_rate": 1.866645041372018e-05,
+      "loss": 0.2176,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6755984958812598,
+      "learning_rate": 1.8665521807442495e-05,
+      "loss": 0.4254,
+      "step": 3331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41328940725588204,
+      "learning_rate": 1.866459290107816e-05,
+      "loss": 0.2699,
+      "step": 3332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5305543815291648,
+      "learning_rate": 1.8663663694659348e-05,
+      "loss": 0.3819,
+      "step": 3333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42991083738123465,
+      "learning_rate": 1.866273418821823e-05,
+      "loss": 0.3441,
+      "step": 3334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6544073955837417,
+      "learning_rate": 1.8661804381787e-05,
+      "loss": 0.2438,
+      "step": 3335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3474028790555719,
+      "learning_rate": 1.8660874275397864e-05,
+      "loss": 0.2269,
+      "step": 3336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44660721208933224,
+      "learning_rate": 1.8659943869083016e-05,
+      "loss": 0.3194,
+      "step": 3337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36479851194376395,
+      "learning_rate": 1.865901316287469e-05,
+      "loss": 0.3026,
+      "step": 3338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7251793067592559,
+      "learning_rate": 1.8658082156805105e-05,
+      "loss": 0.4025,
+      "step": 3339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1762657605614177,
+      "learning_rate": 1.8657150850906515e-05,
+      "loss": 0.4898,
+      "step": 3340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44822935372518946,
+      "learning_rate": 1.8656219245211157e-05,
+      "loss": 0.3101,
+      "step": 3341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3095470714000383,
+      "learning_rate": 1.86552873397513e-05,
+      "loss": 0.2329,
+      "step": 3342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8687453149208805,
+      "learning_rate": 1.8654355134559216e-05,
+      "loss": 0.5129,
+      "step": 3343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3021584131565959,
+      "learning_rate": 1.8653422629667183e-05,
+      "loss": 0.2725,
+      "step": 3344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5873873067313393,
+      "learning_rate": 1.8652489825107497e-05,
+      "loss": 0.358,
+      "step": 3345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46313571761420985,
+      "learning_rate": 1.865155672091246e-05,
+      "loss": 0.3272,
+      "step": 3346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42763704537288455,
+      "learning_rate": 1.865062331711439e-05,
+      "loss": 0.2979,
+      "step": 3347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8943885070194173,
+      "learning_rate": 1.8649689613745605e-05,
+      "loss": 0.4268,
+      "step": 3348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4519340597314842,
+      "learning_rate": 1.864875561083844e-05,
+      "loss": 0.3254,
+      "step": 3349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34699922308137227,
+      "learning_rate": 1.864782130842524e-05,
+      "loss": 0.2848,
+      "step": 3350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6473095340940566,
+      "learning_rate": 1.8646886706538358e-05,
+      "loss": 0.471,
+      "step": 3351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.337720021022626,
+      "learning_rate": 1.8645951805210164e-05,
+      "loss": 0.1635,
+      "step": 3352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515168665181641,
+      "learning_rate": 1.864501660447303e-05,
+      "loss": 0.363,
+      "step": 3353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.45899933383969893,
+      "learning_rate": 1.8644081104359343e-05,
+      "loss": 0.3239,
+      "step": 3354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4341098182538872,
+      "learning_rate": 1.8643145304901497e-05,
+      "loss": 0.2779,
+      "step": 3355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43247406577763553,
+      "learning_rate": 1.8642209206131902e-05,
+      "loss": 0.3562,
+      "step": 3356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.315059317047122,
+      "learning_rate": 1.8641272808082975e-05,
+      "loss": 0.3091,
+      "step": 3357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24709989347560327,
+      "learning_rate": 1.864033611078714e-05,
+      "loss": 0.0748,
+      "step": 3358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4234610157900327,
+      "learning_rate": 1.863939911427684e-05,
+      "loss": 0.3157,
+      "step": 3359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0117601153578217,
+      "learning_rate": 1.8638461818584517e-05,
+      "loss": 0.672,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4147187717202964,
+      "learning_rate": 1.8637524223742636e-05,
+      "loss": 0.3426,
+      "step": 3361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35245150890474214,
+      "learning_rate": 1.863658632978366e-05,
+      "loss": 0.2728,
+      "step": 3362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4678697988608274,
+      "learning_rate": 1.8635648136740072e-05,
+      "loss": 0.3828,
+      "step": 3363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4667194994023077,
+      "learning_rate": 1.863470964464436e-05,
+      "loss": 0.2872,
+      "step": 3364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32564335685434603,
+      "learning_rate": 1.8633770853529025e-05,
+      "loss": 0.236,
+      "step": 3365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5608166440455324,
+      "learning_rate": 1.8632831763426574e-05,
+      "loss": 0.3899,
+      "step": 3366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6682512540559473,
+      "learning_rate": 1.863189237436953e-05,
+      "loss": 0.4906,
+      "step": 3367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35298988092787115,
+      "learning_rate": 1.863095268639043e-05,
+      "loss": 0.2482,
+      "step": 3368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42340797367221217,
+      "learning_rate": 1.8630012699521806e-05,
+      "loss": 0.3356,
+      "step": 3369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2767430952636719,
+      "learning_rate": 1.8629072413796213e-05,
+      "loss": 0.1825,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3605012936700501,
+      "learning_rate": 1.862813182924621e-05,
+      "loss": 0.2341,
+      "step": 3371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.153440050494576,
+      "learning_rate": 1.8627190945904382e-05,
+      "loss": 0.5661,
+      "step": 3372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48897754156527273,
+      "learning_rate": 1.8626249763803295e-05,
+      "loss": 0.3302,
+      "step": 3373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46612356766072266,
+      "learning_rate": 1.8625308282975552e-05,
+      "loss": 0.3312,
+      "step": 3374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3181013883121376,
+      "learning_rate": 1.8624366503453752e-05,
+      "loss": 0.2266,
+      "step": 3375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4652146982144647,
+      "learning_rate": 1.8623424425270514e-05,
+      "loss": 0.2808,
+      "step": 3376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39602293510398723,
+      "learning_rate": 1.8622482048458454e-05,
+      "loss": 0.3131,
+      "step": 3377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44220445658577895,
+      "learning_rate": 1.8621539373050218e-05,
+      "loss": 0.3288,
+      "step": 3378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7299981826510856,
+      "learning_rate": 1.862059639907844e-05,
+      "loss": 0.569,
+      "step": 3379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39958092415064894,
+      "learning_rate": 1.861965312657578e-05,
+      "loss": 0.2999,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3869027931462021,
+      "learning_rate": 1.8618709555574903e-05,
+      "loss": 0.277,
+      "step": 3381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3501211008611354,
+      "learning_rate": 1.8617765686108486e-05,
+      "loss": 0.2184,
+      "step": 3382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3614144002216407,
+      "learning_rate": 1.8616821518209213e-05,
+      "loss": 0.2989,
+      "step": 3383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8955308958623391,
+      "learning_rate": 1.8615877051909783e-05,
+      "loss": 0.4199,
+      "step": 3384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.420971422861682,
+      "learning_rate": 1.8614932287242897e-05,
+      "loss": 0.3649,
+      "step": 3385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39696752016884945,
+      "learning_rate": 1.8613987224241283e-05,
+      "loss": 0.2856,
+      "step": 3386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1534536864598737,
+      "learning_rate": 1.8613041862937656e-05,
+      "loss": 0.6503,
+      "step": 3387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26372459169274415,
+      "learning_rate": 1.861209620336476e-05,
+      "loss": 0.0746,
+      "step": 3388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5163757820951047,
+      "learning_rate": 1.8611150245555345e-05,
+      "loss": 0.3125,
+      "step": 3389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5196639264351134,
+      "learning_rate": 1.861020398954217e-05,
+      "loss": 0.3546,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8214877192095381,
+      "learning_rate": 1.8609257435357995e-05,
+      "loss": 0.387,
+      "step": 3391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.45946932605048707,
+      "learning_rate": 1.8608310583035607e-05,
+      "loss": 0.3055,
+      "step": 3392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.452824130541817,
+      "learning_rate": 1.8607363432607793e-05,
+      "loss": 0.3342,
+      "step": 3393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33934441895880785,
+      "learning_rate": 1.8606415984107357e-05,
+      "loss": 0.1221,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4336853637931559,
+      "learning_rate": 1.8605468237567103e-05,
+      "loss": 0.2819,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0742880039688079,
+      "learning_rate": 1.8604520193019855e-05,
+      "loss": 0.5185,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39427788076090214,
+      "learning_rate": 1.860357185049844e-05,
+      "loss": 0.2925,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39690461047371556,
+      "learning_rate": 1.8602623210035707e-05,
+      "loss": 0.307,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1140528804897003,
+      "learning_rate": 1.8601674271664497e-05,
+      "loss": 0.7088,
+      "step": 3399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2886567836009131,
+      "learning_rate": 1.8600725035417678e-05,
+      "loss": 0.1846,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3227616596528429,
+      "learning_rate": 1.8599775501328125e-05,
+      "loss": 0.2425,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.848213881735556,
+      "learning_rate": 1.8598825669428713e-05,
+      "loss": 0.473,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.837716610668294,
+      "learning_rate": 1.8597875539752337e-05,
+      "loss": 0.562,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33372430563348676,
+      "learning_rate": 1.85969251123319e-05,
+      "loss": 0.2439,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49703820168497725,
+      "learning_rate": 1.859597438720032e-05,
+      "loss": 0.3647,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4523814047602288,
+      "learning_rate": 1.8595023364390515e-05,
+      "loss": 0.3094,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3932121628983667,
+      "learning_rate": 1.8594072043935418e-05,
+      "loss": 0.241,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.530507341323851,
+      "learning_rate": 1.8593120425867977e-05,
+      "loss": 0.3333,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45189749977195065,
+      "learning_rate": 1.859216851022115e-05,
+      "loss": 0.3763,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31896053252137,
+      "learning_rate": 1.859121629702789e-05,
+      "loss": 0.2049,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4510347587273766,
+      "learning_rate": 1.8590263786321182e-05,
+      "loss": 0.7991,
+      "step": 3411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5061445451316303,
+      "learning_rate": 1.858931097813401e-05,
+      "loss": 0.4186,
+      "step": 3412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35254918453174994,
+      "learning_rate": 1.8588357872499364e-05,
+      "loss": 0.2737,
+      "step": 3413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30901119829012413,
+      "learning_rate": 1.8587404469450256e-05,
+      "loss": 0.1988,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1748732561056525,
+      "learning_rate": 1.85864507690197e-05,
+      "loss": 0.7539,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4472144634212105,
+      "learning_rate": 1.8585496771240726e-05,
+      "loss": 0.3451,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3288085382989217,
+      "learning_rate": 1.8584542476146364e-05,
+      "loss": 0.2777,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.701787075717411,
+      "learning_rate": 1.8583587883769668e-05,
+      "loss": 0.5152,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4124720953337286,
+      "learning_rate": 1.8582632994143693e-05,
+      "loss": 0.3517,
+      "step": 3419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26448985422784416,
+      "learning_rate": 1.8581677807301507e-05,
+      "loss": 0.1267,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4416407925509442,
+      "learning_rate": 1.8580722323276186e-05,
+      "loss": 0.35,
+      "step": 3421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36003377851432783,
+      "learning_rate": 1.857976654210082e-05,
+      "loss": 0.2706,
+      "step": 3422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0498458699375062,
+      "learning_rate": 1.857881046380851e-05,
+      "loss": 0.3784,
+      "step": 3423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3688550683980165,
+      "learning_rate": 1.8577854088432355e-05,
+      "loss": 0.3505,
+      "step": 3424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.44942001956730154,
+      "learning_rate": 1.8576897416005487e-05,
+      "loss": 0.3227,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6148292590435895,
+      "learning_rate": 1.857594044656103e-05,
+      "loss": 0.3691,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26088148061349903,
+      "learning_rate": 1.8574983180132128e-05,
+      "loss": 0.139,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6143931862134248,
+      "learning_rate": 1.8574025616751923e-05,
+      "loss": 0.2772,
+      "step": 3428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40106184091905833,
+      "learning_rate": 1.8573067756453578e-05,
+      "loss": 0.312,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6215169188275277,
+      "learning_rate": 1.8572109599270266e-05,
+      "loss": 0.4106,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4565998521916527,
+      "learning_rate": 1.857115114523517e-05,
+      "loss": 0.3022,
+      "step": 3431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3754630031717685,
+      "learning_rate": 1.857019239438148e-05,
+      "loss": 0.2459,
+      "step": 3432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3584316941904439,
+      "learning_rate": 1.8569233346742392e-05,
+      "loss": 0.2165,
+      "step": 3433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4805254322830583,
+      "learning_rate": 1.856827400235112e-05,
+      "loss": 0.3607,
+      "step": 3434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5382648282007101,
+      "learning_rate": 1.8567314361240893e-05,
+      "loss": 0.3951,
+      "step": 3435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4865226514408291,
+      "learning_rate": 1.8566354423444933e-05,
+      "loss": 0.3781,
+      "step": 3436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.358374646552267,
+      "learning_rate": 1.856539418899649e-05,
+      "loss": 0.2488,
+      "step": 3437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5303949361998881,
+      "learning_rate": 1.8564433657928815e-05,
+      "loss": 0.3902,
+      "step": 3438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4547723658493619,
+      "learning_rate": 1.8563472830275172e-05,
+      "loss": 0.3408,
+      "step": 3439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.270919077914169,
+      "learning_rate": 1.856251170606883e-05,
+      "loss": 0.1925,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38320462625120033,
+      "learning_rate": 1.8561550285343077e-05,
+      "loss": 0.3415,
+      "step": 3441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7710149460976287,
+      "learning_rate": 1.8560588568131205e-05,
+      "loss": 0.5157,
+      "step": 3442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45201487807738494,
+      "learning_rate": 1.8559626554466523e-05,
+      "loss": 0.1742,
+      "step": 3443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6001960036849652,
+      "learning_rate": 1.8558664244382338e-05,
+      "loss": 0.3938,
+      "step": 3444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4399877397096689,
+      "learning_rate": 1.8557701637911978e-05,
+      "loss": 0.3467,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6694405301341718,
+      "learning_rate": 1.855673873508878e-05,
+      "loss": 0.2563,
+      "step": 3446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5429342439640811,
+      "learning_rate": 1.855577553594609e-05,
+      "loss": 0.3684,
+      "step": 3447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3009787386986737,
+      "learning_rate": 1.8554812040517255e-05,
+      "loss": 0.2322,
+      "step": 3448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.782021777816558,
+      "learning_rate": 1.855384824883565e-05,
+      "loss": 0.4354,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3773608366203985,
+      "learning_rate": 1.8552884160934647e-05,
+      "loss": 0.2431,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2925400215861826,
+      "learning_rate": 1.8551919776847634e-05,
+      "loss": 0.837,
+      "step": 3451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43774430556639365,
+      "learning_rate": 1.8550955096608007e-05,
+      "loss": 0.3146,
+      "step": 3452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4075921947601884,
+      "learning_rate": 1.8549990120249174e-05,
+      "loss": 0.3066,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3309237068426338,
+      "learning_rate": 1.8549024847804547e-05,
+      "loss": 0.1845,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5641842287581473,
+      "learning_rate": 1.854805927930756e-05,
+      "loss": 0.3114,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4169525329222009,
+      "learning_rate": 1.854709341479165e-05,
+      "loss": 0.2763,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49289840010942465,
+      "learning_rate": 1.8546127254290257e-05,
+      "loss": 0.3636,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5541013067086684,
+      "learning_rate": 1.8545160797836847e-05,
+      "loss": 0.4148,
+      "step": 3458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4141408240338715,
+      "learning_rate": 1.8544194045464888e-05,
+      "loss": 0.2903,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28105441189183067,
+      "learning_rate": 1.8543226997207854e-05,
+      "loss": 0.2312,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6534352410994891,
+      "learning_rate": 1.8542259653099236e-05,
+      "loss": 0.2852,
+      "step": 3461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4909083697020356,
+      "learning_rate": 1.8541292013172538e-05,
+      "loss": 0.3315,
+      "step": 3462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.924048730821875,
+      "learning_rate": 1.854032407746126e-05,
+      "loss": 0.4172,
+      "step": 3463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36777627302291016,
+      "learning_rate": 1.853935584599893e-05,
+      "loss": 0.2993,
+      "step": 3464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4041852347714912,
+      "learning_rate": 1.8538387318819074e-05,
+      "loss": 0.3167,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2281237706488959,
+      "learning_rate": 1.853741849595523e-05,
+      "loss": 0.0881,
+      "step": 3466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8221586813021399,
+      "learning_rate": 1.853644937744095e-05,
+      "loss": 0.486,
+      "step": 3467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3859475003282859,
+      "learning_rate": 1.8535479963309796e-05,
+      "loss": 0.313,
+      "step": 3468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6530897529329045,
+      "learning_rate": 1.853451025359534e-05,
+      "loss": 0.3369,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.042977278524951,
+      "learning_rate": 1.8533540248331162e-05,
+      "loss": 0.4573,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41146027977536204,
+      "learning_rate": 1.8532569947550846e-05,
+      "loss": 0.3255,
+      "step": 3471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35765902247522974,
+      "learning_rate": 1.8531599351288007e-05,
+      "loss": 0.2252,
+      "step": 3472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3855611438390921,
+      "learning_rate": 1.8530628459576243e-05,
+      "loss": 0.2276,
+      "step": 3473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5204397928422537,
+      "learning_rate": 1.8529657272449186e-05,
+      "loss": 0.3064,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.810437510701928,
+      "learning_rate": 1.8528685789940463e-05,
+      "loss": 0.5076,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40869087933670045,
+      "learning_rate": 1.8527714012083718e-05,
+      "loss": 0.2685,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43618505181314465,
+      "learning_rate": 1.8526741938912605e-05,
+      "loss": 0.3277,
+      "step": 3477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.42187847427639297,
+      "learning_rate": 1.8525769570460783e-05,
+      "loss": 0.2454,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3165643524146173,
+      "learning_rate": 1.8524796906761928e-05,
+      "loss": 0.1485,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.46732317429125,
+      "learning_rate": 1.8523823947849722e-05,
+      "loss": 0.2923,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1275700154976764,
+      "learning_rate": 1.8522850693757865e-05,
+      "loss": 0.4158,
+      "step": 3481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7907367906006263,
+      "learning_rate": 1.8521877144520047e-05,
+      "loss": 0.2819,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.44558259936120503,
+      "learning_rate": 1.8520903300169993e-05,
+      "loss": 0.3128,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31241687566734866,
+      "learning_rate": 1.8519929160741427e-05,
+      "loss": 0.2827,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31618469870226495,
+      "learning_rate": 1.8518954726268076e-05,
+      "loss": 0.1135,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3964958438859382,
+      "learning_rate": 1.851797999678369e-05,
+      "loss": 0.3282,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8788401444161869,
+      "learning_rate": 1.8517004972322022e-05,
+      "loss": 0.5349,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5842037117974624,
+      "learning_rate": 1.851602965291684e-05,
+      "loss": 0.3523,
+      "step": 3488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4250093402378405,
+      "learning_rate": 1.851505403860192e-05,
+      "loss": 0.2307,
+      "step": 3489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1648122285456606,
+      "learning_rate": 1.8514078129411045e-05,
+      "loss": 0.5946,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33870419487266124,
+      "learning_rate": 1.8513101925378006e-05,
+      "loss": 0.2547,
+      "step": 3491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3289610462955457,
+      "learning_rate": 1.8512125426536617e-05,
+      "loss": 0.2312,
+      "step": 3492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7856786137489506,
+      "learning_rate": 1.8511148632920685e-05,
+      "loss": 0.4921,
+      "step": 3493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9243725044925123,
+      "learning_rate": 1.851017154456405e-05,
+      "loss": 0.5751,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38206441570861877,
+      "learning_rate": 1.8509194161500536e-05,
+      "loss": 0.24,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4583245897267825,
+      "learning_rate": 1.8508216483763993e-05,
+      "loss": 0.3524,
+      "step": 3496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2942176231528043,
+      "learning_rate": 1.850723851138828e-05,
+      "loss": 0.2291,
+      "step": 3497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3675286195522216,
+      "learning_rate": 1.850626024440726e-05,
+      "loss": 0.2646,
+      "step": 3498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4881052772922558,
+      "learning_rate": 1.850528168285482e-05,
+      "loss": 0.4122,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5091923599798945,
+      "learning_rate": 1.8504302826764835e-05,
+      "loss": 0.3891,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3633056005786531,
+      "learning_rate": 1.8503323676171212e-05,
+      "loss": 0.2949,
+      "step": 3501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5991201857527949,
+      "learning_rate": 1.8502344231107855e-05,
+      "loss": 0.3974,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40415730403043115,
+      "learning_rate": 1.8501364491608683e-05,
+      "loss": 0.3875,
+      "step": 3503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34569731298708684,
+      "learning_rate": 1.8500384457707625e-05,
+      "loss": 0.285,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2568026864658739,
+      "learning_rate": 1.8499404129438617e-05,
+      "loss": 0.1563,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1707986100418522,
+      "learning_rate": 1.8498423506835613e-05,
+      "loss": 0.7772,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38602275849489903,
+      "learning_rate": 1.8497442589932568e-05,
+      "loss": 0.2805,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4079578462025017,
+      "learning_rate": 1.8496461378763445e-05,
+      "loss": 0.2396,
+      "step": 3508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5885481273942008,
+      "learning_rate": 1.8495479873362237e-05,
+      "loss": 0.4813,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28505728074008135,
+      "learning_rate": 1.8494498073762924e-05,
+      "loss": 0.2308,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3259515702055285,
+      "learning_rate": 1.8493515979999508e-05,
+      "loss": 0.2147,
+      "step": 3511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5081829044842984,
+      "learning_rate": 1.8492533592105998e-05,
+      "loss": 0.3786,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40211123448914077,
+      "learning_rate": 1.8491550910116415e-05,
+      "loss": 0.2937,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5797803004312481,
+      "learning_rate": 1.8490567934064788e-05,
+      "loss": 0.3893,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35474516854979343,
+      "learning_rate": 1.848958466398516e-05,
+      "loss": 0.311,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3743499556350666,
+      "learning_rate": 1.8488601099911582e-05,
+      "loss": 0.2879,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34838342234863295,
+      "learning_rate": 1.8487617241878114e-05,
+      "loss": 0.2501,
+      "step": 3517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9330455844434704,
+      "learning_rate": 1.8486633089918823e-05,
+      "loss": 0.5301,
+      "step": 3518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3876398473412494,
+      "learning_rate": 1.848564864406779e-05,
+      "loss": 0.2736,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38224800897615363,
+      "learning_rate": 1.8484663904359112e-05,
+      "loss": 0.3232,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7342180011930438,
+      "learning_rate": 1.848367887082689e-05,
+      "loss": 0.4027,
+      "step": 3521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37253887582059153,
+      "learning_rate": 1.848269354350523e-05,
+      "loss": 0.2922,
+      "step": 3522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4486634748559754,
+      "learning_rate": 1.848170792242826e-05,
+      "loss": 0.3188,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41124068257992336,
+      "learning_rate": 1.848072200763011e-05,
+      "loss": 0.3629,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.252569055610881,
+      "learning_rate": 1.8479735799144917e-05,
+      "loss": 0.1877,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6926210684337959,
+      "learning_rate": 1.847874929700684e-05,
+      "loss": 0.5067,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5730030339821481,
+      "learning_rate": 1.8477762501250046e-05,
+      "loss": 0.3562,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36594434314465024,
+      "learning_rate": 1.847677541190869e-05,
+      "loss": 0.2255,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5439482028879972,
+      "learning_rate": 1.8475788029016974e-05,
+      "loss": 0.3023,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5147312308129324,
+      "learning_rate": 1.847480035260908e-05,
+      "loss": 0.3325,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.396855562434019,
+      "learning_rate": 1.8473812382719215e-05,
+      "loss": 0.2331,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3650937303546053,
+      "learning_rate": 1.8472824119381592e-05,
+      "loss": 0.3324,
+      "step": 3532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8069519781469283,
+      "learning_rate": 1.8471835562630435e-05,
+      "loss": 0.5369,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3224016264255935,
+      "learning_rate": 1.8470846712499977e-05,
+      "loss": 0.2151,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49887651442034286,
+      "learning_rate": 1.846985756902446e-05,
+      "loss": 0.4021,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.372999385227418,
+      "learning_rate": 1.8468868132238138e-05,
+      "loss": 0.3151,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4155513944432335,
+      "learning_rate": 1.8467878402175278e-05,
+      "loss": 0.3341,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38115414788893204,
+      "learning_rate": 1.8466888378870155e-05,
+      "loss": 0.2816,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3971013745780681,
+      "learning_rate": 1.8465898062357048e-05,
+      "loss": 0.3233,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.405786263334606,
+      "learning_rate": 1.846490745267026e-05,
+      "loss": 0.3123,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.524142882921503,
+      "learning_rate": 1.8463916549844084e-05,
+      "loss": 0.2972,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43131354308913267,
+      "learning_rate": 1.846292535391285e-05,
+      "loss": 0.3471,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3522374006752674,
+      "learning_rate": 1.846193386491087e-05,
+      "loss": 0.316,
+      "step": 3543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4367897212599001,
+      "learning_rate": 1.846094208287248e-05,
+      "loss": 0.3275,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4318912036953102,
+      "learning_rate": 1.845995000783204e-05,
+      "loss": 0.2809,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37576132450278826,
+      "learning_rate": 1.8458957639823887e-05,
+      "loss": 0.271,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.44073497833225533,
+      "learning_rate": 1.8457964978882397e-05,
+      "loss": 0.2809,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4402566585814481,
+      "learning_rate": 1.8456972025041943e-05,
+      "loss": 0.411,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5817384512608036,
+      "learning_rate": 1.845597877833691e-05,
+      "loss": 0.3939,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.347563565667777,
+      "learning_rate": 1.84549852388017e-05,
+      "loss": 0.2921,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3335895526495127,
+      "learning_rate": 1.845399140647071e-05,
+      "loss": 0.2274,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7074274468152657,
+      "learning_rate": 1.8452997281378364e-05,
+      "loss": 0.3816,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4984396584595269,
+      "learning_rate": 1.8452002863559086e-05,
+      "loss": 0.3568,
+      "step": 3553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6284443435298703,
+      "learning_rate": 1.845100815304731e-05,
+      "loss": 0.4117,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3490991839784079,
+      "learning_rate": 1.845001314987749e-05,
+      "loss": 0.28,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4250866389209264,
+      "learning_rate": 1.8449017854084072e-05,
+      "loss": 0.3298,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3043618837957513,
+      "learning_rate": 1.844802226570153e-05,
+      "loss": 0.1234,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4379559170905638,
+      "learning_rate": 1.8447026384764343e-05,
+      "loss": 0.3207,
+      "step": 3558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3965722955626275,
+      "learning_rate": 1.8446030211306993e-05,
+      "loss": 0.3391,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.024955890124149,
+      "learning_rate": 1.844503374536398e-05,
+      "loss": 0.427,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41957172094011896,
+      "learning_rate": 1.8444036986969814e-05,
+      "loss": 0.336,
+      "step": 3561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5542877434696505,
+      "learning_rate": 1.8443039936159007e-05,
+      "loss": 0.3461,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2729193047980405,
+      "learning_rate": 1.8442042592966095e-05,
+      "loss": 0.2327,
+      "step": 3563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3549744193912546,
+      "learning_rate": 1.8441044957425608e-05,
+      "loss": 0.1783,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5966312774507957,
+      "learning_rate": 1.8440047029572094e-05,
+      "loss": 0.41,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0779161799481367,
+      "learning_rate": 1.843904880944012e-05,
+      "loss": 0.5309,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3425507945907625,
+      "learning_rate": 1.843805029706425e-05,
+      "loss": 0.2269,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43024803622830704,
+      "learning_rate": 1.8437051492479053e-05,
+      "loss": 0.3461,
+      "step": 3568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.32182440028828346,
+      "learning_rate": 1.843605239571913e-05,
+      "loss": 0.1903,
+      "step": 3569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.43595402051191434,
+      "learning_rate": 1.8435053006819073e-05,
+      "loss": 0.1819,
+      "step": 3570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4289935360872228,
+      "learning_rate": 1.8434053325813495e-05,
+      "loss": 0.3026,
+      "step": 3571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.259600352356555,
+      "learning_rate": 1.8433053352737014e-05,
+      "loss": 0.5352,
+      "step": 3572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46454768580118233,
+      "learning_rate": 1.8432053087624258e-05,
+      "loss": 0.2123,
+      "step": 3573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3683549795511345,
+      "learning_rate": 1.8431052530509866e-05,
+      "loss": 0.2625,
+      "step": 3574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3434082020790198,
+      "learning_rate": 1.8430051681428486e-05,
+      "loss": 0.2504,
+      "step": 3575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6314261221401793,
+      "learning_rate": 1.842905054041478e-05,
+      "loss": 0.3683,
+      "step": 3576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7178093309482125,
+      "learning_rate": 1.8428049107503417e-05,
+      "loss": 0.3267,
+      "step": 3577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1845693433728908,
+      "learning_rate": 1.8427047382729073e-05,
+      "loss": 0.606,
+      "step": 3578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3857144192125777,
+      "learning_rate": 1.842604536612644e-05,
+      "loss": 0.3092,
+      "step": 3579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.422890021608481,
+      "learning_rate": 1.842504305773022e-05,
+      "loss": 0.2626,
+      "step": 3580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.319692052216907,
+      "learning_rate": 1.8424040457575124e-05,
+      "loss": 0.1854,
+      "step": 3581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.609801137363293,
+      "learning_rate": 1.8423037565695864e-05,
+      "loss": 0.3804,
+      "step": 3582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3952230768238028,
+      "learning_rate": 1.842203438212718e-05,
+      "loss": 0.2583,
+      "step": 3583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9995316109812836,
+      "learning_rate": 1.8421030906903805e-05,
+      "loss": 0.5573,
+      "step": 3584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8080682129475155,
+      "learning_rate": 1.8420027140060493e-05,
+      "loss": 0.464,
+      "step": 3585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38502970294226097,
+      "learning_rate": 1.8419023081632e-05,
+      "loss": 0.2237,
+      "step": 3586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3003903655050074,
+      "learning_rate": 1.8418018731653106e-05,
+      "loss": 0.2376,
+      "step": 3587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8598542283516147,
+      "learning_rate": 1.841701409015858e-05,
+      "loss": 0.4417,
+      "step": 3588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4722555501097505,
+      "learning_rate": 1.841600915718322e-05,
+      "loss": 0.3618,
+      "step": 3589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8780361475667137,
+      "learning_rate": 1.8415003932761823e-05,
+      "loss": 0.5438,
+      "step": 3590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36609700316137517,
+      "learning_rate": 1.8413998416929205e-05,
+      "loss": 0.3317,
+      "step": 3591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4232364971031735,
+      "learning_rate": 1.8412992609720183e-05,
+      "loss": 0.3091,
+      "step": 3592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2582320073159369,
+      "learning_rate": 1.8411986511169585e-05,
+      "loss": 0.0721,
+      "step": 3593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.615145234861627,
+      "learning_rate": 1.8410980121312258e-05,
+      "loss": 0.41,
+      "step": 3594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3981913852643038,
+      "learning_rate": 1.8409973440183054e-05,
+      "loss": 0.3132,
+      "step": 3595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2713381305155262,
+      "learning_rate": 1.840896646781683e-05,
+      "loss": 0.5107,
+      "step": 3596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5843553386069017,
+      "learning_rate": 1.8407959204248455e-05,
+      "loss": 0.3627,
+      "step": 3597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46489119063537054,
+      "learning_rate": 1.8406951649512817e-05,
+      "loss": 0.3248,
+      "step": 3598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47103184457362807,
+      "learning_rate": 1.8405943803644803e-05,
+      "loss": 0.3158,
+      "step": 3599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7760715327257453,
+      "learning_rate": 1.840493566667932e-05,
+      "loss": 0.3387,
+      "step": 3600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3663667134586227,
+      "learning_rate": 1.8403927238651274e-05,
+      "loss": 0.2674,
+      "step": 3601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5303169306539498,
+      "learning_rate": 1.8402918519595592e-05,
+      "loss": 0.3238,
+      "step": 3602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47049291683057637,
+      "learning_rate": 1.8401909509547196e-05,
+      "loss": 0.2613,
+      "step": 3603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4636995547912467,
+      "learning_rate": 1.8400900208541045e-05,
+      "loss": 0.3415,
+      "step": 3604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.610816875773527,
+      "learning_rate": 1.8399890616612073e-05,
+      "loss": 0.4726,
+      "step": 3605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3932357164783332,
+      "learning_rate": 1.8398880733795253e-05,
+      "loss": 0.2688,
+      "step": 3606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33115183414057575,
+      "learning_rate": 1.8397870560125554e-05,
+      "loss": 0.2524,
+      "step": 3607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5466588959281407,
+      "learning_rate": 1.839686009563796e-05,
+      "loss": 0.368,
+      "step": 3608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3873845994650159,
+      "learning_rate": 1.839584934036746e-05,
+      "loss": 0.0684,
+      "step": 3609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5054505521879232,
+      "learning_rate": 1.8394838294349058e-05,
+      "loss": 0.3073,
+      "step": 3610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.595052581816832,
+      "learning_rate": 1.839382695761777e-05,
+      "loss": 0.3465,
+      "step": 3611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1129720778207792,
+      "learning_rate": 1.839281533020861e-05,
+      "loss": 0.4943,
+      "step": 3612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33785668181825734,
+      "learning_rate": 1.839180341215662e-05,
+      "loss": 0.1835,
+      "step": 3613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3938903417041624,
+      "learning_rate": 1.8390791203496842e-05,
+      "loss": 0.8968,
+      "step": 3614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3179001508479327,
+      "learning_rate": 1.838977870426432e-05,
+      "loss": 0.2777,
+      "step": 3615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38387065537886916,
+      "learning_rate": 1.8388765914494124e-05,
+      "loss": 0.1989,
+      "step": 3616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9955530352807715,
+      "learning_rate": 1.8387752834221326e-05,
+      "loss": 0.4931,
+      "step": 3617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45334877775759963,
+      "learning_rate": 1.8386739463481004e-05,
+      "loss": 0.2833,
+      "step": 3618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4419384557222577,
+      "learning_rate": 1.838572580230826e-05,
+      "loss": 0.2239,
+      "step": 3619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2042963328857403,
+      "learning_rate": 1.838471185073819e-05,
+      "loss": 0.7506,
+      "step": 3620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28333197613712585,
+      "learning_rate": 1.8383697608805907e-05,
+      "loss": 0.2277,
+      "step": 3621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35663776056728297,
+      "learning_rate": 1.838268307654654e-05,
+      "loss": 0.2561,
+      "step": 3622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38137141453365087,
+      "learning_rate": 1.8381668253995216e-05,
+      "loss": 0.3414,
+      "step": 3623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3007390949271715,
+      "learning_rate": 1.8380653141187084e-05,
+      "loss": 0.5905,
+      "step": 3624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3586717879521889,
+      "learning_rate": 1.837963773815729e-05,
+      "loss": 0.2828,
+      "step": 3625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4624648099073319,
+      "learning_rate": 1.8378622044941007e-05,
+      "loss": 0.2992,
+      "step": 3626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28370620018557996,
+      "learning_rate": 1.8377606061573398e-05,
+      "loss": 0.1702,
+      "step": 3627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4113606179544061,
+      "learning_rate": 1.8376589788089655e-05,
+      "loss": 0.2951,
+      "step": 3628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.771704015761994,
+      "learning_rate": 1.837557322452496e-05,
+      "loss": 0.4089,
+      "step": 3629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4695330219200323,
+      "learning_rate": 1.8374556370914533e-05,
+      "loss": 0.3328,
+      "step": 3630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3891106792486114,
+      "learning_rate": 1.8373539227293576e-05,
+      "loss": 0.2798,
+      "step": 3631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9231594968954966,
+      "learning_rate": 1.8372521793697317e-05,
+      "loss": 0.5186,
+      "step": 3632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5334306794574006,
+      "learning_rate": 1.8371504070160985e-05,
+      "loss": 0.3294,
+      "step": 3633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4208129570730506,
+      "learning_rate": 1.8370486056719828e-05,
+      "loss": 0.2981,
+      "step": 3634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3440651749124253,
+      "learning_rate": 1.8369467753409102e-05,
+      "loss": 0.2406,
+      "step": 3635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0039452445703967,
+      "learning_rate": 1.8368449160264064e-05,
+      "loss": 0.5453,
+      "step": 3636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44073239184364055,
+      "learning_rate": 1.836743027731999e-05,
+      "loss": 0.3189,
+      "step": 3637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40024632916236896,
+      "learning_rate": 1.8366411104612168e-05,
+      "loss": 0.3527,
+      "step": 3638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4302314709355559,
+      "learning_rate": 1.8365391642175892e-05,
+      "loss": 0.3616,
+      "step": 3639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3902773433049301,
+      "learning_rate": 1.836437189004646e-05,
+      "loss": 0.2706,
+      "step": 3640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3034132980546468,
+      "learning_rate": 1.836335184825919e-05,
+      "loss": 0.2261,
+      "step": 3641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4224483560026623,
+      "learning_rate": 1.8362331516849405e-05,
+      "loss": 0.2859,
+      "step": 3642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3997927453552849,
+      "learning_rate": 1.8361310895852437e-05,
+      "loss": 0.2909,
+      "step": 3643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8073075120292302,
+      "learning_rate": 1.8360289985303637e-05,
+      "loss": 0.5745,
+      "step": 3644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5995548870672561,
+      "learning_rate": 1.8359268785238348e-05,
+      "loss": 0.4058,
+      "step": 3645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4135190586223736,
+      "learning_rate": 1.8358247295691946e-05,
+      "loss": 0.2841,
+      "step": 3646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2933792066466321,
+      "learning_rate": 1.8357225516699797e-05,
+      "loss": 0.2371,
+      "step": 3647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8306758061657067,
+      "learning_rate": 1.835620344829729e-05,
+      "loss": 0.3256,
+      "step": 3648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40272226138516737,
+      "learning_rate": 1.8355181090519814e-05,
+      "loss": 0.2961,
+      "step": 3649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7342811633805951,
+      "learning_rate": 1.8354158443402777e-05,
+      "loss": 0.4417,
+      "step": 3650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5508644077439324,
+      "learning_rate": 1.835313550698159e-05,
+      "loss": 0.3978,
+      "step": 3651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3936970414946991,
+      "learning_rate": 1.8352112281291683e-05,
+      "loss": 0.2251,
+      "step": 3652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30868325393237483,
+      "learning_rate": 1.8351088766368487e-05,
+      "loss": 0.1808,
+      "step": 3653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4717999561118713,
+      "learning_rate": 1.8350064962247443e-05,
+      "loss": 0.3688,
+      "step": 3654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3690124510455886,
+      "learning_rate": 1.8349040868964012e-05,
+      "loss": 0.1872,
+      "step": 3655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.516836148527607,
+      "learning_rate": 1.8348016486553653e-05,
+      "loss": 0.4212,
+      "step": 3656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4174942641382965,
+      "learning_rate": 1.8346991815051844e-05,
+      "loss": 0.4136,
+      "step": 3657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34246783846879647,
+      "learning_rate": 1.8345966854494065e-05,
+      "loss": 0.2218,
+      "step": 3658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29464001657216854,
+      "learning_rate": 1.8344941604915813e-05,
+      "loss": 0.2341,
+      "step": 3659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0963469580919971,
+      "learning_rate": 1.8343916066352593e-05,
+      "loss": 0.5888,
+      "step": 3660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39006202105175725,
+      "learning_rate": 1.834289023883992e-05,
+      "loss": 0.2244,
+      "step": 3661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45351816412520496,
+      "learning_rate": 1.8341864122413313e-05,
+      "loss": 0.3389,
+      "step": 3662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.013846492560218,
+      "learning_rate": 1.8340837717108312e-05,
+      "loss": 0.6265,
+      "step": 3663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44933729211335915,
+      "learning_rate": 1.8339811022960458e-05,
+      "loss": 0.3057,
+      "step": 3664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31780800408881466,
+      "learning_rate": 1.833878404000531e-05,
+      "loss": 0.1626,
+      "step": 3665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3529354991105561,
+      "learning_rate": 1.8337756768278425e-05,
+      "loss": 0.2648,
+      "step": 3666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4312121030833976,
+      "learning_rate": 1.8336729207815386e-05,
+      "loss": 0.2812,
+      "step": 3667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7354723969211455,
+      "learning_rate": 1.833570135865177e-05,
+      "loss": 0.3566,
+      "step": 3668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0099131930682428,
+      "learning_rate": 1.8334673220823175e-05,
+      "loss": 0.6111,
+      "step": 3669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31425998270147343,
+      "learning_rate": 1.8333644794365205e-05,
+      "loss": 0.27,
+      "step": 3670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3286391464636836,
+      "learning_rate": 1.8332616079313473e-05,
+      "loss": 0.2264,
+      "step": 3671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47239060853182213,
+      "learning_rate": 1.8331587075703607e-05,
+      "loss": 0.302,
+      "step": 3672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5911532765330798,
+      "learning_rate": 1.833055778357124e-05,
+      "loss": 0.3735,
+      "step": 3673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3816260176156264,
+      "learning_rate": 1.8329528202952013e-05,
+      "loss": 0.2915,
+      "step": 3674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2602046875475506,
+      "learning_rate": 1.832849833388158e-05,
+      "loss": 0.8317,
+      "step": 3675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5802384065602598,
+      "learning_rate": 1.8327468176395614e-05,
+      "loss": 0.378,
+      "step": 3676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3233690864917701,
+      "learning_rate": 1.8326437730529778e-05,
+      "loss": 0.2625,
+      "step": 3677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30650130488543587,
+      "learning_rate": 1.8325406996319762e-05,
+      "loss": 0.2177,
+      "step": 3678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4361961196440014,
+      "learning_rate": 1.8324375973801262e-05,
+      "loss": 0.3167,
+      "step": 3679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.540582742403159,
+      "learning_rate": 1.8323344663009976e-05,
+      "loss": 0.4255,
+      "step": 3680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.26530181020778,
+      "learning_rate": 1.8322313063981628e-05,
+      "loss": 0.5116,
+      "step": 3681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3019632406735647,
+      "learning_rate": 1.8321281176751932e-05,
+      "loss": 0.2601,
+      "step": 3682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5169685558732896,
+      "learning_rate": 1.8320249001356627e-05,
+      "loss": 0.3703,
+      "step": 3683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40742847831985723,
+      "learning_rate": 1.831921653783146e-05,
+      "loss": 0.1924,
+      "step": 3684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5064412702318805,
+      "learning_rate": 1.8318183786212177e-05,
+      "loss": 0.3476,
+      "step": 3685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39189664074458824,
+      "learning_rate": 1.8317150746534553e-05,
+      "loss": 0.3303,
+      "step": 3686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.386780392774275,
+      "learning_rate": 1.8316117418834352e-05,
+      "loss": 0.8824,
+      "step": 3687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3476158374729174,
+      "learning_rate": 1.831508380314736e-05,
+      "loss": 0.1713,
+      "step": 3688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5407610308721493,
+      "learning_rate": 1.831404989950938e-05,
+      "loss": 0.4049,
+      "step": 3689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34332848216555034,
+      "learning_rate": 1.8313015707956205e-05,
+      "loss": 0.2995,
+      "step": 3690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4909483803893434,
+      "learning_rate": 1.831198122852366e-05,
+      "loss": 0.2905,
+      "step": 3691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34866275708306377,
+      "learning_rate": 1.8310946461247553e-05,
+      "loss": 0.2568,
+      "step": 3692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5144534607864611,
+      "learning_rate": 1.8309911406163736e-05,
+      "loss": 0.3944,
+      "step": 3693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4451881757530133,
+      "learning_rate": 1.830887606330804e-05,
+      "loss": 0.2252,
+      "step": 3694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46182446045835396,
+      "learning_rate": 1.8307840432716323e-05,
+      "loss": 0.3298,
+      "step": 3695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7256437850629539,
+      "learning_rate": 1.830680451442445e-05,
+      "loss": 0.4658,
+      "step": 3696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31246873377998513,
+      "learning_rate": 1.8305768308468294e-05,
+      "loss": 0.1741,
+      "step": 3697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3792276788301899,
+      "learning_rate": 1.830473181488374e-05,
+      "loss": 0.2723,
+      "step": 3698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4475052756468612,
+      "learning_rate": 1.8303695033706675e-05,
+      "loss": 0.8954,
+      "step": 3699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4832360785231713,
+      "learning_rate": 1.8302657964973014e-05,
+      "loss": 0.3362,
+      "step": 3700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38449186292934234,
+      "learning_rate": 1.830162060871866e-05,
+      "loss": 0.2838,
+      "step": 3701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4419538620503951,
+      "learning_rate": 1.8300582964979544e-05,
+      "loss": 0.3564,
+      "step": 3702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36778425880138244,
+      "learning_rate": 1.8299545033791596e-05,
+      "loss": 0.2228,
+      "step": 3703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30802703243687346,
+      "learning_rate": 1.829850681519076e-05,
+      "loss": 0.1985,
+      "step": 3704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6655387140633007,
+      "learning_rate": 1.8297468309212994e-05,
+      "loss": 0.4198,
+      "step": 3705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45041933695514075,
+      "learning_rate": 1.8296429515894255e-05,
+      "loss": 0.3048,
+      "step": 3706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41420386663343106,
+      "learning_rate": 1.8295390435270516e-05,
+      "loss": 0.2809,
+      "step": 3707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6850026949422853,
+      "learning_rate": 1.8294351067377762e-05,
+      "loss": 0.4826,
+      "step": 3708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3475258356965287,
+      "learning_rate": 1.829331141225199e-05,
+      "loss": 0.2416,
+      "step": 3709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39414644413001365,
+      "learning_rate": 1.8292271469929202e-05,
+      "loss": 0.2489,
+      "step": 3710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46620981413281065,
+      "learning_rate": 1.829123124044541e-05,
+      "loss": 0.3112,
+      "step": 3711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8393024501634222,
+      "learning_rate": 1.8290190723836632e-05,
+      "loss": 0.4927,
+      "step": 3712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5442988866000577,
+      "learning_rate": 1.828914992013891e-05,
+      "loss": 0.3368,
+      "step": 3713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5338386638537052,
+      "learning_rate": 1.828810882938828e-05,
+      "loss": 0.2955,
+      "step": 3714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7721046804040257,
+      "learning_rate": 1.8287067451620796e-05,
+      "loss": 0.4498,
+      "step": 3715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39359198704409026,
+      "learning_rate": 1.8286025786872526e-05,
+      "loss": 0.3215,
+      "step": 3716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.309872202825924,
+      "learning_rate": 1.8284983835179536e-05,
+      "loss": 0.2202,
+      "step": 3717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3891445865569338,
+      "learning_rate": 1.8283941596577917e-05,
+      "loss": 0.2658,
+      "step": 3718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5257520788875862,
+      "learning_rate": 1.8282899071103755e-05,
+      "loss": 0.3476,
+      "step": 3719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9016773666981592,
+      "learning_rate": 1.8281856258793155e-05,
+      "loss": 0.364,
+      "step": 3720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40787992390494676,
+      "learning_rate": 1.8280813159682226e-05,
+      "loss": 0.3125,
+      "step": 3721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39215624227848944,
+      "learning_rate": 1.8279769773807094e-05,
+      "loss": 0.295,
+      "step": 3722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.19942256276475304,
+      "learning_rate": 1.8278726101203892e-05,
+      "loss": 0.1365,
+      "step": 3723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0897560592528546,
+      "learning_rate": 1.8277682141908763e-05,
+      "loss": 0.4017,
+      "step": 3724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45932270538036235,
+      "learning_rate": 1.8276637895957853e-05,
+      "loss": 0.331,
+      "step": 3725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38873202435458054,
+      "learning_rate": 1.827559336338733e-05,
+      "loss": 0.3411,
+      "step": 3726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7523879935802436,
+      "learning_rate": 1.8274548544233367e-05,
+      "loss": 0.2851,
+      "step": 3727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3509208008137019,
+      "learning_rate": 1.827350343853214e-05,
+      "loss": 0.2867,
+      "step": 3728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4536301444246971,
+      "learning_rate": 1.8272458046319848e-05,
+      "loss": 0.3437,
+      "step": 3729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.733104281686962,
+      "learning_rate": 1.8271412367632688e-05,
+      "loss": 0.4302,
+      "step": 3730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.294900982042248,
+      "learning_rate": 1.8270366402506872e-05,
+      "loss": 0.223,
+      "step": 3731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45382345157960424,
+      "learning_rate": 1.8269320150978625e-05,
+      "loss": 0.312,
+      "step": 3732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4332203720271532,
+      "learning_rate": 1.8268273613084177e-05,
+      "loss": 0.2502,
+      "step": 3733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4650150296518048,
+      "learning_rate": 1.826722678885977e-05,
+      "loss": 0.3157,
+      "step": 3734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.571029287761417,
+      "learning_rate": 1.8266179678341654e-05,
+      "loss": 0.7587,
+      "step": 3735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.048302589229643,
+      "learning_rate": 1.826513228156609e-05,
+      "loss": 0.3022,
+      "step": 3736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3862146200874309,
+      "learning_rate": 1.826408459856935e-05,
+      "loss": 0.2914,
+      "step": 3737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5222607559671937,
+      "learning_rate": 1.826303662938772e-05,
+      "loss": 0.2394,
+      "step": 3738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3267918828905096,
+      "learning_rate": 1.826198837405748e-05,
+      "loss": 0.709,
+      "step": 3739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39077333658956925,
+      "learning_rate": 1.8260939832614942e-05,
+      "loss": 0.2071,
+      "step": 3740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44790985884301965,
+      "learning_rate": 1.8259891005096414e-05,
+      "loss": 0.3655,
+      "step": 3741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.667719602409911,
+      "learning_rate": 1.8258841891538214e-05,
+      "loss": 0.4282,
+      "step": 3742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.516883514093155,
+      "learning_rate": 1.8257792491976676e-05,
+      "loss": 0.2252,
+      "step": 3743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42556191862763,
+      "learning_rate": 1.825674280644814e-05,
+      "loss": 0.192,
+      "step": 3744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42746714546213527,
+      "learning_rate": 1.8255692834988952e-05,
+      "loss": 0.3717,
+      "step": 3745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3455175925005218,
+      "learning_rate": 1.8254642577635478e-05,
+      "loss": 0.2232,
+      "step": 3746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5540250899869666,
+      "learning_rate": 1.8253592034424085e-05,
+      "loss": 0.5053,
+      "step": 3747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0421298320869656,
+      "learning_rate": 1.8252541205391155e-05,
+      "loss": 0.6174,
+      "step": 3748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44154945291697445,
+      "learning_rate": 1.825149009057308e-05,
+      "loss": 0.2326,
+      "step": 3749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3709396033941532,
+      "learning_rate": 1.8250438690006257e-05,
+      "loss": 0.2199,
+      "step": 3750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3866050733970163,
+      "learning_rate": 1.8249387003727097e-05,
+      "loss": 0.7559,
+      "step": 3751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.384792848713734,
+      "learning_rate": 1.824833503177202e-05,
+      "loss": 0.3101,
+      "step": 3752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3965591271182984,
+      "learning_rate": 1.8247282774177456e-05,
+      "loss": 0.3009,
+      "step": 3753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7462894283297303,
+      "learning_rate": 1.824623023097984e-05,
+      "loss": 0.4997,
+      "step": 3754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3943149843852122,
+      "learning_rate": 1.824517740221563e-05,
+      "loss": 0.2825,
+      "step": 3755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2177870980472017,
+      "learning_rate": 1.824412428792128e-05,
+      "loss": 0.1226,
+      "step": 3756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.499144612415864,
+      "learning_rate": 1.8243070888133262e-05,
+      "loss": 0.361,
+      "step": 3757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5001785109282243,
+      "learning_rate": 1.8242017202888053e-05,
+      "loss": 0.3159,
+      "step": 3758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7388789713211021,
+      "learning_rate": 1.824096323222214e-05,
+      "loss": 0.3556,
+      "step": 3759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.534150495177783,
+      "learning_rate": 1.8239908976172027e-05,
+      "loss": 0.4201,
+      "step": 3760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4950814070548895,
+      "learning_rate": 1.823885443477422e-05,
+      "loss": 0.2938,
+      "step": 3761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28446685490046,
+      "learning_rate": 1.8237799608065238e-05,
+      "loss": 0.2129,
+      "step": 3762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6083754544167314,
+      "learning_rate": 1.823674449608161e-05,
+      "loss": 0.405,
+      "step": 3763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3692492992721875,
+      "learning_rate": 1.8235689098859874e-05,
+      "loss": 0.2551,
+      "step": 3764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39185364731901257,
+      "learning_rate": 1.823463341643658e-05,
+      "loss": 0.3614,
+      "step": 3765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8683819485532979,
+      "learning_rate": 1.8233577448848283e-05,
+      "loss": 0.4439,
+      "step": 3766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3771140986653881,
+      "learning_rate": 1.8232521196131552e-05,
+      "loss": 0.3055,
+      "step": 3767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35751520286149197,
+      "learning_rate": 1.823146465832297e-05,
+      "loss": 0.2238,
+      "step": 3768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4560483343446017,
+      "learning_rate": 1.823040783545912e-05,
+      "loss": 0.309,
+      "step": 3769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35578254331441966,
+      "learning_rate": 1.8229350727576597e-05,
+      "loss": 0.2929,
+      "step": 3770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7910099204556784,
+      "learning_rate": 1.8228293334712015e-05,
+      "loss": 0.5764,
+      "step": 3771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47524677083124306,
+      "learning_rate": 1.822723565690199e-05,
+      "loss": 0.3022,
+      "step": 3772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3292122348325114,
+      "learning_rate": 1.8226177694183144e-05,
+      "loss": 0.2735,
+      "step": 3773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5184964716025334,
+      "learning_rate": 1.8225119446592122e-05,
+      "loss": 0.2828,
+      "step": 3774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3960302570284166,
+      "learning_rate": 1.8224060914165564e-05,
+      "loss": 0.3376,
+      "step": 3775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3777909633108905,
+      "learning_rate": 1.8223002096940133e-05,
+      "loss": 0.2277,
+      "step": 3776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3777145561989262,
+      "learning_rate": 1.822194299495249e-05,
+      "loss": 0.334,
+      "step": 3777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9906118795976717,
+      "learning_rate": 1.8220883608239317e-05,
+      "loss": 0.6105,
+      "step": 3778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3680252029129673,
+      "learning_rate": 1.82198239368373e-05,
+      "loss": 0.2007,
+      "step": 3779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29715685119899815,
+      "learning_rate": 1.8218763980783127e-05,
+      "loss": 0.2653,
+      "step": 3780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3725279137918228,
+      "learning_rate": 1.8217703740113518e-05,
+      "loss": 0.3547,
+      "step": 3781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36171919213090536,
+      "learning_rate": 1.8216643214865176e-05,
+      "loss": 0.1849,
+      "step": 3782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5560972723081516,
+      "learning_rate": 1.8215582405074838e-05,
+      "loss": 0.4323,
+      "step": 3783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6986156985423673,
+      "learning_rate": 1.821452131077923e-05,
+      "loss": 0.7669,
+      "step": 3784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2939380110603323,
+      "learning_rate": 1.8213459932015104e-05,
+      "loss": 0.2367,
+      "step": 3785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5918622124773595,
+      "learning_rate": 1.8212398268819214e-05,
+      "loss": 0.4855,
+      "step": 3786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4214369828365935,
+      "learning_rate": 1.8211336321228326e-05,
+      "loss": 0.3198,
+      "step": 3787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3423353082604522,
+      "learning_rate": 1.8210274089279214e-05,
+      "loss": 0.2273,
+      "step": 3788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48273982176060215,
+      "learning_rate": 1.8209211573008663e-05,
+      "loss": 0.2885,
+      "step": 3789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.081969336795844,
+      "learning_rate": 1.8208148772453466e-05,
+      "loss": 0.6395,
+      "step": 3790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4540363009130433,
+      "learning_rate": 1.8207085687650433e-05,
+      "loss": 0.3198,
+      "step": 3791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6965030051365356,
+      "learning_rate": 1.8206022318636375e-05,
+      "loss": 0.3283,
+      "step": 3792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36981946890013234,
+      "learning_rate": 1.8204958665448116e-05,
+      "loss": 0.3409,
+      "step": 3793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30072224494974714,
+      "learning_rate": 1.8203894728122492e-05,
+      "loss": 0.2236,
+      "step": 3794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.489290813598683,
+      "learning_rate": 1.8202830506696346e-05,
+      "loss": 0.3333,
+      "step": 3795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3308604611705892,
+      "learning_rate": 1.8201766001206533e-05,
+      "loss": 0.2731,
+      "step": 3796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5543044671117653,
+      "learning_rate": 1.8200701211689915e-05,
+      "loss": 0.3272,
+      "step": 3797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49671661520822036,
+      "learning_rate": 1.819963613818337e-05,
+      "loss": 0.3211,
+      "step": 3798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39700198461844927,
+      "learning_rate": 1.8198570780723773e-05,
+      "loss": 0.2987,
+      "step": 3799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6450937703071342,
+      "learning_rate": 1.8197505139348023e-05,
+      "loss": 0.3396,
+      "step": 3800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34925975121207203,
+      "learning_rate": 1.8196439214093023e-05,
+      "loss": 0.2853,
+      "step": 3801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3075989600836906,
+      "learning_rate": 1.819537300499569e-05,
+      "loss": 0.1727,
+      "step": 3802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40942585125745,
+      "learning_rate": 1.8194306512092938e-05,
+      "loss": 0.3184,
+      "step": 3803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3781360749319683,
+      "learning_rate": 1.8193239735421703e-05,
+      "loss": 0.3548,
+      "step": 3804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5636306621999936,
+      "learning_rate": 1.819217267501893e-05,
+      "loss": 0.2625,
+      "step": 3805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44361177405533797,
+      "learning_rate": 1.819110533092157e-05,
+      "loss": 0.3215,
+      "step": 3806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6164138406659124,
+      "learning_rate": 1.8190037703166585e-05,
+      "loss": 0.4649,
+      "step": 3807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25182141423743754,
+      "learning_rate": 1.8188969791790946e-05,
+      "loss": 0.1842,
+      "step": 3808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4578954704514088,
+      "learning_rate": 1.8187901596831638e-05,
+      "loss": 0.3233,
+      "step": 3809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5217483050301579,
+      "learning_rate": 1.8186833118325645e-05,
+      "loss": 0.4206,
+      "step": 3810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5431293205707014,
+      "learning_rate": 1.8185764356309975e-05,
+      "loss": 0.4136,
+      "step": 3811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45279199338795334,
+      "learning_rate": 1.8184695310821635e-05,
+      "loss": 0.2989,
+      "step": 3812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5020020584142632,
+      "learning_rate": 1.8183625981897653e-05,
+      "loss": 0.3023,
+      "step": 3813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34007283638326713,
+      "learning_rate": 1.818255636957505e-05,
+      "loss": 0.2279,
+      "step": 3814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5940995910555453,
+      "learning_rate": 1.818148647389088e-05,
+      "loss": 0.2751,
+      "step": 3815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3728257688155502,
+      "learning_rate": 1.8180416294882178e-05,
+      "loss": 0.3428,
+      "step": 3816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.540902906588058,
+      "learning_rate": 1.817934583258601e-05,
+      "loss": 0.413,
+      "step": 3817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32082684762622404,
+      "learning_rate": 1.8178275087039452e-05,
+      "loss": 0.1668,
+      "step": 3818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4394101437520785,
+      "learning_rate": 1.8177204058279577e-05,
+      "loss": 0.3368,
+      "step": 3819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41746611985218063,
+      "learning_rate": 1.817613274634348e-05,
+      "loss": 0.2987,
+      "step": 3820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5320382789127321,
+      "learning_rate": 1.8175061151268255e-05,
+      "loss": 0.3093,
+      "step": 3821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2989541511419468,
+      "learning_rate": 1.8173989273091014e-05,
+      "loss": 0.2661,
+      "step": 3822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.724527721383952,
+      "learning_rate": 1.8172917111848878e-05,
+      "loss": 0.8488,
+      "step": 3823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3618052569029262,
+      "learning_rate": 1.817184466757897e-05,
+      "loss": 0.2378,
+      "step": 3824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4104423447288719,
+      "learning_rate": 1.8170771940318437e-05,
+      "loss": 0.3514,
+      "step": 3825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.600402119776531,
+      "learning_rate": 1.816969893010442e-05,
+      "loss": 0.496,
+      "step": 3826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3670048381276231,
+      "learning_rate": 1.8168625636974085e-05,
+      "loss": 0.2748,
+      "step": 3827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3009020200838833,
+      "learning_rate": 1.816755206096459e-05,
+      "loss": 0.1996,
+      "step": 3828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.504312784568991,
+      "learning_rate": 1.816647820211312e-05,
+      "loss": 0.38,
+      "step": 3829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6567393314755094,
+      "learning_rate": 1.8165404060456863e-05,
+      "loss": 0.3679,
+      "step": 3830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3984924468921483,
+      "learning_rate": 1.8164329636033012e-05,
+      "loss": 0.252,
+      "step": 3831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4239600262528248,
+      "learning_rate": 1.8163254928878777e-05,
+      "loss": 0.3663,
+      "step": 3832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4164722216338981,
+      "learning_rate": 1.8162179939031377e-05,
+      "loss": 0.299,
+      "step": 3833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35787438325680954,
+      "learning_rate": 1.8161104666528033e-05,
+      "loss": 0.1849,
+      "step": 3834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.982908517879726,
+      "learning_rate": 1.8160029111405986e-05,
+      "loss": 0.6916,
+      "step": 3835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46631107207101113,
+      "learning_rate": 1.8158953273702486e-05,
+      "loss": 0.3176,
+      "step": 3836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3537925294977268,
+      "learning_rate": 1.8157877153454785e-05,
+      "loss": 0.273,
+      "step": 3837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6615628600913611,
+      "learning_rate": 1.8156800750700143e-05,
+      "loss": 0.4955,
+      "step": 3838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41214597949670184,
+      "learning_rate": 1.8155724065475845e-05,
+      "loss": 0.2243,
+      "step": 3839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3755937652787498,
+      "learning_rate": 1.8154647097819172e-05,
+      "loss": 0.2564,
+      "step": 3840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48876484681974797,
+      "learning_rate": 1.8153569847767423e-05,
+      "loss": 0.2991,
+      "step": 3841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7458949562080135,
+      "learning_rate": 1.8152492315357902e-05,
+      "loss": 0.5104,
+      "step": 3842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3735505680224295,
+      "learning_rate": 1.815141450062792e-05,
+      "loss": 0.282,
+      "step": 3843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41540487618060473,
+      "learning_rate": 1.8150336403614804e-05,
+      "loss": 0.3188,
+      "step": 3844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5712946769237918,
+      "learning_rate": 1.814925802435589e-05,
+      "loss": 0.3764,
+      "step": 3845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3572232718406907,
+      "learning_rate": 1.814817936288852e-05,
+      "loss": 0.2627,
+      "step": 3846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3055017102946289,
+      "learning_rate": 1.814710041925005e-05,
+      "loss": 0.1336,
+      "step": 3847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3538207884429109,
+      "learning_rate": 1.8146021193477846e-05,
+      "loss": 0.3007,
+      "step": 3848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4314766190553378,
+      "learning_rate": 1.8144941685609273e-05,
+      "loss": 0.3486,
+      "step": 3849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6441772312208831,
+      "learning_rate": 1.8143861895681723e-05,
+      "loss": 0.4833,
+      "step": 3850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30729483425364906,
+      "learning_rate": 1.8142781823732582e-05,
+      "loss": 0.1791,
+      "step": 3851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3808537248820102,
+      "learning_rate": 1.814170146979926e-05,
+      "loss": 0.2803,
+      "step": 3852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37558207046729103,
+      "learning_rate": 1.8140620833919165e-05,
+      "loss": 0.2885,
+      "step": 3853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6154231825728195,
+      "learning_rate": 1.813953991612972e-05,
+      "loss": 0.4008,
+      "step": 3854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35044717507297185,
+      "learning_rate": 1.813845871646836e-05,
+      "loss": 0.2794,
+      "step": 3855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.412504774677019,
+      "learning_rate": 1.813737723497252e-05,
+      "loss": 0.3406,
+      "step": 3856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5714736784254149,
+      "learning_rate": 1.8136295471679662e-05,
+      "loss": 0.2334,
+      "step": 3857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3209375714349231,
+      "learning_rate": 1.8135213426627237e-05,
+      "loss": 0.2582,
+      "step": 3858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6265957239815656,
+      "learning_rate": 1.8134131099852724e-05,
+      "loss": 0.2952,
+      "step": 3859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4414280891315469,
+      "learning_rate": 1.81330484913936e-05,
+      "loss": 0.284,
+      "step": 3860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41773899450768254,
+      "learning_rate": 1.8131965601287357e-05,
+      "loss": 0.3228,
+      "step": 3861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8234390961731449,
+      "learning_rate": 1.8130882429571496e-05,
+      "loss": 0.5967,
+      "step": 3862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5027296598506996,
+      "learning_rate": 1.8129798976283522e-05,
+      "loss": 0.4022,
+      "step": 3863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4194025271798038,
+      "learning_rate": 1.812871524146096e-05,
+      "loss": 0.2134,
+      "step": 3864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3110021600253895,
+      "learning_rate": 1.812763122514134e-05,
+      "loss": 0.2123,
+      "step": 3865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7499695380435911,
+      "learning_rate": 1.8126546927362204e-05,
+      "loss": 0.4432,
+      "step": 3866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35946689971793444,
+      "learning_rate": 1.8125462348161093e-05,
+      "loss": 0.2403,
+      "step": 3867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4073612075479078,
+      "learning_rate": 1.812437748757557e-05,
+      "loss": 0.3583,
+      "step": 3868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1775040168059026,
+      "learning_rate": 1.8123292345643203e-05,
+      "loss": 0.7311,
+      "step": 3869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3359065562504141,
+      "learning_rate": 1.8122206922401573e-05,
+      "loss": 0.2049,
+      "step": 3870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3035022075371208,
+      "learning_rate": 1.8121121217888268e-05,
+      "loss": 0.2616,
+      "step": 3871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4404755357892392,
+      "learning_rate": 1.8120035232140884e-05,
+      "loss": 0.3895,
+      "step": 3872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3304185548692813,
+      "learning_rate": 1.8118948965197027e-05,
+      "loss": 0.2378,
+      "step": 3873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1038158285453556,
+      "learning_rate": 1.8117862417094318e-05,
+      "loss": 0.8049,
+      "step": 3874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47637000688014897,
+      "learning_rate": 1.811677558787038e-05,
+      "loss": 0.4025,
+      "step": 3875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3109465314784538,
+      "learning_rate": 1.8115688477562855e-05,
+      "loss": 0.27,
+      "step": 3876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2238118146029211,
+      "learning_rate": 1.8114601086209387e-05,
+      "loss": 0.137,
+      "step": 3877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6389128213112062,
+      "learning_rate": 1.8113513413847634e-05,
+      "loss": 0.4794,
+      "step": 3878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9375943093100196,
+      "learning_rate": 1.811242546051526e-05,
+      "loss": 0.2699,
+      "step": 3879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4320199364520138,
+      "learning_rate": 1.811133722624994e-05,
+      "loss": 0.2924,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0738585193981018,
+      "learning_rate": 1.811024871108936e-05,
+      "loss": 0.5673,
+      "step": 3881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4571929492755686,
+      "learning_rate": 1.8109159915071215e-05,
+      "loss": 0.3033,
+      "step": 3882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40435453226191476,
+      "learning_rate": 1.810807083823321e-05,
+      "loss": 0.3142,
+      "step": 3883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3231601205057292,
+      "learning_rate": 1.8106981480613063e-05,
+      "loss": 0.2424,
+      "step": 3884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47123062190681675,
+      "learning_rate": 1.8105891842248496e-05,
+      "loss": 0.3023,
+      "step": 3885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5011503701732182,
+      "learning_rate": 1.810480192317724e-05,
+      "loss": 0.3053,
+      "step": 3886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5743534405718205,
+      "learning_rate": 1.8103711723437048e-05,
+      "loss": 0.4042,
+      "step": 3887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3408418367503883,
+      "learning_rate": 1.8102621243065665e-05,
+      "loss": 0.2877,
+      "step": 3888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6381779781359501,
+      "learning_rate": 1.8101530482100855e-05,
+      "loss": 0.4795,
+      "step": 3889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2908090921032573,
+      "learning_rate": 1.8100439440580393e-05,
+      "loss": 0.1469,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4231337495748093,
+      "learning_rate": 1.809934811854206e-05,
+      "loss": 0.3074,
+      "step": 3891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3200805794748924,
+      "learning_rate": 1.8098256516023654e-05,
+      "loss": 0.2839,
+      "step": 3892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.011602052709592,
+      "learning_rate": 1.809716463306297e-05,
+      "loss": 0.2713,
+      "step": 3893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3933605424410423,
+      "learning_rate": 1.8096072469697822e-05,
+      "loss": 0.324,
+      "step": 3894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7690843943555367,
+      "learning_rate": 1.8094980025966036e-05,
+      "loss": 0.4747,
+      "step": 3895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3765206416318077,
+      "learning_rate": 1.809388730190544e-05,
+      "loss": 0.2389,
+      "step": 3896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3835611886266919,
+      "learning_rate": 1.8092794297553873e-05,
+      "loss": 0.28,
+      "step": 3897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4180873590762182,
+      "learning_rate": 1.8091701012949187e-05,
+      "loss": 0.2725,
+      "step": 3898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44449937407460755,
+      "learning_rate": 1.8090607448129244e-05,
+      "loss": 0.3152,
+      "step": 3899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43700527552620777,
+      "learning_rate": 1.8089513603131913e-05,
+      "loss": 0.301,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6587341669479327,
+      "learning_rate": 1.808841947799507e-05,
+      "loss": 0.4034,
+      "step": 3901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2622215351655142,
+      "learning_rate": 1.808732507275661e-05,
+      "loss": 0.7012,
+      "step": 3902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35778822861958404,
+      "learning_rate": 1.8086230387454434e-05,
+      "loss": 0.2038,
+      "step": 3903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41095801057868664,
+      "learning_rate": 1.8085135422126448e-05,
+      "loss": 0.2571,
+      "step": 3904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5132522096870761,
+      "learning_rate": 1.8084040176810567e-05,
+      "loss": 0.349,
+      "step": 3905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4841833694973949,
+      "learning_rate": 1.808294465154472e-05,
+      "loss": 0.2723,
+      "step": 3906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4857845170840667,
+      "learning_rate": 1.8081848846366852e-05,
+      "loss": 0.3875,
+      "step": 3907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5247363261757332,
+      "learning_rate": 1.8080752761314904e-05,
+      "loss": 0.3969,
+      "step": 3908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34761712484213264,
+      "learning_rate": 1.8079656396426834e-05,
+      "loss": 0.2346,
+      "step": 3909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2940198797850371,
+      "learning_rate": 1.807855975174061e-05,
+      "loss": 0.1858,
+      "step": 3910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40981413692161606,
+      "learning_rate": 1.8077462827294214e-05,
+      "loss": 0.3515,
+      "step": 3911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3987787782958305,
+      "learning_rate": 1.8076365623125625e-05,
+      "loss": 0.2753,
+      "step": 3912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8541449017563671,
+      "learning_rate": 1.8075268139272842e-05,
+      "loss": 0.5343,
+      "step": 3913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.122921649883041,
+      "learning_rate": 1.8074170375773867e-05,
+      "loss": 0.7047,
+      "step": 3914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.337497593773467,
+      "learning_rate": 1.8073072332666723e-05,
+      "loss": 0.2768,
+      "step": 3915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4214408990448417,
+      "learning_rate": 1.807197400998943e-05,
+      "loss": 0.257,
+      "step": 3916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6049790373030124,
+      "learning_rate": 1.8070875407780026e-05,
+      "loss": 0.355,
+      "step": 3917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30708101652710945,
+      "learning_rate": 1.806977652607655e-05,
+      "loss": 0.1988,
+      "step": 3918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4084061861810741,
+      "learning_rate": 1.8068677364917063e-05,
+      "loss": 0.283,
+      "step": 3919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.626383104086622,
+      "learning_rate": 1.806757792433962e-05,
+      "loss": 0.4173,
+      "step": 3920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6969864507969111,
+      "learning_rate": 1.806647820438231e-05,
+      "loss": 0.4274,
+      "step": 3921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3530384228236583,
+      "learning_rate": 1.8065378205083202e-05,
+      "loss": 0.316,
+      "step": 3922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4005606977157132,
+      "learning_rate": 1.8064277926480392e-05,
+      "loss": 0.3583,
+      "step": 3923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31165153005202517,
+      "learning_rate": 1.8063177368611988e-05,
+      "loss": 0.214,
+      "step": 3924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33150213382715915,
+      "learning_rate": 1.8062076531516094e-05,
+      "loss": 0.2387,
+      "step": 3925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8408216515743814,
+      "learning_rate": 1.806097541523084e-05,
+      "loss": 0.4223,
+      "step": 3926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43962821527072266,
+      "learning_rate": 1.8059874019794352e-05,
+      "loss": 0.3006,
+      "step": 3927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4049589074514208,
+      "learning_rate": 1.8058772345244775e-05,
+      "loss": 0.3666,
+      "step": 3928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6436948491887067,
+      "learning_rate": 1.8057670391620258e-05,
+      "loss": 0.3893,
+      "step": 3929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26980658996097057,
+      "learning_rate": 1.8056568158958958e-05,
+      "loss": 0.194,
+      "step": 3930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5857657539673209,
+      "learning_rate": 1.8055465647299052e-05,
+      "loss": 0.4342,
+      "step": 3931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3769739875049705,
+      "learning_rate": 1.805436285667872e-05,
+      "loss": 0.2869,
+      "step": 3932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44323285132468504,
+      "learning_rate": 1.8053259787136144e-05,
+      "loss": 0.3503,
+      "step": 3933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5264863084482408,
+      "learning_rate": 1.8052156438709527e-05,
+      "loss": 0.3917,
+      "step": 3934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3824530834229984,
+      "learning_rate": 1.805105281143708e-05,
+      "loss": 0.3065,
+      "step": 3935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39071495256736205,
+      "learning_rate": 1.8049948905357023e-05,
+      "loss": 0.1898,
+      "step": 3936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3091528912278944,
+      "learning_rate": 1.804884472050758e-05,
+      "loss": 0.2312,
+      "step": 3937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9481919338379736,
+      "learning_rate": 1.8047740256926993e-05,
+      "loss": 0.6748,
+      "step": 3938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35422805559962023,
+      "learning_rate": 1.8046635514653505e-05,
+      "loss": 0.2324,
+      "step": 3939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.412781947395747,
+      "learning_rate": 1.8045530493725375e-05,
+      "loss": 0.3477,
+      "step": 3940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7610604686105258,
+      "learning_rate": 1.8044425194180868e-05,
+      "loss": 0.5374,
+      "step": 3941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23494787925537458,
+      "learning_rate": 1.8043319616058266e-05,
+      "loss": 0.0743,
+      "step": 3942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32367312358901884,
+      "learning_rate": 1.804221375939585e-05,
+      "loss": 0.2955,
+      "step": 3943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9361586658175016,
+      "learning_rate": 1.8041107624231916e-05,
+      "loss": 0.6321,
+      "step": 3944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4736112307448519,
+      "learning_rate": 1.804000121060477e-05,
+      "loss": 0.3214,
+      "step": 3945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4152969262120841,
+      "learning_rate": 1.803889451855273e-05,
+      "loss": 0.346,
+      "step": 3946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.425406663947165,
+      "learning_rate": 1.8037787548114122e-05,
+      "loss": 0.3407,
+      "step": 3947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3436781987502028,
+      "learning_rate": 1.8036680299327273e-05,
+      "loss": 0.1941,
+      "step": 3948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2759805308240463,
+      "learning_rate": 1.8035572772230526e-05,
+      "loss": 0.1915,
+      "step": 3949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7343999781835354,
+      "learning_rate": 1.8034464966862247e-05,
+      "loss": 0.5131,
+      "step": 3950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3709341545619263,
+      "learning_rate": 1.8033356883260786e-05,
+      "loss": 0.3157,
+      "step": 3951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3961664624622881,
+      "learning_rate": 1.8032248521464526e-05,
+      "loss": 0.301,
+      "step": 3952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9898774268867264,
+      "learning_rate": 1.8031139881511844e-05,
+      "loss": 0.6261,
+      "step": 3953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39230522301924536,
+      "learning_rate": 1.8030030963441133e-05,
+      "loss": 0.2662,
+      "step": 3954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28249473186971336,
+      "learning_rate": 1.8028921767290796e-05,
+      "loss": 0.2525,
+      "step": 3955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5699095328371475,
+      "learning_rate": 1.802781229309924e-05,
+      "loss": 0.3755,
+      "step": 3956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7814378890464383,
+      "learning_rate": 1.8026702540904893e-05,
+      "loss": 0.4934,
+      "step": 3957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31906069339089543,
+      "learning_rate": 1.802559251074618e-05,
+      "loss": 0.2279,
+      "step": 3958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4300544687060639,
+      "learning_rate": 1.8024482202661544e-05,
+      "loss": 0.3191,
+      "step": 3959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7882393206382694,
+      "learning_rate": 1.8023371616689437e-05,
+      "loss": 0.4986,
+      "step": 3960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2288574775693125,
+      "learning_rate": 1.8022260752868314e-05,
+      "loss": 0.1694,
+      "step": 3961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.129884549557266,
+      "learning_rate": 1.8021149611236644e-05,
+      "loss": 0.5047,
+      "step": 3962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3673895981661131,
+      "learning_rate": 1.8020038191832912e-05,
+      "loss": 0.3243,
+      "step": 3963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3960014025969493,
+      "learning_rate": 1.80189264946956e-05,
+      "loss": 0.3041,
+      "step": 3964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7265493417525126,
+      "learning_rate": 1.8017814519863206e-05,
+      "loss": 0.2853,
+      "step": 3965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4288534438453486,
+      "learning_rate": 1.8016702267374243e-05,
+      "loss": 0.3599,
+      "step": 3966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3331264382746232,
+      "learning_rate": 1.8015589737267226e-05,
+      "loss": 0.2876,
+      "step": 3967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43022925309376164,
+      "learning_rate": 1.801447692958068e-05,
+      "loss": 0.2281,
+      "step": 3968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5957577630302175,
+      "learning_rate": 1.801336384435314e-05,
+      "loss": 0.3927,
+      "step": 3969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4640169255196903,
+      "learning_rate": 1.8012250481623158e-05,
+      "loss": 0.3406,
+      "step": 3970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39945224996755907,
+      "learning_rate": 1.8011136841429285e-05,
+      "loss": 0.2894,
+      "step": 3971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8005686154322458,
+      "learning_rate": 1.801002292381009e-05,
+      "loss": 0.4909,
+      "step": 3972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33287331011769583,
+      "learning_rate": 1.800890872880414e-05,
+      "loss": 0.2381,
+      "step": 3973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36297847553286217,
+      "learning_rate": 1.8007794256450027e-05,
+      "loss": 0.3012,
+      "step": 3974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5849863340592387,
+      "learning_rate": 1.800667950678635e-05,
+      "loss": 0.4239,
+      "step": 3975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.378342194413626,
+      "learning_rate": 1.8005564479851697e-05,
+      "loss": 0.2682,
+      "step": 3976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7952623423604379,
+      "learning_rate": 1.8004449175684697e-05,
+      "loss": 0.6006,
+      "step": 3977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4097070482739936,
+      "learning_rate": 1.8003333594323962e-05,
+      "loss": 0.2736,
+      "step": 3978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3728346283618812,
+      "learning_rate": 1.800221773580813e-05,
+      "loss": 0.2866,
+      "step": 3979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5766114897185256,
+      "learning_rate": 1.8001101600175843e-05,
+      "loss": 0.4659,
+      "step": 3980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26613403625490034,
+      "learning_rate": 1.799998518746575e-05,
+      "loss": 0.1474,
+      "step": 3981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31031290429059133,
+      "learning_rate": 1.7998868497716516e-05,
+      "loss": 0.2421,
+      "step": 3982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4392924725322208,
+      "learning_rate": 1.7997751530966806e-05,
+      "loss": 0.3419,
+      "step": 3983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5462229687452343,
+      "learning_rate": 1.799663428725531e-05,
+      "loss": 0.2927,
+      "step": 3984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39202509019051524,
+      "learning_rate": 1.7995516766620706e-05,
+      "loss": 0.3135,
+      "step": 3985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7239484639028205,
+      "learning_rate": 1.7994398969101704e-05,
+      "loss": 0.5246,
+      "step": 3986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3566107377592294,
+      "learning_rate": 1.799328089473701e-05,
+      "loss": 0.2878,
+      "step": 3987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4206515109068827,
+      "learning_rate": 1.799216254356534e-05,
+      "loss": 0.3109,
+      "step": 3988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.337123993090019,
+      "learning_rate": 1.7991043915625427e-05,
+      "loss": 0.2135,
+      "step": 3989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.34315829448210505,
+      "learning_rate": 1.798992501095601e-05,
+      "loss": 0.3353,
+      "step": 3990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36382392161356036,
+      "learning_rate": 1.7988805829595825e-05,
+      "loss": 0.2506,
+      "step": 3991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6713822788963748,
+      "learning_rate": 1.7987686371583643e-05,
+      "loss": 0.501,
+      "step": 3992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.326313126631407,
+      "learning_rate": 1.7986566636958228e-05,
+      "loss": 0.8042,
+      "step": 3993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27371657393729876,
+      "learning_rate": 1.798544662575835e-05,
+      "loss": 0.188,
+      "step": 3994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31879060982136537,
+      "learning_rate": 1.7984326338022797e-05,
+      "loss": 0.2921,
+      "step": 3995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6619585978080218,
+      "learning_rate": 1.798320577379037e-05,
+      "loss": 0.4928,
+      "step": 3996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.353273027323354,
+      "learning_rate": 1.7982084933099868e-05,
+      "loss": 0.2369,
+      "step": 3997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1077945691153839,
+      "learning_rate": 1.798096381599011e-05,
+      "loss": 0.6902,
+      "step": 3998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3950848276115181,
+      "learning_rate": 1.7979842422499917e-05,
+      "loss": 0.3657,
+      "step": 3999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.298536345873338,
+      "learning_rate": 1.7978720752668123e-05,
+      "loss": 0.2263,
+      "step": 4000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3103745986561787,
+      "learning_rate": 1.7977598806533575e-05,
+      "loss": 0.1444,
+      "step": 4001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46143426008863875,
+      "learning_rate": 1.797647658413512e-05,
+      "loss": 0.3617,
+      "step": 4002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35964308521923766,
+      "learning_rate": 1.7975354085511627e-05,
+      "loss": 0.3244,
+      "step": 4003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0071718247568406,
+      "learning_rate": 1.7974231310701964e-05,
+      "loss": 0.0483,
+      "step": 4004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8956530240852543,
+      "learning_rate": 1.7973108259745012e-05,
+      "loss": 0.7035,
+      "step": 4005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4631104114709448,
+      "learning_rate": 1.7971984932679663e-05,
+      "loss": 0.2886,
+      "step": 4006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25131716487061634,
+      "learning_rate": 1.7970861329544823e-05,
+      "loss": 0.2199,
+      "step": 4007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6088567020101945,
+      "learning_rate": 1.7969737450379395e-05,
+      "loss": 0.4963,
+      "step": 4008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6053261675860214,
+      "learning_rate": 1.7968613295222304e-05,
+      "loss": 0.2756,
+      "step": 4009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3335128304816215,
+      "learning_rate": 1.7967488864112473e-05,
+      "loss": 0.2663,
+      "step": 4010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4790286393120566,
+      "learning_rate": 1.7966364157088853e-05,
+      "loss": 0.8433,
+      "step": 4011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3681446596417048,
+      "learning_rate": 1.7965239174190376e-05,
+      "loss": 0.2871,
+      "step": 4012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6855409947925746,
+      "learning_rate": 1.7964113915456013e-05,
+      "loss": 0.5071,
+      "step": 4013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3131770028920056,
+      "learning_rate": 1.7962988380924727e-05,
+      "loss": 0.2426,
+      "step": 4014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39953225148360916,
+      "learning_rate": 1.7961862570635496e-05,
+      "loss": 0.2161,
+      "step": 4015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0905675166530646,
+      "learning_rate": 1.7960736484627306e-05,
+      "loss": 0.5986,
+      "step": 4016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8487839708425405,
+      "learning_rate": 1.7959610122939155e-05,
+      "loss": 0.521,
+      "step": 4017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35305893350636597,
+      "learning_rate": 1.7958483485610048e-05,
+      "loss": 0.275,
+      "step": 4018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5051109731237906,
+      "learning_rate": 1.7957356572678998e-05,
+      "loss": 0.389,
+      "step": 4019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37882249996544415,
+      "learning_rate": 1.7956229384185036e-05,
+      "loss": 0.217,
+      "step": 4020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43823490781549507,
+      "learning_rate": 1.7955101920167188e-05,
+      "loss": 0.2772,
+      "step": 4021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5312447099067517,
+      "learning_rate": 1.7953974180664504e-05,
+      "loss": 0.3589,
+      "step": 4022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5289342208062883,
+      "learning_rate": 1.7952846165716038e-05,
+      "loss": 0.3424,
+      "step": 4023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44434750769308423,
+      "learning_rate": 1.795171787536085e-05,
+      "loss": 0.2691,
+      "step": 4024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5755076448856171,
+      "learning_rate": 1.7950589309638014e-05,
+      "loss": 0.4052,
+      "step": 4025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30817817030080696,
+      "learning_rate": 1.794946046858661e-05,
+      "loss": 0.2699,
+      "step": 4026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3306458946911386,
+      "learning_rate": 1.7948331352245736e-05,
+      "loss": 0.1414,
+      "step": 4027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.509175609376385,
+      "learning_rate": 1.7947201960654488e-05,
+      "loss": 0.3813,
+      "step": 4028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8703972842173666,
+      "learning_rate": 1.7946072293851976e-05,
+      "loss": 0.6002,
+      "step": 4029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.343542528326212,
+      "learning_rate": 1.794494235187732e-05,
+      "loss": 0.2338,
+      "step": 4030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4128671296922721,
+      "learning_rate": 1.7943812134769656e-05,
+      "loss": 0.4134,
+      "step": 4031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.041845445196055,
+      "learning_rate": 1.794268164256812e-05,
+      "loss": 0.628,
+      "step": 4032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23769051134991365,
+      "learning_rate": 1.7941550875311858e-05,
+      "loss": 0.1505,
+      "step": 4033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45355094491900144,
+      "learning_rate": 1.794041983304003e-05,
+      "loss": 0.3421,
+      "step": 4034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1868507744408294,
+      "learning_rate": 1.7939288515791806e-05,
+      "loss": 0.7895,
+      "step": 4035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.395618032596114,
+      "learning_rate": 1.7938156923606362e-05,
+      "loss": 0.2637,
+      "step": 4036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7611232233618405,
+      "learning_rate": 1.7937025056522884e-05,
+      "loss": 0.5185,
+      "step": 4037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4522028536625917,
+      "learning_rate": 1.7935892914580572e-05,
+      "loss": 0.3422,
+      "step": 4038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32600638350104655,
+      "learning_rate": 1.7934760497818628e-05,
+      "loss": 0.1793,
+      "step": 4039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5263969279202462,
+      "learning_rate": 1.7933627806276267e-05,
+      "loss": 0.2869,
+      "step": 4040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5985454643455146,
+      "learning_rate": 1.7932494839992723e-05,
+      "loss": 0.4168,
+      "step": 4041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40800954663605743,
+      "learning_rate": 1.793136159900722e-05,
+      "loss": 0.286,
+      "step": 4042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4796678193727095,
+      "learning_rate": 1.7930228083359002e-05,
+      "loss": 0.3264,
+      "step": 4043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4774420141970719,
+      "learning_rate": 1.792909429308733e-05,
+      "loss": 0.3023,
+      "step": 4044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33552706676028116,
+      "learning_rate": 1.792796022823146e-05,
+      "loss": 0.1879,
+      "step": 4045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.34466688409543883,
+      "learning_rate": 1.7926825888830673e-05,
+      "loss": 0.2726,
+      "step": 4046
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6822063282086891,
+      "learning_rate": 1.7925691274924247e-05,
+      "loss": 0.5477,
+      "step": 4047
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4588359590822482,
+      "learning_rate": 1.7924556386551472e-05,
+      "loss": 0.4077,
+      "step": 4048
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39472767234202216,
+      "learning_rate": 1.792342122375165e-05,
+      "loss": 0.289,
+      "step": 4049
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4249909103766952,
+      "learning_rate": 1.792228578656409e-05,
+      "loss": 0.3337,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40114167914557236,
+      "learning_rate": 1.7921150075028112e-05,
+      "loss": 0.2137,
+      "step": 4051
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3296340901766964,
+      "learning_rate": 1.792001408918305e-05,
+      "loss": 0.2957,
+      "step": 4052
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8150910975190974,
+      "learning_rate": 1.791887782906824e-05,
+      "loss": 0.4098,
+      "step": 4053
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3098451298308945,
+      "learning_rate": 1.7917741294723035e-05,
+      "loss": 0.2909,
+      "step": 4054
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6086232025078487,
+      "learning_rate": 1.7916604486186786e-05,
+      "loss": 0.4586,
+      "step": 4055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7056291363756321,
+      "learning_rate": 1.7915467403498864e-05,
+      "loss": 0.3649,
+      "step": 4056
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.232611976797289,
+      "learning_rate": 1.791433004669865e-05,
+      "loss": 0.1749,
+      "step": 4057
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3738788987291126,
+      "learning_rate": 1.791319241582552e-05,
+      "loss": 0.3219,
+      "step": 4058
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0108750504208646,
+      "learning_rate": 1.791205451091888e-05,
+      "loss": 0.4338,
+      "step": 4059
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8253458918236195,
+      "learning_rate": 1.7910916332018137e-05,
+      "loss": 0.4663,
+      "step": 4060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4535609620382996,
+      "learning_rate": 1.7909777879162695e-05,
+      "loss": 0.2864,
+      "step": 4061
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5264572392672097,
+      "learning_rate": 1.7908639152391988e-05,
+      "loss": 0.3096,
+      "step": 4062
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6313130758765294,
+      "learning_rate": 1.790750015174545e-05,
+      "loss": 0.3589,
+      "step": 4063
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4027259179702243,
+      "learning_rate": 1.7906360877262515e-05,
+      "loss": 0.3104,
+      "step": 4064
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8796907969909685,
+      "learning_rate": 1.7905221328982647e-05,
+      "loss": 0.5063,
+      "step": 4065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3771326162949683,
+      "learning_rate": 1.7904081506945304e-05,
+      "loss": 0.25,
+      "step": 4066
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37190974661172604,
+      "learning_rate": 1.790294141118996e-05,
+      "loss": 0.2481,
+      "step": 4067
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5489614559790572,
+      "learning_rate": 1.790180104175609e-05,
+      "loss": 0.7663,
+      "step": 4068
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5644846432613142,
+      "learning_rate": 1.7900660398683192e-05,
+      "loss": 0.2647,
+      "step": 4069
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5649799604812463,
+      "learning_rate": 1.789951948201077e-05,
+      "loss": 0.3422,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5752127497152826,
+      "learning_rate": 1.789837829177832e-05,
+      "loss": 0.3297,
+      "step": 4071
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2515960625460765,
+      "learning_rate": 1.7897236828025373e-05,
+      "loss": 0.13,
+      "step": 4072
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.454993611244226,
+      "learning_rate": 1.7896095090791452e-05,
+      "loss": 0.3405,
+      "step": 4073
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46244441347114756,
+      "learning_rate": 1.7894953080116102e-05,
+      "loss": 0.3233,
+      "step": 4074
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46040709971171495,
+      "learning_rate": 1.7893810796038862e-05,
+      "loss": 0.2484,
+      "step": 4075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7576076875799639,
+      "learning_rate": 1.7892668238599293e-05,
+      "loss": 0.4109,
+      "step": 4076
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6283333218980784,
+      "learning_rate": 1.7891525407836967e-05,
+      "loss": 0.4061,
+      "step": 4077
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49411712691782517,
+      "learning_rate": 1.789038230379145e-05,
+      "loss": 0.3519,
+      "step": 4078
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25958003257867746,
+      "learning_rate": 1.7889238926502336e-05,
+      "loss": 0.1599,
+      "step": 4079
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6904502163139321,
+      "learning_rate": 1.788809527600922e-05,
+      "loss": 0.5345,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42273087648124946,
+      "learning_rate": 1.78869513523517e-05,
+      "loss": 0.3553,
+      "step": 4081
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41321642610780296,
+      "learning_rate": 1.7885807155569395e-05,
+      "loss": 0.2611,
+      "step": 4082
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.29593229800086,
+      "learning_rate": 1.7884662685701927e-05,
+      "loss": 0.797,
+      "step": 4083
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4185943883142081,
+      "learning_rate": 1.788351794278893e-05,
+      "loss": 0.2627,
+      "step": 4084
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30586527586360296,
+      "learning_rate": 1.7882372926870045e-05,
+      "loss": 0.2179,
+      "step": 4085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5378611831339616,
+      "learning_rate": 1.7881227637984922e-05,
+      "loss": 0.391,
+      "step": 4086
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6274700701055269,
+      "learning_rate": 1.788008207617323e-05,
+      "loss": 0.4243,
+      "step": 4087
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3917913457312723,
+      "learning_rate": 1.787893624147463e-05,
+      "loss": 0.3136,
+      "step": 4088
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4596357049274881,
+      "learning_rate": 1.7877790133928807e-05,
+      "loss": 0.2903,
+      "step": 4089
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4524210604969984,
+      "learning_rate": 1.7876643753575457e-05,
+      "loss": 0.3614,
+      "step": 4090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.382573208203039,
+      "learning_rate": 1.7875497100454266e-05,
+      "loss": 0.3108,
+      "step": 4091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3418759654466254,
+      "learning_rate": 1.787435017460495e-05,
+      "loss": 0.1814,
+      "step": 4092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48679353512366635,
+      "learning_rate": 1.7873202976067225e-05,
+      "loss": 0.3678,
+      "step": 4093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3769792537925658,
+      "learning_rate": 1.787205550488082e-05,
+      "loss": 0.3096,
+      "step": 4094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.744428708424359,
+      "learning_rate": 1.7870907761085474e-05,
+      "loss": 0.4643,
+      "step": 4095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6447660292616412,
+      "learning_rate": 1.786975974472093e-05,
+      "loss": 0.4391,
+      "step": 4096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3339067669230319,
+      "learning_rate": 1.7868611455826942e-05,
+      "loss": 0.2451,
+      "step": 4097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3580989985318797,
+      "learning_rate": 1.7867462894443283e-05,
+      "loss": 0.2714,
+      "step": 4098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9081337673125298,
+      "learning_rate": 1.7866314060609714e-05,
+      "loss": 0.6394,
+      "step": 4099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4002506801084894,
+      "learning_rate": 1.7865164954366033e-05,
+      "loss": 0.2894,
+      "step": 4100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4251875388623584,
+      "learning_rate": 1.7864015575752026e-05,
+      "loss": 0.353,
+      "step": 4101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.063061765139635,
+      "learning_rate": 1.78628659248075e-05,
+      "loss": 0.4411,
+      "step": 4102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29534226202391967,
+      "learning_rate": 1.7861716001572262e-05,
+      "loss": 0.2349,
+      "step": 4103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6987250149273035,
+      "learning_rate": 1.7860565806086142e-05,
+      "loss": 0.4517,
+      "step": 4104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31222378317392696,
+      "learning_rate": 1.7859415338388963e-05,
+      "loss": 0.228,
+      "step": 4105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4311047635232939,
+      "learning_rate": 1.7858264598520568e-05,
+      "loss": 0.3151,
+      "step": 4106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4992911334787213,
+      "learning_rate": 1.7857113586520806e-05,
+      "loss": 0.7821,
+      "step": 4107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5395303710203192,
+      "learning_rate": 1.7855962302429542e-05,
+      "loss": 0.2653,
+      "step": 4108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40037400934122036,
+      "learning_rate": 1.785481074628664e-05,
+      "loss": 0.3089,
+      "step": 4109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4222237834089705,
+      "learning_rate": 1.785365891813198e-05,
+      "loss": 0.3479,
+      "step": 4110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26589948850227196,
+      "learning_rate": 1.785250681800545e-05,
+      "loss": 0.1399,
+      "step": 4111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41721052652204943,
+      "learning_rate": 1.7851354445946944e-05,
+      "loss": 0.3075,
+      "step": 4112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4215018631260107,
+      "learning_rate": 1.785020180199637e-05,
+      "loss": 0.3456,
+      "step": 4113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3790817111716487,
+      "learning_rate": 1.7849048886193648e-05,
+      "loss": 0.8667,
+      "step": 4114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.354931172130831,
+      "learning_rate": 1.7847895698578702e-05,
+      "loss": 0.2496,
+      "step": 4115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.423749864426112,
+      "learning_rate": 1.7846742239191464e-05,
+      "loss": 0.3579,
+      "step": 4116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3605983029128115,
+      "learning_rate": 1.784558850807188e-05,
+      "loss": 0.2686,
+      "step": 4117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3667287214717069,
+      "learning_rate": 1.7844434505259904e-05,
+      "loss": 0.2296,
+      "step": 4118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9766307941377927,
+      "learning_rate": 1.7843280230795496e-05,
+      "loss": 0.6913,
+      "step": 4119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7464374748135114,
+      "learning_rate": 1.784212568471863e-05,
+      "loss": 0.5132,
+      "step": 4120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34975514825247495,
+      "learning_rate": 1.7840970867069293e-05,
+      "loss": 0.2485,
+      "step": 4121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7385557641533902,
+      "learning_rate": 1.7839815777887472e-05,
+      "loss": 0.511,
+      "step": 4122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31425441408801236,
+      "learning_rate": 1.7838660417213166e-05,
+      "loss": 0.179,
+      "step": 4123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3292178319221814,
+      "learning_rate": 1.7837504785086386e-05,
+      "loss": 0.2168,
+      "step": 4124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36915751123007257,
+      "learning_rate": 1.7836348881547153e-05,
+      "loss": 0.3212,
+      "step": 4125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9386073864728668,
+      "learning_rate": 1.7835192706635494e-05,
+      "loss": 0.5632,
+      "step": 4126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38900805335201366,
+      "learning_rate": 1.783403626039145e-05,
+      "loss": 0.2911,
+      "step": 4127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7257634073735822,
+      "learning_rate": 1.7832879542855067e-05,
+      "loss": 0.4026,
+      "step": 4128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.260931509801551,
+      "learning_rate": 1.7831722554066403e-05,
+      "loss": 0.2349,
+      "step": 4129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41780863397346774,
+      "learning_rate": 1.7830565294065522e-05,
+      "loss": 0.2964,
+      "step": 4130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5120862661410127,
+      "learning_rate": 1.7829407762892504e-05,
+      "loss": 0.2983,
+      "step": 4131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9249334484384528,
+      "learning_rate": 1.7828249960587428e-05,
+      "loss": 0.4811,
+      "step": 4132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36244641485753243,
+      "learning_rate": 1.7827091887190396e-05,
+      "loss": 0.3057,
+      "step": 4133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36789794513151913,
+      "learning_rate": 1.7825933542741506e-05,
+      "loss": 0.2761,
+      "step": 4134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2799037207296845,
+      "learning_rate": 1.7824774927280877e-05,
+      "loss": 0.1626,
+      "step": 4135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3573932183986537,
+      "learning_rate": 1.7823616040848625e-05,
+      "loss": 0.3016,
+      "step": 4136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4498035506977487,
+      "learning_rate": 1.782245688348489e-05,
+      "loss": 0.3051,
+      "step": 4137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1538000821845913,
+      "learning_rate": 1.7821297455229807e-05,
+      "loss": 0.4959,
+      "step": 4138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3316262304163246,
+      "learning_rate": 1.7820137756123527e-05,
+      "loss": 0.2842,
+      "step": 4139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.53751881341363,
+      "learning_rate": 1.7818977786206217e-05,
+      "loss": 0.8152,
+      "step": 4140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32222675266258977,
+      "learning_rate": 1.7817817545518045e-05,
+      "loss": 0.2308,
+      "step": 4141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3286628540441941,
+      "learning_rate": 1.7816657034099182e-05,
+      "loss": 0.2449,
+      "step": 4142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4467305285964456,
+      "learning_rate": 1.781549625198982e-05,
+      "loss": 0.8061,
+      "step": 4143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.870823301196831,
+      "learning_rate": 1.7814335199230164e-05,
+      "loss": 0.3788,
+      "step": 4144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.336748819198341,
+      "learning_rate": 1.7813173875860416e-05,
+      "loss": 0.2872,
+      "step": 4145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6111350426253821,
+      "learning_rate": 1.781201228192079e-05,
+      "loss": 0.4169,
+      "step": 4146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.21527619658183722,
+      "learning_rate": 1.7810850417451517e-05,
+      "loss": 0.1193,
+      "step": 4147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4550362260441464,
+      "learning_rate": 1.780968828249283e-05,
+      "loss": 0.2929,
+      "step": 4148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49114266037537563,
+      "learning_rate": 1.780852587708497e-05,
+      "loss": 0.3229,
+      "step": 4149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1197232619356512,
+      "learning_rate": 1.78073632012682e-05,
+      "loss": 0.4345,
+      "step": 4150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4730494255334301,
+      "learning_rate": 1.780620025508277e-05,
+      "loss": 0.3158,
+      "step": 4151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5063609235250967,
+      "learning_rate": 1.7805037038568972e-05,
+      "loss": 0.4172,
+      "step": 4152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41048674198046375,
+      "learning_rate": 1.780387355176707e-05,
+      "loss": 0.2985,
+      "step": 4153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.21889835566971555,
+      "learning_rate": 1.7802709794717363e-05,
+      "loss": 0.1212,
+      "step": 4154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0661163910225424,
+      "learning_rate": 1.780154576746015e-05,
+      "loss": 0.4338,
+      "step": 4155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3205369104367537,
+      "learning_rate": 1.7800381470035745e-05,
+      "loss": 0.7398,
+      "step": 4156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36004560177244327,
+      "learning_rate": 1.7799216902484465e-05,
+      "loss": 0.2448,
+      "step": 4157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.496192252384227,
+      "learning_rate": 1.7798052064846637e-05,
+      "loss": 0.805,
+      "step": 4158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8488803818598749,
+      "learning_rate": 1.7796886957162603e-05,
+      "loss": 0.5812,
+      "step": 4159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2631792065123535,
+      "learning_rate": 1.7795721579472712e-05,
+      "loss": 0.2027,
+      "step": 4160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39486478303427447,
+      "learning_rate": 1.7794555931817314e-05,
+      "loss": 0.3416,
+      "step": 4161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7292174481348954,
+      "learning_rate": 1.779339001423678e-05,
+      "loss": 0.5763,
+      "step": 4162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.277092082065919,
+      "learning_rate": 1.7792223826771484e-05,
+      "loss": 0.1583,
+      "step": 4163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.648421763994709,
+      "learning_rate": 1.779105736946181e-05,
+      "loss": 0.4098,
+      "step": 4164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4148689107380746,
+      "learning_rate": 1.778989064234816e-05,
+      "loss": 0.3432,
+      "step": 4165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5232083805274033,
+      "learning_rate": 1.7788723645470928e-05,
+      "loss": 0.3272,
+      "step": 4166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2818994881344073,
+      "learning_rate": 1.7787556378870534e-05,
+      "loss": 0.2582,
+      "step": 4167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.457924102148931,
+      "learning_rate": 1.7786388842587397e-05,
+      "loss": 0.377,
+      "step": 4168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5522352742313344,
+      "learning_rate": 1.7785221036661945e-05,
+      "loss": 0.3301,
+      "step": 4169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32880274033610146,
+      "learning_rate": 1.778405296113463e-05,
+      "loss": 0.2008,
+      "step": 4170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7913590840489569,
+      "learning_rate": 1.7782884616045892e-05,
+      "loss": 0.5754,
+      "step": 4171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5348610835612185,
+      "learning_rate": 1.7781716001436192e-05,
+      "loss": 0.3692,
+      "step": 4172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3764624857271168,
+      "learning_rate": 1.7780547117346005e-05,
+      "loss": 0.2682,
+      "step": 4173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2520872298678765,
+      "learning_rate": 1.7779377963815804e-05,
+      "loss": 0.7315,
+      "step": 4174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3392114778810658,
+      "learning_rate": 1.7778208540886082e-05,
+      "loss": 0.1731,
+      "step": 4175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5068710149305249,
+      "learning_rate": 1.777703884859733e-05,
+      "loss": 0.3636,
+      "step": 4176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.393948244622996,
+      "learning_rate": 1.7775868886990056e-05,
+      "loss": 0.3052,
+      "step": 4177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4345252768576852,
+      "learning_rate": 1.7774698656104778e-05,
+      "loss": 0.343,
+      "step": 4178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5250465194272104,
+      "learning_rate": 1.777352815598202e-05,
+      "loss": 0.3784,
+      "step": 4179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34170118548309036,
+      "learning_rate": 1.7772357386662316e-05,
+      "loss": 0.2921,
+      "step": 4180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3091716890840008,
+      "learning_rate": 1.777118634818621e-05,
+      "loss": 0.2455,
+      "step": 4181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32875149163461864,
+      "learning_rate": 1.7770015040594256e-05,
+      "loss": 0.2709,
+      "step": 4182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5657387770353896,
+      "learning_rate": 1.7768843463927012e-05,
+      "loss": 0.4015,
+      "step": 4183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4578594861031997,
+      "learning_rate": 1.776767161822506e-05,
+      "loss": 0.3508,
+      "step": 4184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3722925933401782,
+      "learning_rate": 1.7766499503528965e-05,
+      "loss": 0.293,
+      "step": 4185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8542895200701599,
+      "learning_rate": 1.776532711987933e-05,
+      "loss": 0.561,
+      "step": 4186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3903147749744142,
+      "learning_rate": 1.7764154467316753e-05,
+      "loss": 0.206,
+      "step": 4187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3327376950139663,
+      "learning_rate": 1.776298154588184e-05,
+      "loss": 0.2483,
+      "step": 4188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6250386359397111,
+      "learning_rate": 1.7761808355615207e-05,
+      "loss": 0.4006,
+      "step": 4189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49450962581882085,
+      "learning_rate": 1.7760634896557483e-05,
+      "loss": 0.2286,
+      "step": 4190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42356405355596793,
+      "learning_rate": 1.775946116874931e-05,
+      "loss": 0.3589,
+      "step": 4191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5552779771392584,
+      "learning_rate": 1.7758287172231333e-05,
+      "loss": 0.4024,
+      "step": 4192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37558968188355335,
+      "learning_rate": 1.77571129070442e-05,
+      "loss": 0.2241,
+      "step": 4193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25518944264314564,
+      "learning_rate": 1.775593837322858e-05,
+      "loss": 0.2089,
+      "step": 4194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.00903247293503,
+      "learning_rate": 1.775476357082515e-05,
+      "loss": 0.4944,
+      "step": 4195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2975944794423369,
+      "learning_rate": 1.7753588499874592e-05,
+      "loss": 0.2273,
+      "step": 4196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44397822897049244,
+      "learning_rate": 1.7752413160417597e-05,
+      "loss": 0.357,
+      "step": 4197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1675312656602383,
+      "learning_rate": 1.7751237552494867e-05,
+      "loss": 0.862,
+      "step": 4198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2532797489976462,
+      "learning_rate": 1.7750061676147114e-05,
+      "loss": 0.108,
+      "step": 4199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4934846294571999,
+      "learning_rate": 1.774888553141506e-05,
+      "loss": 0.3664,
+      "step": 4200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3683129833102259,
+      "learning_rate": 1.7747709118339428e-05,
+      "loss": 0.2941,
+      "step": 4201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1727462583908863,
+      "learning_rate": 1.7746532436960965e-05,
+      "loss": 0.653,
+      "step": 4202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3065183783693479,
+      "learning_rate": 1.7745355487320418e-05,
+      "loss": 0.2344,
+      "step": 4203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41090849776968874,
+      "learning_rate": 1.7744178269458547e-05,
+      "loss": 0.3612,
+      "step": 4204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6804604082844093,
+      "learning_rate": 1.774300078341611e-05,
+      "loss": 0.5032,
+      "step": 4205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3564856518489956,
+      "learning_rate": 1.7741823029233892e-05,
+      "loss": 0.2573,
+      "step": 4206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3123261819071861,
+      "learning_rate": 1.7740645006952674e-05,
+      "loss": 0.2185,
+      "step": 4207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36540726586784045,
+      "learning_rate": 1.773946671661325e-05,
+      "loss": 0.2873,
+      "step": 4208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3851981339848704,
+      "learning_rate": 1.773828815825643e-05,
+      "loss": 0.2368,
+      "step": 4209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.403401900227148,
+      "learning_rate": 1.773710933192302e-05,
+      "loss": 0.8898,
+      "step": 4210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8506443084440741,
+      "learning_rate": 1.7735930237653853e-05,
+      "loss": 0.4175,
+      "step": 4211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3401308629226619,
+      "learning_rate": 1.773475087548975e-05,
+      "loss": 0.2571,
+      "step": 4212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3457458099433843,
+      "learning_rate": 1.7733571245471557e-05,
+      "loss": 0.2495,
+      "step": 4213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3963265595086062,
+      "learning_rate": 1.7732391347640125e-05,
+      "loss": 0.2667,
+      "step": 4214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41455830175953723,
+      "learning_rate": 1.7731211182036312e-05,
+      "loss": 0.3137,
+      "step": 4215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40500875371408623,
+      "learning_rate": 1.773003074870099e-05,
+      "loss": 0.299,
+      "step": 4216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5403576844166071,
+      "learning_rate": 1.7728850047675035e-05,
+      "loss": 0.35,
+      "step": 4217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4087263684779261,
+      "learning_rate": 1.7727669078999336e-05,
+      "loss": 0.3484,
+      "step": 4218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2268995265041368,
+      "learning_rate": 1.772648784271479e-05,
+      "loss": 0.1586,
+      "step": 4219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38627405176100876,
+      "learning_rate": 1.7725306338862298e-05,
+      "loss": 0.3399,
+      "step": 4220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34508423192036813,
+      "learning_rate": 1.7724124567482782e-05,
+      "loss": 0.2773,
+      "step": 4221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5635387172753441,
+      "learning_rate": 1.7722942528617163e-05,
+      "loss": 0.443,
+      "step": 4222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6278867989446979,
+      "learning_rate": 1.772176022230638e-05,
+      "loss": 0.4841,
+      "step": 4223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3090845892934344,
+      "learning_rate": 1.7720577648591368e-05,
+      "loss": 0.2715,
+      "step": 4224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43203694883147464,
+      "learning_rate": 1.771939480751309e-05,
+      "loss": 0.2774,
+      "step": 4225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28886362206526994,
+      "learning_rate": 1.7718211699112496e-05,
+      "loss": 0.1862,
+      "step": 4226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35675105970492776,
+      "learning_rate": 1.7717028323430562e-05,
+      "loss": 0.2986,
+      "step": 4227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4192480731244511,
+      "learning_rate": 1.7715844680508273e-05,
+      "loss": 0.3561,
+      "step": 4228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.954576314104161,
+      "learning_rate": 1.7714660770386615e-05,
+      "loss": 0.3737,
+      "step": 4229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37465111339648155,
+      "learning_rate": 1.771347659310658e-05,
+      "loss": 0.3111,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.154325732369229,
+      "learning_rate": 1.7712292148709188e-05,
+      "loss": 0.667,
+      "step": 4231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23526929972174873,
+      "learning_rate": 1.7711107437235453e-05,
+      "loss": 0.1821,
+      "step": 4232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4190335124027483,
+      "learning_rate": 1.7709922458726395e-05,
+      "loss": 0.3299,
+      "step": 4233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8046301903471953,
+      "learning_rate": 1.770873721322305e-05,
+      "loss": 0.5668,
+      "step": 4234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3958648304902932,
+      "learning_rate": 1.7707551700766474e-05,
+      "loss": 0.2981,
+      "step": 4235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4148428047558691,
+      "learning_rate": 1.770636592139771e-05,
+      "loss": 0.3152,
+      "step": 4236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4855183491267279,
+      "learning_rate": 1.7705179875157826e-05,
+      "loss": 0.3586,
+      "step": 4237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24397917134812547,
+      "learning_rate": 1.7703993562087895e-05,
+      "loss": 0.1169,
+      "step": 4238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3776877212817177,
+      "learning_rate": 1.7702806982229e-05,
+      "loss": 0.2973,
+      "step": 4239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5154970926323544,
+      "learning_rate": 1.7701620135622228e-05,
+      "loss": 0.3764,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.042901070896944,
+      "learning_rate": 1.7700433022308684e-05,
+      "loss": 0.635,
+      "step": 4241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3494430310945817,
+      "learning_rate": 1.7699245642329473e-05,
+      "loss": 0.2064,
+      "step": 4242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1972989287454718,
+      "learning_rate": 1.7698057995725717e-05,
+      "loss": 0.6586,
+      "step": 4243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3358791750280596,
+      "learning_rate": 1.7696870082538544e-05,
+      "loss": 0.2379,
+      "step": 4244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4175936922148545,
+      "learning_rate": 1.769568190280909e-05,
+      "loss": 0.2467,
+      "step": 4245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2074876580005733,
+      "learning_rate": 1.7694493456578503e-05,
+      "loss": 0.4767,
+      "step": 4246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4768857304128634,
+      "learning_rate": 1.769330474388794e-05,
+      "loss": 0.3806,
+      "step": 4247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3347435221937712,
+      "learning_rate": 1.7692115764778564e-05,
+      "loss": 0.2254,
+      "step": 4248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2119371591283326,
+      "learning_rate": 1.7690926519291548e-05,
+      "loss": 0.6777,
+      "step": 4249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5732297049816935,
+      "learning_rate": 1.7689737007468082e-05,
+      "loss": 0.3127,
+      "step": 4250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4110880892865378,
+      "learning_rate": 1.768854722934935e-05,
+      "loss": 0.2711,
+      "step": 4251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5346347752222697,
+      "learning_rate": 1.7687357184976558e-05,
+      "loss": 0.304,
+      "step": 4252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4209249612683654,
+      "learning_rate": 1.7686166874390916e-05,
+      "loss": 0.3113,
+      "step": 4253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4029524455386135,
+      "learning_rate": 1.768497629763365e-05,
+      "loss": 0.284,
+      "step": 4254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39416868752265927,
+      "learning_rate": 1.7683785454745983e-05,
+      "loss": 0.267,
+      "step": 4255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4629564240237894,
+      "learning_rate": 1.768259434576916e-05,
+      "loss": 0.3873,
+      "step": 4256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39823641313634994,
+      "learning_rate": 1.768140297074442e-05,
+      "loss": 0.3366,
+      "step": 4257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44095663559634946,
+      "learning_rate": 1.7680211329713027e-05,
+      "loss": 0.3567,
+      "step": 4258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3111575942167663,
+      "learning_rate": 1.7679019422716244e-05,
+      "loss": 0.2723,
+      "step": 4259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32667092720747753,
+      "learning_rate": 1.767782724979535e-05,
+      "loss": 0.2803,
+      "step": 4260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3594877209168396,
+      "learning_rate": 1.767663481099163e-05,
+      "loss": 0.15,
+      "step": 4261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5838500192099852,
+      "learning_rate": 1.7675442106346377e-05,
+      "loss": 0.4736,
+      "step": 4262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34204796905833723,
+      "learning_rate": 1.7674249135900892e-05,
+      "loss": 0.3,
+      "step": 4263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4358076489958484,
+      "learning_rate": 1.767305589969649e-05,
+      "loss": 0.3492,
+      "step": 4264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.21427413315468624,
+      "learning_rate": 1.7671862397774494e-05,
+      "loss": 0.1329,
+      "step": 4265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35988453024304295,
+      "learning_rate": 1.767066863017623e-05,
+      "loss": 0.2818,
+      "step": 4266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9682883111015376,
+      "learning_rate": 1.766947459694304e-05,
+      "loss": 0.5893,
+      "step": 4267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3709235509398682,
+      "learning_rate": 1.766828029811628e-05,
+      "loss": 0.3169,
+      "step": 4268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5406561017554826,
+      "learning_rate": 1.7667085733737298e-05,
+      "loss": 0.3701,
+      "step": 4269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4233765597253248,
+      "learning_rate": 1.7665890903847468e-05,
+      "loss": 0.3483,
+      "step": 4270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2603354062606371,
+      "learning_rate": 1.7664695808488164e-05,
+      "loss": 0.2296,
+      "step": 4271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4590253911487976,
+      "learning_rate": 1.766350044770078e-05,
+      "loss": 0.2895,
+      "step": 4272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3741916057341281,
+      "learning_rate": 1.76623048215267e-05,
+      "loss": 0.3108,
+      "step": 4273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7421170186002759,
+      "learning_rate": 1.7661108930007334e-05,
+      "loss": 0.3889,
+      "step": 4274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45537247750720905,
+      "learning_rate": 1.7659912773184095e-05,
+      "loss": 0.2839,
+      "step": 4275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3989938266714445,
+      "learning_rate": 1.7658716351098407e-05,
+      "loss": 0.3258,
+      "step": 4276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.318451272676308,
+      "learning_rate": 1.76575196637917e-05,
+      "loss": 0.7469,
+      "step": 4277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22641040750112476,
+      "learning_rate": 1.7656322711305417e-05,
+      "loss": 0.1216,
+      "step": 4278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3801159653441589,
+      "learning_rate": 1.7655125493681012e-05,
+      "loss": 0.3474,
+      "step": 4279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5049163226978106,
+      "learning_rate": 1.7653928010959936e-05,
+      "loss": 0.3951,
+      "step": 4280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3293722513867542,
+      "learning_rate": 1.765273026318366e-05,
+      "loss": 0.2225,
+      "step": 4281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.349210674604087,
+      "learning_rate": 1.7651532250393666e-05,
+      "loss": 0.7544,
+      "step": 4282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3886610404525146,
+      "learning_rate": 1.7650333972631443e-05,
+      "loss": 0.3501,
+      "step": 4283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24857630478813894,
+      "learning_rate": 1.7649135429938477e-05,
+      "loss": 0.1807,
+      "step": 4284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4471098422892603,
+      "learning_rate": 1.764793662235628e-05,
+      "loss": 0.3165,
+      "step": 4285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6148974597430013,
+      "learning_rate": 1.7646737549926376e-05,
+      "loss": 0.4766,
+      "step": 4286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37646323777327495,
+      "learning_rate": 1.764553821269027e-05,
+      "loss": 0.216,
+      "step": 4287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39924535706099146,
+      "learning_rate": 1.764433861068951e-05,
+      "loss": 0.3246,
+      "step": 4288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0115253786277327,
+      "learning_rate": 1.764313874396563e-05,
+      "loss": 0.6713,
+      "step": 4289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37546513680467763,
+      "learning_rate": 1.7641938612560182e-05,
+      "loss": 0.2153,
+      "step": 4290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31499422778096564,
+      "learning_rate": 1.7640738216514733e-05,
+      "loss": 0.293,
+      "step": 4291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.341004193409986,
+      "learning_rate": 1.7639537555870844e-05,
+      "loss": 0.2931,
+      "step": 4292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5975754148604239,
+      "learning_rate": 1.7638336630670102e-05,
+      "loss": 0.3508,
+      "step": 4293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4147059353112628,
+      "learning_rate": 1.763713544095409e-05,
+      "loss": 0.299,
+      "step": 4294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3928079898041313,
+      "learning_rate": 1.7635933986764403e-05,
+      "loss": 0.3508,
+      "step": 4295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.559386737857805,
+      "learning_rate": 1.7634732268142652e-05,
+      "loss": 0.3621,
+      "step": 4296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2935297955911469,
+      "learning_rate": 1.7633530285130452e-05,
+      "loss": 0.2541,
+      "step": 4297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4767343833347403,
+      "learning_rate": 1.7632328037769423e-05,
+      "loss": 0.2457,
+      "step": 4298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3746108706148988,
+      "learning_rate": 1.7631125526101206e-05,
+      "loss": 0.3049,
+      "step": 4299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4020633558516471,
+      "learning_rate": 1.7629922750167437e-05,
+      "loss": 0.2688,
+      "step": 4300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8783609175595337,
+      "learning_rate": 1.7628719710009777e-05,
+      "loss": 0.6493,
+      "step": 4301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3907321374100405,
+      "learning_rate": 1.7627516405669876e-05,
+      "loss": 0.3237,
+      "step": 4302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49895888305836794,
+      "learning_rate": 1.7626312837189412e-05,
+      "loss": 0.4026,
+      "step": 4303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25989485619288133,
+      "learning_rate": 1.7625109004610065e-05,
+      "loss": 0.1937,
+      "step": 4304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7321965943129197,
+      "learning_rate": 1.7623904907973515e-05,
+      "loss": 0.395,
+      "step": 4305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.43686726649567725,
+      "learning_rate": 1.762270054732147e-05,
+      "loss": 0.3508,
+      "step": 4306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37833068558361926,
+      "learning_rate": 1.7621495922695633e-05,
+      "loss": 0.307,
+      "step": 4307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6682080526735688,
+      "learning_rate": 1.7620291034137718e-05,
+      "loss": 0.4271,
+      "step": 4308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3964426570352359,
+      "learning_rate": 1.7619085881689454e-05,
+      "loss": 0.3013,
+      "step": 4309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23222205472935048,
+      "learning_rate": 1.761788046539257e-05,
+      "loss": 0.1052,
+      "step": 4310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3912585076504096,
+      "learning_rate": 1.7616674785288815e-05,
+      "loss": 0.301,
+      "step": 4311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.47490678407484405,
+      "learning_rate": 1.761546884141994e-05,
+      "loss": 0.3829,
+      "step": 4312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7637011922535032,
+      "learning_rate": 1.761426263382771e-05,
+      "loss": 0.3716,
+      "step": 4313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.506704534728562,
+      "learning_rate": 1.761305616255389e-05,
+      "loss": 0.3931,
+      "step": 4314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36459501700346764,
+      "learning_rate": 1.761184942764026e-05,
+      "loss": 0.3025,
+      "step": 4315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.38492087075628456,
+      "learning_rate": 1.761064242912861e-05,
+      "loss": 0.2247,
+      "step": 4316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2590479572690357,
+      "learning_rate": 1.7609435167060745e-05,
+      "loss": 0.1378,
+      "step": 4317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.505463936850372,
+      "learning_rate": 1.7608227641478467e-05,
+      "loss": 0.3552,
+      "step": 4318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4637726016282319,
+      "learning_rate": 1.760701985242359e-05,
+      "loss": 0.3487,
+      "step": 4319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9181606038737803,
+      "learning_rate": 1.7605811799937946e-05,
+      "loss": 0.3159,
+      "step": 4320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4582402750612981,
+      "learning_rate": 1.7604603484063363e-05,
+      "loss": 0.325,
+      "step": 4321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3085161695929028,
+      "learning_rate": 1.760339490484169e-05,
+      "loss": 0.2247,
+      "step": 4322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3913027721852565,
+      "learning_rate": 1.760218606231478e-05,
+      "loss": 0.2511,
+      "step": 4323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5046921571413169,
+      "learning_rate": 1.7600976956524493e-05,
+      "loss": 0.3392,
+      "step": 4324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4373319914356162,
+      "learning_rate": 1.7599767587512698e-05,
+      "loss": 0.4993,
+      "step": 4325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5257980257453084,
+      "learning_rate": 1.7598557955321282e-05,
+      "loss": 0.2709,
+      "step": 4326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32672331423731715,
+      "learning_rate": 1.7597348059992128e-05,
+      "loss": 0.3039,
+      "step": 4327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2592941345566562,
+      "learning_rate": 1.7596137901567138e-05,
+      "loss": 0.1403,
+      "step": 4328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4062233302214164,
+      "learning_rate": 1.759492748008822e-05,
+      "loss": 0.8632,
+      "step": 4329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6436373842109813,
+      "learning_rate": 1.759371679559729e-05,
+      "loss": 0.2666,
+      "step": 4330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5413600815168292,
+      "learning_rate": 1.759250584813627e-05,
+      "loss": 0.3359,
+      "step": 4331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6795446930850362,
+      "learning_rate": 1.7591294637747104e-05,
+      "loss": 0.4545,
+      "step": 4332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34797795924646585,
+      "learning_rate": 1.7590083164471728e-05,
+      "loss": 0.2473,
+      "step": 4333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4751834716100019,
+      "learning_rate": 1.75888714283521e-05,
+      "loss": 0.2814,
+      "step": 4334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4792771622907987,
+      "learning_rate": 1.758765942943018e-05,
+      "loss": 0.3511,
+      "step": 4335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4409034335516681,
+      "learning_rate": 1.7586447167747943e-05,
+      "loss": 0.2339,
+      "step": 4336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0216680780688296,
+      "learning_rate": 1.7585234643347363e-05,
+      "loss": 0.4759,
+      "step": 4337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3495006405637813,
+      "learning_rate": 1.7584021856270435e-05,
+      "loss": 0.2516,
+      "step": 4338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.399030800710068,
+      "learning_rate": 1.7582808806559154e-05,
+      "loss": 0.3012,
+      "step": 4339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33002710223861065,
+      "learning_rate": 1.7581595494255533e-05,
+      "loss": 0.2138,
+      "step": 4340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6027956584907409,
+      "learning_rate": 1.7580381919401586e-05,
+      "loss": 0.4314,
+      "step": 4341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3981500370131176,
+      "learning_rate": 1.757916808203934e-05,
+      "loss": 0.3312,
+      "step": 4342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36799863739960603,
+      "learning_rate": 1.757795398221083e-05,
+      "loss": 0.3211,
+      "step": 4343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44567616701273444,
+      "learning_rate": 1.7576739619958096e-05,
+      "loss": 0.2847,
+      "step": 4344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37317146489865405,
+      "learning_rate": 1.7575524995323195e-05,
+      "loss": 0.2739,
+      "step": 4345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.38858376631551034,
+      "learning_rate": 1.757431010834819e-05,
+      "loss": 0.2789,
+      "step": 4346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7434345776855413,
+      "learning_rate": 1.7573094959075148e-05,
+      "loss": 0.5022,
+      "step": 4347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3514566700372044,
+      "learning_rate": 1.757187954754616e-05,
+      "loss": 0.2869,
+      "step": 4348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7291519504367259,
+      "learning_rate": 1.7570663873803305e-05,
+      "loss": 0.4462,
+      "step": 4349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2717965523574744,
+      "learning_rate": 1.7569447937888686e-05,
+      "loss": 0.2532,
+      "step": 4350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35466411632539085,
+      "learning_rate": 1.756823173984441e-05,
+      "loss": 0.29,
+      "step": 4351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4631381574190476,
+      "learning_rate": 1.7567015279712598e-05,
+      "loss": 0.7348,
+      "step": 4352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8172752249272174,
+      "learning_rate": 1.756579855753537e-05,
+      "loss": 0.3552,
+      "step": 4353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3753765285233513,
+      "learning_rate": 1.756458157335486e-05,
+      "loss": 0.3274,
+      "step": 4354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39089010889022724,
+      "learning_rate": 1.756336432721322e-05,
+      "loss": 0.3154,
+      "step": 4355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23367638912187802,
+      "learning_rate": 1.7562146819152595e-05,
+      "loss": 0.1237,
+      "step": 4356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4181181779260803,
+      "learning_rate": 1.7560929049215155e-05,
+      "loss": 0.3227,
+      "step": 4357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8631313679687664,
+      "learning_rate": 1.7559711017443062e-05,
+      "loss": 0.4784,
+      "step": 4358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41323073775920643,
+      "learning_rate": 1.7558492723878507e-05,
+      "loss": 0.3094,
+      "step": 4359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3782931761755461,
+      "learning_rate": 1.755727416856367e-05,
+      "loss": 0.3099,
+      "step": 4360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2000243266515298,
+      "learning_rate": 1.7556055351540757e-05,
+      "loss": 0.8228,
+      "step": 4361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2778527231030494,
+      "learning_rate": 1.7554836272851967e-05,
+      "loss": 0.1797,
+      "step": 4362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34868729040736435,
+      "learning_rate": 1.7553616932539522e-05,
+      "loss": 0.2861,
+      "step": 4363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0895145164950082,
+      "learning_rate": 1.7552397330645654e-05,
+      "loss": 0.5217,
+      "step": 4364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6910131523731271,
+      "learning_rate": 1.7551177467212585e-05,
+      "loss": 0.508,
+      "step": 4365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3367083352550415,
+      "learning_rate": 1.7549957342282567e-05,
+      "loss": 0.2234,
+      "step": 4366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4083921331597858,
+      "learning_rate": 1.7548736955897852e-05,
+      "loss": 0.3233,
+      "step": 4367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.310682312374531,
+      "learning_rate": 1.75475163081007e-05,
+      "loss": 0.1883,
+      "step": 4368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35150586064186157,
+      "learning_rate": 1.7546295398933383e-05,
+      "loss": 0.2042,
+      "step": 4369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1970685231744054,
+      "learning_rate": 1.754507422843818e-05,
+      "loss": 0.5545,
+      "step": 4370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.390038478824965,
+      "learning_rate": 1.7543852796657382e-05,
+      "loss": 0.3434,
+      "step": 4371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3332006509863933,
+      "learning_rate": 1.7542631103633284e-05,
+      "loss": 0.1989,
+      "step": 4372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2245310547498793,
+      "learning_rate": 1.7541409149408198e-05,
+      "loss": 0.8196,
+      "step": 4373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2594242254872551,
+      "learning_rate": 1.7540186934024434e-05,
+      "loss": 0.2391,
+      "step": 4374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3663500648240596,
+      "learning_rate": 1.7538964457524326e-05,
+      "loss": 0.1984,
+      "step": 4375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3797280623086665,
+      "learning_rate": 1.7537741719950197e-05,
+      "loss": 0.4802,
+      "step": 4376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9288280071064241,
+      "learning_rate": 1.75365187213444e-05,
+      "loss": 0.5739,
+      "step": 4377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4592788301354964,
+      "learning_rate": 1.7535295461749285e-05,
+      "loss": 0.3199,
+      "step": 4378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.371263615108485,
+      "learning_rate": 1.753407194120721e-05,
+      "loss": 0.2718,
+      "step": 4379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4478970966584819,
+      "learning_rate": 1.753284815976055e-05,
+      "loss": 0.2777,
+      "step": 4380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.43708047320580123,
+      "learning_rate": 1.7531624117451678e-05,
+      "loss": 0.263,
+      "step": 4381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7647331876742203,
+      "learning_rate": 1.753039981432299e-05,
+      "loss": 0.3161,
+      "step": 4382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2246597599600573,
+      "learning_rate": 1.7529175250416878e-05,
+      "loss": 0.5649,
+      "step": 4383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4019441310451681,
+      "learning_rate": 1.7527950425775753e-05,
+      "loss": 0.2567,
+      "step": 4384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.932022146410723,
+      "learning_rate": 1.7526725340442028e-05,
+      "loss": 0.5759,
+      "step": 4385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31841220554720956,
+      "learning_rate": 1.7525499994458124e-05,
+      "loss": 0.2613,
+      "step": 4386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37423694836917065,
+      "learning_rate": 1.7524274387866483e-05,
+      "loss": 0.2822,
+      "step": 4387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5350548955670364,
+      "learning_rate": 1.7523048520709543e-05,
+      "loss": 0.2867,
+      "step": 4388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8947754710817244,
+      "learning_rate": 1.7521822393029758e-05,
+      "loss": 0.5852,
+      "step": 4389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3998881128220424,
+      "learning_rate": 1.7520596004869584e-05,
+      "loss": 0.2702,
+      "step": 4390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5361072708652229,
+      "learning_rate": 1.7519369356271492e-05,
+      "loss": 0.3904,
+      "step": 4391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8123518823570961,
+      "learning_rate": 1.751814244727797e-05,
+      "loss": 0.3895,
+      "step": 4392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6059125687840098,
+      "learning_rate": 1.751691527793149e-05,
+      "loss": 0.339,
+      "step": 4393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3313177646320267,
+      "learning_rate": 1.7515687848274562e-05,
+      "loss": 0.2728,
+      "step": 4394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35552239325197443,
+      "learning_rate": 1.7514460158349686e-05,
+      "loss": 0.2089,
+      "step": 4395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4505421213963791,
+      "learning_rate": 1.7513232208199378e-05,
+      "loss": 0.3099,
+      "step": 4396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6644199425595185,
+      "learning_rate": 1.751200399786616e-05,
+      "loss": 0.4381,
+      "step": 4397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4423052272917962,
+      "learning_rate": 1.7510775527392566e-05,
+      "loss": 0.2984,
+      "step": 4398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6068964287449121,
+      "learning_rate": 1.7509546796821144e-05,
+      "loss": 0.3403,
+      "step": 4399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44069931145334174,
+      "learning_rate": 1.7508317806194436e-05,
+      "loss": 0.316,
+      "step": 4400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30685445519676646,
+      "learning_rate": 1.7507088555555003e-05,
+      "loss": 0.1484,
+      "step": 4401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.319991819668123,
+      "learning_rate": 1.750585904494542e-05,
+      "loss": 0.2779,
+      "step": 4402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5975686561638618,
+      "learning_rate": 1.7504629274408257e-05,
+      "loss": 0.4789,
+      "step": 4403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8063943453403252,
+      "learning_rate": 1.750339924398611e-05,
+      "loss": 0.4771,
+      "step": 4404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3169912413687381,
+      "learning_rate": 1.7502168953721564e-05,
+      "loss": 0.2506,
+      "step": 4405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5907726571115687,
+      "learning_rate": 1.7500938403657235e-05,
+      "loss": 0.4054,
+      "step": 4406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28948065757597125,
+      "learning_rate": 1.7499707593835728e-05,
+      "loss": 0.1955,
+      "step": 4407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.346293298851469,
+      "learning_rate": 1.749847652429967e-05,
+      "loss": 0.2249,
+      "step": 4408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1109798245103102,
+      "learning_rate": 1.7497245195091694e-05,
+      "loss": 0.6656,
+      "step": 4409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49764755117641957,
+      "learning_rate": 1.749601360625444e-05,
+      "loss": 0.3518,
+      "step": 4410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6586021135670197,
+      "learning_rate": 1.7494781757830554e-05,
+      "loss": 0.2586,
+      "step": 4411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.415911984220887,
+      "learning_rate": 1.74935496498627e-05,
+      "loss": 0.3387,
+      "step": 4412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3528941062652166,
+      "learning_rate": 1.7492317282393543e-05,
+      "loss": 0.1973,
+      "step": 4413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5379242200023283,
+      "learning_rate": 1.7491084655465762e-05,
+      "loss": 0.3619,
+      "step": 4414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4547219643409678,
+      "learning_rate": 1.748985176912204e-05,
+      "loss": 0.2714,
+      "step": 4415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9862381216448781,
+      "learning_rate": 1.7488618623405075e-05,
+      "loss": 0.5093,
+      "step": 4416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6014511226082759,
+      "learning_rate": 1.748738521835757e-05,
+      "loss": 0.3081,
+      "step": 4417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42678893832728415,
+      "learning_rate": 1.7486151554022233e-05,
+      "loss": 0.266,
+      "step": 4418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29200160850606793,
+      "learning_rate": 1.748491763044179e-05,
+      "loss": 0.1875,
+      "step": 4419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.609705572125106,
+      "learning_rate": 1.7483683447658976e-05,
+      "loss": 0.3825,
+      "step": 4420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6734514732700916,
+      "learning_rate": 1.748244900571652e-05,
+      "loss": 0.3149,
+      "step": 4421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5674551833098238,
+      "learning_rate": 1.748121430465718e-05,
+      "loss": 0.3457,
+      "step": 4422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5692437298539499,
+      "learning_rate": 1.747997934452371e-05,
+      "loss": 0.3981,
+      "step": 4423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40695746244632547,
+      "learning_rate": 1.7478744125358877e-05,
+      "loss": 0.2523,
+      "step": 4424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26111116306772175,
+      "learning_rate": 1.7477508647205456e-05,
+      "loss": 0.2255,
+      "step": 4425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3739424274051894,
+      "learning_rate": 1.7476272910106233e-05,
+      "loss": 0.2915,
+      "step": 4426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5242746749994995,
+      "learning_rate": 1.7475036914104e-05,
+      "loss": 0.4205,
+      "step": 4427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5970732747562592,
+      "learning_rate": 1.747380065924156e-05,
+      "loss": 0.409,
+      "step": 4428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7671707463296635,
+      "learning_rate": 1.7472564145561725e-05,
+      "loss": 0.4088,
+      "step": 4429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3501468041889073,
+      "learning_rate": 1.7471327373107317e-05,
+      "loss": 0.2827,
+      "step": 4430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3703347723536773,
+      "learning_rate": 1.747009034192116e-05,
+      "loss": 0.1713,
+      "step": 4431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6832880354899038,
+      "learning_rate": 1.7468853052046095e-05,
+      "loss": 0.4288,
+      "step": 4432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.385508552267273,
+      "learning_rate": 1.7467615503524973e-05,
+      "loss": 0.3287,
+      "step": 4433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2565484503974381,
+      "learning_rate": 1.7466377696400646e-05,
+      "loss": 0.2168,
+      "step": 4434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6083624161878634,
+      "learning_rate": 1.746513963071598e-05,
+      "loss": 0.374,
+      "step": 4435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4290866249248875,
+      "learning_rate": 1.746390130651385e-05,
+      "loss": 0.3343,
+      "step": 4436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4210646856691826,
+      "learning_rate": 1.746266272383714e-05,
+      "loss": 0.1923,
+      "step": 4437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31382665898340373,
+      "learning_rate": 1.746142388272874e-05,
+      "loss": 0.2997,
+      "step": 4438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40780261576434734,
+      "learning_rate": 1.746018478323155e-05,
+      "loss": 0.3683,
+      "step": 4439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42413178269441665,
+      "learning_rate": 1.7458945425388484e-05,
+      "loss": 0.2996,
+      "step": 4440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3334174550196039,
+      "learning_rate": 1.7457705809242455e-05,
+      "loss": 0.1964,
+      "step": 4441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3571659636822857,
+      "learning_rate": 1.74564659348364e-05,
+      "loss": 0.3006,
+      "step": 4442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8945360836656251,
+      "learning_rate": 1.7455225802213246e-05,
+      "loss": 0.5754,
+      "step": 4443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5110983783626839,
+      "learning_rate": 1.7453985411415945e-05,
+      "loss": 0.3305,
+      "step": 4444
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40930073398439676,
+      "learning_rate": 1.745274476248745e-05,
+      "loss": 0.3161,
+      "step": 4445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2968822772265499,
+      "learning_rate": 1.7451503855470722e-05,
+      "loss": 0.2828,
+      "step": 4446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27604539059250005,
+      "learning_rate": 1.745026269040874e-05,
+      "loss": 0.1201,
+      "step": 4447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4091950250734807,
+      "learning_rate": 1.744902126734448e-05,
+      "loss": 0.305,
+      "step": 4448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4688130768266721,
+      "learning_rate": 1.744777958632093e-05,
+      "loss": 0.3995,
+      "step": 4449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4199210326535034,
+      "learning_rate": 1.74465376473811e-05,
+      "loss": 0.2857,
+      "step": 4450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.36946539820255947,
+      "learning_rate": 1.7445295450567985e-05,
+      "loss": 0.3241,
+      "step": 4451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30417635766535905,
+      "learning_rate": 1.7444052995924612e-05,
+      "loss": 0.2137,
+      "step": 4452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.388316362774309,
+      "learning_rate": 1.7442810283494002e-05,
+      "loss": 0.3331,
+      "step": 4453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.33429814252358525,
+      "learning_rate": 1.7441567313319194e-05,
+      "loss": 0.2181,
+      "step": 4454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7626510299733518,
+      "learning_rate": 1.7440324085443227e-05,
+      "loss": 0.4604,
+      "step": 4455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5831532138070068,
+      "learning_rate": 1.7439080599909163e-05,
+      "loss": 0.4221,
+      "step": 4456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29553524419263844,
+      "learning_rate": 1.743783685676005e-05,
+      "loss": 0.2498,
+      "step": 4457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49518512762179856,
+      "learning_rate": 1.743659285603897e-05,
+      "loss": 0.3983,
+      "step": 4458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25390198843159745,
+      "learning_rate": 1.7435348597789e-05,
+      "loss": 0.1813,
+      "step": 4459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.36903051665820324,
+      "learning_rate": 1.7434104082053227e-05,
+      "loss": 0.2264,
+      "step": 4460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.848917048411256,
+      "learning_rate": 1.743285930887475e-05,
+      "loss": 0.4192,
+      "step": 4461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5239600697847725,
+      "learning_rate": 1.7431614278296672e-05,
+      "loss": 0.3618,
+      "step": 4462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39962990775437407,
+      "learning_rate": 1.7430368990362114e-05,
+      "loss": 0.2368,
+      "step": 4463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41735657916516267,
+      "learning_rate": 1.7429123445114196e-05,
+      "loss": 0.2857,
+      "step": 4464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32698671191140255,
+      "learning_rate": 1.7427877642596053e-05,
+      "loss": 0.2875,
+      "step": 4465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3624786033802996,
+      "learning_rate": 1.7426631582850827e-05,
+      "loss": 0.2762,
+      "step": 4466
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5845800939907507,
+      "learning_rate": 1.742538526592167e-05,
+      "loss": 0.4098,
+      "step": 4467
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9870741959193776,
+      "learning_rate": 1.742413869185174e-05,
+      "loss": 0.6744,
+      "step": 4468
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42257891785206114,
+      "learning_rate": 1.7422891860684202e-05,
+      "loss": 0.2697,
+      "step": 4469
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32188754467568104,
+      "learning_rate": 1.7421644772462247e-05,
+      "loss": 0.2059,
+      "step": 4470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4304055386861345,
+      "learning_rate": 1.7420397427229045e-05,
+      "loss": 0.2905,
+      "step": 4471
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.33978838217773394,
+      "learning_rate": 1.7419149825027802e-05,
+      "loss": 0.3002,
+      "step": 4472
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3788118884308564,
+      "learning_rate": 1.7417901965901717e-05,
+      "loss": 0.3153,
+      "step": 4473
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.521900293581829,
+      "learning_rate": 1.7416653849894008e-05,
+      "loss": 0.3964,
+      "step": 4474
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3665116506708937,
+      "learning_rate": 1.7415405477047895e-05,
+      "loss": 0.2777,
+      "step": 4475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9883665529741489,
+      "learning_rate": 1.741415684740661e-05,
+      "loss": 0.4042,
+      "step": 4476
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24629002200117167,
+      "learning_rate": 1.741290796101339e-05,
+      "loss": 0.2254,
+      "step": 4477
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3537787580716675,
+      "learning_rate": 1.7411658817911487e-05,
+      "loss": 0.3092,
+      "step": 4478
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6388605821733183,
+      "learning_rate": 1.741040941814416e-05,
+      "loss": 0.4883,
+      "step": 4479
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6188260255813043,
+      "learning_rate": 1.740915976175467e-05,
+      "loss": 0.3133,
+      "step": 4480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4176908201497853,
+      "learning_rate": 1.74079098487863e-05,
+      "loss": 0.3111,
+      "step": 4481
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45129474320870905,
+      "learning_rate": 1.7406659679282326e-05,
+      "loss": 0.3424,
+      "step": 4482
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4174295494281297,
+      "learning_rate": 1.740540925328605e-05,
+      "loss": 0.2623,
+      "step": 4483
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3767565914818703,
+      "learning_rate": 1.7404158570840765e-05,
+      "loss": 0.282,
+      "step": 4484
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2783113500296502,
+      "learning_rate": 1.7402907631989793e-05,
+      "loss": 0.2917,
+      "step": 4485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9873955581325996,
+      "learning_rate": 1.7401656436776445e-05,
+      "loss": 0.4259,
+      "step": 4486
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3845423744262796,
+      "learning_rate": 1.740040498524405e-05,
+      "loss": 0.2788,
+      "step": 4487
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7272016165043925,
+      "learning_rate": 1.7399153277435954e-05,
+      "loss": 0.5691,
+      "step": 4488
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40153871463464896,
+      "learning_rate": 1.7397901313395497e-05,
+      "loss": 0.2958,
+      "step": 4489
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39987319741953187,
+      "learning_rate": 1.7396649093166034e-05,
+      "loss": 0.3051,
+      "step": 4490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35067447545845665,
+      "learning_rate": 1.7395396616790932e-05,
+      "loss": 0.1451,
+      "step": 4491
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9550529008649429,
+      "learning_rate": 1.7394143884313562e-05,
+      "loss": 0.7368,
+      "step": 4492
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30621419928245114,
+      "learning_rate": 1.7392890895777305e-05,
+      "loss": 0.2344,
+      "step": 4493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4312493233871788,
+      "learning_rate": 1.7391637651225556e-05,
+      "loss": 0.7849,
+      "step": 4494
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6076999334668871,
+      "learning_rate": 1.7390384150701715e-05,
+      "loss": 0.4489,
+      "step": 4495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38957330875145113,
+      "learning_rate": 1.738913039424919e-05,
+      "loss": 0.1873,
+      "step": 4496
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34020086053360765,
+      "learning_rate": 1.7387876381911395e-05,
+      "loss": 0.2616,
+      "step": 4497
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43665721845984257,
+      "learning_rate": 1.7386622113731758e-05,
+      "loss": 0.3332,
+      "step": 4498
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35317580614898686,
+      "learning_rate": 1.7385367589753714e-05,
+      "loss": 0.2111,
+      "step": 4499
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8179148641200064,
+      "learning_rate": 1.738411281002071e-05,
+      "loss": 0.4621,
+      "step": 4500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37262485183921623,
+      "learning_rate": 1.7382857774576195e-05,
+      "loss": 0.3425,
+      "step": 4501
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4434924061456271,
+      "learning_rate": 1.7381602483463637e-05,
+      "loss": 0.2924,
+      "step": 4502
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23825968934381375,
+      "learning_rate": 1.73803469367265e-05,
+      "loss": 0.1482,
+      "step": 4503
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3993363887214298,
+      "learning_rate": 1.7379091134408265e-05,
+      "loss": 0.9296,
+      "step": 4504
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3323388490174505,
+      "learning_rate": 1.737783507655242e-05,
+      "loss": 0.2951,
+      "step": 4505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.596493427646773,
+      "learning_rate": 1.7376578763202465e-05,
+      "loss": 0.3465,
+      "step": 4506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.85841984621207,
+      "learning_rate": 1.7375322194401905e-05,
+      "loss": 0.5163,
+      "step": 4507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3979185219008631,
+      "learning_rate": 1.7374065370194253e-05,
+      "loss": 0.3023,
+      "step": 4508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3792909090305572,
+      "learning_rate": 1.7372808290623034e-05,
+      "loss": 0.2717,
+      "step": 4509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3288322471843398,
+      "learning_rate": 1.7371550955731786e-05,
+      "loss": 0.2334,
+      "step": 4510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3603072373462248,
+      "learning_rate": 1.737029336556404e-05,
+      "loss": 0.2721,
+      "step": 4511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0258413937489468,
+      "learning_rate": 1.7369035520163355e-05,
+      "loss": 0.4195,
+      "step": 4512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38788937506634363,
+      "learning_rate": 1.7367777419573285e-05,
+      "loss": 0.3458,
+      "step": 4513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3663495306117589,
+      "learning_rate": 1.73665190638374e-05,
+      "loss": 0.2827,
+      "step": 4514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35904037745450235,
+      "learning_rate": 1.7365260452999277e-05,
+      "loss": 0.239,
+      "step": 4515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.368230106992383,
+      "learning_rate": 1.73640015871025e-05,
+      "loss": 0.2299,
+      "step": 4516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4124820648794057,
+      "learning_rate": 1.7362742466190668e-05,
+      "loss": 0.2929,
+      "step": 4517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6313760443488636,
+      "learning_rate": 1.736148309030738e-05,
+      "loss": 0.3801,
+      "step": 4518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8649835515367721,
+      "learning_rate": 1.7360223459496244e-05,
+      "loss": 0.3975,
+      "step": 4519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3995323752107863,
+      "learning_rate": 1.735896357380089e-05,
+      "loss": 0.2723,
+      "step": 4520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34544221499812,
+      "learning_rate": 1.7357703433264945e-05,
+      "loss": 0.3229,
+      "step": 4521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2902986551221321,
+      "learning_rate": 1.7356443037932046e-05,
+      "loss": 0.1643,
+      "step": 4522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4838021066394585,
+      "learning_rate": 1.7355182387845843e-05,
+      "loss": 0.3209,
+      "step": 4523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5004591776306092,
+      "learning_rate": 1.7353921483049985e-05,
+      "loss": 0.3296,
+      "step": 4524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4258313917846841,
+      "learning_rate": 1.7352660323588146e-05,
+      "loss": 0.2727,
+      "step": 4525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4033155476729041,
+      "learning_rate": 1.7351398909503995e-05,
+      "loss": 0.3081,
+      "step": 4526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5300104630284196,
+      "learning_rate": 1.7350137240841218e-05,
+      "loss": 0.4131,
+      "step": 4527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6412206824998394,
+      "learning_rate": 1.73488753176435e-05,
+      "loss": 0.5222,
+      "step": 4528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3077706381580906,
+      "learning_rate": 1.7347613139954548e-05,
+      "loss": 0.2402,
+      "step": 4529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6688065829605588,
+      "learning_rate": 1.734635070781807e-05,
+      "loss": 0.538,
+      "step": 4530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35909638887144074,
+      "learning_rate": 1.734508802127778e-05,
+      "loss": 0.2608,
+      "step": 4531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41579088397011704,
+      "learning_rate": 1.734382508037741e-05,
+      "loss": 0.1842,
+      "step": 4532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4211404512665711,
+      "learning_rate": 1.7342561885160694e-05,
+      "loss": 0.3147,
+      "step": 4533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8256111760795807,
+      "learning_rate": 1.7341298435671373e-05,
+      "loss": 0.5434,
+      "step": 4534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5292156928559316,
+      "learning_rate": 1.7340034731953204e-05,
+      "loss": 0.2707,
+      "step": 4535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3626707079391625,
+      "learning_rate": 1.7338770774049948e-05,
+      "loss": 0.2969,
+      "step": 4536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2987530938150695,
+      "learning_rate": 1.733750656200538e-05,
+      "loss": 0.2727,
+      "step": 4537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41984643592449417,
+      "learning_rate": 1.733624209586327e-05,
+      "loss": 0.0769,
+      "step": 4538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4498582260094257,
+      "learning_rate": 1.733497737566741e-05,
+      "loss": 0.3627,
+      "step": 4539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7896624131435914,
+      "learning_rate": 1.7333712401461602e-05,
+      "loss": 0.5321,
+      "step": 4540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.33277012471811757,
+      "learning_rate": 1.7332447173289648e-05,
+      "loss": 0.3177,
+      "step": 4541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4780594140246693,
+      "learning_rate": 1.7331181691195364e-05,
+      "loss": 0.2849,
+      "step": 4542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35576731933438077,
+      "learning_rate": 1.7329915955222578e-05,
+      "loss": 0.2063,
+      "step": 4543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5907181564894486,
+      "learning_rate": 1.7328649965415108e-05,
+      "loss": 0.358,
+      "step": 4544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31900993570485137,
+      "learning_rate": 1.732738372181681e-05,
+      "loss": 0.2571,
+      "step": 4545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8407790010179014,
+      "learning_rate": 1.7326117224471534e-05,
+      "loss": 0.4832,
+      "step": 4546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6475543922825446,
+      "learning_rate": 1.7324850473423124e-05,
+      "loss": 0.4714,
+      "step": 4547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4103670098764781,
+      "learning_rate": 1.7323583468715464e-05,
+      "loss": 0.2612,
+      "step": 4548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25746760920249057,
+      "learning_rate": 1.732231621039242e-05,
+      "loss": 0.2225,
+      "step": 4549
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45385734178505405,
+      "learning_rate": 1.732104869849788e-05,
+      "loss": 0.3283,
+      "step": 4550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5715738690411651,
+      "learning_rate": 1.731978093307574e-05,
+      "loss": 0.3138,
+      "step": 4551
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47645140136250524,
+      "learning_rate": 1.7318512914169903e-05,
+      "loss": 0.3329,
+      "step": 4552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8370401653202683,
+      "learning_rate": 1.7317244641824275e-05,
+      "loss": 0.5012,
+      "step": 4553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40217735496250034,
+      "learning_rate": 1.731597611608278e-05,
+      "loss": 0.319,
+      "step": 4554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22407629826832084,
+      "learning_rate": 1.731470733698935e-05,
+      "loss": 0.137,
+      "step": 4555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43215683904816754,
+      "learning_rate": 1.7313438304587918e-05,
+      "loss": 0.3071,
+      "step": 4556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.318869649065026,
+      "learning_rate": 1.731216901892243e-05,
+      "loss": 0.2901,
+      "step": 4557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8355303688845711,
+      "learning_rate": 1.7310899480036845e-05,
+      "loss": 0.3845,
+      "step": 4558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8346211453513296,
+      "learning_rate": 1.7309629687975126e-05,
+      "loss": 0.5678,
+      "step": 4559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3490837116866384,
+      "learning_rate": 1.730835964278124e-05,
+      "loss": 0.2892,
+      "step": 4560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32214562875779884,
+      "learning_rate": 1.7307089344499178e-05,
+      "loss": 0.2011,
+      "step": 4561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4094468270950029,
+      "learning_rate": 1.730581879317293e-05,
+      "loss": 0.2334,
+      "step": 4562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6224042650550641,
+      "learning_rate": 1.730454798884649e-05,
+      "loss": 0.3563,
+      "step": 4563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.379128591299934,
+      "learning_rate": 1.7303276931563862e-05,
+      "loss": 0.2706,
+      "step": 4564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6789909142350193,
+      "learning_rate": 1.730200562136907e-05,
+      "loss": 0.4433,
+      "step": 4565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4124711177122947,
+      "learning_rate": 1.7300734058306138e-05,
+      "loss": 0.335,
+      "step": 4566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2858262013205675,
+      "learning_rate": 1.72994622424191e-05,
+      "loss": 0.2076,
+      "step": 4567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30957498941420897,
+      "learning_rate": 1.7298190173751996e-05,
+      "loss": 0.2559,
+      "step": 4568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45178314832303773,
+      "learning_rate": 1.7296917852348882e-05,
+      "loss": 0.342,
+      "step": 4569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6733044811452652,
+      "learning_rate": 1.7295645278253817e-05,
+      "loss": 0.5276,
+      "step": 4570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49577571076452776,
+      "learning_rate": 1.729437245151087e-05,
+      "loss": 0.2828,
+      "step": 4571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3413429533842971,
+      "learning_rate": 1.7293099372164118e-05,
+      "loss": 0.2894,
+      "step": 4572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3407676552622252,
+      "learning_rate": 1.729182604025765e-05,
+      "loss": 0.2752,
+      "step": 4573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.561291828414926,
+      "learning_rate": 1.729055245583556e-05,
+      "loss": 0.3268,
+      "step": 4574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29530320708055313,
+      "learning_rate": 1.728927861894195e-05,
+      "loss": 0.2669,
+      "step": 4575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3881094912526957,
+      "learning_rate": 1.7288004529620935e-05,
+      "loss": 0.3326,
+      "step": 4576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1701078373770246,
+      "learning_rate": 1.7286730187916635e-05,
+      "loss": 0.8335,
+      "step": 4577
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31692692063010275,
+      "learning_rate": 1.7285455593873183e-05,
+      "loss": 0.2234,
+      "step": 4578
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7275364835658681,
+      "learning_rate": 1.728418074753472e-05,
+      "loss": 0.5326,
+      "step": 4579
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3451992181340388,
+      "learning_rate": 1.7282905648945386e-05,
+      "loss": 0.342,
+      "step": 4580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2693554960138304,
+      "learning_rate": 1.7281630298149346e-05,
+      "loss": 0.2137,
+      "step": 4581
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4373966659922116,
+      "learning_rate": 1.728035469519076e-05,
+      "loss": 0.2423,
+      "step": 4582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8356130794756954,
+      "learning_rate": 1.7279078840113805e-05,
+      "loss": 0.4754,
+      "step": 4583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3556758936236917,
+      "learning_rate": 1.7277802732962662e-05,
+      "loss": 0.2365,
+      "step": 4584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3887816940254315,
+      "learning_rate": 1.7276526373781525e-05,
+      "loss": 0.3398,
+      "step": 4585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6248091341949249,
+      "learning_rate": 1.7275249762614592e-05,
+      "loss": 0.4088,
+      "step": 4586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2780554084812607,
+      "learning_rate": 1.727397289950607e-05,
+      "loss": 0.1789,
+      "step": 4587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35914471939866316,
+      "learning_rate": 1.7272695784500185e-05,
+      "loss": 0.2958,
+      "step": 4588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1927746460401998,
+      "learning_rate": 1.7271418417641153e-05,
+      "loss": 0.7851,
+      "step": 4589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3336997350745605,
+      "learning_rate": 1.7270140798973215e-05,
+      "loss": 0.259,
+      "step": 4590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7934746268294893,
+      "learning_rate": 1.7268862928540616e-05,
+      "loss": 0.3916,
+      "step": 4591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.440992452385509,
+      "learning_rate": 1.7267584806387604e-05,
+      "loss": 0.3606,
+      "step": 4592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3643981449840152,
+      "learning_rate": 1.726630643255844e-05,
+      "loss": 0.2713,
+      "step": 4593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23831222638239447,
+      "learning_rate": 1.7265027807097402e-05,
+      "loss": 0.1099,
+      "step": 4594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9794922046894605,
+      "learning_rate": 1.726374893004876e-05,
+      "loss": 0.641,
+      "step": 4595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3302361301129314,
+      "learning_rate": 1.7262469801456806e-05,
+      "loss": 0.2753,
+      "step": 4596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6913170287228856,
+      "learning_rate": 1.7261190421365836e-05,
+      "loss": 0.3405,
+      "step": 4597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7638555754563725,
+      "learning_rate": 1.7259910789820152e-05,
+      "loss": 0.5113,
+      "step": 4598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30601965328502595,
+      "learning_rate": 1.7258630906864068e-05,
+      "loss": 0.203,
+      "step": 4599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35651461621894115,
+      "learning_rate": 1.7257350772541914e-05,
+      "loss": 0.2329,
+      "step": 4600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8782920891777145,
+      "learning_rate": 1.725607038689801e-05,
+      "loss": 0.6243,
+      "step": 4601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42622194472487185,
+      "learning_rate": 1.7254789749976703e-05,
+      "loss": 0.2775,
+      "step": 4602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1987544832356745,
+      "learning_rate": 1.7253508861822338e-05,
+      "loss": 0.4799,
+      "step": 4603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37401071978627676,
+      "learning_rate": 1.725222772247927e-05,
+      "loss": 0.2316,
+      "step": 4604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4714683560992018,
+      "learning_rate": 1.725094633199187e-05,
+      "loss": 0.279,
+      "step": 4605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3252260992389096,
+      "learning_rate": 1.7249664690404514e-05,
+      "loss": 0.1989,
+      "step": 4606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2681138371986997,
+      "learning_rate": 1.7248382797761576e-05,
+      "loss": 0.4433,
+      "step": 4607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38382751971823414,
+      "learning_rate": 1.7247100654107458e-05,
+      "loss": 0.2707,
+      "step": 4608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5398894230168891,
+      "learning_rate": 1.724581825948655e-05,
+      "loss": 0.3814,
+      "step": 4609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7840013858936152,
+      "learning_rate": 1.7244535613943273e-05,
+      "loss": 0.3339,
+      "step": 4610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4177543238883827,
+      "learning_rate": 1.7243252717522037e-05,
+      "loss": 0.2759,
+      "step": 4611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2814456169503759,
+      "learning_rate": 1.724196957026727e-05,
+      "loss": 0.2205,
+      "step": 4612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.286573728865289,
+      "learning_rate": 1.724068617222341e-05,
+      "loss": 0.4222,
+      "step": 4613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3835480722641368,
+      "learning_rate": 1.7239402523434898e-05,
+      "loss": 0.2947,
+      "step": 4614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9076811236266914,
+      "learning_rate": 1.723811862394619e-05,
+      "loss": 0.5795,
+      "step": 4615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4121511463726725,
+      "learning_rate": 1.7236834473801744e-05,
+      "loss": 0.344,
+      "step": 4616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33727558721149664,
+      "learning_rate": 1.723555007304603e-05,
+      "loss": 0.2264,
+      "step": 4617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49247949403688007,
+      "learning_rate": 1.7234265421723528e-05,
+      "loss": 0.2619,
+      "step": 4618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7509746626101724,
+      "learning_rate": 1.7232980519878727e-05,
+      "loss": 0.4011,
+      "step": 4619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3204159675138603,
+      "learning_rate": 1.723169536755612e-05,
+      "loss": 0.2003,
+      "step": 4620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41050620070576016,
+      "learning_rate": 1.7230409964800215e-05,
+      "loss": 0.3201,
+      "step": 4621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.120953694972054,
+      "learning_rate": 1.7229124311655524e-05,
+      "loss": 0.5876,
+      "step": 4622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34506829258410193,
+      "learning_rate": 1.722783840816657e-05,
+      "loss": 0.2215,
+      "step": 4623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39872366387719055,
+      "learning_rate": 1.7226552254377883e-05,
+      "loss": 0.3462,
+      "step": 4624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5939456678866099,
+      "learning_rate": 1.7225265850333997e-05,
+      "loss": 0.4869,
+      "step": 4625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3497056047001658,
+      "learning_rate": 1.7223979196079466e-05,
+      "loss": 0.2402,
+      "step": 4626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31503886763827527,
+      "learning_rate": 1.7222692291658853e-05,
+      "loss": 0.2371,
+      "step": 4627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5476162155781016,
+      "learning_rate": 1.7221405137116712e-05,
+      "loss": 0.3956,
+      "step": 4628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4289308831084935,
+      "learning_rate": 1.722011773249762e-05,
+      "loss": 0.2724,
+      "step": 4629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.45957358140657234,
+      "learning_rate": 1.7218830077846164e-05,
+      "loss": 0.329,
+      "step": 4630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4495152254009832,
+      "learning_rate": 1.7217542173206932e-05,
+      "loss": 0.3848,
+      "step": 4631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3251634536660676,
+      "learning_rate": 1.7216254018624524e-05,
+      "loss": 0.2706,
+      "step": 4632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36095585230014704,
+      "learning_rate": 1.7214965614143554e-05,
+      "loss": 0.1692,
+      "step": 4633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42598816677891027,
+      "learning_rate": 1.721367695980863e-05,
+      "loss": 0.2535,
+      "step": 4634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.37269976982163927,
+      "learning_rate": 1.7212388055664385e-05,
+      "loss": 0.2948,
+      "step": 4635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3768645284351786,
+      "learning_rate": 1.7211098901755453e-05,
+      "loss": 0.2551,
+      "step": 4636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7238591197734653,
+      "learning_rate": 1.7209809498126473e-05,
+      "loss": 0.4616,
+      "step": 4637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.46787631269308716,
+      "learning_rate": 1.72085198448221e-05,
+      "loss": 0.327,
+      "step": 4638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28015550313024745,
+      "learning_rate": 1.7207229941887e-05,
+      "loss": 0.2313,
+      "step": 4639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32694743935340276,
+      "learning_rate": 1.7205939789365834e-05,
+      "loss": 0.2899,
+      "step": 4640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.57415873988465,
+      "learning_rate": 1.720464938730328e-05,
+      "loss": 0.3279,
+      "step": 4641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4253211319708633,
+      "learning_rate": 1.720335873574403e-05,
+      "loss": 0.3365,
+      "step": 4642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4051995236013618,
+      "learning_rate": 1.7202067834732778e-05,
+      "loss": 0.3115,
+      "step": 4643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4540323810623343,
+      "learning_rate": 1.7200776684314226e-05,
+      "loss": 0.3126,
+      "step": 4644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3539122334412321,
+      "learning_rate": 1.7199485284533088e-05,
+      "loss": 0.2621,
+      "step": 4645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26353112650494664,
+      "learning_rate": 1.7198193635434083e-05,
+      "loss": 0.0729,
+      "step": 4646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40396435081770415,
+      "learning_rate": 1.719690173706194e-05,
+      "loss": 0.3095,
+      "step": 4647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33135791809707316,
+      "learning_rate": 1.71956095894614e-05,
+      "loss": 0.3202,
+      "step": 4648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8182839067779433,
+      "learning_rate": 1.719431719267721e-05,
+      "loss": 0.4072,
+      "step": 4649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4000599760224488,
+      "learning_rate": 1.7193024546754125e-05,
+      "loss": 0.3128,
+      "step": 4650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5156932368052041,
+      "learning_rate": 1.719173165173691e-05,
+      "loss": 0.406,
+      "step": 4651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2688587493026527,
+      "learning_rate": 1.7190438507670337e-05,
+      "loss": 0.1869,
+      "step": 4652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36163058971972184,
+      "learning_rate": 1.7189145114599188e-05,
+      "loss": 0.3013,
+      "step": 4653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.787712368896139,
+      "learning_rate": 1.718785147256825e-05,
+      "loss": 0.4917,
+      "step": 4654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3618209143003872,
+      "learning_rate": 1.7186557581622327e-05,
+      "loss": 0.3338,
+      "step": 4655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32715965622544163,
+      "learning_rate": 1.7185263441806227e-05,
+      "loss": 0.2166,
+      "step": 4656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5405807570579455,
+      "learning_rate": 1.7183969053164757e-05,
+      "loss": 0.3873,
+      "step": 4657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33734823448871526,
+      "learning_rate": 1.718267441574275e-05,
+      "loss": 0.2136,
+      "step": 4658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41124631225255276,
+      "learning_rate": 1.718137952958504e-05,
+      "loss": 0.2576,
+      "step": 4659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.372867877130881,
+      "learning_rate": 1.7180084394736464e-05,
+      "loss": 0.3317,
+      "step": 4660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2657038079052518,
+      "learning_rate": 1.717878901124187e-05,
+      "loss": 0.8484,
+      "step": 4661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30325582500593035,
+      "learning_rate": 1.7177493379146123e-05,
+      "loss": 0.1239,
+      "step": 4662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32165604472772885,
+      "learning_rate": 1.717619749849409e-05,
+      "loss": 0.296,
+      "step": 4663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30992102992511755,
+      "learning_rate": 1.7174901369330648e-05,
+      "loss": 0.2242,
+      "step": 4664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.48324249109187434,
+      "learning_rate": 1.7173604991700678e-05,
+      "loss": 0.363,
+      "step": 4665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.403222549687474,
+      "learning_rate": 1.7172308365649077e-05,
+      "loss": 0.2954,
+      "step": 4666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.45450122410769483,
+      "learning_rate": 1.7171011491220744e-05,
+      "loss": 0.3464,
+      "step": 4667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6262524047569294,
+      "learning_rate": 1.7169714368460593e-05,
+      "loss": 0.3987,
+      "step": 4668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40588895977274775,
+      "learning_rate": 1.716841699741354e-05,
+      "loss": 0.241,
+      "step": 4669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7130249951084894,
+      "learning_rate": 1.7167119378124516e-05,
+      "loss": 0.4351,
+      "step": 4670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3442235779286332,
+      "learning_rate": 1.7165821510638456e-05,
+      "loss": 0.2933,
+      "step": 4671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23626658265469405,
+      "learning_rate": 1.7164523395000304e-05,
+      "loss": 0.1589,
+      "step": 4672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8478012547554327,
+      "learning_rate": 1.7163225031255018e-05,
+      "loss": 0.6467,
+      "step": 4673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5128732762763075,
+      "learning_rate": 1.7161926419447555e-05,
+      "loss": 0.3303,
+      "step": 4674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3949461464032854,
+      "learning_rate": 1.7160627559622888e-05,
+      "loss": 0.2547,
+      "step": 4675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5768859794534354,
+      "learning_rate": 1.7159328451825995e-05,
+      "loss": 0.3725,
+      "step": 4676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4045747712340561,
+      "learning_rate": 1.7158029096101868e-05,
+      "loss": 0.2743,
+      "step": 4677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31346151516334575,
+      "learning_rate": 1.71567294924955e-05,
+      "loss": 0.2506,
+      "step": 4678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38541238161229885,
+      "learning_rate": 1.71554296410519e-05,
+      "loss": 0.281,
+      "step": 4679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8739729047054263,
+      "learning_rate": 1.7154129541816078e-05,
+      "loss": 0.509,
+      "step": 4680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38755430202848945,
+      "learning_rate": 1.7152829194833054e-05,
+      "loss": 0.2878,
+      "step": 4681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7597051840219803,
+      "learning_rate": 1.7151528600147868e-05,
+      "loss": 0.3923,
+      "step": 4682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3355241879573539,
+      "learning_rate": 1.7150227757805552e-05,
+      "loss": 0.3151,
+      "step": 4683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25231565932966504,
+      "learning_rate": 1.7148926667851156e-05,
+      "loss": 0.1974,
+      "step": 4684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9743567662925537,
+      "learning_rate": 1.7147625330329734e-05,
+      "loss": 0.5156,
+      "step": 4685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5801427580892613,
+      "learning_rate": 1.714632374528636e-05,
+      "loss": 0.4265,
+      "step": 4686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38310225366593875,
+      "learning_rate": 1.7145021912766096e-05,
+      "loss": 0.3218,
+      "step": 4687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5825425991209423,
+      "learning_rate": 1.7143719832814034e-05,
+      "loss": 0.3381,
+      "step": 4688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3340408898626788,
+      "learning_rate": 1.714241750547526e-05,
+      "loss": 0.239,
+      "step": 4689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3168336149753657,
+      "learning_rate": 1.7141114930794876e-05,
+      "loss": 0.2373,
+      "step": 4690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40724845843177176,
+      "learning_rate": 1.7139812108817988e-05,
+      "loss": 0.3455,
+      "step": 4691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5094991595317109,
+      "learning_rate": 1.7138509039589713e-05,
+      "loss": 0.2222,
+      "step": 4692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4787797053735826,
+      "learning_rate": 1.7137205723155178e-05,
+      "loss": 0.2925,
+      "step": 4693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7871752853141931,
+      "learning_rate": 1.7135902159559518e-05,
+      "loss": 0.4657,
+      "step": 4694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3929226671181522,
+      "learning_rate": 1.713459834884787e-05,
+      "loss": 0.2591,
+      "step": 4695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31143988193266453,
+      "learning_rate": 1.7133294291065387e-05,
+      "loss": 0.2568,
+      "step": 4696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5147095929397764,
+      "learning_rate": 1.7131989986257233e-05,
+      "loss": 0.3835,
+      "step": 4697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8803237290957555,
+      "learning_rate": 1.713068543446857e-05,
+      "loss": 0.337,
+      "step": 4698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32761454354627956,
+      "learning_rate": 1.7129380635744578e-05,
+      "loss": 0.2859,
+      "step": 4699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6789050208924245,
+      "learning_rate": 1.712807559013044e-05,
+      "loss": 0.5107,
+      "step": 4700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.580704421074495,
+      "learning_rate": 1.7126770297671353e-05,
+      "loss": 0.1709,
+      "step": 4701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3050190429982249,
+      "learning_rate": 1.7125464758412517e-05,
+      "loss": 0.2517,
+      "step": 4702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4061107982630693,
+      "learning_rate": 1.7124158972399142e-05,
+      "loss": 0.3583,
+      "step": 4703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1006064066544996,
+      "learning_rate": 1.7122852939676448e-05,
+      "loss": 0.4624,
+      "step": 4704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41879601303777364,
+      "learning_rate": 1.712154666028966e-05,
+      "loss": 0.2789,
+      "step": 4705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4464716584986142,
+      "learning_rate": 1.712024013428402e-05,
+      "loss": 0.2636,
+      "step": 4706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3530264257968865,
+      "learning_rate": 1.7118933361704773e-05,
+      "loss": 0.297,
+      "step": 4707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4282023284180023,
+      "learning_rate": 1.7117626342597168e-05,
+      "loss": 0.2344,
+      "step": 4708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4580585344409273,
+      "learning_rate": 1.711631907700647e-05,
+      "loss": 0.3537,
+      "step": 4709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7468136572361314,
+      "learning_rate": 1.711501156497794e-05,
+      "loss": 0.3764,
+      "step": 4710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.37408046619978336,
+      "learning_rate": 1.7113703806556875e-05,
+      "loss": 0.2427,
+      "step": 4711
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49604510274681346,
+      "learning_rate": 1.711239580178855e-05,
+      "loss": 0.3652,
+      "step": 4712
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1377881037810844,
+      "learning_rate": 1.7111087550718265e-05,
+      "loss": 0.6517,
+      "step": 4713
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3778912354875147,
+      "learning_rate": 1.7109779053391322e-05,
+      "loss": 0.2191,
+      "step": 4714
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3796078306581816,
+      "learning_rate": 1.710847030985304e-05,
+      "loss": 0.3385,
+      "step": 4715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6783617967841638,
+      "learning_rate": 1.710716132014873e-05,
+      "loss": 0.4694,
+      "step": 4716
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3984794551445479,
+      "learning_rate": 1.7105852084323736e-05,
+      "loss": 0.2969,
+      "step": 4717
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28023283128766,
+      "learning_rate": 1.7104542602423385e-05,
+      "loss": 0.0739,
+      "step": 4718
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.365961154130174,
+      "learning_rate": 1.710323287449303e-05,
+      "loss": 0.3137,
+      "step": 4719
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3846054233012314,
+      "learning_rate": 1.710192290057803e-05,
+      "loss": 0.2914,
+      "step": 4720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7279304764082829,
+      "learning_rate": 1.7100612680723744e-05,
+      "loss": 0.4087,
+      "step": 4721
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35474730728888954,
+      "learning_rate": 1.7099302214975545e-05,
+      "loss": 0.3523,
+      "step": 4722
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35652743317062197,
+      "learning_rate": 1.7097991503378812e-05,
+      "loss": 0.2756,
+      "step": 4723
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17111219321848872,
+      "learning_rate": 1.7096680545978946e-05,
+      "loss": 0.0714,
+      "step": 4724
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7528476749497117,
+      "learning_rate": 1.709536934282133e-05,
+      "loss": 0.4266,
+      "step": 4725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4257309461122103,
+      "learning_rate": 1.7094057893951385e-05,
+      "loss": 0.306,
+      "step": 4726
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40249958959659726,
+      "learning_rate": 1.709274619941452e-05,
+      "loss": 0.3084,
+      "step": 4727
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6821814944692812,
+      "learning_rate": 1.7091434259256155e-05,
+      "loss": 0.4543,
+      "step": 4728
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4068026712533653,
+      "learning_rate": 1.7090122073521726e-05,
+      "loss": 0.2961,
+      "step": 4729
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30780879761690466,
+      "learning_rate": 1.7088809642256677e-05,
+      "loss": 0.2363,
+      "step": 4730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4450334908612994,
+      "learning_rate": 1.7087496965506457e-05,
+      "loss": 0.2771,
+      "step": 4731
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3757471843854514,
+      "learning_rate": 1.7086184043316518e-05,
+      "loss": 0.2816,
+      "step": 4732
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.718898235374059,
+      "learning_rate": 1.7084870875732332e-05,
+      "loss": 0.4816,
+      "step": 4733
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35620390057471335,
+      "learning_rate": 1.708355746279937e-05,
+      "loss": 0.2784,
+      "step": 4734
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3257456059274286,
+      "learning_rate": 1.7082243804563123e-05,
+      "loss": 0.2776,
+      "step": 4735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2670452463661605,
+      "learning_rate": 1.7080929901069076e-05,
+      "loss": 0.163,
+      "step": 4736
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6344362497021302,
+      "learning_rate": 1.7079615752362727e-05,
+      "loss": 0.3661,
+      "step": 4737
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35318412416104933,
+      "learning_rate": 1.707830135848959e-05,
+      "loss": 0.2906,
+      "step": 4738
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.415746007004765,
+      "learning_rate": 1.7076986719495184e-05,
+      "loss": 0.3621,
+      "step": 4739
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9097433398595413,
+      "learning_rate": 1.7075671835425032e-05,
+      "loss": 0.6311,
+      "step": 4740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30654017642837456,
+      "learning_rate": 1.7074356706324668e-05,
+      "loss": 0.22,
+      "step": 4741
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2861658906239352,
+      "learning_rate": 1.7073041332239634e-05,
+      "loss": 0.1787,
+      "step": 4742
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38805970077000235,
+      "learning_rate": 1.7071725713215483e-05,
+      "loss": 0.3573,
+      "step": 4743
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35965465826234444,
+      "learning_rate": 1.7070409849297774e-05,
+      "loss": 0.216,
+      "step": 4744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7013740357821047,
+      "learning_rate": 1.7069093740532083e-05,
+      "loss": 0.4085,
+      "step": 4745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3788364821404023,
+      "learning_rate": 1.706777738696397e-05,
+      "loss": 0.3364,
+      "step": 4746
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33608417673074076,
+      "learning_rate": 1.7066460788639035e-05,
+      "loss": 0.2014,
+      "step": 4747
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2977740305003241,
+      "learning_rate": 1.7065143945602867e-05,
+      "loss": 0.1881,
+      "step": 4748
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6092323032794157,
+      "learning_rate": 1.7063826857901066e-05,
+      "loss": 0.4919,
+      "step": 4749
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35754157093924055,
+      "learning_rate": 1.7062509525579244e-05,
+      "loss": 0.2416,
+      "step": 4750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3915972606090248,
+      "learning_rate": 1.7061191948683024e-05,
+      "loss": 0.3557,
+      "step": 4751
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2231230334847767,
+      "learning_rate": 1.7059874127258028e-05,
+      "loss": 0.7646,
+      "step": 4752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3921839884714695,
+      "learning_rate": 1.7058556061349894e-05,
+      "loss": 0.3111,
+      "step": 4753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2249213774799527,
+      "learning_rate": 1.705723775100427e-05,
+      "loss": 0.1823,
+      "step": 4754
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.46286782233072077,
+      "learning_rate": 1.7055919196266806e-05,
+      "loss": 0.3736,
+      "step": 4755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.441673470781393,
+      "learning_rate": 1.7054600397183162e-05,
+      "loss": 0.3233,
+      "step": 4756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5357999621963078,
+      "learning_rate": 1.705328135379901e-05,
+      "loss": 0.2646,
+      "step": 4757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4294620287968328,
+      "learning_rate": 1.7051962066160027e-05,
+      "loss": 0.3657,
+      "step": 4758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36051927885640783,
+      "learning_rate": 1.7050642534311904e-05,
+      "loss": 0.2862,
+      "step": 4759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3010743262062142,
+      "learning_rate": 1.704932275830033e-05,
+      "loss": 0.1729,
+      "step": 4760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4334278974070254,
+      "learning_rate": 1.704800273817101e-05,
+      "loss": 0.3478,
+      "step": 4761
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4264062843363303,
+      "learning_rate": 1.7046682473969664e-05,
+      "loss": 0.2949,
+      "step": 4762
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4010993186213262,
+      "learning_rate": 1.7045361965742004e-05,
+      "loss": 0.2649,
+      "step": 4763
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2493476218292963,
+      "learning_rate": 1.704404121353376e-05,
+      "loss": 0.7307,
+      "step": 4764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5374611777732092,
+      "learning_rate": 1.7042720217390677e-05,
+      "loss": 0.3441,
+      "step": 4765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35668982665157084,
+      "learning_rate": 1.7041398977358494e-05,
+      "loss": 0.2937,
+      "step": 4766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5150162023118233,
+      "learning_rate": 1.7040077493482964e-05,
+      "loss": 0.3482,
+      "step": 4767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3524313571951868,
+      "learning_rate": 1.7038755765809857e-05,
+      "loss": 0.2136,
+      "step": 4768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3518307329267539,
+      "learning_rate": 1.7037433794384938e-05,
+      "loss": 0.2492,
+      "step": 4769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.37417425107718427,
+      "learning_rate": 1.7036111579253992e-05,
+      "loss": 0.2956,
+      "step": 4770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3943110391556108,
+      "learning_rate": 1.70347891204628e-05,
+      "loss": 0.3086,
+      "step": 4771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4517027039596723,
+      "learning_rate": 1.7033466418057166e-05,
+      "loss": 0.3716,
+      "step": 4772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44044704690746816,
+      "learning_rate": 1.7032143472082893e-05,
+      "loss": 0.2414,
+      "step": 4773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4506278934272334,
+      "learning_rate": 1.7030820282585795e-05,
+      "loss": 0.256,
+      "step": 4774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39868821005548993,
+      "learning_rate": 1.7029496849611687e-05,
+      "loss": 0.2626,
+      "step": 4775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3150474383898678,
+      "learning_rate": 1.702817317320641e-05,
+      "loss": 0.5298,
+      "step": 4776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3613155978762499,
+      "learning_rate": 1.70268492534158e-05,
+      "loss": 0.2521,
+      "step": 4777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.460323545686999,
+      "learning_rate": 1.70255250902857e-05,
+      "loss": 0.3628,
+      "step": 4778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5444376532180347,
+      "learning_rate": 1.702420068386197e-05,
+      "loss": 0.3987,
+      "step": 4779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26635586243885684,
+      "learning_rate": 1.7022876034190468e-05,
+      "loss": 0.0774,
+      "step": 4780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3994512255967334,
+      "learning_rate": 1.7021551141317075e-05,
+      "loss": 0.2613,
+      "step": 4781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5373762784077355,
+      "learning_rate": 1.7020226005287665e-05,
+      "loss": 0.3608,
+      "step": 4782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5289576791227809,
+      "learning_rate": 1.701890062614813e-05,
+      "loss": 0.19,
+      "step": 4783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40705610275602827,
+      "learning_rate": 1.7017575003944374e-05,
+      "loss": 0.3555,
+      "step": 4784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8059748782716618,
+      "learning_rate": 1.7016249138722295e-05,
+      "loss": 0.5116,
+      "step": 4785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3090383888410909,
+      "learning_rate": 1.7014923030527808e-05,
+      "loss": 0.1888,
+      "step": 4786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3782301763328678,
+      "learning_rate": 1.701359667940684e-05,
+      "loss": 0.284,
+      "step": 4787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3901516550893385,
+      "learning_rate": 1.7012270085405317e-05,
+      "loss": 0.8495,
+      "step": 4788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4533944763836812,
+      "learning_rate": 1.7010943248569185e-05,
+      "loss": 0.2694,
+      "step": 4789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4358691412921533,
+      "learning_rate": 1.700961616894439e-05,
+      "loss": 0.3246,
+      "step": 4790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2521917339608342,
+      "learning_rate": 1.7008288846576886e-05,
+      "loss": 0.5989,
+      "step": 4791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3808308096968665,
+      "learning_rate": 1.7006961281512645e-05,
+      "loss": 0.251,
+      "step": 4792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46592574712611845,
+      "learning_rate": 1.7005633473797632e-05,
+      "loss": 0.2673,
+      "step": 4793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3932673619699627,
+      "learning_rate": 1.7004305423477835e-05,
+      "loss": 0.3008,
+      "step": 4794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6720359832254608,
+      "learning_rate": 1.700297713059924e-05,
+      "loss": 0.374,
+      "step": 4795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38595536528371815,
+      "learning_rate": 1.700164859520785e-05,
+      "loss": 0.2642,
+      "step": 4796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.525812046298573,
+      "learning_rate": 1.7000319817349673e-05,
+      "loss": 0.3478,
+      "step": 4797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39726507148120443,
+      "learning_rate": 1.699899079707072e-05,
+      "loss": 0.3034,
+      "step": 4798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38302101200138,
+      "learning_rate": 1.6997661534417015e-05,
+      "loss": 0.2645,
+      "step": 4799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45339021595960577,
+      "learning_rate": 1.699633202943459e-05,
+      "loss": 0.3409,
+      "step": 4800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6208504710150762,
+      "learning_rate": 1.699500228216949e-05,
+      "loss": 0.4191,
+      "step": 4801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3185182893708112,
+      "learning_rate": 1.6993672292667766e-05,
+      "loss": 0.2207,
+      "step": 4802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.197840212134981,
+      "learning_rate": 1.6992342060975467e-05,
+      "loss": 0.6374,
+      "step": 4803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7354342772061968,
+      "learning_rate": 1.6991011587138665e-05,
+      "loss": 0.4787,
+      "step": 4804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4419399786931295,
+      "learning_rate": 1.698968087120343e-05,
+      "loss": 0.3406,
+      "step": 4805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5150312372588404,
+      "learning_rate": 1.6988349913215848e-05,
+      "loss": 0.2817,
+      "step": 4806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6398701761513857,
+      "learning_rate": 1.698701871322201e-05,
+      "loss": 0.3152,
+      "step": 4807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2889587602038658,
+      "learning_rate": 1.698568727126801e-05,
+      "loss": 0.2323,
+      "step": 4808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46663731677113146,
+      "learning_rate": 1.6984355587399964e-05,
+      "loss": 0.284,
+      "step": 4809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44856779416870474,
+      "learning_rate": 1.6983023661663987e-05,
+      "loss": 0.3398,
+      "step": 4810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5213901240015865,
+      "learning_rate": 1.6981691494106196e-05,
+      "loss": 0.3272,
+      "step": 4811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.055086036269047,
+      "learning_rate": 1.698035908477273e-05,
+      "loss": 0.4127,
+      "step": 4812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4546873145751855,
+      "learning_rate": 1.697902643370973e-05,
+      "loss": 0.3177,
+      "step": 4813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4027812289256814,
+      "learning_rate": 1.6977693540963347e-05,
+      "loss": 0.3157,
+      "step": 4814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2686362672041927,
+      "learning_rate": 1.6976360406579734e-05,
+      "loss": 0.1518,
+      "step": 4815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7597687415584575,
+      "learning_rate": 1.697502703060506e-05,
+      "loss": 0.4421,
+      "step": 4816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44251032893440007,
+      "learning_rate": 1.69736934130855e-05,
+      "loss": 0.2877,
+      "step": 4817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5865777905310624,
+      "learning_rate": 1.6972359554067237e-05,
+      "loss": 0.3693,
+      "step": 4818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8136374248949291,
+      "learning_rate": 1.6971025453596463e-05,
+      "loss": 0.3066,
+      "step": 4819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2770444473086494,
+      "learning_rate": 1.6969691111719377e-05,
+      "loss": 0.2053,
+      "step": 4820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5373847607276323,
+      "learning_rate": 1.6968356528482187e-05,
+      "loss": 0.399,
+      "step": 4821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4172327112811106,
+      "learning_rate": 1.696702170393111e-05,
+      "loss": 0.225,
+      "step": 4822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4781253488053807,
+      "learning_rate": 1.6965686638112373e-05,
+      "loss": 0.308,
+      "step": 4823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2234511064967395,
+      "learning_rate": 1.6964351331072205e-05,
+      "loss": 0.4584,
+      "step": 4824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4180514861148252,
+      "learning_rate": 1.696301578285685e-05,
+      "loss": 0.2807,
+      "step": 4825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2841258483758155,
+      "learning_rate": 1.696167999351256e-05,
+      "loss": 0.2448,
+      "step": 4826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.459516630611072,
+      "learning_rate": 1.6960343963085587e-05,
+      "loss": 0.2917,
+      "step": 4827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0542058454902612,
+      "learning_rate": 1.6959007691622206e-05,
+      "loss": 0.5435,
+      "step": 4828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3935580375131141,
+      "learning_rate": 1.6957671179168687e-05,
+      "loss": 0.2066,
+      "step": 4829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5260916371309337,
+      "learning_rate": 1.695633442577131e-05,
+      "loss": 0.349,
+      "step": 4830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6822341191380212,
+      "learning_rate": 1.6954997431476376e-05,
+      "loss": 0.4439,
+      "step": 4831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.34931542652822906,
+      "learning_rate": 1.695366019633018e-05,
+      "loss": 0.2223,
+      "step": 4832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28818157760600444,
+      "learning_rate": 1.695232272037903e-05,
+      "loss": 0.2245,
+      "step": 4833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44886495313474906,
+      "learning_rate": 1.695098500366924e-05,
+      "loss": 0.3747,
+      "step": 4834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.376489095537162,
+      "learning_rate": 1.694964704624714e-05,
+      "loss": 0.257,
+      "step": 4835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6917611206680756,
+      "learning_rate": 1.6948308848159064e-05,
+      "loss": 0.5613,
+      "step": 4836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41105937337798043,
+      "learning_rate": 1.694697040945135e-05,
+      "loss": 0.3446,
+      "step": 4837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3334884198947588,
+      "learning_rate": 1.694563173017035e-05,
+      "loss": 0.225,
+      "step": 4838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3046497158528754,
+      "learning_rate": 1.694429281036242e-05,
+      "loss": 0.1881,
+      "step": 4839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6787872188789502,
+      "learning_rate": 1.6942953650073926e-05,
+      "loss": 0.4572,
+      "step": 4840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3605281610226177,
+      "learning_rate": 1.6941614249351252e-05,
+      "loss": 0.3004,
+      "step": 4841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4345068829504192,
+      "learning_rate": 1.6940274608240773e-05,
+      "loss": 0.2796,
+      "step": 4842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0070853290732158,
+      "learning_rate": 1.693893472678888e-05,
+      "loss": 0.6626,
+      "step": 4843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3629286253161658,
+      "learning_rate": 1.693759460504198e-05,
+      "loss": 0.2631,
+      "step": 4844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2677448034312148,
+      "learning_rate": 1.6936254243046472e-05,
+      "loss": 0.152,
+      "step": 4845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6535965345558263,
+      "learning_rate": 1.6934913640848782e-05,
+      "loss": 0.3904,
+      "step": 4846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3450106487612924,
+      "learning_rate": 1.6933572798495327e-05,
+      "loss": 0.2696,
+      "step": 4847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8901525449936295,
+      "learning_rate": 1.6932231716032548e-05,
+      "loss": 0.463,
+      "step": 4848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.35316268141975193,
+      "learning_rate": 1.6930890393506882e-05,
+      "loss": 0.3056,
+      "step": 4849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37333332234721345,
+      "learning_rate": 1.692954883096478e-05,
+      "loss": 0.314,
+      "step": 4850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26962638573625075,
+      "learning_rate": 1.6928207028452698e-05,
+      "loss": 0.1461,
+      "step": 4851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7124643097673313,
+      "learning_rate": 1.6926864986017105e-05,
+      "loss": 0.4647,
+      "step": 4852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3938996265220362,
+      "learning_rate": 1.6925522703704475e-05,
+      "loss": 0.3013,
+      "step": 4853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4226078356649217,
+      "learning_rate": 1.6924180181561297e-05,
+      "loss": 0.3757,
+      "step": 4854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.993544391815923,
+      "learning_rate": 1.6922837419634052e-05,
+      "loss": 0.4262,
+      "step": 4855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37675799254882103,
+      "learning_rate": 1.6921494417969245e-05,
+      "loss": 0.2918,
+      "step": 4856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3003610120862018,
+      "learning_rate": 1.6920151176613383e-05,
+      "loss": 0.2703,
+      "step": 4857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5360807413545341,
+      "learning_rate": 1.6918807695612984e-05,
+      "loss": 0.2786,
+      "step": 4858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37804769698458474,
+      "learning_rate": 1.6917463975014575e-05,
+      "loss": 0.2754,
+      "step": 4859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1356673660352166,
+      "learning_rate": 1.691612001486468e-05,
+      "loss": 0.6455,
+      "step": 4860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3493864639681745,
+      "learning_rate": 1.6914775815209853e-05,
+      "loss": 0.2859,
+      "step": 4861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37809936495877855,
+      "learning_rate": 1.6913431376096633e-05,
+      "loss": 0.2852,
+      "step": 4862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7970041802176591,
+      "learning_rate": 1.6912086697571584e-05,
+      "loss": 0.5109,
+      "step": 4863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2716805871139541,
+      "learning_rate": 1.6910741779681264e-05,
+      "loss": 0.2215,
+      "step": 4864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4244076834143325,
+      "learning_rate": 1.690939662247226e-05,
+      "loss": 0.2942,
+      "step": 4865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5023134654075716,
+      "learning_rate": 1.6908051225991146e-05,
+      "loss": 0.3721,
+      "step": 4866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2878202571798292,
+      "learning_rate": 1.6906705590284517e-05,
+      "loss": 0.8234,
+      "step": 4867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37248189574940166,
+      "learning_rate": 1.690535971539897e-05,
+      "loss": 0.1709,
+      "step": 4868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4067382861706183,
+      "learning_rate": 1.690401360138111e-05,
+      "loss": 0.3374,
+      "step": 4869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.316105557965605,
+      "learning_rate": 1.6902667248277557e-05,
+      "loss": 0.2383,
+      "step": 4870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38753037141289065,
+      "learning_rate": 1.6901320656134935e-05,
+      "loss": 0.2109,
+      "step": 4871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9724087834349608,
+      "learning_rate": 1.6899973824999872e-05,
+      "loss": 0.7267,
+      "step": 4872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3812345170205305,
+      "learning_rate": 1.6898626754919018e-05,
+      "loss": 0.3212,
+      "step": 4873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3219187996146285,
+      "learning_rate": 1.6897279445939012e-05,
+      "loss": 0.1968,
+      "step": 4874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6624262246155297,
+      "learning_rate": 1.6895931898106517e-05,
+      "loss": 0.4522,
+      "step": 4875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47937324578610235,
+      "learning_rate": 1.6894584111468196e-05,
+      "loss": 0.2651,
+      "step": 4876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2998688492477607,
+      "learning_rate": 1.6893236086070722e-05,
+      "loss": 0.2302,
+      "step": 4877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4740529224435299,
+      "learning_rate": 1.6891887821960783e-05,
+      "loss": 0.299,
+      "step": 4878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7311170213555515,
+      "learning_rate": 1.689053931918506e-05,
+      "loss": 0.5443,
+      "step": 4879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41467295139387256,
+      "learning_rate": 1.688919057779026e-05,
+      "loss": 0.3303,
+      "step": 4880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3324366508784158,
+      "learning_rate": 1.6887841597823088e-05,
+      "loss": 0.2817,
+      "step": 4881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4272022615661138,
+      "learning_rate": 1.6886492379330254e-05,
+      "loss": 0.2808,
+      "step": 4882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29255807880365126,
+      "learning_rate": 1.6885142922358486e-05,
+      "loss": 0.2237,
+      "step": 4883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8529213592258498,
+      "learning_rate": 1.6883793226954516e-05,
+      "loss": 0.5751,
+      "step": 4884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5232124412222744,
+      "learning_rate": 1.6882443293165083e-05,
+      "loss": 0.3541,
+      "step": 4885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.513732509312485,
+      "learning_rate": 1.6881093121036933e-05,
+      "loss": 0.3429,
+      "step": 4886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43596299045855724,
+      "learning_rate": 1.6879742710616826e-05,
+      "loss": 0.3057,
+      "step": 4887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33911296956631076,
+      "learning_rate": 1.6878392061951525e-05,
+      "loss": 0.2385,
+      "step": 4888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40006287993866735,
+      "learning_rate": 1.6877041175087802e-05,
+      "loss": 0.2909,
+      "step": 4889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43421895746675027,
+      "learning_rate": 1.6875690050072435e-05,
+      "loss": 0.3087,
+      "step": 4890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4299156983242811,
+      "learning_rate": 1.687433868695222e-05,
+      "loss": 0.3169,
+      "step": 4891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38431969838964786,
+      "learning_rate": 1.687298708577395e-05,
+      "loss": 0.3265,
+      "step": 4892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4932553865657155,
+      "learning_rate": 1.687163524658444e-05,
+      "loss": 0.3143,
+      "step": 4893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6671294183699192,
+      "learning_rate": 1.687028316943049e-05,
+      "loss": 0.3774,
+      "step": 4894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5398718612910458,
+      "learning_rate": 1.686893085435893e-05,
+      "loss": 0.2871,
+      "step": 4895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5618030739271445,
+      "learning_rate": 1.686757830141659e-05,
+      "loss": 0.435,
+      "step": 4896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4098895772036004,
+      "learning_rate": 1.6866225510650312e-05,
+      "loss": 0.3044,
+      "step": 4897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3133046926958272,
+      "learning_rate": 1.686487248210694e-05,
+      "loss": 0.2384,
+      "step": 4898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38071433829603696,
+      "learning_rate": 1.6863519215833327e-05,
+      "loss": 0.2787,
+      "step": 4899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4427576937958323,
+      "learning_rate": 1.686216571187634e-05,
+      "loss": 0.2594,
+      "step": 4900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37213118554343044,
+      "learning_rate": 1.6860811970282844e-05,
+      "loss": 0.2937,
+      "step": 4901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5833133836609231,
+      "learning_rate": 1.685945799109973e-05,
+      "loss": 0.3999,
+      "step": 4902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6780494801302973,
+      "learning_rate": 1.6858103774373877e-05,
+      "loss": 0.5117,
+      "step": 4903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.24541000634477722,
+      "learning_rate": 1.685674932015219e-05,
+      "loss": 0.0744,
+      "step": 4904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3375440873846795,
+      "learning_rate": 1.6855394628481565e-05,
+      "loss": 0.3181,
+      "step": 4905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8924703502539542,
+      "learning_rate": 1.6854039699408923e-05,
+      "loss": 0.5844,
+      "step": 4906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4013162255792,
+      "learning_rate": 1.6852684532981176e-05,
+      "loss": 0.1698,
+      "step": 4907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.442763641896375,
+      "learning_rate": 1.685132912924526e-05,
+      "loss": 0.328,
+      "step": 4908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.48415946618162076,
+      "learning_rate": 1.684997348824811e-05,
+      "loss": 0.3534,
+      "step": 4909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27868722935079016,
+      "learning_rate": 1.6848617610036676e-05,
+      "loss": 0.0971,
+      "step": 4910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32980783132441815,
+      "learning_rate": 1.6847261494657904e-05,
+      "loss": 0.2478,
+      "step": 4911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1178058794694627,
+      "learning_rate": 1.6845905142158764e-05,
+      "loss": 0.6226,
+      "step": 4912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2863922274104645,
+      "learning_rate": 1.6844548552586225e-05,
+      "loss": 0.2348,
+      "step": 4913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49945881324068725,
+      "learning_rate": 1.684319172598726e-05,
+      "loss": 0.3969,
+      "step": 4914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7076134237344921,
+      "learning_rate": 1.684183466240886e-05,
+      "loss": 0.5169,
+      "step": 4915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3167088144334443,
+      "learning_rate": 1.684047736189802e-05,
+      "loss": 0.2335,
+      "step": 4916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3104910368581827,
+      "learning_rate": 1.6839119824501747e-05,
+      "loss": 0.2116,
+      "step": 4917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9284231793627837,
+      "learning_rate": 1.6837762050267044e-05,
+      "loss": 0.567,
+      "step": 4918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.596611993917205,
+      "learning_rate": 1.683640403924094e-05,
+      "loss": 0.3658,
+      "step": 4919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3944841548350714,
+      "learning_rate": 1.6835045791470453e-05,
+      "loss": 0.2971,
+      "step": 4920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3928615059353819,
+      "learning_rate": 1.683368730700263e-05,
+      "loss": 0.3528,
+      "step": 4921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8844377675062224,
+      "learning_rate": 1.6832328585884505e-05,
+      "loss": 0.4563,
+      "step": 4922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.24744733076868877,
+      "learning_rate": 1.6830969628163134e-05,
+      "loss": 0.1611,
+      "step": 4923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37799077238329093,
+      "learning_rate": 1.6829610433885583e-05,
+      "loss": 0.308,
+      "step": 4924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7391018917542396,
+      "learning_rate": 1.6828251003098913e-05,
+      "loss": 0.448,
+      "step": 4925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4267263829375562,
+      "learning_rate": 1.6826891335850205e-05,
+      "loss": 0.2828,
+      "step": 4926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9901949114209159,
+      "learning_rate": 1.6825531432186545e-05,
+      "loss": 0.6693,
+      "step": 4927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4338305088608627,
+      "learning_rate": 1.682417129215502e-05,
+      "loss": 0.3244,
+      "step": 4928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3122226230150543,
+      "learning_rate": 1.682281091580274e-05,
+      "loss": 0.2397,
+      "step": 4929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2536689856422566,
+      "learning_rate": 1.682145030317681e-05,
+      "loss": 0.1118,
+      "step": 4930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2644511808907366,
+      "learning_rate": 1.6820089454324355e-05,
+      "loss": 0.4325,
+      "step": 4931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3830284678075912,
+      "learning_rate": 1.681872836929249e-05,
+      "loss": 0.297,
+      "step": 4932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44403371374329265,
+      "learning_rate": 1.6817367048128357e-05,
+      "loss": 0.2896,
+      "step": 4933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6483549831795611,
+      "learning_rate": 1.6816005490879096e-05,
+      "loss": 0.3426,
+      "step": 4934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32440225504314685,
+      "learning_rate": 1.6814643697591857e-05,
+      "loss": 0.2452,
+      "step": 4935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39596665611969095,
+      "learning_rate": 1.68132816683138e-05,
+      "loss": 0.2387,
+      "step": 4936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9595030167316971,
+      "learning_rate": 1.6811919403092093e-05,
+      "loss": 0.3894,
+      "step": 4937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4468030719392794,
+      "learning_rate": 1.6810556901973907e-05,
+      "loss": 0.3355,
+      "step": 4938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8717707766987532,
+      "learning_rate": 1.680919416500643e-05,
+      "loss": 0.5147,
+      "step": 4939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38028104901258003,
+      "learning_rate": 1.6807831192236855e-05,
+      "loss": 0.3398,
+      "step": 4940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3862054255947419,
+      "learning_rate": 1.6806467983712375e-05,
+      "loss": 0.2686,
+      "step": 4941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40855243616372805,
+      "learning_rate": 1.68051045394802e-05,
+      "loss": 0.2499,
+      "step": 4942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7503261455917112,
+      "learning_rate": 1.680374085958755e-05,
+      "loss": 0.3199,
+      "step": 4943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47228792715288803,
+      "learning_rate": 1.680237694408165e-05,
+      "loss": 0.3086,
+      "step": 4944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5675653321231017,
+      "learning_rate": 1.6801012793009725e-05,
+      "loss": 0.4409,
+      "step": 4945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.657462471357182,
+      "learning_rate": 1.6799648406419018e-05,
+      "loss": 0.2353,
+      "step": 4946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2388900030606124,
+      "learning_rate": 1.679828378435678e-05,
+      "loss": 0.1982,
+      "step": 4947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4841571640687985,
+      "learning_rate": 1.6796918926870266e-05,
+      "loss": 0.3666,
+      "step": 4948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5190146104941877,
+      "learning_rate": 1.6795553834006746e-05,
+      "loss": 0.2917,
+      "step": 4949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44852090196045463,
+      "learning_rate": 1.6794188505813486e-05,
+      "loss": 0.3222,
+      "step": 4950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.985136552799657,
+      "learning_rate": 1.679282294233777e-05,
+      "loss": 0.622,
+      "step": 4951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32555809213851533,
+      "learning_rate": 1.6791457143626887e-05,
+      "loss": 0.2532,
+      "step": 4952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3561321537682771,
+      "learning_rate": 1.679009110972814e-05,
+      "loss": 0.2804,
+      "step": 4953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6560341934071064,
+      "learning_rate": 1.6788724840688823e-05,
+      "loss": 0.4473,
+      "step": 4954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31007613366907594,
+      "learning_rate": 1.678735833655626e-05,
+      "loss": 0.1873,
+      "step": 4955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.326241133662786,
+      "learning_rate": 1.6785991597377772e-05,
+      "loss": 0.243,
+      "step": 4956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5177649120868529,
+      "learning_rate": 1.6784624623200684e-05,
+      "loss": 0.4381,
+      "step": 4957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.240446473986299,
+      "learning_rate": 1.6783257414072336e-05,
+      "loss": 0.6064,
+      "step": 4958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33109360065083926,
+      "learning_rate": 1.678188997004008e-05,
+      "loss": 0.2334,
+      "step": 4959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39985346698702146,
+      "learning_rate": 1.678052229115126e-05,
+      "loss": 0.3504,
+      "step": 4960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5592780579901474,
+      "learning_rate": 1.6779154377453248e-05,
+      "loss": 0.1822,
+      "step": 4961
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3496638175036046,
+      "learning_rate": 1.677778622899341e-05,
+      "loss": 0.2214,
+      "step": 4962
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0124778899205107,
+      "learning_rate": 1.677641784581913e-05,
+      "loss": 0.6795,
+      "step": 4963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41286864504725357,
+      "learning_rate": 1.6775049227977788e-05,
+      "loss": 0.3245,
+      "step": 4964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.37057078830092144,
+      "learning_rate": 1.6773680375516784e-05,
+      "loss": 0.2586,
+      "step": 4965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6333735841850971,
+      "learning_rate": 1.6772311288483517e-05,
+      "loss": 0.4853,
+      "step": 4966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4697963107259023,
+      "learning_rate": 1.6770941966925404e-05,
+      "loss": 0.2266,
+      "step": 4967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33424175052387356,
+      "learning_rate": 1.676957241088986e-05,
+      "loss": 0.2609,
+      "step": 4968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.485965282810177,
+      "learning_rate": 1.6768202620424317e-05,
+      "loss": 0.2656,
+      "step": 4969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6859894051180101,
+      "learning_rate": 1.6766832595576205e-05,
+      "loss": 0.4445,
+      "step": 4970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41946713395899093,
+      "learning_rate": 1.6765462336392975e-05,
+      "loss": 0.2915,
+      "step": 4971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29292304755912535,
+      "learning_rate": 1.6764091842922076e-05,
+      "loss": 0.2806,
+      "step": 4972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41222303207216854,
+      "learning_rate": 1.6762721115210964e-05,
+      "loss": 0.3055,
+      "step": 4973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.420655109877888,
+      "learning_rate": 1.676135015330711e-05,
+      "loss": 0.308,
+      "step": 4974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29319502490689703,
+      "learning_rate": 1.675997895725799e-05,
+      "loss": 0.1811,
+      "step": 4975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4219624190782678,
+      "learning_rate": 1.675860752711109e-05,
+      "loss": 0.33,
+      "step": 4976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3887698623250924,
+      "learning_rate": 1.675723586291391e-05,
+      "loss": 0.2888,
+      "step": 4977
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6488987519940098,
+      "learning_rate": 1.6755863964713934e-05,
+      "loss": 0.3712,
+      "step": 4978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42002949206495394,
+      "learning_rate": 1.675449183255868e-05,
+      "loss": 0.3128,
+      "step": 4979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33873593445127587,
+      "learning_rate": 1.6753119466495667e-05,
+      "loss": 0.2737,
+      "step": 4980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.334940371790023,
+      "learning_rate": 1.6751746866572415e-05,
+      "loss": 0.2823,
+      "step": 4981
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9608719470015646,
+      "learning_rate": 1.6750374032836462e-05,
+      "loss": 0.3832,
+      "step": 4982
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33594948667433394,
+      "learning_rate": 1.6749000965335344e-05,
+      "loss": 0.2858,
+      "step": 4983
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3677272022726247,
+      "learning_rate": 1.6747627664116615e-05,
+      "loss": 0.3131,
+      "step": 4984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8046638072008366,
+      "learning_rate": 1.674625412922783e-05,
+      "loss": 0.4111,
+      "step": 4985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32686647655558276,
+      "learning_rate": 1.6744880360716555e-05,
+      "loss": 0.2732,
+      "step": 4986
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41757090613796255,
+      "learning_rate": 1.6743506358630358e-05,
+      "loss": 0.2837,
+      "step": 4987
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35935865239680986,
+      "learning_rate": 1.674213212301683e-05,
+      "loss": 0.268,
+      "step": 4988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4316263565610041,
+      "learning_rate": 1.6740757653923557e-05,
+      "loss": 0.2988,
+      "step": 4989
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5887593298772523,
+      "learning_rate": 1.673938295139814e-05,
+      "loss": 0.3779,
+      "step": 4990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5006344489216323,
+      "learning_rate": 1.6738008015488174e-05,
+      "loss": 0.3939,
+      "step": 4991
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32402544400838923,
+      "learning_rate": 1.6736632846241282e-05,
+      "loss": 0.2105,
+      "step": 4992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3172635891197995,
+      "learning_rate": 1.6735257443705085e-05,
+      "loss": 0.2832,
+      "step": 4993
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3766031153954571,
+      "learning_rate": 1.6733881807927212e-05,
+      "loss": 0.3244,
+      "step": 4994
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33773156689449546,
+      "learning_rate": 1.67325059389553e-05,
+      "loss": 0.1948,
+      "step": 4995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4008554402803368,
+      "learning_rate": 1.6731129836837e-05,
+      "loss": 0.3248,
+      "step": 4996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4279087807348623,
+      "learning_rate": 1.672975350161996e-05,
+      "loss": 0.7799,
+      "step": 4997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32856664750277276,
+      "learning_rate": 1.672837693335185e-05,
+      "loss": 0.2259,
+      "step": 4998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3994700065203357,
+      "learning_rate": 1.6727000132080332e-05,
+      "loss": 0.3668,
+      "step": 4999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4195276798735785,
+      "learning_rate": 1.6725623097853092e-05,
+      "loss": 0.3758,
+      "step": 5000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.19978298158962446,
+      "learning_rate": 1.672424583071781e-05,
+      "loss": 0.1467,
+      "step": 5001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.198150816066066,
+      "learning_rate": 1.672286833072219e-05,
+      "loss": 0.6234,
+      "step": 5002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.45414769727547355,
+      "learning_rate": 1.6721490597913922e-05,
+      "loss": 0.3633,
+      "step": 5003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38110366694566206,
+      "learning_rate": 1.672011263234073e-05,
+      "loss": 0.2942,
+      "step": 5004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43945156180002193,
+      "learning_rate": 1.671873443405032e-05,
+      "loss": 0.3225,
+      "step": 5005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44775285754769784,
+      "learning_rate": 1.671735600309043e-05,
+      "loss": 0.3277,
+      "step": 5006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31152176309096835,
+      "learning_rate": 1.6715977339508787e-05,
+      "loss": 0.2134,
+      "step": 5007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.335916270130865,
+      "learning_rate": 1.6714598443353137e-05,
+      "loss": 0.2523,
+      "step": 5008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2214895870371527,
+      "learning_rate": 1.6713219314671235e-05,
+      "loss": 0.7403,
+      "step": 5009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5000389590232583,
+      "learning_rate": 1.6711839953510835e-05,
+      "loss": 0.3313,
+      "step": 5010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.371776258828232,
+      "learning_rate": 1.6710460359919708e-05,
+      "loss": 0.2937,
+      "step": 5011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3811309362080807,
+      "learning_rate": 1.6709080533945624e-05,
+      "loss": 0.3059,
+      "step": 5012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3476381176969048,
+      "learning_rate": 1.6707700475636372e-05,
+      "loss": 0.1721,
+      "step": 5013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41573768957075496,
+      "learning_rate": 1.6706320185039742e-05,
+      "loss": 0.2663,
+      "step": 5014
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5511130497075707,
+      "learning_rate": 1.670493966220353e-05,
+      "loss": 0.3951,
+      "step": 5015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34083421472885944,
+      "learning_rate": 1.6703558907175548e-05,
+      "loss": 0.3184,
+      "step": 5016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47507592384082353,
+      "learning_rate": 1.6702177920003607e-05,
+      "loss": 0.4127,
+      "step": 5017
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8809556008590094,
+      "learning_rate": 1.6700796700735534e-05,
+      "loss": 0.3604,
+      "step": 5018
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3058900980432858,
+      "learning_rate": 1.669941524941916e-05,
+      "loss": 0.1977,
+      "step": 5019
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4092102952384807,
+      "learning_rate": 1.6698033566102322e-05,
+      "loss": 0.2895,
+      "step": 5020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7002227774841857,
+      "learning_rate": 1.6696651650832874e-05,
+      "loss": 0.3808,
+      "step": 5021
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39936961207401184,
+      "learning_rate": 1.6695269503658663e-05,
+      "loss": 0.339,
+      "step": 5022
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5260095728189574,
+      "learning_rate": 1.6693887124627556e-05,
+      "loss": 0.3984,
+      "step": 5023
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41866801996607933,
+      "learning_rate": 1.6692504513787432e-05,
+      "loss": 0.2768,
+      "step": 5024
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2600843617318522,
+      "learning_rate": 1.669112167118616e-05,
+      "loss": 0.1399,
+      "step": 5025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44756851536480885,
+      "learning_rate": 1.668973859687163e-05,
+      "loss": 0.3325,
+      "step": 5026
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42835429389794066,
+      "learning_rate": 1.6688355290891746e-05,
+      "loss": 0.308,
+      "step": 5027
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5913982247983119,
+      "learning_rate": 1.66869717532944e-05,
+      "loss": 0.3365,
+      "step": 5028
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40328321757099445,
+      "learning_rate": 1.6685587984127513e-05,
+      "loss": 0.3689,
+      "step": 5029
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0759732739079926,
+      "learning_rate": 1.6684203983439e-05,
+      "loss": 0.606,
+      "step": 5030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2783514820703303,
+      "learning_rate": 1.668281975127679e-05,
+      "loss": 0.1537,
+      "step": 5031
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33181140217007127,
+      "learning_rate": 1.6681435287688823e-05,
+      "loss": 0.2664,
+      "step": 5032
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8302507762797569,
+      "learning_rate": 1.6680050592723038e-05,
+      "loss": 0.5131,
+      "step": 5033
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6098011206817671,
+      "learning_rate": 1.6678665666427387e-05,
+      "loss": 0.3049,
+      "step": 5034
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35164039268183567,
+      "learning_rate": 1.6677280508849828e-05,
+      "loss": 0.3084,
+      "step": 5035
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.951993401806805,
+      "learning_rate": 1.667589512003834e-05,
+      "loss": 0.6228,
+      "step": 5036
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34895682350666707,
+      "learning_rate": 1.6674509500040885e-05,
+      "loss": 0.2188,
+      "step": 5037
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3204560482309398,
+      "learning_rate": 1.6673123648905454e-05,
+      "loss": 0.2321,
+      "step": 5038
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4577995928154056,
+      "learning_rate": 1.667173756668004e-05,
+      "loss": 0.337,
+      "step": 5039
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3938689527075941,
+      "learning_rate": 1.667035125341264e-05,
+      "loss": 0.2516,
+      "step": 5040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.572570413679517,
+      "learning_rate": 1.6668964709151265e-05,
+      "loss": 0.4199,
+      "step": 5041
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4299646855354586,
+      "learning_rate": 1.6667577933943925e-05,
+      "loss": 0.7393,
+      "step": 5042
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33121075066613553,
+      "learning_rate": 1.666619092783865e-05,
+      "loss": 0.3047,
+      "step": 5043
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24035738339207335,
+      "learning_rate": 1.6664803690883465e-05,
+      "loss": 0.1529,
+      "step": 5044
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5652684265469679,
+      "learning_rate": 1.666341622312642e-05,
+      "loss": 0.4583,
+      "step": 5045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8551662769154373,
+      "learning_rate": 1.666202852461556e-05,
+      "loss": 0.5423,
+      "step": 5046
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3637742889707305,
+      "learning_rate": 1.6660640595398934e-05,
+      "loss": 0.2563,
+      "step": 5047
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49084959630933184,
+      "learning_rate": 1.6659252435524613e-05,
+      "loss": 0.3588,
+      "step": 5048
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9077022712537721,
+      "learning_rate": 1.6657864045040665e-05,
+      "loss": 0.4459,
+      "step": 5049
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44606011433230913,
+      "learning_rate": 1.665647542399517e-05,
+      "loss": 0.2837,
+      "step": 5050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40943914248299257,
+      "learning_rate": 1.665508657243622e-05,
+      "loss": 0.3512,
+      "step": 5051
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30297531819075785,
+      "learning_rate": 1.6653697490411905e-05,
+      "loss": 0.187,
+      "step": 5052
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3876810435050498,
+      "learning_rate": 1.6652308177970335e-05,
+      "loss": 0.2396,
+      "step": 5053
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1355168231759345,
+      "learning_rate": 1.665091863515962e-05,
+      "loss": 0.6173,
+      "step": 5054
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3430108180818135,
+      "learning_rate": 1.6649528862027877e-05,
+      "loss": 0.3208,
+      "step": 5055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3775947002380297,
+      "learning_rate": 1.6648138858623236e-05,
+      "loss": 0.3046,
+      "step": 5056
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3376050104963141,
+      "learning_rate": 1.664674862499383e-05,
+      "loss": 0.1998,
+      "step": 5057
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5965276651128877,
+      "learning_rate": 1.664535816118781e-05,
+      "loss": 0.4102,
+      "step": 5058
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4126818965088,
+      "learning_rate": 1.6643967467253316e-05,
+      "loss": 0.3416,
+      "step": 5059
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35265629901301665,
+      "learning_rate": 1.664257654323852e-05,
+      "loss": 0.282,
+      "step": 5060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46529035889049486,
+      "learning_rate": 1.664118538919158e-05,
+      "loss": 0.3418,
+      "step": 5061
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.386042563376412,
+      "learning_rate": 1.6639794005160677e-05,
+      "loss": 0.3206,
+      "step": 5062
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.347541240252572,
+      "learning_rate": 1.6638402391193993e-05,
+      "loss": 0.234,
+      "step": 5063
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3657479293256798,
+      "learning_rate": 1.663701054733972e-05,
+      "loss": 0.8305,
+      "step": 5064
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35412391052650544,
+      "learning_rate": 1.6635618473646058e-05,
+      "loss": 0.2778,
+      "step": 5065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32667171769057474,
+      "learning_rate": 1.6634226170161213e-05,
+      "loss": 0.2391,
+      "step": 5066
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43478785542869364,
+      "learning_rate": 1.66328336369334e-05,
+      "loss": 0.3431,
+      "step": 5067
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33734429475445615,
+      "learning_rate": 1.6631440874010842e-05,
+      "loss": 0.3147,
+      "step": 5068
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8787246640463405,
+      "learning_rate": 1.6630047881441774e-05,
+      "loss": 0.6012,
+      "step": 5069
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49363055525613264,
+      "learning_rate": 1.6628654659274433e-05,
+      "loss": 0.3192,
+      "step": 5070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30082111207502216,
+      "learning_rate": 1.6627261207557068e-05,
+      "loss": 0.2643,
+      "step": 5071
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26296980966438765,
+      "learning_rate": 1.6625867526337928e-05,
+      "loss": 0.1485,
+      "step": 5072
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6053081587769884,
+      "learning_rate": 1.662447361566528e-05,
+      "loss": 0.3424,
+      "step": 5073
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41307754279842807,
+      "learning_rate": 1.6623079475587403e-05,
+      "loss": 0.3124,
+      "step": 5074
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40972224319212935,
+      "learning_rate": 1.6621685106152564e-05,
+      "loss": 0.3288,
+      "step": 5075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49432112174734544,
+      "learning_rate": 1.6620290507409053e-05,
+      "loss": 0.3193,
+      "step": 5076
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.380041557575946,
+      "learning_rate": 1.6618895679405165e-05,
+      "loss": 0.283,
+      "step": 5077
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35160063555106125,
+      "learning_rate": 1.6617500622189208e-05,
+      "loss": 0.246,
+      "step": 5078
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3811836486280641,
+      "learning_rate": 1.6616105335809487e-05,
+      "loss": 0.3454,
+      "step": 5079
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33824258296520193,
+      "learning_rate": 1.6614709820314323e-05,
+      "loss": 0.2238,
+      "step": 5080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8228245625272075,
+      "learning_rate": 1.6613314075752044e-05,
+      "loss": 0.5268,
+      "step": 5081
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1414700434465859,
+      "learning_rate": 1.661191810217098e-05,
+      "loss": 0.7977,
+      "step": 5082
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.294198134495597,
+      "learning_rate": 1.661052189961948e-05,
+      "loss": 0.2376,
+      "step": 5083
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46694318435537846,
+      "learning_rate": 1.660912546814589e-05,
+      "loss": 0.3321,
+      "step": 5084
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40253889624722783,
+      "learning_rate": 1.6607728807798568e-05,
+      "loss": 0.2855,
+      "step": 5085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3838630594646114,
+      "learning_rate": 1.660633191862588e-05,
+      "loss": 0.2163,
+      "step": 5086
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4452069568209437,
+      "learning_rate": 1.6604934800676207e-05,
+      "loss": 0.3252,
+      "step": 5087
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.205226316107024,
+      "learning_rate": 1.660353745399792e-05,
+      "loss": 0.7247,
+      "step": 5088
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32386036990488415,
+      "learning_rate": 1.6602139878639417e-05,
+      "loss": 0.2436,
+      "step": 5089
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7146860073301247,
+      "learning_rate": 1.6600742074649095e-05,
+      "loss": 0.4543,
+      "step": 5090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2521102701136926,
+      "learning_rate": 1.659934404207536e-05,
+      "loss": 0.2253,
+      "step": 5091
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3929148102982469,
+      "learning_rate": 1.6597945780966626e-05,
+      "loss": 0.2923,
+      "step": 5092
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7744747631107772,
+      "learning_rate": 1.659654729137131e-05,
+      "loss": 0.3298,
+      "step": 5093
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.385373842207475,
+      "learning_rate": 1.6595148573337843e-05,
+      "loss": 0.3322,
+      "step": 5094
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44188848071352976,
+      "learning_rate": 1.6593749626914665e-05,
+      "loss": 0.3031,
+      "step": 5095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.596587404876109,
+      "learning_rate": 1.6592350452150223e-05,
+      "loss": 0.34,
+      "step": 5096
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2930106307014018,
+      "learning_rate": 1.6590951049092966e-05,
+      "loss": 0.1554,
+      "step": 5097
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.45593236462612596,
+      "learning_rate": 1.658955141779136e-05,
+      "loss": 0.3053,
+      "step": 5098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5457068512880814,
+      "learning_rate": 1.6588151558293874e-05,
+      "loss": 0.2644,
+      "step": 5099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2365592496960516,
+      "learning_rate": 1.658675147064898e-05,
+      "loss": 0.8412,
+      "step": 5100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4162610077770308,
+      "learning_rate": 1.6585351154905163e-05,
+      "loss": 0.2924,
+      "step": 5101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4220963462240051,
+      "learning_rate": 1.6583950611110923e-05,
+      "loss": 0.3159,
+      "step": 5102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44314877537383723,
+      "learning_rate": 1.6582549839314756e-05,
+      "loss": 0.2598,
+      "step": 5103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2803168276990925,
+      "learning_rate": 1.658114883956517e-05,
+      "loss": 0.2255,
+      "step": 5104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9679407388981311,
+      "learning_rate": 1.6579747611910684e-05,
+      "loss": 0.557,
+      "step": 5105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7293098670982834,
+      "learning_rate": 1.657834615639982e-05,
+      "loss": 0.2901,
+      "step": 5106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3672744421536344,
+      "learning_rate": 1.6576944473081112e-05,
+      "loss": 0.2756,
+      "step": 5107
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8130875641546872,
+      "learning_rate": 1.6575542562003097e-05,
+      "loss": 0.5067,
+      "step": 5108
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2011438979448081,
+      "learning_rate": 1.6574140423214327e-05,
+      "loss": 0.1124,
+      "step": 5109
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38553771059667385,
+      "learning_rate": 1.657273805676336e-05,
+      "loss": 0.2765,
+      "step": 5110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4494145707717537,
+      "learning_rate": 1.6571335462698755e-05,
+      "loss": 0.3405,
+      "step": 5111
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1128740224578688,
+      "learning_rate": 1.6569932641069083e-05,
+      "loss": 0.4035,
+      "step": 5112
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40172625854201377,
+      "learning_rate": 1.656852959192293e-05,
+      "loss": 0.325,
+      "step": 5113
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4042048518785761,
+      "learning_rate": 1.656712631530888e-05,
+      "loss": 0.3685,
+      "step": 5114
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25272424394886184,
+      "learning_rate": 1.6565722811275526e-05,
+      "loss": 0.1761,
+      "step": 5115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3961321724738942,
+      "learning_rate": 1.6564319079871472e-05,
+      "loss": 0.2789,
+      "step": 5116
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5680165426058105,
+      "learning_rate": 1.656291512114533e-05,
+      "loss": 0.3786,
+      "step": 5117
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5206323825700394,
+      "learning_rate": 1.656151093514572e-05,
+      "loss": 0.3638,
+      "step": 5118
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3543282139228944,
+      "learning_rate": 1.6560106521921272e-05,
+      "loss": 0.248,
+      "step": 5119
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9717910872721999,
+      "learning_rate": 1.6558701881520616e-05,
+      "loss": 0.6354,
+      "step": 5120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4011467519529105,
+      "learning_rate": 1.6557297013992395e-05,
+      "loss": 0.2849,
+      "step": 5121
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2840900794225836,
+      "learning_rate": 1.6555891919385262e-05,
+      "loss": 0.1773,
+      "step": 5122
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4765438374827972,
+      "learning_rate": 1.655448659774787e-05,
+      "loss": 0.3306,
+      "step": 5123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0038008012167339,
+      "learning_rate": 1.6553081049128894e-05,
+      "loss": 0.4335,
+      "step": 5124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3737795700734047,
+      "learning_rate": 1.6551675273577e-05,
+      "loss": 0.2289,
+      "step": 5125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0662857043698486,
+      "learning_rate": 1.6550269271140872e-05,
+      "loss": 0.6902,
+      "step": 5126
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3832966114148078,
+      "learning_rate": 1.6548863041869203e-05,
+      "loss": 0.3521,
+      "step": 5127
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2462215752704896,
+      "learning_rate": 1.6547456585810687e-05,
+      "loss": 0.1567,
+      "step": 5128
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9104250748758096,
+      "learning_rate": 1.6546049903014034e-05,
+      "loss": 0.5161,
+      "step": 5129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39445391596511015,
+      "learning_rate": 1.6544642993527952e-05,
+      "loss": 0.3519,
+      "step": 5130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4248670156930363,
+      "learning_rate": 1.6543235857401163e-05,
+      "loss": 0.3151,
+      "step": 5131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5492970992711453,
+      "learning_rate": 1.6541828494682398e-05,
+      "loss": 0.3257,
+      "step": 5132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4673011597855598,
+      "learning_rate": 1.6540420905420395e-05,
+      "loss": 0.2742,
+      "step": 5133
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39281271146739444,
+      "learning_rate": 1.6539013089663897e-05,
+      "loss": 0.3005,
+      "step": 5134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4087028084684856,
+      "learning_rate": 1.6537605047461654e-05,
+      "loss": 0.2509,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8586628812462002,
+      "learning_rate": 1.6536196778862433e-05,
+      "loss": 0.5085,
+      "step": 5136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5709002766581852,
+      "learning_rate": 1.6534788283915e-05,
+      "loss": 0.3931,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3001544493179158,
+      "learning_rate": 1.6533379562668126e-05,
+      "loss": 0.261,
+      "step": 5138
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2209468608505851,
+      "learning_rate": 1.6531970615170602e-05,
+      "loss": 0.7915,
+      "step": 5139
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29863282667681906,
+      "learning_rate": 1.6530561441471215e-05,
+      "loss": 0.2372,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6536704998804858,
+      "learning_rate": 1.6529152041618767e-05,
+      "loss": 0.3358,
+      "step": 5141
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2971516691014197,
+      "learning_rate": 1.652774241566206e-05,
+      "loss": 0.2563,
+      "step": 5142
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4049693944771069,
+      "learning_rate": 1.652633256364992e-05,
+      "loss": 0.3354,
+      "step": 5143
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.569984221282444,
+      "learning_rate": 1.652492248563116e-05,
+      "loss": 0.452,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6023169787669885,
+      "learning_rate": 1.6523512181654616e-05,
+      "loss": 0.379,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3194650614374405,
+      "learning_rate": 1.6522101651769124e-05,
+      "loss": 0.2582,
+      "step": 5146
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6151729058432179,
+      "learning_rate": 1.6520690896023536e-05,
+      "loss": 0.3939,
+      "step": 5147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2655700172516779,
+      "learning_rate": 1.6519279914466703e-05,
+      "loss": 0.1816,
+      "step": 5148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8313908290203407,
+      "learning_rate": 1.6517868707147484e-05,
+      "loss": 0.461,
+      "step": 5149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40243904423512905,
+      "learning_rate": 1.651645727411475e-05,
+      "loss": 0.3019,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48595523304443944,
+      "learning_rate": 1.6515045615417385e-05,
+      "loss": 0.2973,
+      "step": 5151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6055444378343406,
+      "learning_rate": 1.6513633731104268e-05,
+      "loss": 0.3708,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4178836687928149,
+      "learning_rate": 1.6512221621224296e-05,
+      "loss": 0.336,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2420107469251138,
+      "learning_rate": 1.651080928582637e-05,
+      "loss": 0.2302,
+      "step": 5154
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2784278043315081,
+      "learning_rate": 1.6509396724959396e-05,
+      "loss": 0.1629,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4202735086060693,
+      "learning_rate": 1.6507983938672295e-05,
+      "loss": 0.3557,
+      "step": 5156
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7828694849798539,
+      "learning_rate": 1.650657092701399e-05,
+      "loss": 0.5474,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30049067721571665,
+      "learning_rate": 1.6505157690033417e-05,
+      "loss": 0.2597,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3824140483311469,
+      "learning_rate": 1.650374422777951e-05,
+      "loss": 0.3437,
+      "step": 5159
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42131426536676336,
+      "learning_rate": 1.6502330540301217e-05,
+      "loss": 0.266,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.337804821268556,
+      "learning_rate": 1.6500916627647498e-05,
+      "loss": 0.224,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30612728307066595,
+      "learning_rate": 1.6499502489867318e-05,
+      "loss": 0.2456,
+      "step": 5162
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5060885778707491,
+      "learning_rate": 1.6498088127009647e-05,
+      "loss": 0.4094,
+      "step": 5163
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34297861726048556,
+      "learning_rate": 1.649667353912346e-05,
+      "loss": 0.2678,
+      "step": 5164
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5259607161237364,
+      "learning_rate": 1.649525872625775e-05,
+      "loss": 0.3798,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3577087368747908,
+      "learning_rate": 1.649384368846151e-05,
+      "loss": 0.3183,
+      "step": 5166
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8135960641640962,
+      "learning_rate": 1.649242842578374e-05,
+      "loss": 0.504,
+      "step": 5167
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.22855591149860482,
+      "learning_rate": 1.6491012938273457e-05,
+      "loss": 0.1531,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3865190212108506,
+      "learning_rate": 1.6489597225979673e-05,
+      "loss": 0.3684,
+      "step": 5169
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45872722550496825,
+      "learning_rate": 1.6488181288951416e-05,
+      "loss": 0.4088,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2849367931072605,
+      "learning_rate": 1.648676512723772e-05,
+      "loss": 0.2268,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.246603368276133,
+      "learning_rate": 1.6485348740887624e-05,
+      "loss": 0.7727,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3887406473005966,
+      "learning_rate": 1.6483932129950183e-05,
+      "loss": 0.2651,
+      "step": 5173
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24155302974837958,
+      "learning_rate": 1.648251529447445e-05,
+      "loss": 0.2227,
+      "step": 5174
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5747540021102897,
+      "learning_rate": 1.6481098234509493e-05,
+      "loss": 0.4604,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6065999774862946,
+      "learning_rate": 1.647968095010438e-05,
+      "loss": 0.4313,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3560299438233281,
+      "learning_rate": 1.6478263441308197e-05,
+      "loss": 0.2519,
+      "step": 5177
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4235898813948726,
+      "learning_rate": 1.6476845708170025e-05,
+      "loss": 0.3201,
+      "step": 5178
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7045373601265343,
+      "learning_rate": 1.647542775073897e-05,
+      "loss": 0.45,
+      "step": 5179
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41738227489163404,
+      "learning_rate": 1.647400956906413e-05,
+      "loss": 0.3245,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3797306952733853,
+      "learning_rate": 1.6472591163194613e-05,
+      "loss": 0.2625,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2729805774158222,
+      "learning_rate": 1.6471172533179545e-05,
+      "loss": 0.2373,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42957414092760055,
+      "learning_rate": 1.646975367906805e-05,
+      "loss": 0.3196,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.040323100353576,
+      "learning_rate": 1.6468334600909265e-05,
+      "loss": 0.4741,
+      "step": 5184
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4346814146354387,
+      "learning_rate": 1.646691529875233e-05,
+      "loss": 0.3474,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.36473624376665237,
+      "learning_rate": 1.6465495772646395e-05,
+      "loss": 0.3155,
+      "step": 5186
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3374164626344427,
+      "learning_rate": 1.646407602264062e-05,
+      "loss": 0.2212,
+      "step": 5187
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9076271617776354,
+      "learning_rate": 1.646265604878417e-05,
+      "loss": 0.5641,
+      "step": 5188
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34881563109518027,
+      "learning_rate": 1.6461235851126217e-05,
+      "loss": 0.2696,
+      "step": 5189
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4391047335953409,
+      "learning_rate": 1.6459815429715947e-05,
+      "loss": 0.2938,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6558038337089468,
+      "learning_rate": 1.6458394784602548e-05,
+      "loss": 0.4387,
+      "step": 5191
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3632607082060281,
+      "learning_rate": 1.6456973915835216e-05,
+      "loss": 0.3324,
+      "step": 5192
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8758092746992635,
+      "learning_rate": 1.645555282346315e-05,
+      "loss": 0.5511,
+      "step": 5193
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25760201740560185,
+      "learning_rate": 1.645413150753557e-05,
+      "loss": 0.1809,
+      "step": 5194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37353304142667226,
+      "learning_rate": 1.645270996810169e-05,
+      "loss": 0.2963,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.087573636596399,
+      "learning_rate": 1.645128820521075e-05,
+      "loss": 0.7903,
+      "step": 5196
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33864893891719766,
+      "learning_rate": 1.644986621891197e-05,
+      "loss": 0.281,
+      "step": 5197
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39698780680277895,
+      "learning_rate": 1.64484440092546e-05,
+      "loss": 0.2983,
+      "step": 5198
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8776485906722473,
+      "learning_rate": 1.6447021576287893e-05,
+      "loss": 0.5644,
+      "step": 5199
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18605182562974185,
+      "learning_rate": 1.6445598920061104e-05,
+      "loss": 0.1143,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3799839090356079,
+      "learning_rate": 1.6444176040623506e-05,
+      "loss": 0.2902,
+      "step": 5201
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3898545043893843,
+      "learning_rate": 1.6442752938024367e-05,
+      "loss": 0.3586,
+      "step": 5202
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7637000427048503,
+      "learning_rate": 1.644132961231297e-05,
+      "loss": 0.3395,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41850232975506924,
+      "learning_rate": 1.6439906063538602e-05,
+      "loss": 0.32,
+      "step": 5204
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5203465807048746,
+      "learning_rate": 1.6438482291750567e-05,
+      "loss": 0.4057,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2428540144139758,
+      "learning_rate": 1.6437058296998168e-05,
+      "loss": 0.1995,
+      "step": 5206
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3433948371999074,
+      "learning_rate": 1.643563407933072e-05,
+      "loss": 0.2046,
+      "step": 5207
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7815039800380226,
+      "learning_rate": 1.6434209638797535e-05,
+      "loss": 0.5257,
+      "step": 5208
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41070430570121524,
+      "learning_rate": 1.643278497544795e-05,
+      "loss": 0.3437,
+      "step": 5209
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3379383058006753,
+      "learning_rate": 1.6431360089331297e-05,
+      "loss": 0.2553,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.943370279969184,
+      "learning_rate": 1.642993498049692e-05,
+      "loss": 0.6497,
+      "step": 5211
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2744195839372174,
+      "learning_rate": 1.6428509648994172e-05,
+      "loss": 0.1808,
+      "step": 5212
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2973940784991242,
+      "learning_rate": 1.6427084094872413e-05,
+      "loss": 0.2217,
+      "step": 5213
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5296153676122131,
+      "learning_rate": 1.6425658318181007e-05,
+      "loss": 0.4247,
+      "step": 5214
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7738880370457544,
+      "learning_rate": 1.6424232318969327e-05,
+      "loss": 0.4883,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3653662339834992,
+      "learning_rate": 1.642280609728676e-05,
+      "loss": 0.2146,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.532355787670923,
+      "learning_rate": 1.6421379653182695e-05,
+      "loss": 0.3656,
+      "step": 5217
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2788563700674229,
+      "learning_rate": 1.6419952986706523e-05,
+      "loss": 0.2249,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42198391043135125,
+      "learning_rate": 1.641852609790766e-05,
+      "loss": 0.3224,
+      "step": 5219
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8746102140737043,
+      "learning_rate": 1.641709898683552e-05,
+      "loss": 0.3473,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4154761127237117,
+      "learning_rate": 1.641567165353951e-05,
+      "loss": 0.3684,
+      "step": 5221
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41962129121867997,
+      "learning_rate": 1.6414244098069068e-05,
+      "loss": 0.2641,
+      "step": 5222
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46239702845240743,
+      "learning_rate": 1.641281632047363e-05,
+      "loss": 0.2797,
+      "step": 5223
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34177914074807497,
+      "learning_rate": 1.6411388320802637e-05,
+      "loss": 0.2041,
+      "step": 5224
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3992637612794155,
+      "learning_rate": 1.6409960099105543e-05,
+      "loss": 0.2882,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7002086710277914,
+      "learning_rate": 1.6408531655431806e-05,
+      "loss": 0.3284,
+      "step": 5226
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3006242151336662,
+      "learning_rate": 1.6407102989830894e-05,
+      "loss": 0.7902,
+      "step": 5227
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4334420186583708,
+      "learning_rate": 1.640567410235228e-05,
+      "loss": 0.2615,
+      "step": 5228
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6111767264288636,
+      "learning_rate": 1.6404244993045447e-05,
+      "loss": 0.2422,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30863707616793146,
+      "learning_rate": 1.6402815661959886e-05,
+      "loss": 0.227,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38996995551842206,
+      "learning_rate": 1.6401386109145098e-05,
+      "loss": 0.291,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9864271562835062,
+      "learning_rate": 1.639995633465058e-05,
+      "loss": 0.4252,
+      "step": 5232
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5006342644476464,
+      "learning_rate": 1.6398526338525852e-05,
+      "loss": 0.3058,
+      "step": 5233
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3515919149907774,
+      "learning_rate": 1.639709612082043e-05,
+      "loss": 0.2666,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9812781991271189,
+      "learning_rate": 1.6395665681583842e-05,
+      "loss": 0.5665,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5594165984678707,
+      "learning_rate": 1.639423502086563e-05,
+      "loss": 0.3173,
+      "step": 5236
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4214927341840395,
+      "learning_rate": 1.6392804138715334e-05,
+      "loss": 0.28,
+      "step": 5237
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.368125224933628,
+      "learning_rate": 1.6391373035182506e-05,
+      "loss": 0.296,
+      "step": 5238
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29769042288686703,
+      "learning_rate": 1.6389941710316703e-05,
+      "loss": 0.183,
+      "step": 5239
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42317068403499525,
+      "learning_rate": 1.6388510164167492e-05,
+      "loss": 0.2838,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38688120862100955,
+      "learning_rate": 1.6387078396784447e-05,
+      "loss": 0.317,
+      "step": 5241
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9506995811934351,
+      "learning_rate": 1.6385646408217158e-05,
+      "loss": 0.492,
+      "step": 5242
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3817544458955061,
+      "learning_rate": 1.638421419851521e-05,
+      "loss": 0.261,
+      "step": 5243
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3240206473375033,
+      "learning_rate": 1.638278176772819e-05,
+      "loss": 0.2811,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3847891589500509,
+      "learning_rate": 1.6381349115905718e-05,
+      "loss": 0.3365,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3515846322925854,
+      "learning_rate": 1.6379916243097398e-05,
+      "loss": 0.2199,
+      "step": 5246
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4147502570911691,
+      "learning_rate": 1.6378483149352857e-05,
+      "loss": 0.2787,
+      "step": 5247
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7085441353078938,
+      "learning_rate": 1.6377049834721713e-05,
+      "loss": 0.4585,
+      "step": 5248
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3256721748852568,
+      "learning_rate": 1.637561629925361e-05,
+      "loss": 0.2384,
+      "step": 5249
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.850566415646217,
+      "learning_rate": 1.637418254299819e-05,
+      "loss": 0.5317,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4374892514913494,
+      "learning_rate": 1.63727485660051e-05,
+      "loss": 0.3344,
+      "step": 5251
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24873000470380066,
+      "learning_rate": 1.6371314368324002e-05,
+      "loss": 0.136,
+      "step": 5252
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3694156067809314,
+      "learning_rate": 1.6369879950004564e-05,
+      "loss": 0.3051,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8041414855390386,
+      "learning_rate": 1.6368445311096452e-05,
+      "loss": 0.4433,
+      "step": 5254
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5618816209689073,
+      "learning_rate": 1.6367010451649357e-05,
+      "loss": 0.3546,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39348874710374027,
+      "learning_rate": 1.636557537171296e-05,
+      "loss": 0.2843,
+      "step": 5256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3829165507517021,
+      "learning_rate": 1.6364140071336967e-05,
+      "loss": 0.3449,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3896253368807964,
+      "learning_rate": 1.6362704550571073e-05,
+      "loss": 0.2402,
+      "step": 5258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2660992170071291,
+      "learning_rate": 1.6361268809464998e-05,
+      "loss": 0.2279,
+      "step": 5259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7432002175469005,
+      "learning_rate": 1.6359832848068455e-05,
+      "loss": 0.4667,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33013913089948044,
+      "learning_rate": 1.6358396666431176e-05,
+      "loss": 0.269,
+      "step": 5261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39504619567062776,
+      "learning_rate": 1.635696026460289e-05,
+      "loss": 0.2618,
+      "step": 5262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2763020879980969,
+      "learning_rate": 1.6355523642633346e-05,
+      "loss": 0.832,
+      "step": 5263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23495958054537083,
+      "learning_rate": 1.6354086800572287e-05,
+      "loss": 0.1672,
+      "step": 5264
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28905131340359264,
+      "learning_rate": 1.635264973846948e-05,
+      "loss": 0.2705,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7167395323464834,
+      "learning_rate": 1.6351212456374684e-05,
+      "loss": 0.4445,
+      "step": 5266
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4054608197750515,
+      "learning_rate": 1.6349774954337676e-05,
+      "loss": 0.3239,
+      "step": 5267
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5343281535109453,
+      "learning_rate": 1.6348337232408235e-05,
+      "loss": 0.3897,
+      "step": 5268
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3590528787248728,
+      "learning_rate": 1.6346899290636145e-05,
+      "loss": 0.288,
+      "step": 5269
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3249108619545788,
+      "learning_rate": 1.6345461129071207e-05,
+      "loss": 0.2352,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5694464956960685,
+      "learning_rate": 1.6344022747763225e-05,
+      "loss": 0.3867,
+      "step": 5271
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3305233715465694,
+      "learning_rate": 1.6342584146762005e-05,
+      "loss": 0.2424,
+      "step": 5272
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5162607918437077,
+      "learning_rate": 1.634114532611737e-05,
+      "loss": 0.3408,
+      "step": 5273
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40391491962918796,
+      "learning_rate": 1.6339706285879144e-05,
+      "loss": 0.3143,
+      "step": 5274
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9206819584778668,
+      "learning_rate": 1.6338267026097162e-05,
+      "loss": 0.5015,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4979518942637146,
+      "learning_rate": 1.633682754682127e-05,
+      "loss": 0.3239,
+      "step": 5276
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3491657958787403,
+      "learning_rate": 1.6335387848101307e-05,
+      "loss": 0.3165,
+      "step": 5277
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28912270831094544,
+      "learning_rate": 1.6333947929987137e-05,
+      "loss": 0.177,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5085041190375486,
+      "learning_rate": 1.6332507792528626e-05,
+      "loss": 0.3688,
+      "step": 5279
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5023828501731868,
+      "learning_rate": 1.633106743577564e-05,
+      "loss": 0.3515,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4313172948281927,
+      "learning_rate": 1.6329626859778057e-05,
+      "loss": 0.3235,
+      "step": 5281
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4878188145202696,
+      "learning_rate": 1.632818606458577e-05,
+      "loss": 0.2731,
+      "step": 5282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39398917324063887,
+      "learning_rate": 1.6326745050248675e-05,
+      "loss": 0.3514,
+      "step": 5283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3186584290148543,
+      "learning_rate": 1.632530381681667e-05,
+      "loss": 0.1892,
+      "step": 5284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2990432265997818,
+      "learning_rate": 1.6323862364339663e-05,
+      "loss": 0.2251,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5905438488593204,
+      "learning_rate": 1.6322420692867577e-05,
+      "loss": 0.3983,
+      "step": 5286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7228437691044046,
+      "learning_rate": 1.632097880245033e-05,
+      "loss": 0.585,
+      "step": 5287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3268212611120316,
+      "learning_rate": 1.6319536693137862e-05,
+      "loss": 0.2669,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40593698432909653,
+      "learning_rate": 1.631809436498011e-05,
+      "loss": 0.3245,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2983396151975038,
+      "learning_rate": 1.6316651818027024e-05,
+      "loss": 0.1933,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47509655229133435,
+      "learning_rate": 1.6315209052328554e-05,
+      "loss": 0.1699,
+      "step": 5291
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46728137226170324,
+      "learning_rate": 1.6313766067934668e-05,
+      "loss": 0.3261,
+      "step": 5292
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3831760523182167,
+      "learning_rate": 1.6312322864895334e-05,
+      "loss": 0.3321,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.70358765871688,
+      "learning_rate": 1.631087944326053e-05,
+      "loss": 0.4898,
+      "step": 5294
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33376431508160115,
+      "learning_rate": 1.6309435803080244e-05,
+      "loss": 0.2051,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2822733219097903,
+      "learning_rate": 1.6307991944404466e-05,
+      "loss": 0.2165,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4344437630599208,
+      "learning_rate": 1.6306547867283197e-05,
+      "loss": 0.3403,
+      "step": 5297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3531370933554759,
+      "learning_rate": 1.630510357176645e-05,
+      "loss": 0.2542,
+      "step": 5298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8157810100988,
+      "learning_rate": 1.6303659057904232e-05,
+      "loss": 0.4894,
+      "step": 5299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33867142000505207,
+      "learning_rate": 1.6302214325746577e-05,
+      "loss": 0.336,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3516250686888519,
+      "learning_rate": 1.6300769375343508e-05,
+      "loss": 0.2335,
+      "step": 5301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4404752724035265,
+      "learning_rate": 1.6299324206745066e-05,
+      "loss": 0.2584,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3540326459594793,
+      "learning_rate": 1.6297878820001302e-05,
+      "loss": 0.2618,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3477074044921296,
+      "learning_rate": 1.6296433215162258e-05,
+      "loss": 0.2255,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.444796893845932,
+      "learning_rate": 1.629498739227801e-05,
+      "loss": 0.3526,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5521271885013249,
+      "learning_rate": 1.6293541351398616e-05,
+      "loss": 0.3678,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43164109413325197,
+      "learning_rate": 1.6292095092574154e-05,
+      "loss": 0.3293,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37843319285317356,
+      "learning_rate": 1.6290648615854712e-05,
+      "loss": 0.249,
+      "step": 5308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25288135816628304,
+      "learning_rate": 1.6289201921290377e-05,
+      "loss": 0.1774,
+      "step": 5309
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.335350277217877,
+      "learning_rate": 1.6287755008931255e-05,
+      "loss": 0.2844,
+      "step": 5310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7224538732310122,
+      "learning_rate": 1.6286307878827443e-05,
+      "loss": 0.4002,
+      "step": 5311
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4249497381815193,
+      "learning_rate": 1.6284860531029062e-05,
+      "loss": 0.342,
+      "step": 5312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37589361249803893,
+      "learning_rate": 1.6283412965586227e-05,
+      "loss": 0.2856,
+      "step": 5313
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32777900712701974,
+      "learning_rate": 1.6281965182549077e-05,
+      "loss": 0.1539,
+      "step": 5314
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.46467288740435403,
+      "learning_rate": 1.6280517181967733e-05,
+      "loss": 0.3337,
+      "step": 5315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.302574483899151,
+      "learning_rate": 1.6279068963892358e-05,
+      "loss": 0.2726,
+      "step": 5316
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5464579445728996,
+      "learning_rate": 1.6277620528373094e-05,
+      "loss": 0.3922,
+      "step": 5317
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6586890435544829,
+      "learning_rate": 1.6276171875460097e-05,
+      "loss": 0.3371,
+      "step": 5318
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4057018760179316,
+      "learning_rate": 1.6274723005203542e-05,
+      "loss": 0.2928,
+      "step": 5319
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5354467630028948,
+      "learning_rate": 1.6273273917653596e-05,
+      "loss": 0.3619,
+      "step": 5320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24525178033362732,
+      "learning_rate": 1.6271824612860445e-05,
+      "loss": 0.1796,
+      "step": 5321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3806740733460614,
+      "learning_rate": 1.6270375090874276e-05,
+      "loss": 0.2955,
+      "step": 5322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.016911246302538,
+      "learning_rate": 1.626892535174529e-05,
+      "loss": 0.4905,
+      "step": 5323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39313652227601653,
+      "learning_rate": 1.6267475395523686e-05,
+      "loss": 0.265,
+      "step": 5324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42444131470135105,
+      "learning_rate": 1.626602522225968e-05,
+      "loss": 0.31,
+      "step": 5325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1439901110836537,
+      "learning_rate": 1.6264574832003492e-05,
+      "loss": 0.704,
+      "step": 5326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3857249796155631,
+      "learning_rate": 1.6263124224805345e-05,
+      "loss": 0.1889,
+      "step": 5327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3412589948287796,
+      "learning_rate": 1.6261673400715475e-05,
+      "loss": 0.2426,
+      "step": 5328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5851866759927743,
+      "learning_rate": 1.6260222359784123e-05,
+      "loss": 0.4064,
+      "step": 5329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3336728934135882,
+      "learning_rate": 1.6258771102061543e-05,
+      "loss": 0.8571,
+      "step": 5330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3092169640199647,
+      "learning_rate": 1.6257319627597986e-05,
+      "loss": 0.2259,
+      "step": 5331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4063809902015279,
+      "learning_rate": 1.6255867936443724e-05,
+      "loss": 0.3177,
+      "step": 5332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7712406027659143,
+      "learning_rate": 1.625441602864902e-05,
+      "loss": 0.485,
+      "step": 5333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3238377010953509,
+      "learning_rate": 1.625296390426416e-05,
+      "loss": 0.2591,
+      "step": 5334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3062860851213545,
+      "learning_rate": 1.6251511563339426e-05,
+      "loss": 0.1903,
+      "step": 5335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42276890338671497,
+      "learning_rate": 1.6250059005925117e-05,
+      "loss": 0.3612,
+      "step": 5336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3259165375372027,
+      "learning_rate": 1.6248606232071536e-05,
+      "loss": 0.2178,
+      "step": 5337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.165783767802092,
+      "learning_rate": 1.6247153241828985e-05,
+      "loss": 0.6111,
+      "step": 5338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.71001471508266,
+      "learning_rate": 1.624570003524779e-05,
+      "loss": 0.3921,
+      "step": 5339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3474997200766379,
+      "learning_rate": 1.624424661237827e-05,
+      "loss": 0.2082,
+      "step": 5340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3418873789831343,
+      "learning_rate": 1.6242792973270758e-05,
+      "loss": 0.2739,
+      "step": 5341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4231357230178703,
+      "learning_rate": 1.6241339117975596e-05,
+      "loss": 0.2998,
+      "step": 5342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4355566229176757,
+      "learning_rate": 1.6239885046543125e-05,
+      "loss": 0.2823,
+      "step": 5343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35899156253158837,
+      "learning_rate": 1.6238430759023706e-05,
+      "loss": 0.27,
+      "step": 5344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8135718113775179,
+      "learning_rate": 1.6236976255467697e-05,
+      "loss": 0.4483,
+      "step": 5345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4186232071036985,
+      "learning_rate": 1.623552153592547e-05,
+      "loss": 0.3089,
+      "step": 5346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26769754516544547,
+      "learning_rate": 1.6234066600447397e-05,
+      "loss": 0.1907,
+      "step": 5347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37312629158031374,
+      "learning_rate": 1.6232611449083866e-05,
+      "loss": 0.3283,
+      "step": 5348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3136542119516781,
+      "learning_rate": 1.623115608188527e-05,
+      "loss": 0.2588,
+      "step": 5349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0894221985880963,
+      "learning_rate": 1.6229700498902008e-05,
+      "loss": 0.3676,
+      "step": 5350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7124718935045236,
+      "learning_rate": 1.6228244700184484e-05,
+      "loss": 0.5309,
+      "step": 5351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3407741653174107,
+      "learning_rate": 1.622678868578311e-05,
+      "loss": 0.2742,
+      "step": 5352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47774823506024416,
+      "learning_rate": 1.622533245574832e-05,
+      "loss": 0.2726,
+      "step": 5353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33453155351886066,
+      "learning_rate": 1.622387601013053e-05,
+      "loss": 0.2116,
+      "step": 5354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3884129554118678,
+      "learning_rate": 1.622241934898018e-05,
+      "loss": 0.2891,
+      "step": 5355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5698439071171101,
+      "learning_rate": 1.622096247234772e-05,
+      "loss": 0.3661,
+      "step": 5356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6227867700244734,
+      "learning_rate": 1.6219505380283593e-05,
+      "loss": 0.4115,
+      "step": 5357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35548790471839764,
+      "learning_rate": 1.6218048072838265e-05,
+      "loss": 0.2962,
+      "step": 5358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3251708651923235,
+      "learning_rate": 1.62165905500622e-05,
+      "loss": 0.7052,
+      "step": 5359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29768315161623826,
+      "learning_rate": 1.621513281200587e-05,
+      "loss": 0.2041,
+      "step": 5360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.432601921603546,
+      "learning_rate": 1.6213674858719758e-05,
+      "loss": 0.3035,
+      "step": 5361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1288404236236713,
+      "learning_rate": 1.6212216690254353e-05,
+      "loss": 0.4222,
+      "step": 5362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8101163644596073,
+      "learning_rate": 1.621075830666015e-05,
+      "loss": 0.3231,
+      "step": 5363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3462029633898307,
+      "learning_rate": 1.6209299707987656e-05,
+      "loss": 0.2848,
+      "step": 5364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5494931342333284,
+      "learning_rate": 1.6207840894287377e-05,
+      "loss": 0.42,
+      "step": 5365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31147969450902874,
+      "learning_rate": 1.6206381865609836e-05,
+      "loss": 0.1297,
+      "step": 5366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42917601362153945,
+      "learning_rate": 1.620492262200556e-05,
+      "loss": 0.3263,
+      "step": 5367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43424402968714376,
+      "learning_rate": 1.620346316352508e-05,
+      "loss": 0.3371,
+      "step": 5368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43098949036670525,
+      "learning_rate": 1.6202003490218932e-05,
+      "loss": 0.2333,
+      "step": 5369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4145635482591804,
+      "learning_rate": 1.6200543602137676e-05,
+      "loss": 0.2221,
+      "step": 5370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.537768224298209,
+      "learning_rate": 1.619908349933186e-05,
+      "loss": 0.851,
+      "step": 5371
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.513208048312179,
+      "learning_rate": 1.619762318185205e-05,
+      "loss": 0.3555,
+      "step": 5372
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32926945386236683,
+      "learning_rate": 1.6196162649748815e-05,
+      "loss": 0.255,
+      "step": 5373
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6119467962093083,
+      "learning_rate": 1.6194701903072734e-05,
+      "loss": 0.4787,
+      "step": 5374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2881545608075555,
+      "learning_rate": 1.619324094187439e-05,
+      "loss": 0.2352,
+      "step": 5375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3695005519980028,
+      "learning_rate": 1.619177976620438e-05,
+      "loss": 0.2304,
+      "step": 5376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6251592955848971,
+      "learning_rate": 1.6190318376113307e-05,
+      "loss": 0.4482,
+      "step": 5377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7395901583447987,
+      "learning_rate": 1.618885677165177e-05,
+      "loss": 0.4733,
+      "step": 5378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39642035979972523,
+      "learning_rate": 1.6187394952870392e-05,
+      "loss": 0.2308,
+      "step": 5379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3336838668804227,
+      "learning_rate": 1.618593291981979e-05,
+      "loss": 0.305,
+      "step": 5380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28950738150393335,
+      "learning_rate": 1.61844706725506e-05,
+      "loss": 0.1669,
+      "step": 5381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5750376618606018,
+      "learning_rate": 1.6183008211113454e-05,
+      "loss": 0.3297,
+      "step": 5382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3405128455992142,
+      "learning_rate": 1.6181545535559e-05,
+      "loss": 0.277,
+      "step": 5383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5570942141284251,
+      "learning_rate": 1.6180082645937888e-05,
+      "loss": 0.3617,
+      "step": 5384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6443057721616354,
+      "learning_rate": 1.6178619542300783e-05,
+      "loss": 0.3859,
+      "step": 5385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3067421355198817,
+      "learning_rate": 1.617715622469835e-05,
+      "loss": 0.2076,
+      "step": 5386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.341538422997227,
+      "learning_rate": 1.617569269318126e-05,
+      "loss": 0.2534,
+      "step": 5387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3996699166427557,
+      "learning_rate": 1.61742289478002e-05,
+      "loss": 0.2901,
+      "step": 5388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35294345317967296,
+      "learning_rate": 1.6172764988605855e-05,
+      "loss": 0.2503,
+      "step": 5389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.679404700705096,
+      "learning_rate": 1.6171300815648922e-05,
+      "loss": 0.4566,
+      "step": 5390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4050109468817518,
+      "learning_rate": 1.6169836428980108e-05,
+      "loss": 0.3135,
+      "step": 5391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33445459512149867,
+      "learning_rate": 1.6168371828650123e-05,
+      "loss": 0.2532,
+      "step": 5392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29912038151863696,
+      "learning_rate": 1.616690701470969e-05,
+      "loss": 0.2024,
+      "step": 5393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4029994690890542,
+      "learning_rate": 1.6165441987209532e-05,
+      "loss": 0.2963,
+      "step": 5394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5210104635097429,
+      "learning_rate": 1.6163976746200384e-05,
+      "loss": 0.3754,
+      "step": 5395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41946498662119747,
+      "learning_rate": 1.6162511291732984e-05,
+      "loss": 0.3246,
+      "step": 5396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8088520261509311,
+      "learning_rate": 1.616104562385808e-05,
+      "loss": 0.4104,
+      "step": 5397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41364569032392223,
+      "learning_rate": 1.615957974262644e-05,
+      "loss": 0.3169,
+      "step": 5398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25882567121199573,
+      "learning_rate": 1.615811364808881e-05,
+      "loss": 0.1889,
+      "step": 5399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5591145105552626,
+      "learning_rate": 1.6156647340295973e-05,
+      "loss": 0.3205,
+      "step": 5400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39999487861467214,
+      "learning_rate": 1.6155180819298703e-05,
+      "loss": 0.3648,
+      "step": 5401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.669508277334371,
+      "learning_rate": 1.6153714085147783e-05,
+      "loss": 0.3816,
+      "step": 5402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.397360661949208,
+      "learning_rate": 1.6152247137894012e-05,
+      "loss": 0.2861,
+      "step": 5403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3426434244744426,
+      "learning_rate": 1.6150779977588186e-05,
+      "loss": 0.2873,
+      "step": 5404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40844847649718674,
+      "learning_rate": 1.6149312604281115e-05,
+      "loss": 0.2428,
+      "step": 5405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.332255987493901,
+      "learning_rate": 1.6147845018023612e-05,
+      "loss": 0.1821,
+      "step": 5406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35299100766743446,
+      "learning_rate": 1.61463772188665e-05,
+      "loss": 0.3115,
+      "step": 5407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.46449691352497247,
+      "learning_rate": 1.6144909206860607e-05,
+      "loss": 0.3955,
+      "step": 5408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3484300824139732,
+      "learning_rate": 1.6143440982056777e-05,
+      "loss": 0.2251,
+      "step": 5409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6109240447729549,
+      "learning_rate": 1.614197254450585e-05,
+      "loss": 0.378,
+      "step": 5410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29480715347424497,
+      "learning_rate": 1.6140503894258674e-05,
+      "loss": 0.2724,
+      "step": 5411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3103918092644329,
+      "learning_rate": 1.6139035031366116e-05,
+      "loss": 0.169,
+      "step": 5412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3729779854233855,
+      "learning_rate": 1.6137565955879036e-05,
+      "loss": 0.3555,
+      "step": 5413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7389313540620621,
+      "learning_rate": 1.6136096667848313e-05,
+      "loss": 0.5823,
+      "step": 5414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3544691682936791,
+      "learning_rate": 1.6134627167324827e-05,
+      "loss": 0.2543,
+      "step": 5415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3960851188225142,
+      "learning_rate": 1.613315745435946e-05,
+      "loss": 0.2833,
+      "step": 5416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2765012263636344,
+      "learning_rate": 1.613168752900312e-05,
+      "loss": 0.1917,
+      "step": 5417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6375209628011956,
+      "learning_rate": 1.61302173913067e-05,
+      "loss": 0.4634,
+      "step": 5418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2715428312700595,
+      "learning_rate": 1.612874704132112e-05,
+      "loss": 0.2351,
+      "step": 5419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7184165179152175,
+      "learning_rate": 1.612727647909729e-05,
+      "loss": 0.5226,
+      "step": 5420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7367803499422515,
+      "learning_rate": 1.612580570468614e-05,
+      "loss": 0.481,
+      "step": 5421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40185485568900076,
+      "learning_rate": 1.6124334718138602e-05,
+      "loss": 0.2303,
+      "step": 5422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4198601147325941,
+      "learning_rate": 1.6122863519505618e-05,
+      "loss": 0.3464,
+      "step": 5423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4956845797294979,
+      "learning_rate": 1.6121392108838132e-05,
+      "loss": 0.3011,
+      "step": 5424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26916066161630525,
+      "learning_rate": 1.6119920486187102e-05,
+      "loss": 0.2054,
+      "step": 5425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1799061771549846,
+      "learning_rate": 1.611844865160349e-05,
+      "loss": 0.9037,
+      "step": 5426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34547055929760423,
+      "learning_rate": 1.611697660513826e-05,
+      "loss": 0.3116,
+      "step": 5427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3723557442890717,
+      "learning_rate": 1.6115504346842393e-05,
+      "loss": 0.215,
+      "step": 5428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7354679497674211,
+      "learning_rate": 1.6114031876766877e-05,
+      "loss": 0.4834,
+      "step": 5429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5397253363814973,
+      "learning_rate": 1.61125591949627e-05,
+      "loss": 0.385,
+      "step": 5430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31820906837570323,
+      "learning_rate": 1.611108630148086e-05,
+      "loss": 0.2479,
+      "step": 5431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43775281349809764,
+      "learning_rate": 1.610961319637236e-05,
+      "loss": 0.3013,
+      "step": 5432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48523621853194515,
+      "learning_rate": 1.610813987968822e-05,
+      "loss": 0.3288,
+      "step": 5433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3286365653300218,
+      "learning_rate": 1.6106666351479462e-05,
+      "loss": 0.2691,
+      "step": 5434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4073669030943272,
+      "learning_rate": 1.610519261179711e-05,
+      "loss": 0.2907,
+      "step": 5435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.86365452708418,
+      "learning_rate": 1.61037186606922e-05,
+      "loss": 0.5174,
+      "step": 5436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34139315066637166,
+      "learning_rate": 1.610224449821577e-05,
+      "loss": 0.2648,
+      "step": 5437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.781900670758003,
+      "learning_rate": 1.6100770124418882e-05,
+      "loss": 0.5627,
+      "step": 5438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2584851055369354,
+      "learning_rate": 1.6099295539352583e-05,
+      "loss": 0.2251,
+      "step": 5439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32668801819835613,
+      "learning_rate": 1.6097820743067945e-05,
+      "loss": 0.2867,
+      "step": 5440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8179062477321604,
+      "learning_rate": 1.6096345735616036e-05,
+      "loss": 0.3766,
+      "step": 5441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5923473411548568,
+      "learning_rate": 1.6094870517047937e-05,
+      "loss": 0.4198,
+      "step": 5442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31872599364875503,
+      "learning_rate": 1.609339508741473e-05,
+      "loss": 0.2893,
+      "step": 5443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.379638334873006,
+      "learning_rate": 1.6091919446767517e-05,
+      "loss": 0.8334,
+      "step": 5444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2200588702556372,
+      "learning_rate": 1.6090443595157396e-05,
+      "loss": 0.0865,
+      "step": 5445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3663693385434264,
+      "learning_rate": 1.608896753263547e-05,
+      "loss": 0.3005,
+      "step": 5446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4523250755868021,
+      "learning_rate": 1.6087491259252865e-05,
+      "loss": 0.3534,
+      "step": 5447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6353127882984126,
+      "learning_rate": 1.60860147750607e-05,
+      "loss": 0.316,
+      "step": 5448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4157340148684122,
+      "learning_rate": 1.60845380801101e-05,
+      "loss": 0.316,
+      "step": 5449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2008936520366085,
+      "learning_rate": 1.6083061174452214e-05,
+      "loss": 0.8211,
+      "step": 5450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22059778641398187,
+      "learning_rate": 1.6081584058138178e-05,
+      "loss": 0.1828,
+      "step": 5451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38232047263168106,
+      "learning_rate": 1.6080106731219147e-05,
+      "loss": 0.2847,
+      "step": 5452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7296256845720059,
+      "learning_rate": 1.6078629193746283e-05,
+      "loss": 0.4682,
+      "step": 5453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49677717222315976,
+      "learning_rate": 1.607715144577075e-05,
+      "loss": 0.2671,
+      "step": 5454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3413141139616929,
+      "learning_rate": 1.6075673487343725e-05,
+      "loss": 0.2721,
+      "step": 5455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2043097896012789,
+      "learning_rate": 1.6074195318516385e-05,
+      "loss": 0.8261,
+      "step": 5456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3009263002823461,
+      "learning_rate": 1.6072716939339924e-05,
+      "loss": 0.227,
+      "step": 5457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40159765325055435,
+      "learning_rate": 1.607123834986554e-05,
+      "loss": 0.2515,
+      "step": 5458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4338965338706003,
+      "learning_rate": 1.606975955014443e-05,
+      "loss": 0.3357,
+      "step": 5459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1490956748629666,
+      "learning_rate": 1.6068280540227807e-05,
+      "loss": 0.6926,
+      "step": 5460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.36110726958810213,
+      "learning_rate": 1.6066801320166897e-05,
+      "loss": 0.2308,
+      "step": 5461
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.998317680190995,
+      "learning_rate": 1.606532189001291e-05,
+      "loss": 0.6979,
+      "step": 5462
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32793332659947005,
+      "learning_rate": 1.606384224981709e-05,
+      "loss": 0.3374,
+      "step": 5463
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3159672370415879,
+      "learning_rate": 1.6062362399630673e-05,
+      "loss": 0.2052,
+      "step": 5464
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28198983815444506,
+      "learning_rate": 1.6060882339504905e-05,
+      "loss": 0.2491,
+      "step": 5465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3914189720919406,
+      "learning_rate": 1.6059402069491047e-05,
+      "loss": 0.3358,
+      "step": 5466
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31173129053212056,
+      "learning_rate": 1.605792158964035e-05,
+      "loss": 0.2001,
+      "step": 5467
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0947564593865875,
+      "learning_rate": 1.6056440900004094e-05,
+      "loss": 0.7421,
+      "step": 5468
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5952027400139973,
+      "learning_rate": 1.6054960000633545e-05,
+      "loss": 0.4776,
+      "step": 5469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.420934632151069,
+      "learning_rate": 1.6053478891579993e-05,
+      "loss": 0.3136,
+      "step": 5470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24668825200323996,
+      "learning_rate": 1.605199757289473e-05,
+      "loss": 0.1935,
+      "step": 5471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1178765441595595,
+      "learning_rate": 1.605051604462905e-05,
+      "loss": 0.646,
+      "step": 5472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38129148872885377,
+      "learning_rate": 1.6049034306834258e-05,
+      "loss": 0.275,
+      "step": 5473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4828124601149025,
+      "learning_rate": 1.6047552359561672e-05,
+      "loss": 0.3161,
+      "step": 5474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6728617355740534,
+      "learning_rate": 1.6046070202862605e-05,
+      "loss": 0.378,
+      "step": 5475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37208712058090737,
+      "learning_rate": 1.6044587836788385e-05,
+      "loss": 0.2971,
+      "step": 5476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2293316968318131,
+      "learning_rate": 1.6043105261390352e-05,
+      "loss": 0.0965,
+      "step": 5477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4546947048442612,
+      "learning_rate": 1.6041622476719842e-05,
+      "loss": 0.3595,
+      "step": 5478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3668420835351272,
+      "learning_rate": 1.6040139482828207e-05,
+      "loss": 0.2734,
+      "step": 5479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.312312858344449,
+      "learning_rate": 1.60386562797668e-05,
+      "loss": 0.7963,
+      "step": 5480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6925249119934285,
+      "learning_rate": 1.6037172867586984e-05,
+      "loss": 0.3647,
+      "step": 5481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4077235005317406,
+      "learning_rate": 1.6035689246340132e-05,
+      "loss": 0.2688,
+      "step": 5482
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34113054517286356,
+      "learning_rate": 1.603420541607762e-05,
+      "loss": 0.291,
+      "step": 5483
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23609225173467213,
+      "learning_rate": 1.6032721376850837e-05,
+      "loss": 0.1449,
+      "step": 5484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4239941906638344,
+      "learning_rate": 1.603123712871117e-05,
+      "loss": 0.2859,
+      "step": 5485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42335183714015223,
+      "learning_rate": 1.602975267171002e-05,
+      "loss": 0.3501,
+      "step": 5486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5602162541545916,
+      "learning_rate": 1.6028268005898798e-05,
+      "loss": 0.3304,
+      "step": 5487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4283245085092912,
+      "learning_rate": 1.6026783131328915e-05,
+      "loss": 0.292,
+      "step": 5488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27342802197048033,
+      "learning_rate": 1.6025298048051784e-05,
+      "loss": 0.229,
+      "step": 5489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34821181236786064,
+      "learning_rate": 1.6023812756118845e-05,
+      "loss": 0.2565,
+      "step": 5490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4254165247173863,
+      "learning_rate": 1.602232725558153e-05,
+      "loss": 0.3036,
+      "step": 5491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4819874115221115,
+      "learning_rate": 1.6020841546491278e-05,
+      "loss": 0.3834,
+      "step": 5492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8487213879795744,
+      "learning_rate": 1.6019355628899545e-05,
+      "loss": 0.5686,
+      "step": 5493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33225687121251357,
+      "learning_rate": 1.6017869502857785e-05,
+      "loss": 0.2232,
+      "step": 5494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3973590452255858,
+      "learning_rate": 1.6016383168417463e-05,
+      "loss": 0.3365,
+      "step": 5495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29339420160733515,
+      "learning_rate": 1.601489662563005e-05,
+      "loss": 0.1986,
+      "step": 5496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3666754590084301,
+      "learning_rate": 1.6013409874547026e-05,
+      "loss": 0.2422,
+      "step": 5497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8964912741710412,
+      "learning_rate": 1.6011922915219877e-05,
+      "loss": 0.4598,
+      "step": 5498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38686744503564313,
+      "learning_rate": 1.6010435747700097e-05,
+      "loss": 0.3271,
+      "step": 5499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36402267171787295,
+      "learning_rate": 1.600894837203918e-05,
+      "loss": 0.1334,
+      "step": 5500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.39810639722148555,
+      "learning_rate": 1.6007460788288644e-05,
+      "loss": 0.3626,
+      "step": 5501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26937219839700005,
+      "learning_rate": 1.60059729965e-05,
+      "loss": 0.237,
+      "step": 5502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3572405700250666,
+      "learning_rate": 1.600448499672477e-05,
+      "loss": 0.2023,
+      "step": 5503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7519951161732132,
+      "learning_rate": 1.600299678901448e-05,
+      "loss": 0.4123,
+      "step": 5504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8481384242186841,
+      "learning_rate": 1.6001508373420666e-05,
+      "loss": 0.609,
+      "step": 5505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38602966781075637,
+      "learning_rate": 1.6000019749994882e-05,
+      "loss": 0.3056,
+      "step": 5506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3067806475149186,
+      "learning_rate": 1.5998530918788667e-05,
+      "loss": 0.254,
+      "step": 5507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28563113629324893,
+      "learning_rate": 1.5997041879853585e-05,
+      "loss": 0.1853,
+      "step": 5508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.560627051750689,
+      "learning_rate": 1.5995552633241206e-05,
+      "loss": 0.3737,
+      "step": 5509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31902956156250195,
+      "learning_rate": 1.599406317900309e-05,
+      "loss": 0.2693,
+      "step": 5510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0324425132045223,
+      "learning_rate": 1.5992573517190826e-05,
+      "loss": 0.7321,
+      "step": 5511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5816127897195071,
+      "learning_rate": 1.5991083647856003e-05,
+      "loss": 0.3399,
+      "step": 5512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3638464303346846,
+      "learning_rate": 1.5989593571050207e-05,
+      "loss": 0.2275,
+      "step": 5513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2997715451537564,
+      "learning_rate": 1.5988103286825043e-05,
+      "loss": 0.2977,
+      "step": 5514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3882699857377122,
+      "learning_rate": 1.5986612795232122e-05,
+      "loss": 0.3314,
+      "step": 5515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4607329976592203,
+      "learning_rate": 1.5985122096323053e-05,
+      "loss": 0.3253,
+      "step": 5516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42979471017625737,
+      "learning_rate": 1.5983631190149466e-05,
+      "loss": 0.3066,
+      "step": 5517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3052137957577575,
+      "learning_rate": 1.5982140076762988e-05,
+      "loss": 0.2657,
+      "step": 5518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4134905162494446,
+      "learning_rate": 1.5980648756215256e-05,
+      "loss": 0.3369,
+      "step": 5519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5637451037693646,
+      "learning_rate": 1.597915722855792e-05,
+      "loss": 0.3639,
+      "step": 5520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37421259673932755,
+      "learning_rate": 1.597766549384262e-05,
+      "loss": 0.2469,
+      "step": 5521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3058915736850958,
+      "learning_rate": 1.5976173552121023e-05,
+      "loss": 0.2922,
+      "step": 5522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2841295684463912,
+      "learning_rate": 1.597468140344479e-05,
+      "loss": 0.1977,
+      "step": 5523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6162024062628615,
+      "learning_rate": 1.59731890478656e-05,
+      "loss": 0.2815,
+      "step": 5524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4094893668831604,
+      "learning_rate": 1.5971696485435128e-05,
+      "loss": 0.3378,
+      "step": 5525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40588421773351036,
+      "learning_rate": 1.5970203716205066e-05,
+      "loss": 0.2895,
+      "step": 5526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6938594055918771,
+      "learning_rate": 1.5968710740227106e-05,
+      "loss": 0.3855,
+      "step": 5527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4116993177994943,
+      "learning_rate": 1.5967217557552944e-05,
+      "loss": 0.329,
+      "step": 5528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22779602155734383,
+      "learning_rate": 1.5965724168234304e-05,
+      "loss": 0.086,
+      "step": 5529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3500048899822258,
+      "learning_rate": 1.5964230572322884e-05,
+      "loss": 0.2697,
+      "step": 5530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4157625173818804,
+      "learning_rate": 1.596273676987042e-05,
+      "loss": 0.3202,
+      "step": 5531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8107650750102577,
+      "learning_rate": 1.596124276092864e-05,
+      "loss": 0.4456,
+      "step": 5532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35678553696656035,
+      "learning_rate": 1.595974854554928e-05,
+      "loss": 0.2202,
+      "step": 5533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40454325538013913,
+      "learning_rate": 1.5958254123784077e-05,
+      "loss": 0.3206,
+      "step": 5534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2564749702657718,
+      "learning_rate": 1.5956759495684796e-05,
+      "loss": 0.2145,
+      "step": 5535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37351136197316315,
+      "learning_rate": 1.595526466130319e-05,
+      "loss": 0.1787,
+      "step": 5536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5263171270492955,
+      "learning_rate": 1.595376962069102e-05,
+      "loss": 0.3519,
+      "step": 5537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42044904764267466,
+      "learning_rate": 1.5952274373900067e-05,
+      "loss": 0.3532,
+      "step": 5538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45297528673077875,
+      "learning_rate": 1.5950778920982108e-05,
+      "loss": 0.2163,
+      "step": 5539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3791943745784225,
+      "learning_rate": 1.5949283261988934e-05,
+      "loss": 0.3221,
+      "step": 5540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2635293913337198,
+      "learning_rate": 1.5947787396972332e-05,
+      "loss": 0.2126,
+      "step": 5541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.39755428740194615,
+      "learning_rate": 1.5946291325984108e-05,
+      "loss": 0.2607,
+      "step": 5542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3970521445893526,
+      "learning_rate": 1.5944795049076072e-05,
+      "loss": 0.3215,
+      "step": 5543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6915226148979662,
+      "learning_rate": 1.594329856630004e-05,
+      "loss": 0.5286,
+      "step": 5544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0272727002837614,
+      "learning_rate": 1.5941801877707835e-05,
+      "loss": 0.5231,
+      "step": 5545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2711694648544462,
+      "learning_rate": 1.594030498335129e-05,
+      "loss": 0.2551,
+      "step": 5546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0522809594213387,
+      "learning_rate": 1.5938807883282234e-05,
+      "loss": 0.6938,
+      "step": 5547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2780677764777073,
+      "learning_rate": 1.5937310577552517e-05,
+      "loss": 0.2152,
+      "step": 5548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3764746606296057,
+      "learning_rate": 1.5935813066213988e-05,
+      "loss": 0.254,
+      "step": 5549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.39611418844585977,
+      "learning_rate": 1.593431534931851e-05,
+      "loss": 0.345,
+      "step": 5550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6565438887638069,
+      "learning_rate": 1.5932817426917945e-05,
+      "loss": 0.4974,
+      "step": 5551
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3633914159676232,
+      "learning_rate": 1.593131929906417e-05,
+      "loss": 0.2314,
+      "step": 5552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31526102547006846,
+      "learning_rate": 1.5929820965809063e-05,
+      "loss": 0.2752,
+      "step": 5553
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3736946671960914,
+      "learning_rate": 1.592832242720451e-05,
+      "loss": 0.3219,
+      "step": 5554
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3615427063133629,
+      "learning_rate": 1.5926823683302404e-05,
+      "loss": 0.2956,
+      "step": 5555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40873885213362576,
+      "learning_rate": 1.5925324734154654e-05,
+      "loss": 0.2416,
+      "step": 5556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5019426658331753,
+      "learning_rate": 1.5923825579813158e-05,
+      "loss": 0.4355,
+      "step": 5557
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3081784728369049,
+      "learning_rate": 1.592232622032984e-05,
+      "loss": 0.2679,
+      "step": 5558
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.581079170192961,
+      "learning_rate": 1.5920826655756617e-05,
+      "loss": 0.4,
+      "step": 5559
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5650872586904916,
+      "learning_rate": 1.591932688614542e-05,
+      "loss": 0.4542,
+      "step": 5560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27258563527685675,
+      "learning_rate": 1.5917826911548194e-05,
+      "loss": 0.2233,
+      "step": 5561
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2959203113731466,
+      "learning_rate": 1.591632673201687e-05,
+      "loss": 0.2158,
+      "step": 5562
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2341214992158127,
+      "learning_rate": 1.5914826347603412e-05,
+      "loss": 0.7517,
+      "step": 5563
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33028417983241837,
+      "learning_rate": 1.591332575835977e-05,
+      "loss": 0.2764,
+      "step": 5564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46893872523644764,
+      "learning_rate": 1.591182496433791e-05,
+      "loss": 0.3126,
+      "step": 5565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5536360678259141,
+      "learning_rate": 1.5910323965589803e-05,
+      "loss": 0.3726,
+      "step": 5566
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3735817111568126,
+      "learning_rate": 1.5908822762167435e-05,
+      "loss": 0.3044,
+      "step": 5567
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35011633102273504,
+      "learning_rate": 1.5907321354122788e-05,
+      "loss": 0.2051,
+      "step": 5568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3548591682635962,
+      "learning_rate": 1.5905819741507856e-05,
+      "loss": 0.2592,
+      "step": 5569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3581690176525001,
+      "learning_rate": 1.590431792437464e-05,
+      "loss": 0.3012,
+      "step": 5570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7579819660734486,
+      "learning_rate": 1.590281590277515e-05,
+      "loss": 0.5101,
+      "step": 5571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4774255317314944,
+      "learning_rate": 1.5901313676761397e-05,
+      "loss": 0.3312,
+      "step": 5572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3935810128176364,
+      "learning_rate": 1.5899811246385404e-05,
+      "loss": 0.2783,
+      "step": 5573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2434832696741011,
+      "learning_rate": 1.5898308611699204e-05,
+      "loss": 0.2066,
+      "step": 5574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.838098712521921,
+      "learning_rate": 1.589680577275483e-05,
+      "loss": 0.4197,
+      "step": 5575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38018436720857546,
+      "learning_rate": 1.5895302729604323e-05,
+      "loss": 0.274,
+      "step": 5576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5879533204990789,
+      "learning_rate": 1.589379948229974e-05,
+      "loss": 0.4019,
+      "step": 5577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.469626009070046,
+      "learning_rate": 1.5892296030893134e-05,
+      "loss": 0.2904,
+      "step": 5578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3569116943337263,
+      "learning_rate": 1.5890792375436568e-05,
+      "loss": 0.2826,
+      "step": 5579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30781010648834556,
+      "learning_rate": 1.5889288515982118e-05,
+      "loss": 0.214,
+      "step": 5580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4452871745168325,
+      "learning_rate": 1.5887784452581857e-05,
+      "loss": 0.3477,
+      "step": 5581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3423890676840111,
+      "learning_rate": 1.5886280185287874e-05,
+      "loss": 0.2258,
+      "step": 5582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9397225493977781,
+      "learning_rate": 1.588477571415226e-05,
+      "loss": 0.4489,
+      "step": 5583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8992191750862868,
+      "learning_rate": 1.588327103922712e-05,
+      "loss": 0.5669,
+      "step": 5584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34519578977057513,
+      "learning_rate": 1.5881766160564553e-05,
+      "loss": 0.2008,
+      "step": 5585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2613946128658462,
+      "learning_rate": 1.588026107821668e-05,
+      "loss": 0.2106,
+      "step": 5586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1043781598317808,
+      "learning_rate": 1.5878755792235616e-05,
+      "loss": 0.5675,
+      "step": 5587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3429172657775994,
+      "learning_rate": 1.5877250302673493e-05,
+      "loss": 0.2068,
+      "step": 5588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6522423001767897,
+      "learning_rate": 1.5875744609582444e-05,
+      "loss": 0.3912,
+      "step": 5589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4731252524113648,
+      "learning_rate": 1.587423871301461e-05,
+      "loss": 0.3802,
+      "step": 5590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33128379858329554,
+      "learning_rate": 1.5872732613022147e-05,
+      "loss": 0.2088,
+      "step": 5591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2749688226198136,
+      "learning_rate": 1.5871226309657203e-05,
+      "loss": 0.1898,
+      "step": 5592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.449249676522092,
+      "learning_rate": 1.5869719802971947e-05,
+      "loss": 0.3547,
+      "step": 5593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.545477058578212,
+      "learning_rate": 1.5868213093018543e-05,
+      "loss": 0.3069,
+      "step": 5594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7375781226958218,
+      "learning_rate": 1.5866706179849172e-05,
+      "loss": 0.3483,
+      "step": 5595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0617075868345798,
+      "learning_rate": 1.586519906351602e-05,
+      "loss": 0.6041,
+      "step": 5596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3294184144488115,
+      "learning_rate": 1.586369174407128e-05,
+      "loss": 0.2779,
+      "step": 5597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2594062832915638,
+      "learning_rate": 1.5862184221567144e-05,
+      "loss": 0.1825,
+      "step": 5598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9733876205705159,
+      "learning_rate": 1.586067649605582e-05,
+      "loss": 0.5383,
+      "step": 5599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4270131446780353,
+      "learning_rate": 1.585916856758952e-05,
+      "loss": 0.2909,
+      "step": 5600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5168643875206522,
+      "learning_rate": 1.5857660436220466e-05,
+      "loss": 0.315,
+      "step": 5601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5136780415745466,
+      "learning_rate": 1.5856152102000878e-05,
+      "loss": 0.3597,
+      "step": 5602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3472687358366042,
+      "learning_rate": 1.5854643564983e-05,
+      "loss": 0.2772,
+      "step": 5603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.21984666591055935,
+      "learning_rate": 1.5853134825219066e-05,
+      "loss": 0.1203,
+      "step": 5604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5955025245870108,
+      "learning_rate": 1.5851625882761326e-05,
+      "loss": 0.3752,
+      "step": 5605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3950271910878111,
+      "learning_rate": 1.585011673766203e-05,
+      "loss": 0.2838,
+      "step": 5606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6397240293769488,
+      "learning_rate": 1.5848607389973446e-05,
+      "loss": 0.4141,
+      "step": 5607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4432763685637882,
+      "learning_rate": 1.584709783974784e-05,
+      "loss": 0.2897,
+      "step": 5608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4240833841777455,
+      "learning_rate": 1.5845588087037484e-05,
+      "loss": 0.2904,
+      "step": 5609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43798524592763793,
+      "learning_rate": 1.5844078131894668e-05,
+      "loss": 0.3254,
+      "step": 5610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44403383059135365,
+      "learning_rate": 1.5842567974371672e-05,
+      "loss": 0.2548,
+      "step": 5611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37943493120580635,
+      "learning_rate": 1.5841057614520803e-05,
+      "loss": 0.3146,
+      "step": 5612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3330746925906659,
+      "learning_rate": 1.5839547052394356e-05,
+      "loss": 0.3125,
+      "step": 5613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36957538435382176,
+      "learning_rate": 1.583803628804465e-05,
+      "loss": 0.1807,
+      "step": 5614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3799639377524021,
+      "learning_rate": 1.5836525321523998e-05,
+      "loss": 0.2976,
+      "step": 5615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5407995328754999,
+      "learning_rate": 1.5835014152884722e-05,
+      "loss": 0.371,
+      "step": 5616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4602781375274827,
+      "learning_rate": 1.5833502782179157e-05,
+      "loss": 0.3056,
+      "step": 5617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2703581818132286,
+      "learning_rate": 1.5831991209459646e-05,
+      "loss": 0.2294,
+      "step": 5618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3941351735792076,
+      "learning_rate": 1.583047943477853e-05,
+      "loss": 0.8033,
+      "step": 5619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4269407961412312,
+      "learning_rate": 1.5828967458188155e-05,
+      "loss": 0.2612,
+      "step": 5620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32749877211677897,
+      "learning_rate": 1.5827455279740892e-05,
+      "loss": 0.2406,
+      "step": 5621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7597682432080118,
+      "learning_rate": 1.5825942899489103e-05,
+      "loss": 0.4945,
+      "step": 5622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6648356899162475,
+      "learning_rate": 1.5824430317485163e-05,
+      "loss": 0.4611,
+      "step": 5623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3576858149221899,
+      "learning_rate": 1.582291753378145e-05,
+      "loss": 0.1644,
+      "step": 5624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30960874021564383,
+      "learning_rate": 1.5821404548430352e-05,
+      "loss": 0.2993,
+      "step": 5625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2692899639231665,
+      "learning_rate": 1.5819891361484266e-05,
+      "loss": 0.2074,
+      "step": 5626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.362044309838756,
+      "learning_rate": 1.5818377972995594e-05,
+      "loss": 0.2007,
+      "step": 5627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5932912412651169,
+      "learning_rate": 1.581686438301674e-05,
+      "loss": 0.3933,
+      "step": 5628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37069755753292,
+      "learning_rate": 1.5815350591600124e-05,
+      "loss": 0.3382,
+      "step": 5629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30830834981343785,
+      "learning_rate": 1.5813836598798168e-05,
+      "loss": 0.1911,
+      "step": 5630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36059298843687776,
+      "learning_rate": 1.5812322404663304e-05,
+      "loss": 0.2967,
+      "step": 5631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.385845649979882,
+      "learning_rate": 1.581080800924796e-05,
+      "loss": 0.1999,
+      "step": 5632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33126056841829,
+      "learning_rate": 1.5809293412604584e-05,
+      "loss": 0.299,
+      "step": 5633
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.386058427132325,
+      "learning_rate": 1.580777861478563e-05,
+      "loss": 0.3071,
+      "step": 5634
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7101238664226756,
+      "learning_rate": 1.580626361584355e-05,
+      "loss": 0.5124,
+      "step": 5635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.589818500555523,
+      "learning_rate": 1.5804748415830814e-05,
+      "loss": 0.4467,
+      "step": 5636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2932950338270886,
+      "learning_rate": 1.5803233014799887e-05,
+      "loss": 0.2563,
+      "step": 5637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2636118456350962,
+      "learning_rate": 1.5801717412803246e-05,
+      "loss": 0.2042,
+      "step": 5638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3969761989987564,
+      "learning_rate": 1.580020160989339e-05,
+      "loss": 0.3166,
+      "step": 5639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5113430169982263,
+      "learning_rate": 1.5798685606122795e-05,
+      "loss": 0.3068,
+      "step": 5640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40055815261009947,
+      "learning_rate": 1.579716940154397e-05,
+      "loss": 0.3314,
+      "step": 5641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7927953903429585,
+      "learning_rate": 1.5795652996209416e-05,
+      "loss": 0.5219,
+      "step": 5642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37086620699465855,
+      "learning_rate": 1.579413639017165e-05,
+      "loss": 0.3161,
+      "step": 5643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24236166483538968,
+      "learning_rate": 1.5792619583483183e-05,
+      "loss": 0.1854,
+      "step": 5644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6014945831118405,
+      "learning_rate": 1.5791102576196555e-05,
+      "loss": 0.4099,
+      "step": 5645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3698987883509682,
+      "learning_rate": 1.5789585368364296e-05,
+      "loss": 0.3163,
+      "step": 5646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7593189596779604,
+      "learning_rate": 1.5788067960038942e-05,
+      "loss": 0.3718,
+      "step": 5647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5227360365319026,
+      "learning_rate": 1.5786550351273043e-05,
+      "loss": 0.3476,
+      "step": 5648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34840131645436817,
+      "learning_rate": 1.5785032542119155e-05,
+      "loss": 0.3044,
+      "step": 5649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3537908681190076,
+      "learning_rate": 1.578351453262984e-05,
+      "loss": 0.1915,
+      "step": 5650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5534665540651362,
+      "learning_rate": 1.578199632285766e-05,
+      "loss": 0.3911,
+      "step": 5651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3296603796034221,
+      "learning_rate": 1.57804779128552e-05,
+      "loss": 0.2536,
+      "step": 5652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36719045075931117,
+      "learning_rate": 1.577895930267504e-05,
+      "loss": 0.255,
+      "step": 5653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5549103218709931,
+      "learning_rate": 1.5777440492369764e-05,
+      "loss": 0.4179,
+      "step": 5654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4099773567180406,
+      "learning_rate": 1.5775921481991976e-05,
+      "loss": 0.3489,
+      "step": 5655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9487980274852443,
+      "learning_rate": 1.5774402271594272e-05,
+      "loss": 0.4496,
+      "step": 5656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30146272872090246,
+      "learning_rate": 1.577288286122927e-05,
+      "loss": 0.2396,
+      "step": 5657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32480569594008457,
+      "learning_rate": 1.5771363250949582e-05,
+      "loss": 0.246,
+      "step": 5658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3711344965945374,
+      "learning_rate": 1.5769843440807828e-05,
+      "loss": 0.2793,
+      "step": 5659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3992119007745314,
+      "learning_rate": 1.5768323430856647e-05,
+      "loss": 0.2608,
+      "step": 5660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.335697923487834,
+      "learning_rate": 1.5766803221148676e-05,
+      "loss": 0.2743,
+      "step": 5661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1159080031165294,
+      "learning_rate": 1.576528281173655e-05,
+      "loss": 0.4803,
+      "step": 5662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24838037661375026,
+      "learning_rate": 1.5763762202672933e-05,
+      "loss": 0.1415,
+      "step": 5663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30420554106490244,
+      "learning_rate": 1.576224139401048e-05,
+      "loss": 0.2233,
+      "step": 5664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.422273654707413,
+      "learning_rate": 1.5760720385801855e-05,
+      "loss": 0.3285,
+      "step": 5665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6169292780868955,
+      "learning_rate": 1.575919917809973e-05,
+      "loss": 0.2722,
+      "step": 5666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4049489150331608,
+      "learning_rate": 1.5757677770956785e-05,
+      "loss": 0.2992,
+      "step": 5667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8471162417110218,
+      "learning_rate": 1.5756156164425703e-05,
+      "loss": 0.5039,
+      "step": 5668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4114900323796081,
+      "learning_rate": 1.5754634358559187e-05,
+      "loss": 0.2948,
+      "step": 5669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24786075406450575,
+      "learning_rate": 1.5753112353409928e-05,
+      "loss": 0.1681,
+      "step": 5670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1452188003836672,
+      "learning_rate": 1.5751590149030632e-05,
+      "loss": 0.7431,
+      "step": 5671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4562773003590891,
+      "learning_rate": 1.575006774547402e-05,
+      "loss": 0.3109,
+      "step": 5672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.434282672437878,
+      "learning_rate": 1.5748545142792807e-05,
+      "loss": 0.2306,
+      "step": 5673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0737371125385688,
+      "learning_rate": 1.5747022341039727e-05,
+      "loss": 0.4466,
+      "step": 5674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.719146996338594,
+      "learning_rate": 1.5745499340267508e-05,
+      "loss": 0.4678,
+      "step": 5675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.314472948884982,
+      "learning_rate": 1.5743976140528893e-05,
+      "loss": 0.1734,
+      "step": 5676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40527113405825066,
+      "learning_rate": 1.5742452741876632e-05,
+      "loss": 0.2722,
+      "step": 5677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7245521666416738,
+      "learning_rate": 1.574092914436348e-05,
+      "loss": 0.4212,
+      "step": 5678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4596505412358331,
+      "learning_rate": 1.5739405348042197e-05,
+      "loss": 0.2702,
+      "step": 5679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42284574176881146,
+      "learning_rate": 1.5737881352965556e-05,
+      "loss": 0.34,
+      "step": 5680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7840092327565245,
+      "learning_rate": 1.573635715918633e-05,
+      "loss": 0.7355,
+      "step": 5681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3697786735178063,
+      "learning_rate": 1.5734832766757302e-05,
+      "loss": 0.2946,
+      "step": 5682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26905521183914954,
+      "learning_rate": 1.573330817573126e-05,
+      "loss": 0.1296,
+      "step": 5683
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3968443457830917,
+      "learning_rate": 1.5731783386161007e-05,
+      "loss": 0.3475,
+      "step": 5684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3421390360738499,
+      "learning_rate": 1.5730258398099335e-05,
+      "loss": 0.2744,
+      "step": 5685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7376461566219767,
+      "learning_rate": 1.5728733211599067e-05,
+      "loss": 0.3775,
+      "step": 5686
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2203592124029954,
+      "learning_rate": 1.572720782671301e-05,
+      "loss": 0.6442,
+      "step": 5687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3257697053227165,
+      "learning_rate": 1.5725682243493995e-05,
+      "loss": 0.2935,
+      "step": 5688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24273051299883028,
+      "learning_rate": 1.572415646199485e-05,
+      "loss": 0.1582,
+      "step": 5689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7445388041307233,
+      "learning_rate": 1.5722630482268413e-05,
+      "loss": 0.4049,
+      "step": 5690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3916307018462784,
+      "learning_rate": 1.5721104304367526e-05,
+      "loss": 0.318,
+      "step": 5691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3538495633272354,
+      "learning_rate": 1.5719577928345045e-05,
+      "loss": 0.2772,
+      "step": 5692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7364894330489864,
+      "learning_rate": 1.5718051354253828e-05,
+      "loss": 0.4585,
+      "step": 5693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4082152277198964,
+      "learning_rate": 1.5716524582146734e-05,
+      "loss": 0.304,
+      "step": 5694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31484704632365135,
+      "learning_rate": 1.5714997612076643e-05,
+      "loss": 0.232,
+      "step": 5695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39016338699647013,
+      "learning_rate": 1.571347044409643e-05,
+      "loss": 0.3046,
+      "step": 5696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3233481171244161,
+      "learning_rate": 1.571194307825898e-05,
+      "loss": 0.2492,
+      "step": 5697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.092219467523616,
+      "learning_rate": 1.5710415514617187e-05,
+      "loss": 0.7269,
+      "step": 5698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0398976979337753,
+      "learning_rate": 1.5708887753223953e-05,
+      "loss": 0.2765,
+      "step": 5699
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29904995978400906,
+      "learning_rate": 1.5707359794132178e-05,
+      "loss": 0.2535,
+      "step": 5700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36950212436020996,
+      "learning_rate": 1.5705831637394783e-05,
+      "loss": 0.3008,
+      "step": 5701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5786942167807272,
+      "learning_rate": 1.570430328306468e-05,
+      "loss": 0.3812,
+      "step": 5702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39500817970292046,
+      "learning_rate": 1.5702774731194802e-05,
+      "loss": 0.3126,
+      "step": 5703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3453869948145563,
+      "learning_rate": 1.570124598183808e-05,
+      "loss": 0.2725,
+      "step": 5704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4214946705124231,
+      "learning_rate": 1.569971703504745e-05,
+      "loss": 0.2872,
+      "step": 5705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3257990453502148,
+      "learning_rate": 1.5698187890875867e-05,
+      "loss": 0.2885,
+      "step": 5706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6989314635364635,
+      "learning_rate": 1.5696658549376286e-05,
+      "loss": 0.5029,
+      "step": 5707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34822630078350825,
+      "learning_rate": 1.569512901060166e-05,
+      "loss": 0.3442,
+      "step": 5708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34505897851320166,
+      "learning_rate": 1.569359927460496e-05,
+      "loss": 0.1921,
+      "step": 5709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2695284101966276,
+      "learning_rate": 1.5692069341439164e-05,
+      "loss": 0.2228,
+      "step": 5710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5436723132485777,
+      "learning_rate": 1.5690539211157255e-05,
+      "loss": 0.3901,
+      "step": 5711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33269868149267295,
+      "learning_rate": 1.5689008883812212e-05,
+      "loss": 0.2106,
+      "step": 5712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5032767085391571,
+      "learning_rate": 1.568747835945704e-05,
+      "loss": 0.3967,
+      "step": 5713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7674497920759249,
+      "learning_rate": 1.5685947638144736e-05,
+      "loss": 0.6305,
+      "step": 5714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33746113836252484,
+      "learning_rate": 1.5684416719928314e-05,
+      "loss": 0.2059,
+      "step": 5715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27730558315725434,
+      "learning_rate": 1.568288560486078e-05,
+      "loss": 0.2342,
+      "step": 5716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8937613239506347,
+      "learning_rate": 1.5681354292995164e-05,
+      "loss": 0.5943,
+      "step": 5717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32282729775940155,
+      "learning_rate": 1.5679822784384492e-05,
+      "loss": 0.213,
+      "step": 5718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0217359415141611,
+      "learning_rate": 1.56782910790818e-05,
+      "loss": 0.4414,
+      "step": 5719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37906399149967873,
+      "learning_rate": 1.5676759177140132e-05,
+      "loss": 0.3148,
+      "step": 5720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32548021526366716,
+      "learning_rate": 1.567522707861254e-05,
+      "loss": 0.2696,
+      "step": 5721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27672622569811445,
+      "learning_rate": 1.567369478355208e-05,
+      "loss": 0.1349,
+      "step": 5722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34813866318867676,
+      "learning_rate": 1.567216229201181e-05,
+      "loss": 0.2762,
+      "step": 5723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31379746680097137,
+      "learning_rate": 1.5670629604044804e-05,
+      "loss": 0.2616,
+      "step": 5724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8743307806116454,
+      "learning_rate": 1.566909671970414e-05,
+      "loss": 0.3819,
+      "step": 5725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9740313278256991,
+      "learning_rate": 1.5667563639042904e-05,
+      "loss": 0.6429,
+      "step": 5726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3849691561398392,
+      "learning_rate": 1.5666030362114175e-05,
+      "loss": 0.2787,
+      "step": 5727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29480543187124647,
+      "learning_rate": 1.566449688897106e-05,
+      "loss": 0.2065,
+      "step": 5728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5564381113672089,
+      "learning_rate": 1.5662963219666666e-05,
+      "loss": 0.3209,
+      "step": 5729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6185522440055002,
+      "learning_rate": 1.5661429354254096e-05,
+      "loss": 0.3824,
+      "step": 5730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40049158281522773,
+      "learning_rate": 1.5659895292786474e-05,
+      "loss": 0.3183,
+      "step": 5731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4038008265047527,
+      "learning_rate": 1.565836103531692e-05,
+      "loss": 0.3016,
+      "step": 5732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3987251335173393,
+      "learning_rate": 1.5656826581898563e-05,
+      "loss": 0.2744,
+      "step": 5733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34573611183952435,
+      "learning_rate": 1.565529193258455e-05,
+      "loss": 0.2572,
+      "step": 5734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.392359831338795,
+      "learning_rate": 1.5653757087428015e-05,
+      "loss": 0.2155,
+      "step": 5735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29979714182243516,
+      "learning_rate": 1.5652222046482118e-05,
+      "loss": 0.288,
+      "step": 5736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5553731791527236,
+      "learning_rate": 1.5650686809800016e-05,
+      "loss": 0.388,
+      "step": 5737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.935348590466938,
+      "learning_rate": 1.564915137743487e-05,
+      "loss": 0.5044,
+      "step": 5738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42901141158263,
+      "learning_rate": 1.5647615749439858e-05,
+      "loss": 0.302,
+      "step": 5739
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3289608155747018,
+      "learning_rate": 1.5646079925868152e-05,
+      "loss": 0.2922,
+      "step": 5740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5271177700015318,
+      "learning_rate": 1.564454390677294e-05,
+      "loss": 0.2541,
+      "step": 5741
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3476763157178258,
+      "learning_rate": 1.5643007692207422e-05,
+      "loss": 0.3031,
+      "step": 5742
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4244417596608532,
+      "learning_rate": 1.5641471282224788e-05,
+      "loss": 0.2396,
+      "step": 5743
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4413347170413946,
+      "learning_rate": 1.563993467687824e-05,
+      "loss": 0.3494,
+      "step": 5744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38877102993582696,
+      "learning_rate": 1.5638397876221002e-05,
+      "loss": 0.2112,
+      "step": 5745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4975149211929953,
+      "learning_rate": 1.563686088030629e-05,
+      "loss": 0.4052,
+      "step": 5746
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5640690106411652,
+      "learning_rate": 1.5635323689187323e-05,
+      "loss": 0.3971,
+      "step": 5747
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2221851855752157,
+      "learning_rate": 1.5633786302917343e-05,
+      "loss": 0.1517,
+      "step": 5748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2994501028669219,
+      "learning_rate": 1.5632248721549584e-05,
+      "loss": 0.2548,
+      "step": 5749
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8917217167943163,
+      "learning_rate": 1.5630710945137293e-05,
+      "loss": 0.6478,
+      "step": 5750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31361359549888296,
+      "learning_rate": 1.5629172973733724e-05,
+      "loss": 0.1688,
+      "step": 5751
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35912544773740485,
+      "learning_rate": 1.562763480739214e-05,
+      "loss": 0.2893,
+      "step": 5752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.711575311545721,
+      "learning_rate": 1.56260964461658e-05,
+      "loss": 0.4265,
+      "step": 5753
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2564745105098827,
+      "learning_rate": 1.5624557890107983e-05,
+      "loss": 0.1733,
+      "step": 5754
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3070550130073271,
+      "learning_rate": 1.5623019139271967e-05,
+      "loss": 0.2762,
+      "step": 5755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5205475343496394,
+      "learning_rate": 1.5621480193711046e-05,
+      "loss": 0.427,
+      "step": 5756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6040346459958775,
+      "learning_rate": 1.56199410534785e-05,
+      "loss": 0.3935,
+      "step": 5757
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5393463384567777,
+      "learning_rate": 1.5618401718627644e-05,
+      "loss": 0.2879,
+      "step": 5758
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4889068289885966,
+      "learning_rate": 1.5616862189211774e-05,
+      "loss": 0.3763,
+      "step": 5759
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27513668919227896,
+      "learning_rate": 1.561532246528421e-05,
+      "loss": 0.2058,
+      "step": 5760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2966718812667392,
+      "learning_rate": 1.5613782546898268e-05,
+      "loss": 0.2096,
+      "step": 5761
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9302097875569516,
+      "learning_rate": 1.561224243410728e-05,
+      "loss": 0.6853,
+      "step": 5762
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3034217660889585,
+      "learning_rate": 1.561070212696458e-05,
+      "loss": 0.2851,
+      "step": 5763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3943008222386303,
+      "learning_rate": 1.56091616255235e-05,
+      "loss": 0.2914,
+      "step": 5764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7197890030209596,
+      "learning_rate": 1.5607620929837398e-05,
+      "loss": 0.4985,
+      "step": 5765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3846549791318257,
+      "learning_rate": 1.5606080039959624e-05,
+      "loss": 0.2175,
+      "step": 5766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2644764393345778,
+      "learning_rate": 1.5604538955943542e-05,
+      "loss": 0.2057,
+      "step": 5767
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5406969227154103,
+      "learning_rate": 1.5602997677842515e-05,
+      "loss": 0.4251,
+      "step": 5768
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6047624630317593,
+      "learning_rate": 1.560145620570992e-05,
+      "loss": 0.4523,
+      "step": 5769
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.355995131291561,
+      "learning_rate": 1.5599914539599135e-05,
+      "loss": 0.314,
+      "step": 5770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4253729585784995,
+      "learning_rate": 1.559837267956355e-05,
+      "loss": 0.2929,
+      "step": 5771
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42743246173578286,
+      "learning_rate": 1.559683062565656e-05,
+      "loss": 0.3047,
+      "step": 5772
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2515830454917624,
+      "learning_rate": 1.559528837793157e-05,
+      "loss": 0.2118,
+      "step": 5773
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7515736027063699,
+      "learning_rate": 1.559374593644198e-05,
+      "loss": 0.5555,
+      "step": 5774
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4013581312855905,
+      "learning_rate": 1.559220330124121e-05,
+      "loss": 0.3164,
+      "step": 5775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32182922729328917,
+      "learning_rate": 1.5590660472382682e-05,
+      "loss": 0.3081,
+      "step": 5776
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.853128767784892,
+      "learning_rate": 1.558911744991982e-05,
+      "loss": 0.2577,
+      "step": 5777
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3250498927602875,
+      "learning_rate": 1.5587574233906063e-05,
+      "loss": 0.2265,
+      "step": 5778
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30334383260105524,
+      "learning_rate": 1.5586030824394848e-05,
+      "loss": 0.2567,
+      "step": 5779
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36184984465163766,
+      "learning_rate": 1.5584487221439628e-05,
+      "loss": 0.3016,
+      "step": 5780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46282632685077807,
+      "learning_rate": 1.5582943425093856e-05,
+      "loss": 0.3936,
+      "step": 5781
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40893305897006493,
+      "learning_rate": 1.558139943541099e-05,
+      "loss": 0.3362,
+      "step": 5782
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3743712544882061,
+      "learning_rate": 1.5579855252444506e-05,
+      "loss": 0.3249,
+      "step": 5783
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46996244742050114,
+      "learning_rate": 1.557831087624787e-05,
+      "loss": 0.1574,
+      "step": 5784
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2683925413029673,
+      "learning_rate": 1.5576766306874572e-05,
+      "loss": 0.2282,
+      "step": 5785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6926340821108188,
+      "learning_rate": 1.5575221544378094e-05,
+      "loss": 0.4933,
+      "step": 5786
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31226964320871636,
+      "learning_rate": 1.5573676588811935e-05,
+      "loss": 0.2602,
+      "step": 5787
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39932555162861216,
+      "learning_rate": 1.5572131440229593e-05,
+      "loss": 0.3255,
+      "step": 5788
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3892684727048228,
+      "learning_rate": 1.557058609868458e-05,
+      "loss": 0.2558,
+      "step": 5789
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2385948593781187,
+      "learning_rate": 1.5569040564230414e-05,
+      "loss": 0.1092,
+      "step": 5790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27404269814150173,
+      "learning_rate": 1.556749483692061e-05,
+      "loss": 0.2733,
+      "step": 5791
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8492486232920454,
+      "learning_rate": 1.5565948916808697e-05,
+      "loss": 0.5681,
+      "step": 5792
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6782578354702179,
+      "learning_rate": 1.5564402803948215e-05,
+      "loss": 0.2971,
+      "step": 5793
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3093663841738642,
+      "learning_rate": 1.55628564983927e-05,
+      "loss": 0.2356,
+      "step": 5794
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3650624325489127,
+      "learning_rate": 1.556131000019571e-05,
+      "loss": 0.3337,
+      "step": 5795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.760194714593758,
+      "learning_rate": 1.5559763309410787e-05,
+      "loss": 0.4862,
+      "step": 5796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3294635430927283,
+      "learning_rate": 1.5558216426091505e-05,
+      "loss": 0.2005,
+      "step": 5797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.077418090885813,
+      "learning_rate": 1.5556669350291422e-05,
+      "loss": 0.4722,
+      "step": 5798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3504419610826631,
+      "learning_rate": 1.5555122082064123e-05,
+      "loss": 0.323,
+      "step": 5799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3475869790586581,
+      "learning_rate": 1.5553574621463183e-05,
+      "loss": 0.2275,
+      "step": 5800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3020170894774215,
+      "learning_rate": 1.5552026968542192e-05,
+      "loss": 0.224,
+      "step": 5801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.697556497115101,
+      "learning_rate": 1.555047912335475e-05,
+      "loss": 0.6149,
+      "step": 5802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29416917374336826,
+      "learning_rate": 1.5548931085954448e-05,
+      "loss": 0.2219,
+      "step": 5803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6232842819271314,
+      "learning_rate": 1.5547382856394905e-05,
+      "loss": 0.3939,
+      "step": 5804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1892346356469627,
+      "learning_rate": 1.5545834434729732e-05,
+      "loss": 0.8569,
+      "step": 5805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3967731998369141,
+      "learning_rate": 1.554428582101255e-05,
+      "loss": 0.2805,
+      "step": 5806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34087023673325834,
+      "learning_rate": 1.554273701529699e-05,
+      "loss": 0.1824,
+      "step": 5807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.374281261563989,
+      "learning_rate": 1.5541188017636683e-05,
+      "loss": 0.678,
+      "step": 5808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3489684607123253,
+      "learning_rate": 1.5539638828085278e-05,
+      "loss": 0.2842,
+      "step": 5809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6789014020469017,
+      "learning_rate": 1.5538089446696414e-05,
+      "loss": 0.3743,
+      "step": 5810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4519718080022967,
+      "learning_rate": 1.553653987352375e-05,
+      "loss": 0.3394,
+      "step": 5811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4552790837774947,
+      "learning_rate": 1.553499010862095e-05,
+      "loss": 0.299,
+      "step": 5812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4965276597889446,
+      "learning_rate": 1.553344015204168e-05,
+      "loss": 0.1166,
+      "step": 5813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5101673833679304,
+      "learning_rate": 1.553189000383962e-05,
+      "loss": 0.394,
+      "step": 5814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42668233072512435,
+      "learning_rate": 1.553033966406844e-05,
+      "loss": 0.2904,
+      "step": 5815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3709815294186113,
+      "learning_rate": 1.552878913278184e-05,
+      "loss": 0.3063,
+      "step": 5816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7836478889317833,
+      "learning_rate": 1.5527238410033508e-05,
+      "loss": 0.4855,
+      "step": 5817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43253924637788094,
+      "learning_rate": 1.552568749587715e-05,
+      "loss": 0.2718,
+      "step": 5818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.287688911198083,
+      "learning_rate": 1.5524136390366468e-05,
+      "loss": 0.2535,
+      "step": 5819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.723004816360338,
+      "learning_rate": 1.5522585093555184e-05,
+      "loss": 0.3463,
+      "step": 5820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3808008491185971,
+      "learning_rate": 1.5521033605497013e-05,
+      "loss": 0.2936,
+      "step": 5821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8200187852102342,
+      "learning_rate": 1.5519481926245687e-05,
+      "loss": 0.5662,
+      "step": 5822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36108551149142204,
+      "learning_rate": 1.551793005585494e-05,
+      "loss": 0.2783,
+      "step": 5823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3836885248758884,
+      "learning_rate": 1.5516377994378513e-05,
+      "loss": 0.288,
+      "step": 5824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32215157055186644,
+      "learning_rate": 1.551482574187015e-05,
+      "loss": 0.234,
+      "step": 5825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4798535959282858,
+      "learning_rate": 1.5513273298383607e-05,
+      "loss": 0.3197,
+      "step": 5826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32604749167703956,
+      "learning_rate": 1.551172066397265e-05,
+      "loss": 0.2737,
+      "step": 5827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9180829156266932,
+      "learning_rate": 1.5510167838691047e-05,
+      "loss": 0.5507,
+      "step": 5828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2475558469396781,
+      "learning_rate": 1.550861482259256e-05,
+      "loss": 0.5742,
+      "step": 5829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37812167267697283,
+      "learning_rate": 1.5507061615730986e-05,
+      "loss": 0.2834,
+      "step": 5830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31942624860729857,
+      "learning_rate": 1.5505508218160103e-05,
+      "loss": 0.2459,
+      "step": 5831
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5164865863370692,
+      "learning_rate": 1.5503954629933707e-05,
+      "loss": 0.3329,
+      "step": 5832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4604448154652938,
+      "learning_rate": 1.5502400851105603e-05,
+      "loss": 0.1969,
+      "step": 5833
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.57149449313321,
+      "learning_rate": 1.5500846881729587e-05,
+      "loss": 0.4065,
+      "step": 5834
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4424655995824884,
+      "learning_rate": 1.5499292721859483e-05,
+      "loss": 0.3512,
+      "step": 5835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3409734791754797,
+      "learning_rate": 1.5497738371549108e-05,
+      "loss": 0.2024,
+      "step": 5836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44336289462689166,
+      "learning_rate": 1.549618383085229e-05,
+      "loss": 0.3014,
+      "step": 5837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3706424088281798,
+      "learning_rate": 1.549462909982286e-05,
+      "loss": 0.2969,
+      "step": 5838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3273551180286123,
+      "learning_rate": 1.5493074178514665e-05,
+      "loss": 0.2201,
+      "step": 5839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4602400150554773,
+      "learning_rate": 1.5491519066981547e-05,
+      "loss": 0.7239,
+      "step": 5840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2934303316229814,
+      "learning_rate": 1.5489963765277356e-05,
+      "loss": 0.8344,
+      "step": 5841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37170829852311305,
+      "learning_rate": 1.548840827345596e-05,
+      "loss": 0.2075,
+      "step": 5842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3690170784055627,
+      "learning_rate": 1.5486852591571217e-05,
+      "loss": 0.3312,
+      "step": 5843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42868695087431713,
+      "learning_rate": 1.5485296719677005e-05,
+      "loss": 0.2912,
+      "step": 5844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.32516213319606574,
+      "learning_rate": 1.5483740657827205e-05,
+      "loss": 0.2423,
+      "step": 5845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8624921342883979,
+      "learning_rate": 1.5482184406075705e-05,
+      "loss": 0.3653,
+      "step": 5846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43311550871258736,
+      "learning_rate": 1.5480627964476392e-05,
+      "loss": 0.3782,
+      "step": 5847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3466225539727474,
+      "learning_rate": 1.547907133308317e-05,
+      "loss": 0.2827,
+      "step": 5848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6526142864376929,
+      "learning_rate": 1.547751451194994e-05,
+      "loss": 0.3642,
+      "step": 5849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3194788764576168,
+      "learning_rate": 1.5475957501130622e-05,
+      "loss": 0.3314,
+      "step": 5850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33247814679611337,
+      "learning_rate": 1.5474400300679128e-05,
+      "loss": 0.2097,
+      "step": 5851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34086337965321845,
+      "learning_rate": 1.5472842910649387e-05,
+      "loss": 0.2012,
+      "step": 5852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8793044262444761,
+      "learning_rate": 1.5471285331095334e-05,
+      "loss": 0.6676,
+      "step": 5853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5278240555166073,
+      "learning_rate": 1.5469727562070904e-05,
+      "loss": 0.3511,
+      "step": 5854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3723179875730455,
+      "learning_rate": 1.5468169603630045e-05,
+      "loss": 0.2741,
+      "step": 5855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4558277232143511,
+      "learning_rate": 1.5466611455826703e-05,
+      "loss": 0.2971,
+      "step": 5856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3657369354981838,
+      "learning_rate": 1.5465053118714846e-05,
+      "loss": 0.2267,
+      "step": 5857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40254068196433923,
+      "learning_rate": 1.5463494592348435e-05,
+      "loss": 0.3381,
+      "step": 5858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4389789142234693,
+      "learning_rate": 1.5461935876781436e-05,
+      "loss": 0.3111,
+      "step": 5859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42792874304101514,
+      "learning_rate": 1.5460376972067837e-05,
+      "loss": 0.3259,
+      "step": 5860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5828730662973869,
+      "learning_rate": 1.5458817878261617e-05,
+      "loss": 0.3779,
+      "step": 5861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35919316933562373,
+      "learning_rate": 1.5457258595416766e-05,
+      "loss": 0.2507,
+      "step": 5862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33609490374293677,
+      "learning_rate": 1.5455699123587286e-05,
+      "loss": 0.225,
+      "step": 5863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34888688858881917,
+      "learning_rate": 1.5454139462827183e-05,
+      "loss": 0.2434,
+      "step": 5864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.743990986427087,
+      "learning_rate": 1.5452579613190462e-05,
+      "loss": 0.4317,
+      "step": 5865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3496844176665509,
+      "learning_rate": 1.5451019574731147e-05,
+      "loss": 0.3032,
+      "step": 5866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39178507409648256,
+      "learning_rate": 1.5449459347503255e-05,
+      "loss": 0.3358,
+      "step": 5867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30121681347934715,
+      "learning_rate": 1.5447898931560824e-05,
+      "loss": 0.2011,
+      "step": 5868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2932285924621865,
+      "learning_rate": 1.544633832695788e-05,
+      "loss": 0.205,
+      "step": 5869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3773063886258503,
+      "learning_rate": 1.544477753374848e-05,
+      "loss": 0.3348,
+      "step": 5870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4860936216259758,
+      "learning_rate": 1.5443216551986667e-05,
+      "loss": 0.4027,
+      "step": 5871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5365140122310895,
+      "learning_rate": 1.5441655381726496e-05,
+      "loss": 0.2675,
+      "step": 5872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5126169170705192,
+      "learning_rate": 1.5440094023022035e-05,
+      "loss": 0.347,
+      "step": 5873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3880198754870328,
+      "learning_rate": 1.5438532475927354e-05,
+      "loss": 0.3277,
+      "step": 5874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3398774666350326,
+      "learning_rate": 1.5436970740496527e-05,
+      "loss": 0.1709,
+      "step": 5875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2934353428437412,
+      "learning_rate": 1.5435408816783635e-05,
+      "loss": 0.2122,
+      "step": 5876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1948470060235745,
+      "learning_rate": 1.543384670484277e-05,
+      "loss": 0.5015,
+      "step": 5877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37606014000496846,
+      "learning_rate": 1.5432284404728027e-05,
+      "loss": 0.256,
+      "step": 5878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4058687646461349,
+      "learning_rate": 1.5430721916493507e-05,
+      "loss": 0.3533,
+      "step": 5879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8713355870009596,
+      "learning_rate": 1.542915924019332e-05,
+      "loss": 0.6042,
+      "step": 5880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2912589294582795,
+      "learning_rate": 1.5427596375881587e-05,
+      "loss": 0.1711,
+      "step": 5881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6110141021008527,
+      "learning_rate": 1.5426033323612425e-05,
+      "loss": 0.3838,
+      "step": 5882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8743221394886532,
+      "learning_rate": 1.5424470083439958e-05,
+      "loss": 0.3455,
+      "step": 5883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6854045258290827,
+      "learning_rate": 1.5422906655418327e-05,
+      "loss": 0.4075,
+      "step": 5884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28645214108364625,
+      "learning_rate": 1.5421343039601672e-05,
+      "loss": 0.185,
+      "step": 5885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39523662011429733,
+      "learning_rate": 1.5419779236044142e-05,
+      "loss": 0.3153,
+      "step": 5886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6709944581548378,
+      "learning_rate": 1.541821524479989e-05,
+      "loss": 0.4063,
+      "step": 5887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3576383977318532,
+      "learning_rate": 1.541665106592307e-05,
+      "loss": 0.2235,
+      "step": 5888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0212539106154426,
+      "learning_rate": 1.5415086699467864e-05,
+      "loss": 0.4622,
+      "step": 5889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39618828216583873,
+      "learning_rate": 1.5413522145488437e-05,
+      "loss": 0.3115,
+      "step": 5890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27461183094712965,
+      "learning_rate": 1.541195740403897e-05,
+      "loss": 0.1857,
+      "step": 5891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9085111519204856,
+      "learning_rate": 1.541039247517365e-05,
+      "loss": 0.5692,
+      "step": 5892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8533930219299408,
+      "learning_rate": 1.5408827358946675e-05,
+      "loss": 0.5139,
+      "step": 5893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4079774306429693,
+      "learning_rate": 1.5407262055412238e-05,
+      "loss": 0.2896,
+      "step": 5894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7506257968661608,
+      "learning_rate": 1.540569656462455e-05,
+      "loss": 0.3278,
+      "step": 5895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6944121449841651,
+      "learning_rate": 1.5404130886637822e-05,
+      "loss": 0.4847,
+      "step": 5896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.289549678372288,
+      "learning_rate": 1.5402565021506273e-05,
+      "loss": 0.2154,
+      "step": 5897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3543644317502926,
+      "learning_rate": 1.540099896928413e-05,
+      "loss": 0.237,
+      "step": 5898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5313370994029551,
+      "learning_rate": 1.5399432730025626e-05,
+      "loss": 0.3474,
+      "step": 5899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3858996506274038,
+      "learning_rate": 1.5397866303784996e-05,
+      "loss": 0.3342,
+      "step": 5900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8419789834381768,
+      "learning_rate": 1.539629969061649e-05,
+      "loss": 0.428,
+      "step": 5901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.32299250679010993,
+      "learning_rate": 1.539473289057436e-05,
+      "loss": 0.3036,
+      "step": 5902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27648283554967706,
+      "learning_rate": 1.5393165903712856e-05,
+      "loss": 0.2395,
+      "step": 5903
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2843375577344682,
+      "learning_rate": 1.5391598730086254e-05,
+      "loss": 0.1435,
+      "step": 5904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5555925119853822,
+      "learning_rate": 1.539003136974882e-05,
+      "loss": 0.3774,
+      "step": 5905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3144089559292263,
+      "learning_rate": 1.5388463822754827e-05,
+      "loss": 0.2653,
+      "step": 5906
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4289801488875986,
+      "learning_rate": 1.538689608915857e-05,
+      "loss": 0.3952,
+      "step": 5907
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4080185198275796,
+      "learning_rate": 1.5385328169014325e-05,
+      "loss": 0.2654,
+      "step": 5908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23617679924423607,
+      "learning_rate": 1.53837600623764e-05,
+      "loss": 0.1897,
+      "step": 5909
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3640319175940855,
+      "learning_rate": 1.5382191769299096e-05,
+      "loss": 0.3338,
+      "step": 5910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5949894824883931,
+      "learning_rate": 1.5380623289836724e-05,
+      "loss": 0.2869,
+      "step": 5911
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33508321042850364,
+      "learning_rate": 1.5379054624043596e-05,
+      "loss": 0.3238,
+      "step": 5912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.042458284916835,
+      "learning_rate": 1.537748577197404e-05,
+      "loss": 0.7016,
+      "step": 5913
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28987327834740845,
+      "learning_rate": 1.537591673368238e-05,
+      "loss": 0.2563,
+      "step": 5914
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25989501170956886,
+      "learning_rate": 1.5374347509222962e-05,
+      "loss": 0.2252,
+      "step": 5915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4262012511260465,
+      "learning_rate": 1.5372778098650115e-05,
+      "loss": 0.2876,
+      "step": 5916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6347305763733735,
+      "learning_rate": 1.5371208502018194e-05,
+      "loss": 0.3217,
+      "step": 5917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.312424850972814,
+      "learning_rate": 1.5369638719381555e-05,
+      "loss": 0.2914,
+      "step": 5918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4014464340135386,
+      "learning_rate": 1.5368068750794557e-05,
+      "loss": 0.7689,
+      "step": 5919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6480541705283047,
+      "learning_rate": 1.5366498596311568e-05,
+      "loss": 0.4657,
+      "step": 5920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2557476620889882,
+      "learning_rate": 1.5364928255986966e-05,
+      "loss": 0.1722,
+      "step": 5921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3457499150185082,
+      "learning_rate": 1.5363357729875126e-05,
+      "loss": 0.2854,
+      "step": 5922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7202941326070144,
+      "learning_rate": 1.536178701803044e-05,
+      "loss": 0.392,
+      "step": 5923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34428868696575277,
+      "learning_rate": 1.5360216120507302e-05,
+      "loss": 0.2324,
+      "step": 5924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5147196984487438,
+      "learning_rate": 1.5358645037360108e-05,
+      "loss": 0.4216,
+      "step": 5925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.537087396199377,
+      "learning_rate": 1.535707376864327e-05,
+      "loss": 0.377,
+      "step": 5926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34006389687556765,
+      "learning_rate": 1.5355502314411194e-05,
+      "loss": 0.1906,
+      "step": 5927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3373315399765145,
+      "learning_rate": 1.5353930674718305e-05,
+      "loss": 0.2455,
+      "step": 5928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.46153518536615695,
+      "learning_rate": 1.5352358849619024e-05,
+      "loss": 0.3767,
+      "step": 5929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3368768244126122,
+      "learning_rate": 1.535078683916779e-05,
+      "loss": 0.2229,
+      "step": 5930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1003227430828704,
+      "learning_rate": 1.5349214643419034e-05,
+      "loss": 0.6771,
+      "step": 5931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4907212686940303,
+      "learning_rate": 1.5347642262427206e-05,
+      "loss": 0.8189,
+      "step": 5932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3184937333207399,
+      "learning_rate": 1.5346069696246758e-05,
+      "loss": 0.2215,
+      "step": 5933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41845407671672097,
+      "learning_rate": 1.534449694493215e-05,
+      "loss": 0.2991,
+      "step": 5934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4774275005408674,
+      "learning_rate": 1.534292400853784e-05,
+      "loss": 0.3517,
+      "step": 5935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.379325144992996,
+      "learning_rate": 1.53413508871183e-05,
+      "loss": 0.2905,
+      "step": 5936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.563366841637593,
+      "learning_rate": 1.5339777580728003e-05,
+      "loss": 0.3085,
+      "step": 5937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.463202999933601,
+      "learning_rate": 1.5338204089421447e-05,
+      "loss": 0.3625,
+      "step": 5938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3695255382677873,
+      "learning_rate": 1.5336630413253108e-05,
+      "loss": 0.2843,
+      "step": 5939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6537639861062193,
+      "learning_rate": 1.533505655227749e-05,
+      "loss": 0.3645,
+      "step": 5940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.255986281866402,
+      "learning_rate": 1.533348250654909e-05,
+      "loss": 0.2218,
+      "step": 5941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40879555334744677,
+      "learning_rate": 1.5331908276122424e-05,
+      "loss": 0.3137,
+      "step": 5942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9387421903416497,
+      "learning_rate": 1.5330333861051998e-05,
+      "loss": 0.5082,
+      "step": 5943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.137603306784779,
+      "learning_rate": 1.5328759261392344e-05,
+      "loss": 0.5119,
+      "step": 5944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3617661594873792,
+      "learning_rate": 1.5327184477197984e-05,
+      "loss": 0.301,
+      "step": 5945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.443054484749188,
+      "learning_rate": 1.5325609508523456e-05,
+      "loss": 0.3465,
+      "step": 5946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2663146536046802,
+      "learning_rate": 1.53240343554233e-05,
+      "loss": 0.1144,
+      "step": 5947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38139187527879836,
+      "learning_rate": 1.532245901795206e-05,
+      "loss": 0.2786,
+      "step": 5948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5305598314897465,
+      "learning_rate": 1.5320883496164295e-05,
+      "loss": 0.3807,
+      "step": 5949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4413082084005958,
+      "learning_rate": 1.5319307790114563e-05,
+      "loss": 0.3002,
+      "step": 5950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3583027015117161,
+      "learning_rate": 1.5317731899857434e-05,
+      "loss": 0.3078,
+      "step": 5951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7092834562172043,
+      "learning_rate": 1.5316155825447476e-05,
+      "loss": 0.4817,
+      "step": 5952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2955321017741597,
+      "learning_rate": 1.5314579566939274e-05,
+      "loss": 0.1988,
+      "step": 5953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37715745593890787,
+      "learning_rate": 1.5313003124387404e-05,
+      "loss": 0.2891,
+      "step": 5954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3060160054166214,
+      "learning_rate": 1.5311426497846466e-05,
+      "loss": 0.7819,
+      "step": 5955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7017722643430202,
+      "learning_rate": 1.530984968737106e-05,
+      "loss": 0.3755,
+      "step": 5956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44304273216374057,
+      "learning_rate": 1.5308272693015785e-05,
+      "loss": 0.2946,
+      "step": 5957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3809476824053261,
+      "learning_rate": 1.530669551483525e-05,
+      "loss": 0.2984,
+      "step": 5958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.32731689524967394,
+      "learning_rate": 1.5305118152884086e-05,
+      "loss": 0.2025,
+      "step": 5959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.32662163125723054,
+      "learning_rate": 1.5303540607216906e-05,
+      "loss": 0.2074,
+      "step": 5960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6125753143165558,
+      "learning_rate": 1.5301962877888338e-05,
+      "loss": 0.4477,
+      "step": 5961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7210324834024456,
+      "learning_rate": 1.5300384964953028e-05,
+      "loss": 0.3894,
+      "step": 5962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3512008862927995,
+      "learning_rate": 1.5298806868465615e-05,
+      "loss": 0.2395,
+      "step": 5963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.203964316197915,
+      "learning_rate": 1.5297228588480744e-05,
+      "loss": 0.6873,
+      "step": 5964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4470807698277849,
+      "learning_rate": 1.5295650125053078e-05,
+      "loss": 0.344,
+      "step": 5965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27900170832730514,
+      "learning_rate": 1.529407147823728e-05,
+      "loss": 0.1535,
+      "step": 5966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2564959101088011,
+      "learning_rate": 1.529249264808801e-05,
+      "loss": 0.9462,
+      "step": 5967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8273397489953469,
+      "learning_rate": 1.5290913634659946e-05,
+      "loss": 0.4594,
+      "step": 5968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38156537104710225,
+      "learning_rate": 1.528933443800777e-05,
+      "loss": 0.2717,
+      "step": 5969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40525374010366605,
+      "learning_rate": 1.5287755058186173e-05,
+      "loss": 0.2764,
+      "step": 5970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44559989652126264,
+      "learning_rate": 1.5286175495249845e-05,
+      "loss": 0.2679,
+      "step": 5971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3800803496165543,
+      "learning_rate": 1.5284595749253486e-05,
+      "loss": 0.2887,
+      "step": 5972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5854777998684502,
+      "learning_rate": 1.5283015820251802e-05,
+      "loss": 0.3128,
+      "step": 5973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6490823546446826,
+      "learning_rate": 1.528143570829951e-05,
+      "loss": 0.3814,
+      "step": 5974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37018842276377245,
+      "learning_rate": 1.5279855413451323e-05,
+      "loss": 0.223,
+      "step": 5975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4312382176580631,
+      "learning_rate": 1.527827493576197e-05,
+      "loss": 0.284,
+      "step": 5976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40708070399001484,
+      "learning_rate": 1.5276694275286188e-05,
+      "loss": 0.3473,
+      "step": 5977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5786187802356816,
+      "learning_rate": 1.5275113432078707e-05,
+      "loss": 0.3675,
+      "step": 5978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5832872436341768,
+      "learning_rate": 1.5273532406194273e-05,
+      "loss": 0.3081,
+      "step": 5979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48552801852112154,
+      "learning_rate": 1.5271951197687642e-05,
+      "loss": 0.3343,
+      "step": 5980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3366837647895135,
+      "learning_rate": 1.5270369806613566e-05,
+      "loss": 0.2745,
+      "step": 5981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30439148010092015,
+      "learning_rate": 1.5268788233026813e-05,
+      "loss": 0.2576,
+      "step": 5982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8410593472550371,
+      "learning_rate": 1.5267206476982143e-05,
+      "loss": 0.2501,
+      "step": 5983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41983616658716516,
+      "learning_rate": 1.5265624538534346e-05,
+      "loss": 0.335,
+      "step": 5984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29918174520908947,
+      "learning_rate": 1.5264042417738198e-05,
+      "loss": 0.3013,
+      "step": 5985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9234263832577031,
+      "learning_rate": 1.5262460114648487e-05,
+      "loss": 0.4014,
+      "step": 5986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23127103984282238,
+      "learning_rate": 1.5260877629320003e-05,
+      "loss": 0.1819,
+      "step": 5987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6068087085849305,
+      "learning_rate": 1.5259294961807557e-05,
+      "loss": 0.3792,
+      "step": 5988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4333690325500259,
+      "learning_rate": 1.5257712112165952e-05,
+      "loss": 0.2803,
+      "step": 5989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3537720037573216,
+      "learning_rate": 1.5256129080450004e-05,
+      "loss": 0.2856,
+      "step": 5990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1232073792582922,
+      "learning_rate": 1.5254545866714531e-05,
+      "loss": 0.5189,
+      "step": 5991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4744517484846774,
+      "learning_rate": 1.5252962471014358e-05,
+      "loss": 0.3488,
+      "step": 5992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3529505990489683,
+      "learning_rate": 1.5251378893404324e-05,
+      "loss": 0.2746,
+      "step": 5993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2696964638067003,
+      "learning_rate": 1.5249795133939262e-05,
+      "loss": 0.2123,
+      "step": 5994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.98685682080623,
+      "learning_rate": 1.5248211192674023e-05,
+      "loss": 0.5288,
+      "step": 5995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6945777224570654,
+      "learning_rate": 1.5246627069663453e-05,
+      "loss": 0.2634,
+      "step": 5996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3390663328402181,
+      "learning_rate": 1.5245042764962416e-05,
+      "loss": 0.3044,
+      "step": 5997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1668116852346306,
+      "learning_rate": 1.5243458278625771e-05,
+      "loss": 0.8487,
+      "step": 5998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20443862509215088,
+      "learning_rate": 1.5241873610708395e-05,
+      "loss": 0.0946,
+      "step": 5999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4356866107324079,
+      "learning_rate": 1.5240288761265158e-05,
+      "loss": 0.3402,
+      "step": 6000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36544328154024314,
+      "learning_rate": 1.523870373035095e-05,
+      "loss": 0.3262,
+      "step": 6001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.684350221463673,
+      "learning_rate": 1.5237118518020656e-05,
+      "loss": 0.2627,
+      "step": 6002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7652539514029776,
+      "learning_rate": 1.5235533124329172e-05,
+      "loss": 0.4287,
+      "step": 6003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4259890730487947,
+      "learning_rate": 1.5233947549331399e-05,
+      "loss": 0.8299,
+      "step": 6004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2552789675583203,
+      "learning_rate": 1.5232361793082251e-05,
+      "loss": 0.2188,
+      "step": 6005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26345923306917046,
+      "learning_rate": 1.5230775855636635e-05,
+      "loss": 0.1973,
+      "step": 6006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7307326078246826,
+      "learning_rate": 1.522918973704948e-05,
+      "loss": 0.4981,
+      "step": 6007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.622572484624821,
+      "learning_rate": 1.5227603437375704e-05,
+      "loss": 0.3457,
+      "step": 6008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3376871552073923,
+      "learning_rate": 1.5226016956670251e-05,
+      "loss": 0.2555,
+      "step": 6009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1334926262165705,
+      "learning_rate": 1.5224430294988055e-05,
+      "loss": 0.7165,
+      "step": 6010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5732166382836338,
+      "learning_rate": 1.522284345238406e-05,
+      "loss": 0.3788,
+      "step": 6011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2403970752189537,
+      "learning_rate": 1.5221256428913225e-05,
+      "loss": 0.1541,
+      "step": 6012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46260415675738403,
+      "learning_rate": 1.52196692246305e-05,
+      "loss": 0.3579,
+      "step": 6013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7041434257174486,
+      "learning_rate": 1.5218081839590855e-05,
+      "loss": 0.4429,
+      "step": 6014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33938295637228305,
+      "learning_rate": 1.5216494273849261e-05,
+      "loss": 0.243,
+      "step": 6015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5118168354209909,
+      "learning_rate": 1.5214906527460695e-05,
+      "loss": 0.3826,
+      "step": 6016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4851060809319056,
+      "learning_rate": 1.5213318600480138e-05,
+      "loss": 0.3629,
+      "step": 6017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.22921600058430264,
+      "learning_rate": 1.5211730492962587e-05,
+      "loss": 0.161,
+      "step": 6018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.675451182016171,
+      "learning_rate": 1.521014220496303e-05,
+      "loss": 0.4479,
+      "step": 6019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5422555876026078,
+      "learning_rate": 1.5208553736536473e-05,
+      "loss": 0.3674,
+      "step": 6020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34458266162583756,
+      "learning_rate": 1.5206965087737922e-05,
+      "loss": 0.2923,
+      "step": 6021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9441003102508054,
+      "learning_rate": 1.5205376258622397e-05,
+      "loss": 0.4712,
+      "step": 6022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5721243311796721,
+      "learning_rate": 1.5203787249244914e-05,
+      "loss": 0.3698,
+      "step": 6023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32675837351870496,
+      "learning_rate": 1.5202198059660504e-05,
+      "loss": 0.2448,
+      "step": 6024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43602889305513376,
+      "learning_rate": 1.5200608689924197e-05,
+      "loss": 0.3166,
+      "step": 6025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3492478650448725,
+      "learning_rate": 1.5199019140091037e-05,
+      "loss": 0.1646,
+      "step": 6026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3940936399332753,
+      "learning_rate": 1.5197429410216065e-05,
+      "loss": 0.3279,
+      "step": 6027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3892155238830433,
+      "learning_rate": 1.5195839500354337e-05,
+      "loss": 0.2918,
+      "step": 6028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6679840842709365,
+      "learning_rate": 1.5194249410560913e-05,
+      "loss": 0.4226,
+      "step": 6029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3529717737800829,
+      "learning_rate": 1.5192659140890851e-05,
+      "loss": 0.3048,
+      "step": 6030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3234022745735802,
+      "learning_rate": 1.5191068691399229e-05,
+      "loss": 0.2089,
+      "step": 6031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4406803384243917,
+      "learning_rate": 1.518947806214112e-05,
+      "loss": 0.356,
+      "step": 6032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3316857147599567,
+      "learning_rate": 1.5187887253171609e-05,
+      "loss": 0.3087,
+      "step": 6033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9447485585303956,
+      "learning_rate": 1.5186296264545787e-05,
+      "loss": 0.6407,
+      "step": 6034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.507392674710051,
+      "learning_rate": 1.5184705096318748e-05,
+      "loss": 0.2305,
+      "step": 6035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31821764679036,
+      "learning_rate": 1.5183113748545595e-05,
+      "loss": 0.2953,
+      "step": 6036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37936593474211566,
+      "learning_rate": 1.5181522221281435e-05,
+      "loss": 0.335,
+      "step": 6037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1908639292571459,
+      "learning_rate": 1.5179930514581383e-05,
+      "loss": 0.1056,
+      "step": 6038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39556981816109443,
+      "learning_rate": 1.517833862850056e-05,
+      "loss": 0.3185,
+      "step": 6039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0851322992534065,
+      "learning_rate": 1.5176746563094092e-05,
+      "loss": 0.6819,
+      "step": 6040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32365313141441393,
+      "learning_rate": 1.5175154318417116e-05,
+      "loss": 0.2474,
+      "step": 6041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44971760340828426,
+      "learning_rate": 1.5173561894524765e-05,
+      "loss": 0.3509,
+      "step": 6042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6905509263632318,
+      "learning_rate": 1.517196929147219e-05,
+      "loss": 0.4647,
+      "step": 6043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24909171253007986,
+      "learning_rate": 1.5170376509314539e-05,
+      "loss": 0.1757,
+      "step": 6044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39274007428317864,
+      "learning_rate": 1.5168783548106976e-05,
+      "loss": 0.3238,
+      "step": 6045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9155198249464075,
+      "learning_rate": 1.5167190407904656e-05,
+      "loss": 0.6702,
+      "step": 6046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.573228632743676,
+      "learning_rate": 1.5165597088762757e-05,
+      "loss": 0.3907,
+      "step": 6047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37380539990802764,
+      "learning_rate": 1.5164003590736452e-05,
+      "loss": 0.2131,
+      "step": 6048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43565239129066136,
+      "learning_rate": 1.5162409913880927e-05,
+      "loss": 0.3494,
+      "step": 6049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31162452510656125,
+      "learning_rate": 1.5160816058251367e-05,
+      "loss": 0.1867,
+      "step": 6050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33539240652800983,
+      "learning_rate": 1.5159222023902969e-05,
+      "loss": 0.2238,
+      "step": 6051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3631277574565755,
+      "learning_rate": 1.5157627810890937e-05,
+      "loss": 0.3246,
+      "step": 6052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.744527008947136,
+      "learning_rate": 1.5156033419270472e-05,
+      "loss": 0.4741,
+      "step": 6053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3080561599256677,
+      "learning_rate": 1.5154438849096791e-05,
+      "loss": 0.2386,
+      "step": 6054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9356964028488047,
+      "learning_rate": 1.5152844100425114e-05,
+      "loss": 0.5416,
+      "step": 6055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24395016905721084,
+      "learning_rate": 1.5151249173310672e-05,
+      "loss": 0.2228,
+      "step": 6056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.36970463621043104,
+      "learning_rate": 1.5149654067808688e-05,
+      "loss": 0.2883,
+      "step": 6057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9048703602569318,
+      "learning_rate": 1.5148058783974407e-05,
+      "loss": 0.4467,
+      "step": 6058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.695872288096352,
+      "learning_rate": 1.5146463321863069e-05,
+      "loss": 0.4602,
+      "step": 6059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37486593573199756,
+      "learning_rate": 1.514486768152993e-05,
+      "loss": 0.2843,
+      "step": 6060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3653532580952253,
+      "learning_rate": 1.5143271863030244e-05,
+      "loss": 0.2839,
+      "step": 6061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2517009406199723,
+      "learning_rate": 1.5141675866419276e-05,
+      "loss": 0.1826,
+      "step": 6062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3704254478483168,
+      "learning_rate": 1.5140079691752293e-05,
+      "loss": 0.2775,
+      "step": 6063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46425654567960384,
+      "learning_rate": 1.5138483339084571e-05,
+      "loss": 0.3136,
+      "step": 6064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.959500141667516,
+      "learning_rate": 1.5136886808471389e-05,
+      "loss": 0.5221,
+      "step": 6065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3435644123187082,
+      "learning_rate": 1.5135290099968043e-05,
+      "loss": 0.2827,
+      "step": 6066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0712411877235641,
+      "learning_rate": 1.5133693213629818e-05,
+      "loss": 0.3403,
+      "step": 6067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2858791739075835,
+      "learning_rate": 1.513209614951202e-05,
+      "loss": 0.2423,
+      "step": 6068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4780296493408544,
+      "learning_rate": 1.5130498907669952e-05,
+      "loss": 0.2785,
+      "step": 6069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8319222954221501,
+      "learning_rate": 1.512890148815893e-05,
+      "loss": 0.479,
+      "step": 6070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32882328520167026,
+      "learning_rate": 1.5127303891034264e-05,
+      "loss": 0.183,
+      "step": 6071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33276901027696554,
+      "learning_rate": 1.5125706116351291e-05,
+      "loss": 0.279,
+      "step": 6072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5268756881469756,
+      "learning_rate": 1.5124108164165333e-05,
+      "loss": 0.3849,
+      "step": 6073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2489127782935394,
+      "learning_rate": 1.512251003453173e-05,
+      "loss": 0.4358,
+      "step": 6074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34742346899867804,
+      "learning_rate": 1.5120911727505822e-05,
+      "loss": 0.2864,
+      "step": 6075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.537616793518178,
+      "learning_rate": 1.5119313243142964e-05,
+      "loss": 0.3793,
+      "step": 6076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4013084106067741,
+      "learning_rate": 1.5117714581498509e-05,
+      "loss": 0.1941,
+      "step": 6077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38191970049336904,
+      "learning_rate": 1.5116115742627815e-05,
+      "loss": 0.2683,
+      "step": 6078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8920627144260248,
+      "learning_rate": 1.5114516726586254e-05,
+      "loss": 0.5596,
+      "step": 6079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4618446581767463,
+      "learning_rate": 1.51129175334292e-05,
+      "loss": 0.2894,
+      "step": 6080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43028981687553913,
+      "learning_rate": 1.5111318163212032e-05,
+      "loss": 0.2758,
+      "step": 6081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6246636674476441,
+      "learning_rate": 1.5109718615990135e-05,
+      "loss": 0.397,
+      "step": 6082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2595488255827491,
+      "learning_rate": 1.51081188918189e-05,
+      "loss": 0.2166,
+      "step": 6083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3366273152459292,
+      "learning_rate": 1.5106518990753731e-05,
+      "loss": 0.2066,
+      "step": 6084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5509473415305383,
+      "learning_rate": 1.5104918912850029e-05,
+      "loss": 0.3952,
+      "step": 6085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.168537237871508,
+      "learning_rate": 1.5103318658163202e-05,
+      "loss": 0.6159,
+      "step": 6086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44113838145483497,
+      "learning_rate": 1.5101718226748673e-05,
+      "loss": 0.2563,
+      "step": 6087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42233647100665583,
+      "learning_rate": 1.5100117618661856e-05,
+      "loss": 0.3352,
+      "step": 6088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.153551365592648,
+      "learning_rate": 1.5098516833958187e-05,
+      "loss": 0.7705,
+      "step": 6089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21605941110243826,
+      "learning_rate": 1.50969158726931e-05,
+      "loss": 0.1453,
+      "step": 6090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1214770160197514,
+      "learning_rate": 1.5095314734922037e-05,
+      "loss": 0.6139,
+      "step": 6091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4166865421164272,
+      "learning_rate": 1.509371342070044e-05,
+      "loss": 0.323,
+      "step": 6092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3283893206173644,
+      "learning_rate": 1.509211193008377e-05,
+      "loss": 0.2456,
+      "step": 6093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0555578416538822,
+      "learning_rate": 1.509051026312748e-05,
+      "loss": 0.6061,
+      "step": 6094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29921174955999685,
+      "learning_rate": 1.508890841988704e-05,
+      "loss": 0.2598,
+      "step": 6095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3149749288748337,
+      "learning_rate": 1.5087306400417921e-05,
+      "loss": 0.2747,
+      "step": 6096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31212173256557457,
+      "learning_rate": 1.5085704204775598e-05,
+      "loss": 0.1728,
+      "step": 6097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5652954029006605,
+      "learning_rate": 1.508410183301556e-05,
+      "loss": 0.4581,
+      "step": 6098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6856421881804132,
+      "learning_rate": 1.508249928519329e-05,
+      "loss": 0.3919,
+      "step": 6099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30656764100593287,
+      "learning_rate": 1.5080896561364293e-05,
+      "loss": 0.2557,
+      "step": 6100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4687395321208568,
+      "learning_rate": 1.5079293661584063e-05,
+      "loss": 0.3314,
+      "step": 6101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3349070250202825,
+      "learning_rate": 1.5077690585908113e-05,
+      "loss": 0.2066,
+      "step": 6102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3640182696226341,
+      "learning_rate": 1.5076087334391957e-05,
+      "loss": 0.2836,
+      "step": 6103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34746788590963384,
+      "learning_rate": 1.5074483907091115e-05,
+      "loss": 0.3276,
+      "step": 6104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3712053354906611,
+      "learning_rate": 1.5072880304061112e-05,
+      "loss": 0.3489,
+      "step": 6105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4718096602897961,
+      "learning_rate": 1.5071276525357486e-05,
+      "loss": 0.3113,
+      "step": 6106
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5120999080003616,
+      "learning_rate": 1.5069672571035766e-05,
+      "loss": 0.3747,
+      "step": 6107
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28354797155566364,
+      "learning_rate": 1.506806844115151e-05,
+      "loss": 0.2414,
+      "step": 6108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37166878535602804,
+      "learning_rate": 1.5066464135760254e-05,
+      "loss": 0.2915,
+      "step": 6109
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5029551308263018,
+      "learning_rate": 1.506485965491757e-05,
+      "loss": 0.3747,
+      "step": 6110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4002329105834612,
+      "learning_rate": 1.5063254998679009e-05,
+      "loss": 0.2825,
+      "step": 6111
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3814977933139438,
+      "learning_rate": 1.5061650167100146e-05,
+      "loss": 0.2968,
+      "step": 6112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.829273859116276,
+      "learning_rate": 1.5060045160236556e-05,
+      "loss": 0.452,
+      "step": 6113
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27937980686236547,
+      "learning_rate": 1.505843997814382e-05,
+      "loss": 0.2292,
+      "step": 6114
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47676045935760386,
+      "learning_rate": 1.5056834620877525e-05,
+      "loss": 0.312,
+      "step": 6115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3451101796926662,
+      "learning_rate": 1.5055229088493264e-05,
+      "loss": 0.2801,
+      "step": 6116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.579516426649944,
+      "learning_rate": 1.5053623381046639e-05,
+      "loss": 0.394,
+      "step": 6117
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44553554966049524,
+      "learning_rate": 1.505201749859325e-05,
+      "loss": 0.3536,
+      "step": 6118
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3590680813627751,
+      "learning_rate": 1.5050411441188714e-05,
+      "loss": 0.2677,
+      "step": 6119
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5367161392325959,
+      "learning_rate": 1.5048805208888651e-05,
+      "loss": 0.3061,
+      "step": 6120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4162731031659161,
+      "learning_rate": 1.5047198801748677e-05,
+      "loss": 0.3569,
+      "step": 6121
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2970681347926114,
+      "learning_rate": 1.5045592219824423e-05,
+      "loss": 0.2119,
+      "step": 6122
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35046893311451455,
+      "learning_rate": 1.5043985463171532e-05,
+      "loss": 0.2086,
+      "step": 6123
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.313277427937686,
+      "learning_rate": 1.5042378531845638e-05,
+      "loss": 0.3118,
+      "step": 6124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2384222887646135,
+      "learning_rate": 1.5040771425902393e-05,
+      "loss": 0.744,
+      "step": 6125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3171223595122585,
+      "learning_rate": 1.503916414539745e-05,
+      "loss": 0.2281,
+      "step": 6126
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.413448931832896,
+      "learning_rate": 1.5037556690386472e-05,
+      "loss": 0.389,
+      "step": 6127
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23777876075584098,
+      "learning_rate": 1.5035949060925118e-05,
+      "loss": 0.2254,
+      "step": 6128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3183163904203893,
+      "learning_rate": 1.5034341257069072e-05,
+      "loss": 0.2173,
+      "step": 6129
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5650146581955648,
+      "learning_rate": 1.5032733278873996e-05,
+      "loss": 0.4094,
+      "step": 6130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.513930176388067,
+      "learning_rate": 1.5031125126395589e-05,
+      "loss": 0.4267,
+      "step": 6131
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3492885607799953,
+      "learning_rate": 1.5029516799689533e-05,
+      "loss": 0.3188,
+      "step": 6132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3884188797338796,
+      "learning_rate": 1.5027908298811527e-05,
+      "loss": 0.2838,
+      "step": 6133
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28370779822743003,
+      "learning_rate": 1.5026299623817273e-05,
+      "loss": 0.2077,
+      "step": 6134
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4015496523870909,
+      "learning_rate": 1.5024690774762478e-05,
+      "loss": 0.3009,
+      "step": 6135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3578652014760878,
+      "learning_rate": 1.5023081751702857e-05,
+      "loss": 0.2493,
+      "step": 6136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7011799483411165,
+      "learning_rate": 1.5021472554694134e-05,
+      "loss": 0.46,
+      "step": 6137
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5617022132112179,
+      "learning_rate": 1.501986318379203e-05,
+      "loss": 0.4276,
+      "step": 6138
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3212717857556724,
+      "learning_rate": 1.501825363905228e-05,
+      "loss": 0.2388,
+      "step": 6139
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25259321483356195,
+      "learning_rate": 1.5016643920530625e-05,
+      "loss": 0.2136,
+      "step": 6140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7643284203192063,
+      "learning_rate": 1.5015034028282802e-05,
+      "loss": 0.4888,
+      "step": 6141
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3205694041270008,
+      "learning_rate": 1.5013423962364571e-05,
+      "loss": 0.2469,
+      "step": 6142
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47535488032774514,
+      "learning_rate": 1.501181372283168e-05,
+      "loss": 0.3845,
+      "step": 6143
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5102639689359593,
+      "learning_rate": 1.5010203309739897e-05,
+      "loss": 0.3412,
+      "step": 6144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34843918812053276,
+      "learning_rate": 1.5008592723144987e-05,
+      "loss": 0.3098,
+      "step": 6145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23040537290013943,
+      "learning_rate": 1.500698196310273e-05,
+      "loss": 0.0713,
+      "step": 6146
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4026091520264639,
+      "learning_rate": 1.5005371029668899e-05,
+      "loss": 0.2918,
+      "step": 6147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3332787714379259,
+      "learning_rate": 1.5003759922899286e-05,
+      "loss": 0.2908,
+      "step": 6148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8116793651226156,
+      "learning_rate": 1.5002148642849683e-05,
+      "loss": 0.3831,
+      "step": 6149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4649050578936953,
+      "learning_rate": 1.5000537189575885e-05,
+      "loss": 0.3492,
+      "step": 6150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.398384632333775,
+      "learning_rate": 1.4998925563133702e-05,
+      "loss": 0.3362,
+      "step": 6151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2859884604619067,
+      "learning_rate": 1.499731376357894e-05,
+      "loss": 0.1982,
+      "step": 6152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2499071609751304,
+      "learning_rate": 1.499570179096742e-05,
+      "loss": 0.8175,
+      "step": 6153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35604390551676335,
+      "learning_rate": 1.499408964535496e-05,
+      "loss": 0.2641,
+      "step": 6154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6096935274842166,
+      "learning_rate": 1.499247732679739e-05,
+      "loss": 0.3215,
+      "step": 6155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4061485551230446,
+      "learning_rate": 1.4990864835350544e-05,
+      "loss": 0.1853,
+      "step": 6156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3753388063151038,
+      "learning_rate": 1.4989252171070265e-05,
+      "loss": 0.2943,
+      "step": 6157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43330743666285093,
+      "learning_rate": 1.4987639334012398e-05,
+      "loss": 0.2916,
+      "step": 6158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3607792270908736,
+      "learning_rate": 1.4986026324232796e-05,
+      "loss": 0.2567,
+      "step": 6159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41735006093173743,
+      "learning_rate": 1.4984413141787312e-05,
+      "loss": 0.3247,
+      "step": 6160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6882683097516364,
+      "learning_rate": 1.498279978673182e-05,
+      "loss": 0.4989,
+      "step": 6161
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33450054220257863,
+      "learning_rate": 1.4981186259122185e-05,
+      "loss": 0.176,
+      "step": 6162
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3138479440645983,
+      "learning_rate": 1.4979572559014284e-05,
+      "loss": 0.2784,
+      "step": 6163
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9177725613411829,
+      "learning_rate": 1.4977958686463998e-05,
+      "loss": 0.6058,
+      "step": 6164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43862792944574786,
+      "learning_rate": 1.497634464152722e-05,
+      "loss": 0.2266,
+      "step": 6165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4316088985564484,
+      "learning_rate": 1.4974730424259836e-05,
+      "loss": 0.3325,
+      "step": 6166
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4754951018407942,
+      "learning_rate": 1.4973116034717754e-05,
+      "loss": 0.3411,
+      "step": 6167
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25015470689457353,
+      "learning_rate": 1.4971501472956875e-05,
+      "loss": 0.1052,
+      "step": 6168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46799978705707523,
+      "learning_rate": 1.4969886739033116e-05,
+      "loss": 0.3381,
+      "step": 6169
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5456881995311484,
+      "learning_rate": 1.4968271833002393e-05,
+      "loss": 0.8595,
+      "step": 6170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35870092364375566,
+      "learning_rate": 1.4966656754920635e-05,
+      "loss": 0.3482,
+      "step": 6171
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32167903780767804,
+      "learning_rate": 1.496504150484376e-05,
+      "loss": 0.2257,
+      "step": 6172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4229197497644949,
+      "learning_rate": 1.4963426082827714e-05,
+      "loss": 0.3368,
+      "step": 6173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47531000042653976,
+      "learning_rate": 1.4961810488928434e-05,
+      "loss": 0.2762,
+      "step": 6174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3133918190580812,
+      "learning_rate": 1.4960194723201873e-05,
+      "loss": 0.2267,
+      "step": 6175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.041910669070392,
+      "learning_rate": 1.4958578785703982e-05,
+      "loss": 0.6656,
+      "step": 6176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.827017268485664,
+      "learning_rate": 1.4956962676490719e-05,
+      "loss": 0.5432,
+      "step": 6177
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3650626346399905,
+      "learning_rate": 1.495534639561805e-05,
+      "loss": 0.2469,
+      "step": 6178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3928206475372948,
+      "learning_rate": 1.4953729943141952e-05,
+      "loss": 0.3242,
+      "step": 6179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.284766364444701,
+      "learning_rate": 1.49521133191184e-05,
+      "loss": 0.1743,
+      "step": 6180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3653187691143623,
+      "learning_rate": 1.4950496523603373e-05,
+      "loss": 0.2148,
+      "step": 6181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9229835999005866,
+      "learning_rate": 1.4948879556652866e-05,
+      "loss": 0.5495,
+      "step": 6182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4713117458300445,
+      "learning_rate": 1.4947262418322872e-05,
+      "loss": 0.3238,
+      "step": 6183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4509127635088528,
+      "learning_rate": 1.4945645108669395e-05,
+      "loss": 0.2793,
+      "step": 6184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4986348814970563,
+      "learning_rate": 1.4944027627748438e-05,
+      "loss": 0.3296,
+      "step": 6185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2946512585424218,
+      "learning_rate": 1.4942409975616019e-05,
+      "loss": 0.1613,
+      "step": 6186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3968498037859521,
+      "learning_rate": 1.4940792152328156e-05,
+      "loss": 0.2815,
+      "step": 6187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3634508359888088,
+      "learning_rate": 1.4939174157940872e-05,
+      "loss": 0.2695,
+      "step": 6188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6800060796414303,
+      "learning_rate": 1.4937555992510198e-05,
+      "loss": 0.4762,
+      "step": 6189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34781887939904654,
+      "learning_rate": 1.4935937656092175e-05,
+      "loss": 0.2857,
+      "step": 6190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3356382407547884,
+      "learning_rate": 1.493431914874284e-05,
+      "loss": 0.2678,
+      "step": 6191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2680360999255305,
+      "learning_rate": 1.4932700470518247e-05,
+      "loss": 0.1794,
+      "step": 6192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36281121646648407,
+      "learning_rate": 1.4931081621474448e-05,
+      "loss": 0.2694,
+      "step": 6193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7202442627048605,
+      "learning_rate": 1.4929462601667504e-05,
+      "loss": 0.3639,
+      "step": 6194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38416910639035173,
+      "learning_rate": 1.4927843411153481e-05,
+      "loss": 0.3403,
+      "step": 6195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33023100678471856,
+      "learning_rate": 1.4926224049988456e-05,
+      "loss": 0.2986,
+      "step": 6196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2815573161505327,
+      "learning_rate": 1.4924604518228503e-05,
+      "loss": 0.8653,
+      "step": 6197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23076577952264316,
+      "learning_rate": 1.4922984815929707e-05,
+      "loss": 0.1842,
+      "step": 6198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2980217129945579,
+      "learning_rate": 1.4921364943148158e-05,
+      "loss": 0.2783,
+      "step": 6199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8018563432059227,
+      "learning_rate": 1.4919744899939952e-05,
+      "loss": 0.4694,
+      "step": 6200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7658024271159686,
+      "learning_rate": 1.4918124686361193e-05,
+      "loss": 0.3918,
+      "step": 6201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40099316107463984,
+      "learning_rate": 1.4916504302467987e-05,
+      "loss": 0.2832,
+      "step": 6202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3396109439104663,
+      "learning_rate": 1.4914883748316448e-05,
+      "loss": 0.3053,
+      "step": 6203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17946159090846295,
+      "learning_rate": 1.4913263023962698e-05,
+      "loss": 0.0944,
+      "step": 6204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42650868983363494,
+      "learning_rate": 1.491164212946286e-05,
+      "loss": 0.2872,
+      "step": 6205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7748315569899414,
+      "learning_rate": 1.4910021064873066e-05,
+      "loss": 0.4013,
+      "step": 6206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33464339880757243,
+      "learning_rate": 1.4908399830249454e-05,
+      "loss": 0.2786,
+      "step": 6207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32860205040007584,
+      "learning_rate": 1.4906778425648165e-05,
+      "loss": 0.2614,
+      "step": 6208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1586441525453701,
+      "learning_rate": 1.4905156851125354e-05,
+      "loss": 0.7866,
+      "step": 6209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46748475889837865,
+      "learning_rate": 1.4903535106737166e-05,
+      "loss": 0.2579,
+      "step": 6210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26132463339166095,
+      "learning_rate": 1.4901913192539773e-05,
+      "loss": 0.2383,
+      "step": 6211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46282006638803863,
+      "learning_rate": 1.4900291108589335e-05,
+      "loss": 0.3461,
+      "step": 6212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4299229249539138,
+      "learning_rate": 1.4898668854942029e-05,
+      "loss": 0.8093,
+      "step": 6213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3654395694243854,
+      "learning_rate": 1.4897046431654028e-05,
+      "loss": 0.2173,
+      "step": 6214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40808297868803545,
+      "learning_rate": 1.4895423838781523e-05,
+      "loss": 0.3156,
+      "step": 6215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8572774804511525,
+      "learning_rate": 1.4893801076380697e-05,
+      "loss": 0.6124,
+      "step": 6216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3163807117672398,
+      "learning_rate": 1.4892178144507754e-05,
+      "loss": 0.2505,
+      "step": 6217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39502399231456026,
+      "learning_rate": 1.4890555043218888e-05,
+      "loss": 0.3094,
+      "step": 6218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.305393097707522,
+      "learning_rate": 1.4888931772570314e-05,
+      "loss": 0.26,
+      "step": 6219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4573475138697167,
+      "learning_rate": 1.4887308332618245e-05,
+      "loss": 0.2775,
+      "step": 6220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5099326579508142,
+      "learning_rate": 1.4885684723418897e-05,
+      "loss": 0.2951,
+      "step": 6221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3622752534920921,
+      "learning_rate": 1.4884060945028495e-05,
+      "loss": 0.3238,
+      "step": 6222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4769109545150248,
+      "learning_rate": 1.4882436997503273e-05,
+      "loss": 0.3067,
+      "step": 6223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31122680816110204,
+      "learning_rate": 1.4880812880899472e-05,
+      "loss": 0.2261,
+      "step": 6224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44219289230671227,
+      "learning_rate": 1.4879188595273326e-05,
+      "loss": 0.3029,
+      "step": 6225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.566041714859357,
+      "learning_rate": 1.487756414068109e-05,
+      "loss": 0.334,
+      "step": 6226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3654004821252147,
+      "learning_rate": 1.4875939517179016e-05,
+      "loss": 0.2639,
+      "step": 6227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.755425769942834,
+      "learning_rate": 1.4874314724823368e-05,
+      "loss": 0.5034,
+      "step": 6228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3716823951023819,
+      "learning_rate": 1.487268976367041e-05,
+      "loss": 0.3237,
+      "step": 6229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2941511105541551,
+      "learning_rate": 1.4871064633776418e-05,
+      "loss": 0.2806,
+      "step": 6230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31363827129766925,
+      "learning_rate": 1.4869439335197661e-05,
+      "loss": 0.1919,
+      "step": 6231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38920044645943885,
+      "learning_rate": 1.4867813867990435e-05,
+      "loss": 0.2882,
+      "step": 6232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5297592578525103,
+      "learning_rate": 1.486618823221102e-05,
+      "loss": 0.4331,
+      "step": 6233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34389016761519975,
+      "learning_rate": 1.4864562427915722e-05,
+      "loss": 0.2951,
+      "step": 6234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4098129452025419,
+      "learning_rate": 1.486293645516083e-05,
+      "loss": 0.3112,
+      "step": 6235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5460341175814037,
+      "learning_rate": 1.4861310314002659e-05,
+      "loss": 0.339,
+      "step": 6236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24296990938486865,
+      "learning_rate": 1.485968400449752e-05,
+      "loss": 0.1272,
+      "step": 6237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4041372596227764,
+      "learning_rate": 1.485805752670174e-05,
+      "loss": 0.2752,
+      "step": 6238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31120053524746066,
+      "learning_rate": 1.4856430880671628e-05,
+      "loss": 0.3153,
+      "step": 6239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5829267478852111,
+      "learning_rate": 1.485480406646353e-05,
+      "loss": 0.3675,
+      "step": 6240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5853392794597301,
+      "learning_rate": 1.485317708413377e-05,
+      "loss": 0.2736,
+      "step": 6241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3633480158282146,
+      "learning_rate": 1.48515499337387e-05,
+      "loss": 0.2897,
+      "step": 6242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3294481476064836,
+      "learning_rate": 1.4849922615334662e-05,
+      "loss": 0.2107,
+      "step": 6243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.596863758511093,
+      "learning_rate": 1.4848295128978016e-05,
+      "loss": 0.3861,
+      "step": 6244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30007973784979886,
+      "learning_rate": 1.4846667474725115e-05,
+      "loss": 0.2746,
+      "step": 6245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4615792309021516,
+      "learning_rate": 1.484503965263233e-05,
+      "loss": 0.3483,
+      "step": 6246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.35022336479364224,
+      "learning_rate": 1.4843411662756028e-05,
+      "loss": 0.2043,
+      "step": 6247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4208558050427091,
+      "learning_rate": 1.484178350515259e-05,
+      "loss": 0.3449,
+      "step": 6248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.038883513719484,
+      "learning_rate": 1.4840155179878398e-05,
+      "loss": 0.7463,
+      "step": 6249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2896134927667209,
+      "learning_rate": 1.4838526686989836e-05,
+      "loss": 0.2265,
+      "step": 6250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2974585469109081,
+      "learning_rate": 1.4836898026543307e-05,
+      "loss": 0.2787,
+      "step": 6251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4006467030665213,
+      "learning_rate": 1.4835269198595206e-05,
+      "loss": 0.3279,
+      "step": 6252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33309182783458663,
+      "learning_rate": 1.483364020320194e-05,
+      "loss": 0.2246,
+      "step": 6253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36433501957512965,
+      "learning_rate": 1.4832011040419922e-05,
+      "loss": 0.3392,
+      "step": 6254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5059114706018922,
+      "learning_rate": 1.4830381710305572e-05,
+      "loss": 0.4148,
+      "step": 6255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4326822693331583,
+      "learning_rate": 1.4828752212915309e-05,
+      "loss": 0.2464,
+      "step": 6256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.346803304093782,
+      "learning_rate": 1.4827122548305566e-05,
+      "loss": 0.3429,
+      "step": 6257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24676986871220055,
+      "learning_rate": 1.4825492716532772e-05,
+      "loss": 0.2387,
+      "step": 6258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8936383019306559,
+      "learning_rate": 1.4823862717653377e-05,
+      "loss": 0.4183,
+      "step": 6259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3613517629923259,
+      "learning_rate": 1.4822232551723824e-05,
+      "loss": 0.2395,
+      "step": 6260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7576742857770312,
+      "learning_rate": 1.4820602218800562e-05,
+      "loss": 0.604,
+      "step": 6261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36124256710193003,
+      "learning_rate": 1.4818971718940053e-05,
+      "loss": 0.282,
+      "step": 6262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3117081152835165,
+      "learning_rate": 1.4817341052198763e-05,
+      "loss": 0.2732,
+      "step": 6263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4077993695451762,
+      "learning_rate": 1.481571021863316e-05,
+      "loss": 0.2519,
+      "step": 6264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36274755930196695,
+      "learning_rate": 1.4814079218299715e-05,
+      "loss": 0.2588,
+      "step": 6265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31755716594371836,
+      "learning_rate": 1.4812448051254914e-05,
+      "loss": 0.2515,
+      "step": 6266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2379686903604816,
+      "learning_rate": 1.4810816717555248e-05,
+      "loss": 0.8207,
+      "step": 6267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6200789147881792,
+      "learning_rate": 1.4809185217257205e-05,
+      "loss": 0.4401,
+      "step": 6268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3576042390007011,
+      "learning_rate": 1.4807553550417281e-05,
+      "loss": 0.236,
+      "step": 6269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3135412551606865,
+      "learning_rate": 1.4805921717091989e-05,
+      "loss": 0.2926,
+      "step": 6270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33504011374279746,
+      "learning_rate": 1.480428971733783e-05,
+      "loss": 0.2396,
+      "step": 6271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.395484364592832,
+      "learning_rate": 1.4802657551211331e-05,
+      "loss": 0.2966,
+      "step": 6272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7334203945138051,
+      "learning_rate": 1.4801025218769001e-05,
+      "loss": 0.3673,
+      "step": 6273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3095981068865218,
+      "learning_rate": 1.4799392720067378e-05,
+      "loss": 0.2805,
+      "step": 6274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37500869393675135,
+      "learning_rate": 1.4797760055162988e-05,
+      "loss": 0.3432,
+      "step": 6275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21332335884529122,
+      "learning_rate": 1.4796127224112378e-05,
+      "loss": 0.1254,
+      "step": 6276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1738522008194672,
+      "learning_rate": 1.4794494226972088e-05,
+      "loss": 0.7157,
+      "step": 6277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2622768681089619,
+      "learning_rate": 1.4792861063798664e-05,
+      "loss": 0.2618,
+      "step": 6278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5248605026141624,
+      "learning_rate": 1.4791227734648672e-05,
+      "loss": 0.3278,
+      "step": 6279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5791380522386198,
+      "learning_rate": 1.4789594239578668e-05,
+      "loss": 0.3969,
+      "step": 6280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40689169491954263,
+      "learning_rate": 1.4787960578645222e-05,
+      "loss": 0.2991,
+      "step": 6281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.35430059420501625,
+      "learning_rate": 1.4786326751904907e-05,
+      "loss": 0.2538,
+      "step": 6282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25326613938208253,
+      "learning_rate": 1.4784692759414303e-05,
+      "loss": 0.1784,
+      "step": 6283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3879921231259588,
+      "learning_rate": 1.4783058601229994e-05,
+      "loss": 0.3079,
+      "step": 6284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7226592658190863,
+      "learning_rate": 1.4781424277408572e-05,
+      "loss": 0.4558,
+      "step": 6285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30693828632115694,
+      "learning_rate": 1.4779789788006632e-05,
+      "loss": 0.2734,
+      "step": 6286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3880005622421795,
+      "learning_rate": 1.4778155133080776e-05,
+      "loss": 0.2958,
+      "step": 6287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1057367174861266,
+      "learning_rate": 1.4776520312687614e-05,
+      "loss": 0.6752,
+      "step": 6288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21447372968479209,
+      "learning_rate": 1.477488532688376e-05,
+      "loss": 0.0726,
+      "step": 6289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.35395443621688044,
+      "learning_rate": 1.4773250175725833e-05,
+      "loss": 0.3048,
+      "step": 6290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5289778545984578,
+      "learning_rate": 1.4771614859270458e-05,
+      "loss": 0.3525,
+      "step": 6291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48440354407701286,
+      "learning_rate": 1.4769979377574264e-05,
+      "loss": 0.2645,
+      "step": 6292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4060500797659063,
+      "learning_rate": 1.4768343730693888e-05,
+      "loss": 0.3174,
+      "step": 6293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3819313122256466,
+      "learning_rate": 1.4766707918685974e-05,
+      "loss": 0.3538,
+      "step": 6294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3470332126651621,
+      "learning_rate": 1.4765071941607172e-05,
+      "loss": 0.2331,
+      "step": 6295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.306715059456966,
+      "learning_rate": 1.4763435799514132e-05,
+      "loss": 0.2315,
+      "step": 6296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5344701290774023,
+      "learning_rate": 1.4761799492463516e-05,
+      "loss": 0.381,
+      "step": 6297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45486931737761455,
+      "learning_rate": 1.4760163020511986e-05,
+      "loss": 0.3881,
+      "step": 6298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31770712221775865,
+      "learning_rate": 1.4758526383716219e-05,
+      "loss": 0.2241,
+      "step": 6299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1779532369599066,
+      "learning_rate": 1.4756889582132886e-05,
+      "loss": 0.7267,
+      "step": 6300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25416641771855647,
+      "learning_rate": 1.4755252615818671e-05,
+      "loss": 0.2218,
+      "step": 6301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28196065175567037,
+      "learning_rate": 1.4753615484830261e-05,
+      "loss": 0.2354,
+      "step": 6302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6902396292135428,
+      "learning_rate": 1.4751978189224354e-05,
+      "loss": 0.4771,
+      "step": 6303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0221054595371608,
+      "learning_rate": 1.4750340729057646e-05,
+      "loss": 0.728,
+      "step": 6304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3009926847405351,
+      "learning_rate": 1.4748703104386843e-05,
+      "loss": 0.215,
+      "step": 6305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32641194854409294,
+      "learning_rate": 1.4747065315268655e-05,
+      "loss": 0.2941,
+      "step": 6306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4241970288360463,
+      "learning_rate": 1.4745427361759801e-05,
+      "loss": 0.3376,
+      "step": 6307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45701611260847547,
+      "learning_rate": 1.4743789243916999e-05,
+      "loss": 0.2903,
+      "step": 6308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2999184950849396,
+      "learning_rate": 1.4742150961796981e-05,
+      "loss": 0.2455,
+      "step": 6309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45425353186377854,
+      "learning_rate": 1.4740512515456479e-05,
+      "loss": 0.3645,
+      "step": 6310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.374441697495103,
+      "learning_rate": 1.4738873904952232e-05,
+      "loss": 0.2927,
+      "step": 6311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8017812489117565,
+      "learning_rate": 1.4737235130340985e-05,
+      "loss": 0.4057,
+      "step": 6312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4483513029265171,
+      "learning_rate": 1.473559619167949e-05,
+      "loss": 0.3609,
+      "step": 6313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30337256483409936,
+      "learning_rate": 1.4733957089024502e-05,
+      "loss": 0.2775,
+      "step": 6314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27788517511323146,
+      "learning_rate": 1.4732317822432782e-05,
+      "loss": 0.1849,
+      "step": 6315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2877189024995006,
+      "learning_rate": 1.4730678391961102e-05,
+      "loss": 0.7865,
+      "step": 6316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3273282971439404,
+      "learning_rate": 1.472903879766623e-05,
+      "loss": 0.2822,
+      "step": 6317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36865587604384104,
+      "learning_rate": 1.4727399039604951e-05,
+      "loss": 0.2766,
+      "step": 6318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6217638120111219,
+      "learning_rate": 1.4725759117834045e-05,
+      "loss": 0.4275,
+      "step": 6319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34169170317511,
+      "learning_rate": 1.4724119032410305e-05,
+      "loss": 0.3136,
+      "step": 6320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30920422479643384,
+      "learning_rate": 1.4722478783390522e-05,
+      "loss": 0.1771,
+      "step": 6321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34775308402691785,
+      "learning_rate": 1.472083837083151e-05,
+      "loss": 0.2888,
+      "step": 6322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3193141427715202,
+      "learning_rate": 1.471919779479006e-05,
+      "loss": 0.256,
+      "step": 6323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1400092958513408,
+      "learning_rate": 1.4717557055322997e-05,
+      "loss": 0.7106,
+      "step": 6324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40323223147712445,
+      "learning_rate": 1.4715916152487135e-05,
+      "loss": 0.3131,
+      "step": 6325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4250005070452697,
+      "learning_rate": 1.47142750863393e-05,
+      "loss": 0.2852,
+      "step": 6326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3343011248837186,
+      "learning_rate": 1.471263385693632e-05,
+      "loss": 0.2307,
+      "step": 6327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3055552288905154,
+      "learning_rate": 1.4710992464335034e-05,
+      "loss": 0.1793,
+      "step": 6328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40306747549933275,
+      "learning_rate": 1.4709350908592281e-05,
+      "loss": 0.2806,
+      "step": 6329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32815408215415465,
+      "learning_rate": 1.4707709189764909e-05,
+      "loss": 0.3094,
+      "step": 6330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6956160911594513,
+      "learning_rate": 1.470606730790977e-05,
+      "loss": 0.356,
+      "step": 6331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3454136921848711,
+      "learning_rate": 1.4704425263083722e-05,
+      "loss": 0.2861,
+      "step": 6332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30035263482807145,
+      "learning_rate": 1.470278305534363e-05,
+      "loss": 0.2579,
+      "step": 6333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4375086593799831,
+      "learning_rate": 1.4701140684746363e-05,
+      "loss": 0.3229,
+      "step": 6334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32172857440332125,
+      "learning_rate": 1.4699498151348797e-05,
+      "loss": 0.2142,
+      "step": 6335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6809092715571192,
+      "learning_rate": 1.469785545520781e-05,
+      "loss": 0.4272,
+      "step": 6336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3487777939185845,
+      "learning_rate": 1.469621259638029e-05,
+      "loss": 0.3407,
+      "step": 6337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29453832274472436,
+      "learning_rate": 1.4694569574923132e-05,
+      "loss": 0.2134,
+      "step": 6338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8308626976224194,
+      "learning_rate": 1.4692926390893234e-05,
+      "loss": 0.5887,
+      "step": 6339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39159026271164604,
+      "learning_rate": 1.469128304434749e-05,
+      "loss": 0.3049,
+      "step": 6340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24883858751617285,
+      "learning_rate": 1.4689639535342823e-05,
+      "loss": 0.1906,
+      "step": 6341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5107631135808215,
+      "learning_rate": 1.4687995863936135e-05,
+      "loss": 0.3797,
+      "step": 6342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6558962140201424,
+      "learning_rate": 1.4686352030184354e-05,
+      "loss": 0.4598,
+      "step": 6343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32642046172243194,
+      "learning_rate": 1.4684708034144403e-05,
+      "loss": 0.174,
+      "step": 6344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3823814532754202,
+      "learning_rate": 1.4683063875873215e-05,
+      "loss": 0.3208,
+      "step": 6345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4802433585379294,
+      "learning_rate": 1.4681419555427727e-05,
+      "loss": 0.4175,
+      "step": 6346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.521892647783289,
+      "learning_rate": 1.467977507286488e-05,
+      "loss": 0.3415,
+      "step": 6347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3458481345611792,
+      "learning_rate": 1.4678130428241623e-05,
+      "loss": 0.2438,
+      "step": 6348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3032108143688598,
+      "learning_rate": 1.4676485621614913e-05,
+      "loss": 0.2495,
+      "step": 6349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5573975802181028,
+      "learning_rate": 1.4674840653041706e-05,
+      "loss": 0.3531,
+      "step": 6350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37910874435426417,
+      "learning_rate": 1.4673195522578967e-05,
+      "loss": 0.2586,
+      "step": 6351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1453819020562075,
+      "learning_rate": 1.467155023028367e-05,
+      "loss": 0.8029,
+      "step": 6352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3225394768457753,
+      "learning_rate": 1.4669904776212786e-05,
+      "loss": 0.301,
+      "step": 6353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4135206494047781,
+      "learning_rate": 1.4668259160423305e-05,
+      "loss": 0.2995,
+      "step": 6354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30632042423602535,
+      "learning_rate": 1.4666613382972205e-05,
+      "loss": 0.2069,
+      "step": 6355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38142377744009864,
+      "learning_rate": 1.4664967443916489e-05,
+      "loss": 0.3089,
+      "step": 6356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3386038426663132,
+      "learning_rate": 1.4663321343313148e-05,
+      "loss": 0.2622,
+      "step": 6357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45717911991857507,
+      "learning_rate": 1.4661675081219191e-05,
+      "loss": 0.3994,
+      "step": 6358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39907584698503695,
+      "learning_rate": 1.4660028657691626e-05,
+      "loss": 0.3111,
+      "step": 6359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4939593127745104,
+      "learning_rate": 1.465838207278747e-05,
+      "loss": 0.3689,
+      "step": 6360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2352579898444018,
+      "learning_rate": 1.4656735326563738e-05,
+      "loss": 0.1747,
+      "step": 6361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.59732590291894,
+      "learning_rate": 1.4655088419077466e-05,
+      "loss": 0.3185,
+      "step": 6362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39752606248880057,
+      "learning_rate": 1.4653441350385682e-05,
+      "loss": 0.3149,
+      "step": 6363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4368499204854432,
+      "learning_rate": 1.4651794120545424e-05,
+      "loss": 0.3275,
+      "step": 6364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3946428941104599,
+      "learning_rate": 1.4650146729613735e-05,
+      "loss": 0.311,
+      "step": 6365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4052584218015384,
+      "learning_rate": 1.4648499177647665e-05,
+      "loss": 0.3519,
+      "step": 6366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2228152253683647,
+      "learning_rate": 1.4646851464704269e-05,
+      "loss": 0.0727,
+      "step": 6367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55650450715017,
+      "learning_rate": 1.4645203590840607e-05,
+      "loss": 0.3699,
+      "step": 6368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2782351877624767,
+      "learning_rate": 1.4643555556113742e-05,
+      "loss": 0.2925,
+      "step": 6369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6856577889773536,
+      "learning_rate": 1.4641907360580749e-05,
+      "loss": 0.3796,
+      "step": 6370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49911463412678736,
+      "learning_rate": 1.4640259004298706e-05,
+      "loss": 0.3838,
+      "step": 6371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3727450596138779,
+      "learning_rate": 1.4638610487324688e-05,
+      "loss": 0.333,
+      "step": 6372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2839276525223571,
+      "learning_rate": 1.4636961809715793e-05,
+      "loss": 0.2624,
+      "step": 6373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24988639619999695,
+      "learning_rate": 1.463531297152911e-05,
+      "loss": 0.1668,
+      "step": 6374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5883872688338556,
+      "learning_rate": 1.4633663972821737e-05,
+      "loss": 0.4016,
+      "step": 6375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.969636159928934,
+      "learning_rate": 1.4632014813650779e-05,
+      "loss": 0.4825,
+      "step": 6376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27400554273030164,
+      "learning_rate": 1.4630365494073348e-05,
+      "loss": 0.2349,
+      "step": 6377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5706865974074818,
+      "learning_rate": 1.4628716014146558e-05,
+      "loss": 0.4436,
+      "step": 6378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28185065914961815,
+      "learning_rate": 1.4627066373927534e-05,
+      "loss": 0.1703,
+      "step": 6379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5918352536787485,
+      "learning_rate": 1.4625416573473397e-05,
+      "loss": 0.2241,
+      "step": 6380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2945602690656717,
+      "learning_rate": 1.4623766612841286e-05,
+      "loss": 0.2792,
+      "step": 6381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.245141698601791,
+      "learning_rate": 1.4622116492088335e-05,
+      "loss": 0.4757,
+      "step": 6382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7575767671387406,
+      "learning_rate": 1.4620466211271686e-05,
+      "loss": 0.4512,
+      "step": 6383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42239162484195003,
+      "learning_rate": 1.4618815770448493e-05,
+      "loss": 0.2349,
+      "step": 6384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28014002852446207,
+      "learning_rate": 1.461716516967591e-05,
+      "loss": 0.2347,
+      "step": 6385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.792038715756322,
+      "learning_rate": 1.4615514409011093e-05,
+      "loss": 0.4452,
+      "step": 6386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7042697472461529,
+      "learning_rate": 1.4613863488511214e-05,
+      "loss": 0.2951,
+      "step": 6387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.090984118378016,
+      "learning_rate": 1.4612212408233438e-05,
+      "loss": 0.5841,
+      "step": 6388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3318163237944138,
+      "learning_rate": 1.4610561168234942e-05,
+      "loss": 0.3015,
+      "step": 6389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36470561968264775,
+      "learning_rate": 1.4608909768572917e-05,
+      "loss": 0.2485,
+      "step": 6390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5483411965792786,
+      "learning_rate": 1.4607258209304542e-05,
+      "loss": 0.2347,
+      "step": 6391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4024074470010744,
+      "learning_rate": 1.4605606490487013e-05,
+      "loss": 0.2603,
+      "step": 6392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3917554003411267,
+      "learning_rate": 1.4603954612177532e-05,
+      "loss": 0.231,
+      "step": 6393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2772938766587616,
+      "learning_rate": 1.4602302574433297e-05,
+      "loss": 0.486,
+      "step": 6394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6668301917465416,
+      "learning_rate": 1.4600650377311523e-05,
+      "loss": 0.4474,
+      "step": 6395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39712967225015333,
+      "learning_rate": 1.4598998020869426e-05,
+      "loss": 0.3148,
+      "step": 6396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28444972237384936,
+      "learning_rate": 1.4597345505164222e-05,
+      "loss": 0.218,
+      "step": 6397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5753183483183651,
+      "learning_rate": 1.4595692830253143e-05,
+      "loss": 0.2875,
+      "step": 6398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5209561480298429,
+      "learning_rate": 1.459403999619342e-05,
+      "loss": 0.3033,
+      "step": 6399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47041709352127614,
+      "learning_rate": 1.4592387003042287e-05,
+      "loss": 0.2843,
+      "step": 6400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4217366843048336,
+      "learning_rate": 1.4590733850856989e-05,
+      "loss": 0.7602,
+      "step": 6401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3419329387584038,
+      "learning_rate": 1.4589080539694778e-05,
+      "loss": 0.2929,
+      "step": 6402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6029781463091936,
+      "learning_rate": 1.4587427069612902e-05,
+      "loss": 0.3747,
+      "step": 6403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3860877146598658,
+      "learning_rate": 1.4585773440668626e-05,
+      "loss": 0.3044,
+      "step": 6404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3332149808196481,
+      "learning_rate": 1.4584119652919213e-05,
+      "loss": 0.2862,
+      "step": 6405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2517107564009192,
+      "learning_rate": 1.4582465706421935e-05,
+      "loss": 0.1132,
+      "step": 6406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9180317484308079,
+      "learning_rate": 1.4580811601234067e-05,
+      "loss": 0.55,
+      "step": 6407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4199511381911743,
+      "learning_rate": 1.4579157337412886e-05,
+      "loss": 0.2734,
+      "step": 6408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40497014891679506,
+      "learning_rate": 1.4577502915015687e-05,
+      "loss": 0.3179,
+      "step": 6409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7870272373752119,
+      "learning_rate": 1.4575848334099756e-05,
+      "loss": 0.3261,
+      "step": 6410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30582763377378736,
+      "learning_rate": 1.4574193594722394e-05,
+      "loss": 0.256,
+      "step": 6411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3369040954406996,
+      "learning_rate": 1.4572538696940908e-05,
+      "loss": 0.2748,
+      "step": 6412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36361392670595094,
+      "learning_rate": 1.4570883640812602e-05,
+      "loss": 0.2604,
+      "step": 6413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3875128108510935,
+      "learning_rate": 1.456922842639479e-05,
+      "loss": 0.2852,
+      "step": 6414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8988967262988671,
+      "learning_rate": 1.4567573053744797e-05,
+      "loss": 0.4492,
+      "step": 6415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40165688387675075,
+      "learning_rate": 1.4565917522919944e-05,
+      "loss": 0.2684,
+      "step": 6416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38594474836512377,
+      "learning_rate": 1.4564261833977563e-05,
+      "loss": 0.3048,
+      "step": 6417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2507980036522095,
+      "learning_rate": 1.4562605986974991e-05,
+      "loss": 0.194,
+      "step": 6418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.872875765589793,
+      "learning_rate": 1.456094998196957e-05,
+      "loss": 0.4096,
+      "step": 6419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.358181410493378,
+      "learning_rate": 1.4559293819018648e-05,
+      "loss": 0.2836,
+      "step": 6420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45407158678411264,
+      "learning_rate": 1.4557637498179577e-05,
+      "loss": 0.346,
+      "step": 6421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9255722275983588,
+      "learning_rate": 1.4555981019509716e-05,
+      "loss": 0.5238,
+      "step": 6422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29691564293618045,
+      "learning_rate": 1.4554324383066427e-05,
+      "loss": 0.2093,
+      "step": 6423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4785704057795955,
+      "learning_rate": 1.4552667588907082e-05,
+      "loss": 0.3371,
+      "step": 6424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2952211128804953,
+      "learning_rate": 1.4551010637089054e-05,
+      "loss": 0.2773,
+      "step": 6425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36035741477242134,
+      "learning_rate": 1.454935352766972e-05,
+      "loss": 0.2268,
+      "step": 6426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8603431400721523,
+      "learning_rate": 1.4547696260706474e-05,
+      "loss": 0.4749,
+      "step": 6427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45518616624458275,
+      "learning_rate": 1.4546038836256698e-05,
+      "loss": 0.3162,
+      "step": 6428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32397378870781374,
+      "learning_rate": 1.4544381254377794e-05,
+      "loss": 0.2011,
+      "step": 6429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48798619337408167,
+      "learning_rate": 1.4542723515127161e-05,
+      "loss": 0.2987,
+      "step": 6430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46757065365840805,
+      "learning_rate": 1.4541065618562211e-05,
+      "loss": 0.301,
+      "step": 6431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39111537126054996,
+      "learning_rate": 1.4539407564740353e-05,
+      "loss": 0.2013,
+      "step": 6432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4772229613888993,
+      "learning_rate": 1.4537749353719006e-05,
+      "loss": 0.3447,
+      "step": 6433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7128445538916925,
+      "learning_rate": 1.4536090985555595e-05,
+      "loss": 0.4374,
+      "step": 6434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37996740838238047,
+      "learning_rate": 1.4534432460307546e-05,
+      "loss": 0.284,
+      "step": 6435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3521223528144348,
+      "learning_rate": 1.4532773778032297e-05,
+      "loss": 0.2757,
+      "step": 6436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4727501869932283,
+      "learning_rate": 1.4531114938787285e-05,
+      "loss": 0.3093,
+      "step": 6437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34669175481499864,
+      "learning_rate": 1.452945594262996e-05,
+      "loss": 0.2832,
+      "step": 6438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3986170161984597,
+      "learning_rate": 1.452779678961777e-05,
+      "loss": 0.2385,
+      "step": 6439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37509714723728876,
+      "learning_rate": 1.4526137479808173e-05,
+      "loss": 0.3058,
+      "step": 6440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3784935823336974,
+      "learning_rate": 1.452447801325863e-05,
+      "loss": 0.2805,
+      "step": 6441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9648337594394374,
+      "learning_rate": 1.4522818390026605e-05,
+      "loss": 0.4174,
+      "step": 6442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36332205508671817,
+      "learning_rate": 1.4521158610169575e-05,
+      "loss": 0.3118,
+      "step": 6443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3043988410895945,
+      "learning_rate": 1.4519498673745019e-05,
+      "loss": 0.2756,
+      "step": 6444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3661883748618258,
+      "learning_rate": 1.4517838580810413e-05,
+      "loss": 0.2444,
+      "step": 6445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8097337394192086,
+      "learning_rate": 1.4516178331423257e-05,
+      "loss": 0.576,
+      "step": 6446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33163644933954645,
+      "learning_rate": 1.4514517925641035e-05,
+      "loss": 0.3033,
+      "step": 6447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5218645713887373,
+      "learning_rate": 1.4512857363521254e-05,
+      "loss": 0.393,
+      "step": 6448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3793598909180065,
+      "learning_rate": 1.4511196645121414e-05,
+      "loss": 0.319,
+      "step": 6449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3896431397236735,
+      "learning_rate": 1.450953577049903e-05,
+      "loss": 0.2771,
+      "step": 6450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2705214298693227,
+      "learning_rate": 1.4507874739711616e-05,
+      "loss": 0.2298,
+      "step": 6451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34851249948794466,
+      "learning_rate": 1.4506213552816694e-05,
+      "loss": 0.2555,
+      "step": 6452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39110676315148374,
+      "learning_rate": 1.4504552209871791e-05,
+      "loss": 0.2659,
+      "step": 6453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5909154208429732,
+      "learning_rate": 1.4502890710934438e-05,
+      "loss": 0.4545,
+      "step": 6454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8661006074737951,
+      "learning_rate": 1.4501229056062174e-05,
+      "loss": 0.36,
+      "step": 6455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2980737565737271,
+      "learning_rate": 1.449956724531254e-05,
+      "loss": 0.2735,
+      "step": 6456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37860627191593116,
+      "learning_rate": 1.4497905278743086e-05,
+      "loss": 0.3064,
+      "step": 6457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4986638746583317,
+      "learning_rate": 1.4496243156411367e-05,
+      "loss": 0.3235,
+      "step": 6458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36881140079123487,
+      "learning_rate": 1.4494580878374942e-05,
+      "loss": 0.2306,
+      "step": 6459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4396370739515429,
+      "learning_rate": 1.4492918444691371e-05,
+      "loss": 0.3544,
+      "step": 6460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4881242453846941,
+      "learning_rate": 1.4491255855418228e-05,
+      "loss": 0.4526,
+      "step": 6461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.31509315415532924,
+      "learning_rate": 1.4489593110613087e-05,
+      "loss": 0.2344,
+      "step": 6462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4766555392210357,
+      "learning_rate": 1.4487930210333532e-05,
+      "loss": 0.29,
+      "step": 6463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3242238172436043,
+      "learning_rate": 1.4486267154637146e-05,
+      "loss": 0.2684,
+      "step": 6464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38454949832911445,
+      "learning_rate": 1.4484603943581523e-05,
+      "loss": 0.1761,
+      "step": 6465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9162557910018109,
+      "learning_rate": 1.4482940577224254e-05,
+      "loss": 0.4149,
+      "step": 6466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9944619174618472,
+      "learning_rate": 1.4481277055622948e-05,
+      "loss": 0.3897,
+      "step": 6467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3133210801429484,
+      "learning_rate": 1.4479613378835211e-05,
+      "loss": 0.2176,
+      "step": 6468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49623665028789815,
+      "learning_rate": 1.4477949546918655e-05,
+      "loss": 0.4053,
+      "step": 6469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28227034323549527,
+      "learning_rate": 1.4476285559930899e-05,
+      "loss": 0.1809,
+      "step": 6470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5943690582874419,
+      "learning_rate": 1.4474621417929566e-05,
+      "loss": 0.2999,
+      "step": 6471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4368804350447025,
+      "learning_rate": 1.4472957120972284e-05,
+      "loss": 0.2854,
+      "step": 6472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0839653605743804,
+      "learning_rate": 1.4471292669116692e-05,
+      "loss": 0.4932,
+      "step": 6473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6103092418262205,
+      "learning_rate": 1.4469628062420427e-05,
+      "loss": 0.3787,
+      "step": 6474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2815049852060284,
+      "learning_rate": 1.4467963300941135e-05,
+      "loss": 0.1699,
+      "step": 6475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42956074118826143,
+      "learning_rate": 1.4466298384736463e-05,
+      "loss": 0.32,
+      "step": 6476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42963144754338745,
+      "learning_rate": 1.4464633313864073e-05,
+      "loss": 0.2869,
+      "step": 6477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5424138447706254,
+      "learning_rate": 1.4462968088381621e-05,
+      "loss": 0.327,
+      "step": 6478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5458547157279315,
+      "learning_rate": 1.4461302708346778e-05,
+      "loss": 0.3884,
+      "step": 6479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32194605311238567,
+      "learning_rate": 1.4459637173817214e-05,
+      "loss": 0.2684,
+      "step": 6480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4947001893698327,
+      "learning_rate": 1.4457971484850604e-05,
+      "loss": 0.2686,
+      "step": 6481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30792978417290967,
+      "learning_rate": 1.4456305641504636e-05,
+      "loss": 0.1929,
+      "step": 6482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39304218999952134,
+      "learning_rate": 1.4454639643836993e-05,
+      "loss": 0.3088,
+      "step": 6483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.350822152919814,
+      "learning_rate": 1.4452973491905372e-05,
+      "loss": 0.3195,
+      "step": 6484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7457583251157986,
+      "learning_rate": 1.4451307185767469e-05,
+      "loss": 0.4278,
+      "step": 6485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9576714232110599,
+      "learning_rate": 1.4449640725480991e-05,
+      "loss": 0.471,
+      "step": 6486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38426895057800664,
+      "learning_rate": 1.4447974111103645e-05,
+      "loss": 0.285,
+      "step": 6487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22401458190615092,
+      "learning_rate": 1.4446307342693149e-05,
+      "loss": 0.2138,
+      "step": 6488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5444407977701945,
+      "learning_rate": 1.4444640420307217e-05,
+      "loss": 0.319,
+      "step": 6489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4500081741351894,
+      "learning_rate": 1.4442973344003582e-05,
+      "loss": 0.3505,
+      "step": 6490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0057233925466538,
+      "learning_rate": 1.4441306113839971e-05,
+      "loss": 0.5737,
+      "step": 6491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34070622377314463,
+      "learning_rate": 1.4439638729874119e-05,
+      "loss": 0.2789,
+      "step": 6492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39274503048589104,
+      "learning_rate": 1.4437971192163768e-05,
+      "loss": 0.3363,
+      "step": 6493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.260970503088254,
+      "learning_rate": 1.4436303500766667e-05,
+      "loss": 0.139,
+      "step": 6494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4353457494525623,
+      "learning_rate": 1.4434635655740566e-05,
+      "loss": 0.3152,
+      "step": 6495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36338683442474146,
+      "learning_rate": 1.4432967657143223e-05,
+      "loss": 0.3243,
+      "step": 6496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4910761065746299,
+      "learning_rate": 1.44312995050324e-05,
+      "loss": 0.8249,
+      "step": 6497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3636983831741691,
+      "learning_rate": 1.4429631199465866e-05,
+      "loss": 0.2165,
+      "step": 6498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6184276802434224,
+      "learning_rate": 1.4427962740501396e-05,
+      "loss": 0.4225,
+      "step": 6499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4139216454007609,
+      "learning_rate": 1.4426294128196763e-05,
+      "loss": 0.3305,
+      "step": 6500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49001496573354486,
+      "learning_rate": 1.4424625362609757e-05,
+      "loss": 0.2682,
+      "step": 6501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27947362207932114,
+      "learning_rate": 1.4422956443798165e-05,
+      "loss": 0.2087,
+      "step": 6502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5741699302860873,
+      "learning_rate": 1.4421287371819781e-05,
+      "loss": 0.4186,
+      "step": 6503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7860385506372445,
+      "learning_rate": 1.4419618146732404e-05,
+      "loss": 0.2432,
+      "step": 6504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4018869658660575,
+      "learning_rate": 1.4417948768593842e-05,
+      "loss": 0.3231,
+      "step": 6505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5974194600768572,
+      "learning_rate": 1.4416279237461903e-05,
+      "loss": 0.4459,
+      "step": 6506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5043837478835471,
+      "learning_rate": 1.4414609553394408e-05,
+      "loss": 0.2003,
+      "step": 6507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3785285725739497,
+      "learning_rate": 1.441293971644917e-05,
+      "loss": 0.2199,
+      "step": 6508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1012730369361647,
+      "learning_rate": 1.4411269726684022e-05,
+      "loss": 0.7594,
+      "step": 6509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8161801633283232,
+      "learning_rate": 1.4409599584156791e-05,
+      "loss": 0.4433,
+      "step": 6510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3898556198374485,
+      "learning_rate": 1.4407929288925316e-05,
+      "loss": 0.2459,
+      "step": 6511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5468015768810729,
+      "learning_rate": 1.440625884104744e-05,
+      "loss": 0.3783,
+      "step": 6512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.734511890686694,
+      "learning_rate": 1.440458824058101e-05,
+      "loss": 0.3978,
+      "step": 6513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2729669946052368,
+      "learning_rate": 1.4402917487583876e-05,
+      "loss": 0.1975,
+      "step": 6514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5058855067236363,
+      "learning_rate": 1.4401246582113904e-05,
+      "loss": 0.4243,
+      "step": 6515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2797481012857069,
+      "learning_rate": 1.4399575524228949e-05,
+      "loss": 0.237,
+      "step": 6516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36853983359838033,
+      "learning_rate": 1.4397904313986881e-05,
+      "loss": 0.2095,
+      "step": 6517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8808245431323464,
+      "learning_rate": 1.439623295144558e-05,
+      "loss": 0.4242,
+      "step": 6518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32676114481551416,
+      "learning_rate": 1.4394561436662917e-05,
+      "loss": 0.282,
+      "step": 6519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32006809977392964,
+      "learning_rate": 1.4392889769696783e-05,
+      "loss": 0.2367,
+      "step": 6520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4047428854960159,
+      "learning_rate": 1.4391217950605063e-05,
+      "loss": 0.3227,
+      "step": 6521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36465854255387276,
+      "learning_rate": 1.4389545979445655e-05,
+      "loss": 0.245,
+      "step": 6522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35563849101387063,
+      "learning_rate": 1.438787385627646e-05,
+      "loss": 0.3014,
+      "step": 6523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3814440985840305,
+      "learning_rate": 1.438620158115538e-05,
+      "loss": 0.2927,
+      "step": 6524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1983742937891095,
+      "learning_rate": 1.438452915414033e-05,
+      "loss": 0.7023,
+      "step": 6525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34701945637472476,
+      "learning_rate": 1.4382856575289223e-05,
+      "loss": 0.2919,
+      "step": 6526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35292559953145647,
+      "learning_rate": 1.438118384465998e-05,
+      "loss": 0.2924,
+      "step": 6527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26027699865809883,
+      "learning_rate": 1.4379510962310532e-05,
+      "loss": 0.1848,
+      "step": 6528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35662767213458074,
+      "learning_rate": 1.4377837928298804e-05,
+      "loss": 0.3079,
+      "step": 6529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6596498854621378,
+      "learning_rate": 1.4376164742682738e-05,
+      "loss": 0.3599,
+      "step": 6530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36191791630131526,
+      "learning_rate": 1.4374491405520274e-05,
+      "loss": 0.2928,
+      "step": 6531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3860443428448384,
+      "learning_rate": 1.4372817916869364e-05,
+      "loss": 0.3045,
+      "step": 6532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9970264359008559,
+      "learning_rate": 1.4371144276787954e-05,
+      "loss": 0.5673,
+      "step": 6533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4185340140823584,
+      "learning_rate": 1.436947048533401e-05,
+      "loss": 0.2696,
+      "step": 6534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37333837957749794,
+      "learning_rate": 1.4367796542565486e-05,
+      "loss": 0.2877,
+      "step": 6535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.38025057980172355,
+      "learning_rate": 1.4366122448540361e-05,
+      "loss": 0.266,
+      "step": 6536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6973816542088787,
+      "learning_rate": 1.4364448203316599e-05,
+      "loss": 0.2136,
+      "step": 6537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.354013344957813,
+      "learning_rate": 1.4362773806952184e-05,
+      "loss": 0.2847,
+      "step": 6538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42870882657679155,
+      "learning_rate": 1.4361099259505102e-05,
+      "loss": 0.3461,
+      "step": 6539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7161912133243448,
+      "learning_rate": 1.4359424561033337e-05,
+      "loss": 0.3663,
+      "step": 6540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32702988657639787,
+      "learning_rate": 1.435774971159489e-05,
+      "loss": 0.2704,
+      "step": 6541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2874840199994187,
+      "learning_rate": 1.4356074711247759e-05,
+      "loss": 0.1933,
+      "step": 6542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3422143523196144,
+      "learning_rate": 1.4354399560049943e-05,
+      "loss": 0.2543,
+      "step": 6543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3620610486020329,
+      "learning_rate": 1.4352724258059461e-05,
+      "loss": 0.2882,
+      "step": 6544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6999998230635959,
+      "learning_rate": 1.4351048805334325e-05,
+      "loss": 0.5029,
+      "step": 6545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7040615510823431,
+      "learning_rate": 1.4349373201932553e-05,
+      "loss": 0.4828,
+      "step": 6546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34328707423243926,
+      "learning_rate": 1.4347697447912176e-05,
+      "loss": 0.2311,
+      "step": 6547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34255042274766273,
+      "learning_rate": 1.4346021543331224e-05,
+      "loss": 0.2109,
+      "step": 6548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9416350640193081,
+      "learning_rate": 1.4344345488247733e-05,
+      "loss": 0.5087,
+      "step": 6549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3524115680822469,
+      "learning_rate": 1.4342669282719741e-05,
+      "loss": 0.2236,
+      "step": 6550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34939936511288033,
+      "learning_rate": 1.4340992926805304e-05,
+      "loss": 0.3067,
+      "step": 6551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9389882276144642,
+      "learning_rate": 1.4339316420562464e-05,
+      "loss": 0.4627,
+      "step": 6552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3572146276189607,
+      "learning_rate": 1.4337639764049285e-05,
+      "loss": 0.2025,
+      "step": 6553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3440940221274594,
+      "learning_rate": 1.4335962957323827e-05,
+      "loss": 0.2088,
+      "step": 6554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43469991652072393,
+      "learning_rate": 1.433428600044416e-05,
+      "loss": 0.3331,
+      "step": 6555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30077868298122107,
+      "learning_rate": 1.4332608893468351e-05,
+      "loss": 0.1945,
+      "step": 6556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6253365617274365,
+      "learning_rate": 1.4330931636454489e-05,
+      "loss": 0.4415,
+      "step": 6557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47640301786281314,
+      "learning_rate": 1.4329254229460645e-05,
+      "loss": 0.3846,
+      "step": 6558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30769115910174644,
+      "learning_rate": 1.4327576672544917e-05,
+      "loss": 0.2608,
+      "step": 6559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24923379904346613,
+      "learning_rate": 1.432589896576539e-05,
+      "loss": 0.152,
+      "step": 6560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1160575763268,
+      "learning_rate": 1.4324221109180173e-05,
+      "loss": 0.6775,
+      "step": 6561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39827529165073156,
+      "learning_rate": 1.4322543102847362e-05,
+      "loss": 0.2889,
+      "step": 6562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4097091571777183,
+      "learning_rate": 1.432086494682507e-05,
+      "loss": 0.3128,
+      "step": 6563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4122864900557979,
+      "learning_rate": 1.4319186641171412e-05,
+      "loss": 0.7525,
+      "step": 6564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.345257798485318,
+      "learning_rate": 1.4317508185944504e-05,
+      "loss": 0.2947,
+      "step": 6565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18780929074416594,
+      "learning_rate": 1.4315829581202474e-05,
+      "loss": 0.0706,
+      "step": 6566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37982000063976756,
+      "learning_rate": 1.431415082700345e-05,
+      "loss": 0.326,
+      "step": 6567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37529555056231606,
+      "learning_rate": 1.4312471923405571e-05,
+      "loss": 0.2817,
+      "step": 6568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7558881857203945,
+      "learning_rate": 1.4310792870466973e-05,
+      "loss": 0.3752,
+      "step": 6569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4981673910095424,
+      "learning_rate": 1.4309113668245804e-05,
+      "loss": 0.4285,
+      "step": 6570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32064816753904357,
+      "learning_rate": 1.4307434316800213e-05,
+      "loss": 0.2705,
+      "step": 6571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30413949280789165,
+      "learning_rate": 1.4305754816188358e-05,
+      "loss": 0.2559,
+      "step": 6572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4309987650378867,
+      "learning_rate": 1.4304075166468396e-05,
+      "loss": 0.2168,
+      "step": 6573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40311342743064493,
+      "learning_rate": 1.43023953676985e-05,
+      "loss": 0.2705,
+      "step": 6574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2954206348253972,
+      "learning_rate": 1.4300715419936834e-05,
+      "loss": 0.3028,
+      "step": 6575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9026406553036119,
+      "learning_rate": 1.4299035323241583e-05,
+      "loss": 0.5688,
+      "step": 6576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3587211236933703,
+      "learning_rate": 1.429735507767092e-05,
+      "loss": 0.2561,
+      "step": 6577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26006833779910876,
+      "learning_rate": 1.4295674683283037e-05,
+      "loss": 0.2293,
+      "step": 6578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3906207143635448,
+      "learning_rate": 1.4293994140136123e-05,
+      "loss": 0.2902,
+      "step": 6579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3336786346711805,
+      "learning_rate": 1.4292313448288377e-05,
+      "loss": 0.2849,
+      "step": 6580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7731830224890437,
+      "learning_rate": 1.4290632607797998e-05,
+      "loss": 0.4842,
+      "step": 6581
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46191721868987656,
+      "learning_rate": 1.4288951618723201e-05,
+      "loss": 0.2651,
+      "step": 6582
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3088949080198425,
+      "learning_rate": 1.428727048112219e-05,
+      "loss": 0.2683,
+      "step": 6583
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0472390701706424,
+      "learning_rate": 1.4285589195053191e-05,
+      "loss": 0.6867,
+      "step": 6584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3409265332982283,
+      "learning_rate": 1.428390776057442e-05,
+      "loss": 0.2472,
+      "step": 6585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35059501414858735,
+      "learning_rate": 1.4282226177744107e-05,
+      "loss": 0.2534,
+      "step": 6586
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3968681655453618,
+      "learning_rate": 1.4280544446620485e-05,
+      "loss": 0.3418,
+      "step": 6587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3699405066952637,
+      "learning_rate": 1.4278862567261796e-05,
+      "loss": 0.8169,
+      "step": 6588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36354057094968956,
+      "learning_rate": 1.4277180539726278e-05,
+      "loss": 0.1807,
+      "step": 6589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37436367140991333,
+      "learning_rate": 1.427549836407218e-05,
+      "loss": 0.274,
+      "step": 6590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4031137876109123,
+      "learning_rate": 1.4273816040357762e-05,
+      "loss": 0.3387,
+      "step": 6591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39913767829709507,
+      "learning_rate": 1.4272133568641273e-05,
+      "loss": 0.2186,
+      "step": 6592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34901425704231603,
+      "learning_rate": 1.4270450948980989e-05,
+      "loss": 0.2849,
+      "step": 6593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43807677930615996,
+      "learning_rate": 1.4268768181435166e-05,
+      "loss": 0.3457,
+      "step": 6594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33170476466487625,
+      "learning_rate": 1.4267085266062088e-05,
+      "loss": 0.1751,
+      "step": 6595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47900650476435713,
+      "learning_rate": 1.4265402202920029e-05,
+      "loss": 0.3583,
+      "step": 6596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6101026898693607,
+      "learning_rate": 1.4263718992067276e-05,
+      "loss": 0.44,
+      "step": 6597
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3100847262447763,
+      "learning_rate": 1.4262035633562117e-05,
+      "loss": 0.2415,
+      "step": 6598
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29612737366854475,
+      "learning_rate": 1.4260352127462848e-05,
+      "loss": 0.2147,
+      "step": 6599
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0381415145756596,
+      "learning_rate": 1.4258668473827766e-05,
+      "loss": 0.7037,
+      "step": 6600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40518343031761545,
+      "learning_rate": 1.4256984672715182e-05,
+      "loss": 0.2977,
+      "step": 6601
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4991942601955228,
+      "learning_rate": 1.4255300724183396e-05,
+      "loss": 0.2844,
+      "step": 6602
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43905524147113645,
+      "learning_rate": 1.4253616628290735e-05,
+      "loss": 0.3412,
+      "step": 6603
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32293505302934256,
+      "learning_rate": 1.425193238509551e-05,
+      "loss": 0.2129,
+      "step": 6604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30523910076856414,
+      "learning_rate": 1.425024799465605e-05,
+      "loss": 0.1843,
+      "step": 6605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7134300357593555,
+      "learning_rate": 1.4248563457030684e-05,
+      "loss": 0.3323,
+      "step": 6606
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3821121733080358,
+      "learning_rate": 1.4246878772277748e-05,
+      "loss": 0.2799,
+      "step": 6607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5362957512770763,
+      "learning_rate": 1.4245193940455583e-05,
+      "loss": 0.2907,
+      "step": 6608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7015232324733971,
+      "learning_rate": 1.4243508961622536e-05,
+      "loss": 0.4711,
+      "step": 6609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36283666869718023,
+      "learning_rate": 1.4241823835836957e-05,
+      "loss": 0.231,
+      "step": 6610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3676225906926481,
+      "learning_rate": 1.4240138563157197e-05,
+      "loss": 0.2852,
+      "step": 6611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3206621409131723,
+      "learning_rate": 1.4238453143641623e-05,
+      "loss": 0.1736,
+      "step": 6612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6024596594608613,
+      "learning_rate": 1.4236767577348597e-05,
+      "loss": 0.3085,
+      "step": 6613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3601264112579988,
+      "learning_rate": 1.4235081864336495e-05,
+      "loss": 0.3118,
+      "step": 6614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4754139334883257,
+      "learning_rate": 1.4233396004663686e-05,
+      "loss": 0.2908,
+      "step": 6615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6780714577012267,
+      "learning_rate": 1.423170999838856e-05,
+      "loss": 0.3664,
+      "step": 6616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4271908689523186,
+      "learning_rate": 1.4230023845569497e-05,
+      "loss": 0.375,
+      "step": 6617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2646098949172002,
+      "learning_rate": 1.422833754626489e-05,
+      "loss": 0.2093,
+      "step": 6618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36022570135160514,
+      "learning_rate": 1.4226651100533136e-05,
+      "loss": 0.2529,
+      "step": 6619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4342374594084135,
+      "learning_rate": 1.4224964508432635e-05,
+      "loss": 0.3454,
+      "step": 6620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0431596454133143,
+      "learning_rate": 1.4223277770021794e-05,
+      "loss": 0.4782,
+      "step": 6621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2630385189545252,
+      "learning_rate": 1.4221590885359029e-05,
+      "loss": 0.2208,
+      "step": 6622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37912554241453955,
+      "learning_rate": 1.421990385450275e-05,
+      "loss": 0.3119,
+      "step": 6623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31620577959594415,
+      "learning_rate": 1.4218216677511383e-05,
+      "loss": 0.2133,
+      "step": 6624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7817419011776064,
+      "learning_rate": 1.4216529354443355e-05,
+      "loss": 0.2797,
+      "step": 6625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3580806273138164,
+      "learning_rate": 1.4214841885357096e-05,
+      "loss": 0.3249,
+      "step": 6626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47803700129219767,
+      "learning_rate": 1.4213154270311043e-05,
+      "loss": 0.4245,
+      "step": 6627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29809480038275465,
+      "learning_rate": 1.421146650936364e-05,
+      "loss": 0.1387,
+      "step": 6628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3552015102693456,
+      "learning_rate": 1.4209778602573332e-05,
+      "loss": 0.2911,
+      "step": 6629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.371671833280386,
+      "learning_rate": 1.4208090549998572e-05,
+      "loss": 0.3407,
+      "step": 6630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.774988864764028,
+      "learning_rate": 1.420640235169782e-05,
+      "loss": 0.3303,
+      "step": 6631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2656032968656245,
+      "learning_rate": 1.420471400772953e-05,
+      "loss": 0.2185,
+      "step": 6632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7574900916209983,
+      "learning_rate": 1.4203025518152178e-05,
+      "loss": 0.4876,
+      "step": 6633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3571697183187376,
+      "learning_rate": 1.420133688302423e-05,
+      "loss": 0.3068,
+      "step": 6634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3627025977269262,
+      "learning_rate": 1.419964810240417e-05,
+      "loss": 0.2454,
+      "step": 6635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9722718639281581,
+      "learning_rate": 1.4197959176350476e-05,
+      "loss": 0.5431,
+      "step": 6636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45319845739485237,
+      "learning_rate": 1.4196270104921637e-05,
+      "loss": 0.2968,
+      "step": 6637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3168864543772265,
+      "learning_rate": 1.4194580888176141e-05,
+      "loss": 0.2416,
+      "step": 6638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4402300020745205,
+      "learning_rate": 1.4192891526172494e-05,
+      "loss": 0.2527,
+      "step": 6639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7619509636775692,
+      "learning_rate": 1.419120201896919e-05,
+      "loss": 0.4639,
+      "step": 6640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4094808720491614,
+      "learning_rate": 1.4189512366624745e-05,
+      "loss": 0.2223,
+      "step": 6641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49203402514236594,
+      "learning_rate": 1.4187822569197662e-05,
+      "loss": 0.3586,
+      "step": 6642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6042575546599518,
+      "learning_rate": 1.4186132626746466e-05,
+      "loss": 0.3776,
+      "step": 6643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2850850597785273,
+      "learning_rate": 1.4184442539329677e-05,
+      "loss": 0.1775,
+      "step": 6644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41097360765078583,
+      "learning_rate": 1.4182752307005822e-05,
+      "loss": 0.2656,
+      "step": 6645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3749280168749956,
+      "learning_rate": 1.4181061929833435e-05,
+      "loss": 0.3093,
+      "step": 6646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36949049484555013,
+      "learning_rate": 1.4179371407871054e-05,
+      "loss": 0.2755,
+      "step": 6647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8305477195457657,
+      "learning_rate": 1.4177680741177217e-05,
+      "loss": 0.3878,
+      "step": 6648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9615488498518213,
+      "learning_rate": 1.4175989929810481e-05,
+      "loss": 0.5805,
+      "step": 6649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2774945282091035,
+      "learning_rate": 1.417429897382939e-05,
+      "loss": 0.2652,
+      "step": 6650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21826789390351636,
+      "learning_rate": 1.4172607873292505e-05,
+      "loss": 0.1444,
+      "step": 6651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7917466348827511,
+      "learning_rate": 1.4170916628258392e-05,
+      "loss": 0.5094,
+      "step": 6652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4402092475426011,
+      "learning_rate": 1.4169225238785611e-05,
+      "loss": 0.3204,
+      "step": 6653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48816511867608736,
+      "learning_rate": 1.4167533704932743e-05,
+      "loss": 0.2935,
+      "step": 6654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7306057653718944,
+      "learning_rate": 1.416584202675836e-05,
+      "loss": 0.44,
+      "step": 6655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32525920456793944,
+      "learning_rate": 1.4164150204321046e-05,
+      "loss": 0.2374,
+      "step": 6656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3296501402587944,
+      "learning_rate": 1.4162458237679389e-05,
+      "loss": 0.1485,
+      "step": 6657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36981045940986673,
+      "learning_rate": 1.4160766126891985e-05,
+      "loss": 0.3043,
+      "step": 6658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3883473725235902,
+      "learning_rate": 1.4159073872017427e-05,
+      "loss": 0.2873,
+      "step": 6659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8316799239970233,
+      "learning_rate": 1.4157381473114323e-05,
+      "loss": 0.4332,
+      "step": 6660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40166375360596235,
+      "learning_rate": 1.4155688930241274e-05,
+      "loss": 0.218,
+      "step": 6661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.375701225006493,
+      "learning_rate": 1.4153996243456898e-05,
+      "loss": 0.2991,
+      "step": 6662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2775566069575302,
+      "learning_rate": 1.4152303412819808e-05,
+      "loss": 0.2017,
+      "step": 6663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7846062358543829,
+      "learning_rate": 1.4150610438388633e-05,
+      "loss": 0.3898,
+      "step": 6664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35540285474583694,
+      "learning_rate": 1.4148917320221992e-05,
+      "loss": 0.2879,
+      "step": 6665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36580621119695167,
+      "learning_rate": 1.4147224058378525e-05,
+      "loss": 0.3285,
+      "step": 6666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8255865521291464,
+      "learning_rate": 1.4145530652916868e-05,
+      "loss": 0.3617,
+      "step": 6667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35219097025016854,
+      "learning_rate": 1.4143837103895663e-05,
+      "loss": 0.278,
+      "step": 6668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29669356912673234,
+      "learning_rate": 1.4142143411373559e-05,
+      "loss": 0.2179,
+      "step": 6669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4085564007164582,
+      "learning_rate": 1.4140449575409203e-05,
+      "loss": 0.2935,
+      "step": 6670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3408342696053366,
+      "learning_rate": 1.4138755596061257e-05,
+      "loss": 0.2819,
+      "step": 6671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8311186813808776,
+      "learning_rate": 1.4137061473388383e-05,
+      "loss": 0.617,
+      "step": 6672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5219159840804157,
+      "learning_rate": 1.4135367207449248e-05,
+      "loss": 0.3674,
+      "step": 6673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29763937422590947,
+      "learning_rate": 1.4133672798302525e-05,
+      "loss": 0.2223,
+      "step": 6674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25470244060159813,
+      "learning_rate": 1.4131978246006892e-05,
+      "loss": 0.1708,
+      "step": 6675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6082879025289717,
+      "learning_rate": 1.4130283550621027e-05,
+      "loss": 0.4386,
+      "step": 6676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3082473816953717,
+      "learning_rate": 1.4128588712203626e-05,
+      "loss": 0.1922,
+      "step": 6677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40828261098657215,
+      "learning_rate": 1.4126893730813369e-05,
+      "loss": 0.3571,
+      "step": 6678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.284813336317904,
+      "learning_rate": 1.4125198606508963e-05,
+      "loss": 0.587,
+      "step": 6679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3249047265979789,
+      "learning_rate": 1.4123503339349105e-05,
+      "loss": 0.2049,
+      "step": 6680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27026754771556555,
+      "learning_rate": 1.4121807929392505e-05,
+      "loss": 0.2416,
+      "step": 6681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4587471969379784,
+      "learning_rate": 1.4120112376697873e-05,
+      "loss": 0.3916,
+      "step": 6682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30805966674421015,
+      "learning_rate": 1.4118416681323925e-05,
+      "loss": 0.1992,
+      "step": 6683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2854672987600315,
+      "learning_rate": 1.4116720843329385e-05,
+      "loss": 0.8468,
+      "step": 6684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5225364984100592,
+      "learning_rate": 1.4115024862772981e-05,
+      "loss": 0.3391,
+      "step": 6685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3050026501666236,
+      "learning_rate": 1.4113328739713442e-05,
+      "loss": 0.2727,
+      "step": 6686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7048532439613371,
+      "learning_rate": 1.4111632474209506e-05,
+      "loss": 0.3857,
+      "step": 6687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2856578199252998,
+      "learning_rate": 1.4109936066319915e-05,
+      "loss": 0.2344,
+      "step": 6688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.38790911949598295,
+      "learning_rate": 1.4108239516103412e-05,
+      "loss": 0.2793,
+      "step": 6689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42389402150691363,
+      "learning_rate": 1.4106542823618754e-05,
+      "loss": 0.2845,
+      "step": 6690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1114232450461785,
+      "learning_rate": 1.4104845988924694e-05,
+      "loss": 0.6795,
+      "step": 6691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31314498273119074,
+      "learning_rate": 1.4103149012079994e-05,
+      "loss": 0.2629,
+      "step": 6692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4220924174883716,
+      "learning_rate": 1.4101451893143418e-05,
+      "loss": 0.2873,
+      "step": 6693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34138109317922455,
+      "learning_rate": 1.4099754632173744e-05,
+      "loss": 0.2751,
+      "step": 6694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.404482535836102,
+      "learning_rate": 1.409805722922974e-05,
+      "loss": 0.2508,
+      "step": 6695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33534481802465277,
+      "learning_rate": 1.409635968437019e-05,
+      "loss": 0.2233,
+      "step": 6696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3605268365663909,
+      "learning_rate": 1.409466199765388e-05,
+      "loss": 0.3063,
+      "step": 6697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39065776412591735,
+      "learning_rate": 1.4092964169139603e-05,
+      "loss": 0.2807,
+      "step": 6698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44905297298310376,
+      "learning_rate": 1.409126619888615e-05,
+      "loss": 0.3731,
+      "step": 6699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34506324518643255,
+      "learning_rate": 1.4089568086952327e-05,
+      "loss": 0.2082,
+      "step": 6700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39996919181229135,
+      "learning_rate": 1.4087869833396936e-05,
+      "loss": 0.2956,
+      "step": 6701
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3419182623503886,
+      "learning_rate": 1.408617143827879e-05,
+      "loss": 0.2966,
+      "step": 6702
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35602932018929834,
+      "learning_rate": 1.40844729016567e-05,
+      "loss": 0.1709,
+      "step": 6703
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33780685153209516,
+      "learning_rate": 1.4082774223589492e-05,
+      "loss": 0.2794,
+      "step": 6704
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3433742555355464,
+      "learning_rate": 1.4081075404135987e-05,
+      "loss": 0.3471,
+      "step": 6705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.698436131545571,
+      "learning_rate": 1.4079376443355016e-05,
+      "loss": 0.0464,
+      "step": 6706
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3513574365202577,
+      "learning_rate": 1.4077677341305414e-05,
+      "loss": 0.2612,
+      "step": 6707
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.17360579573699,
+      "learning_rate": 1.4075978098046022e-05,
+      "loss": 0.7839,
+      "step": 6708
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26317161420211993,
+      "learning_rate": 1.4074278713635683e-05,
+      "loss": 0.2323,
+      "step": 6709
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2971917186585945,
+      "learning_rate": 1.4072579188133247e-05,
+      "loss": 0.2247,
+      "step": 6710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6271548854008792,
+      "learning_rate": 1.407087952159757e-05,
+      "loss": 0.4772,
+      "step": 6711
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.871101430569621,
+      "learning_rate": 1.406917971408751e-05,
+      "loss": 0.6613,
+      "step": 6712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3915942662772745,
+      "learning_rate": 1.4067479765661929e-05,
+      "loss": 0.2208,
+      "step": 6713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33841239049328914,
+      "learning_rate": 1.4065779676379702e-05,
+      "loss": 0.3083,
+      "step": 6714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30049951964028676,
+      "learning_rate": 1.4064079446299699e-05,
+      "loss": 0.2178,
+      "step": 6715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38831455943753335,
+      "learning_rate": 1.4062379075480799e-05,
+      "loss": 0.2218,
+      "step": 6716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.291218423496063,
+      "learning_rate": 1.4060678563981886e-05,
+      "loss": 0.317,
+      "step": 6717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0342262669644686,
+      "learning_rate": 1.4058977911861846e-05,
+      "loss": 0.6031,
+      "step": 6718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35540629091607656,
+      "learning_rate": 1.405727711917958e-05,
+      "loss": 0.2125,
+      "step": 6719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3154034414211785,
+      "learning_rate": 1.405557618599398e-05,
+      "loss": 0.2661,
+      "step": 6720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3914215878276067,
+      "learning_rate": 1.4053875112363953e-05,
+      "loss": 0.3172,
+      "step": 6721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5473122810280361,
+      "learning_rate": 1.40521738983484e-05,
+      "loss": 0.3438,
+      "step": 6722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2795370643280423,
+      "learning_rate": 1.4050472544006243e-05,
+      "loss": 0.2083,
+      "step": 6723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4035101146683626,
+      "learning_rate": 1.4048771049396397e-05,
+      "loss": 0.7028,
+      "step": 6724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2976029002338224,
+      "learning_rate": 1.4047069414577782e-05,
+      "loss": 0.2697,
+      "step": 6725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35788245762048565,
+      "learning_rate": 1.4045367639609326e-05,
+      "loss": 0.265,
+      "step": 6726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8534396855898208,
+      "learning_rate": 1.4043665724549967e-05,
+      "loss": 0.5141,
+      "step": 6727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2832789254792039,
+      "learning_rate": 1.4041963669458633e-05,
+      "loss": 0.2318,
+      "step": 6728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3004873857950981,
+      "learning_rate": 1.4040261474394275e-05,
+      "loss": 0.1911,
+      "step": 6729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48748591817012543,
+      "learning_rate": 1.4038559139415832e-05,
+      "loss": 0.3871,
+      "step": 6730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.66305837104567,
+      "learning_rate": 1.4036856664582263e-05,
+      "loss": 0.4026,
+      "step": 6731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4050811425920618,
+      "learning_rate": 1.403515404995252e-05,
+      "loss": 0.2587,
+      "step": 6732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41824055428183066,
+      "learning_rate": 1.4033451295585565e-05,
+      "loss": 0.3413,
+      "step": 6733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37066586491062015,
+      "learning_rate": 1.4031748401540366e-05,
+      "loss": 0.2305,
+      "step": 6734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29315654585139084,
+      "learning_rate": 1.4030045367875893e-05,
+      "loss": 0.2388,
+      "step": 6735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8745220798442661,
+      "learning_rate": 1.4028342194651123e-05,
+      "loss": 0.4191,
+      "step": 6736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30931597379685477,
+      "learning_rate": 1.4026638881925032e-05,
+      "loss": 0.2733,
+      "step": 6737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37488368146363077,
+      "learning_rate": 1.4024935429756614e-05,
+      "loss": 0.3517,
+      "step": 6738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.548881557197987,
+      "learning_rate": 1.4023231838204854e-05,
+      "loss": 0.372,
+      "step": 6739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31575575686443624,
+      "learning_rate": 1.4021528107328749e-05,
+      "loss": 0.1803,
+      "step": 6740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2777086707952842,
+      "learning_rate": 1.4019824237187296e-05,
+      "loss": 0.2545,
+      "step": 6741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.014625121565217,
+      "learning_rate": 1.4018120227839505e-05,
+      "loss": 0.3794,
+      "step": 6742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5247262023251844,
+      "learning_rate": 1.4016416079344382e-05,
+      "loss": 0.373,
+      "step": 6743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3867050547624269,
+      "learning_rate": 1.4014711791760944e-05,
+      "loss": 0.3268,
+      "step": 6744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35604241188905456,
+      "learning_rate": 1.401300736514821e-05,
+      "loss": 0.2967,
+      "step": 6745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4154383606703178,
+      "learning_rate": 1.4011302799565205e-05,
+      "loss": 0.3106,
+      "step": 6746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.250345897544341,
+      "learning_rate": 1.4009598095070951e-05,
+      "loss": 0.1927,
+      "step": 6747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2188115276047022,
+      "learning_rate": 1.4007893251724491e-05,
+      "loss": 0.7797,
+      "step": 6748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29690370719853887,
+      "learning_rate": 1.400618826958486e-05,
+      "loss": 0.2267,
+      "step": 6749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40303575141635606,
+      "learning_rate": 1.4004483148711101e-05,
+      "loss": 0.3435,
+      "step": 6750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9008436992328396,
+      "learning_rate": 1.4002777889162262e-05,
+      "loss": 0.4848,
+      "step": 6751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2933102711442764,
+      "learning_rate": 1.4001072490997399e-05,
+      "loss": 0.1314,
+      "step": 6752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2828190568585241,
+      "learning_rate": 1.3999366954275566e-05,
+      "loss": 0.2767,
+      "step": 6753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.362682035058518,
+      "learning_rate": 1.3997661279055826e-05,
+      "loss": 0.247,
+      "step": 6754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6611501611701132,
+      "learning_rate": 1.399595546539725e-05,
+      "loss": 0.3083,
+      "step": 6755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3474578684143908,
+      "learning_rate": 1.3994249513358907e-05,
+      "loss": 0.3028,
+      "step": 6756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38528825517022264,
+      "learning_rate": 1.3992543422999876e-05,
+      "loss": 0.3243,
+      "step": 6757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3992245764777032,
+      "learning_rate": 1.3990837194379236e-05,
+      "loss": 0.1723,
+      "step": 6758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24633718428561147,
+      "learning_rate": 1.3989130827556077e-05,
+      "loss": 0.2043,
+      "step": 6759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7078159518251719,
+      "learning_rate": 1.398742432258949e-05,
+      "loss": 0.8635,
+      "step": 6760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.386580417191464,
+      "learning_rate": 1.398571767953857e-05,
+      "loss": 0.3186,
+      "step": 6761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4046253648674059,
+      "learning_rate": 1.3984010898462417e-05,
+      "loss": 0.2438,
+      "step": 6762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8689183740842735,
+      "learning_rate": 1.398230397942014e-05,
+      "loss": 0.5434,
+      "step": 6763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4089398395770051,
+      "learning_rate": 1.3980596922470844e-05,
+      "loss": 0.3337,
+      "step": 6764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29438732424321856,
+      "learning_rate": 1.397888972767365e-05,
+      "loss": 0.2378,
+      "step": 6765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3592290968908871,
+      "learning_rate": 1.3977182395087674e-05,
+      "loss": 0.2221,
+      "step": 6766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6970432495304754,
+      "learning_rate": 1.3975474924772043e-05,
+      "loss": 0.3725,
+      "step": 6767
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4009812575724056,
+      "learning_rate": 1.3973767316785887e-05,
+      "loss": 0.2633,
+      "step": 6768
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37223344888144816,
+      "learning_rate": 1.397205957118834e-05,
+      "loss": 0.3468,
+      "step": 6769
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.57109691824902,
+      "learning_rate": 1.397035168803854e-05,
+      "loss": 0.762,
+      "step": 6770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2339637760874292,
+      "learning_rate": 1.3968643667395634e-05,
+      "loss": 0.1738,
+      "step": 6771
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32562666939438184,
+      "learning_rate": 1.3966935509318766e-05,
+      "loss": 0.2632,
+      "step": 6772
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7221752928764033,
+      "learning_rate": 1.3965227213867093e-05,
+      "loss": 0.4831,
+      "step": 6773
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33874134003021095,
+      "learning_rate": 1.3963518781099774e-05,
+      "loss": 0.2641,
+      "step": 6774
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9198892609799831,
+      "learning_rate": 1.3961810211075965e-05,
+      "loss": 0.4748,
+      "step": 6775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41627147942332504,
+      "learning_rate": 1.3960101503854843e-05,
+      "loss": 0.314,
+      "step": 6776
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32450432247327654,
+      "learning_rate": 1.3958392659495575e-05,
+      "loss": 0.2792,
+      "step": 6777
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3360326187425709,
+      "learning_rate": 1.3956683678057342e-05,
+      "loss": 0.1615,
+      "step": 6778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7677364221803948,
+      "learning_rate": 1.395497455959932e-05,
+      "loss": 0.3848,
+      "step": 6779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3772113181888643,
+      "learning_rate": 1.39532653041807e-05,
+      "loss": 0.2806,
+      "step": 6780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6590453255458756,
+      "learning_rate": 1.3951555911860672e-05,
+      "loss": 0.2858,
+      "step": 6781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8705894982837394,
+      "learning_rate": 1.3949846382698433e-05,
+      "loss": 0.4125,
+      "step": 6782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3819991956597844,
+      "learning_rate": 1.3948136716753183e-05,
+      "loss": 0.2831,
+      "step": 6783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4511179604074903,
+      "learning_rate": 1.394642691408413e-05,
+      "loss": 0.3855,
+      "step": 6784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23375528534045187,
+      "learning_rate": 1.394471697475048e-05,
+      "loss": 0.1753,
+      "step": 6785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3867215053995416,
+      "learning_rate": 1.3943006898811453e-05,
+      "loss": 0.2849,
+      "step": 6786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1350554649970255,
+      "learning_rate": 1.3941296686326266e-05,
+      "loss": 0.8263,
+      "step": 6787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3652869785590026,
+      "learning_rate": 1.3939586337354146e-05,
+      "loss": 0.2795,
+      "step": 6788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3622856983259643,
+      "learning_rate": 1.3937875851954316e-05,
+      "loss": 0.2834,
+      "step": 6789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5852464787006104,
+      "learning_rate": 1.3936165230186018e-05,
+      "loss": 0.4318,
+      "step": 6790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2725903077636448,
+      "learning_rate": 1.3934454472108488e-05,
+      "loss": 0.1255,
+      "step": 6791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3542987477803968,
+      "learning_rate": 1.3932743577780967e-05,
+      "loss": 0.2905,
+      "step": 6792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3643810295443851,
+      "learning_rate": 1.3931032547262707e-05,
+      "loss": 0.2526,
+      "step": 6793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6951650698020321,
+      "learning_rate": 1.3929321380612955e-05,
+      "loss": 0.3213,
+      "step": 6794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3221414003316139,
+      "learning_rate": 1.3927610077890976e-05,
+      "loss": 0.2688,
+      "step": 6795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6841427532021138,
+      "learning_rate": 1.3925898639156028e-05,
+      "loss": 0.4703,
+      "step": 6796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2893061685049914,
+      "learning_rate": 1.3924187064467378e-05,
+      "loss": 0.2869,
+      "step": 6797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3150811635114615,
+      "learning_rate": 1.3922475353884302e-05,
+      "loss": 0.2133,
+      "step": 6798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39697628073105073,
+      "learning_rate": 1.3920763507466071e-05,
+      "loss": 0.2859,
+      "step": 6799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3824519346405898,
+      "learning_rate": 1.3919051525271968e-05,
+      "loss": 0.3316,
+      "step": 6800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3120471764511407,
+      "learning_rate": 1.3917339407361278e-05,
+      "loss": 0.1981,
+      "step": 6801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7719291568422046,
+      "learning_rate": 1.3915627153793294e-05,
+      "loss": 0.4703,
+      "step": 6802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0406178609357577,
+      "learning_rate": 1.3913914764627311e-05,
+      "loss": 0.7213,
+      "step": 6803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3082881655989037,
+      "learning_rate": 1.3912202239922627e-05,
+      "loss": 0.1947,
+      "step": 6804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.269808735942457,
+      "learning_rate": 1.3910489579738548e-05,
+      "loss": 0.2328,
+      "step": 6805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8449664005507034,
+      "learning_rate": 1.3908776784134382e-05,
+      "loss": 0.4975,
+      "step": 6806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3668433609014333,
+      "learning_rate": 1.3907063853169445e-05,
+      "loss": 0.2513,
+      "step": 6807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3828472985391112,
+      "learning_rate": 1.3905350786903053e-05,
+      "loss": 0.3159,
+      "step": 6808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9917346778146909,
+      "learning_rate": 1.3903637585394534e-05,
+      "loss": 0.6777,
+      "step": 6809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34320658920451136,
+      "learning_rate": 1.390192424870321e-05,
+      "loss": 0.2877,
+      "step": 6810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22123271814015424,
+      "learning_rate": 1.3900210776888421e-05,
+      "loss": 0.1037,
+      "step": 6811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43597104711910406,
+      "learning_rate": 1.38984971700095e-05,
+      "loss": 0.3482,
+      "step": 6812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3876067777475436,
+      "learning_rate": 1.3896783428125789e-05,
+      "loss": 0.2848,
+      "step": 6813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.814104832790337,
+      "learning_rate": 1.3895069551296634e-05,
+      "loss": 0.3313,
+      "step": 6814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0191123045537875,
+      "learning_rate": 1.389335553958139e-05,
+      "loss": 0.7098,
+      "step": 6815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3189777727484337,
+      "learning_rate": 1.3891641393039408e-05,
+      "loss": 0.2773,
+      "step": 6816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3103356437264256,
+      "learning_rate": 1.3889927111730056e-05,
+      "loss": 0.1763,
+      "step": 6817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6834230208026675,
+      "learning_rate": 1.3888212695712693e-05,
+      "loss": 0.3407,
+      "step": 6818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4373386910283233,
+      "learning_rate": 1.3886498145046689e-05,
+      "loss": 0.2706,
+      "step": 6819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33425168275995765,
+      "learning_rate": 1.3884783459791427e-05,
+      "loss": 0.2554,
+      "step": 6820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9990408288161097,
+      "learning_rate": 1.3883068640006277e-05,
+      "loss": 0.6442,
+      "step": 6821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3974999866738344,
+      "learning_rate": 1.3881353685750627e-05,
+      "loss": 0.2808,
+      "step": 6822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48249560737044717,
+      "learning_rate": 1.3879638597083864e-05,
+      "loss": 0.3902,
+      "step": 6823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.406224053807835,
+      "learning_rate": 1.387792337406539e-05,
+      "loss": 0.2945,
+      "step": 6824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2699756257137786,
+      "learning_rate": 1.3876208016754589e-05,
+      "loss": 0.2,
+      "step": 6825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35033622402874065,
+      "learning_rate": 1.3874492525210874e-05,
+      "loss": 0.2657,
+      "step": 6826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.923795988793285,
+      "learning_rate": 1.387277689949365e-05,
+      "loss": 0.4945,
+      "step": 6827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29261064970858314,
+      "learning_rate": 1.3871061139662328e-05,
+      "loss": 0.2662,
+      "step": 6828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5097583064881249,
+      "learning_rate": 1.3869345245776326e-05,
+      "loss": 0.3505,
+      "step": 6829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3765657158030522,
+      "learning_rate": 1.3867629217895067e-05,
+      "loss": 0.2203,
+      "step": 6830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3066487278928788,
+      "learning_rate": 1.3865913056077968e-05,
+      "loss": 0.2238,
+      "step": 6831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49241965675255883,
+      "learning_rate": 1.3864196760384471e-05,
+      "loss": 0.3577,
+      "step": 6832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47556465523233393,
+      "learning_rate": 1.3862480330874004e-05,
+      "loss": 0.2933,
+      "step": 6833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3757822961632396,
+      "learning_rate": 1.3860763767606012e-05,
+      "loss": 0.2787,
+      "step": 6834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7440805037762387,
+      "learning_rate": 1.3859047070639933e-05,
+      "loss": 0.4623,
+      "step": 6835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33732921252747394,
+      "learning_rate": 1.3857330240035223e-05,
+      "loss": 0.3338,
+      "step": 6836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18536598268984897,
+      "learning_rate": 1.385561327585133e-05,
+      "loss": 0.0748,
+      "step": 6837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30528617287475135,
+      "learning_rate": 1.3853896178147717e-05,
+      "loss": 0.2542,
+      "step": 6838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48011544732265776,
+      "learning_rate": 1.3852178946983845e-05,
+      "loss": 0.4014,
+      "step": 6839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32213207603181976,
+      "learning_rate": 1.385046158241918e-05,
+      "loss": 0.2102,
+      "step": 6840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6276796859084134,
+      "learning_rate": 1.3848744084513197e-05,
+      "loss": 0.3767,
+      "step": 6841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9645205479290853,
+      "learning_rate": 1.3847026453325371e-05,
+      "loss": 0.6006,
+      "step": 6842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18739949526105895,
+      "learning_rate": 1.3845308688915187e-05,
+      "loss": 0.1148,
+      "step": 6843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29294272479257805,
+      "learning_rate": 1.3843590791342125e-05,
+      "loss": 0.2939,
+      "step": 6844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0560695569212466,
+      "learning_rate": 1.3841872760665682e-05,
+      "loss": 0.7409,
+      "step": 6845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40634206076714197,
+      "learning_rate": 1.3840154596945352e-05,
+      "loss": 0.2672,
+      "step": 6846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5534173099885904,
+      "learning_rate": 1.383843630024063e-05,
+      "loss": 0.4431,
+      "step": 6847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35250494584408637,
+      "learning_rate": 1.3836717870611025e-05,
+      "loss": 0.3131,
+      "step": 6848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3773994520684638,
+      "learning_rate": 1.3834999308116045e-05,
+      "loss": 0.2855,
+      "step": 6849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22077046425320704,
+      "learning_rate": 1.3833280612815204e-05,
+      "loss": 0.1251,
+      "step": 6850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4475654638391578,
+      "learning_rate": 1.383156178476802e-05,
+      "loss": 0.42,
+      "step": 6851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3264987173804153,
+      "learning_rate": 1.3829842824034016e-05,
+      "loss": 0.3119,
+      "step": 6852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4714716770949217,
+      "learning_rate": 1.382812373067272e-05,
+      "loss": 0.3138,
+      "step": 6853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8911497620521466,
+      "learning_rate": 1.3826404504743662e-05,
+      "loss": 0.57,
+      "step": 6854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33170515753882024,
+      "learning_rate": 1.3824685146306385e-05,
+      "loss": 0.2292,
+      "step": 6855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28216547828961547,
+      "learning_rate": 1.3822965655420422e-05,
+      "loss": 0.2139,
+      "step": 6856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.735537001474033,
+      "learning_rate": 1.3821246032145324e-05,
+      "loss": 0.4583,
+      "step": 6857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5961310337286932,
+      "learning_rate": 1.381952627654064e-05,
+      "loss": 0.4259,
+      "step": 6858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3900981399016567,
+      "learning_rate": 1.3817806388665926e-05,
+      "loss": 0.2397,
+      "step": 6859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37331750009126125,
+      "learning_rate": 1.381608636858074e-05,
+      "loss": 0.3157,
+      "step": 6860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5407319177531821,
+      "learning_rate": 1.3814366216344647e-05,
+      "loss": 0.3703,
+      "step": 6861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22703111036296625,
+      "learning_rate": 1.3812645932017217e-05,
+      "loss": 0.1943,
+      "step": 6862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.653693685219927,
+      "learning_rate": 1.3810925515658022e-05,
+      "loss": 0.3258,
+      "step": 6863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4038868654554686,
+      "learning_rate": 1.3809204967326641e-05,
+      "loss": 0.3235,
+      "step": 6864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40131278450352875,
+      "learning_rate": 1.3807484287082655e-05,
+      "loss": 0.3271,
+      "step": 6865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8587126265813383,
+      "learning_rate": 1.3805763474985651e-05,
+      "loss": 0.2415,
+      "step": 6866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3657681750663915,
+      "learning_rate": 1.3804042531095223e-05,
+      "loss": 0.3043,
+      "step": 6867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31855912993435875,
+      "learning_rate": 1.3802321455470967e-05,
+      "loss": 0.2917,
+      "step": 6868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4853379729560478,
+      "learning_rate": 1.3800600248172478e-05,
+      "loss": 0.2178,
+      "step": 6869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39579207655536114,
+      "learning_rate": 1.3798878909259368e-05,
+      "loss": 0.3008,
+      "step": 6870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5117750783727549,
+      "learning_rate": 1.3797157438791244e-05,
+      "loss": 0.4075,
+      "step": 6871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38032603698465695,
+      "learning_rate": 1.3795435836827724e-05,
+      "loss": 0.3377,
+      "step": 6872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5204418610641315,
+      "learning_rate": 1.3793714103428421e-05,
+      "loss": 0.1617,
+      "step": 6873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2955007971124637,
+      "learning_rate": 1.3791992238652965e-05,
+      "loss": 0.2467,
+      "step": 6874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3308877907838149,
+      "learning_rate": 1.3790270242560977e-05,
+      "loss": 0.2977,
+      "step": 6875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38992600679999917,
+      "learning_rate": 1.3788548115212095e-05,
+      "loss": 0.2571,
+      "step": 6876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38401415930593574,
+      "learning_rate": 1.3786825856665958e-05,
+      "loss": 0.3058,
+      "step": 6877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7709559475093668,
+      "learning_rate": 1.3785103466982199e-05,
+      "loss": 0.5874,
+      "step": 6878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3114444710448705,
+      "learning_rate": 1.3783380946220474e-05,
+      "loss": 0.2394,
+      "step": 6879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3432166855056662,
+      "learning_rate": 1.3781658294440427e-05,
+      "loss": 0.3115,
+      "step": 6880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6522356404592244,
+      "learning_rate": 1.3779935511701717e-05,
+      "loss": 0.4908,
+      "step": 6881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21070036258642488,
+      "learning_rate": 1.3778212598064002e-05,
+      "loss": 0.1286,
+      "step": 6882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27838016788335157,
+      "learning_rate": 1.3776489553586949e-05,
+      "loss": 0.2634,
+      "step": 6883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5184621555490639,
+      "learning_rate": 1.3774766378330221e-05,
+      "loss": 0.4013,
+      "step": 6884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.615158987776567,
+      "learning_rate": 1.3773043072353503e-05,
+      "loss": 0.4176,
+      "step": 6885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32917103198871883,
+      "learning_rate": 1.3771319635716459e-05,
+      "loss": 0.2412,
+      "step": 6886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48537442050696666,
+      "learning_rate": 1.376959606847878e-05,
+      "loss": 0.3794,
+      "step": 6887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.437233197013368,
+      "learning_rate": 1.376787237070015e-05,
+      "loss": 0.3628,
+      "step": 6888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3363888104947281,
+      "learning_rate": 1.3766148542440265e-05,
+      "loss": 0.2289,
+      "step": 6889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2641410487640712,
+      "learning_rate": 1.3764424583758816e-05,
+      "loss": 0.1768,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3430245182460297,
+      "learning_rate": 1.3762700494715506e-05,
+      "loss": 0.3178,
+      "step": 6891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3694883870067206,
+      "learning_rate": 1.3760976275370039e-05,
+      "loss": 0.2562,
+      "step": 6892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6870317668549498,
+      "learning_rate": 1.3759251925782127e-05,
+      "loss": 0.4852,
+      "step": 6893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2038599651172182,
+      "learning_rate": 1.3757527446011479e-05,
+      "loss": 0.8427,
+      "step": 6894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2434935746268074,
+      "learning_rate": 1.375580283611782e-05,
+      "loss": 0.1944,
+      "step": 6895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3256197247270627,
+      "learning_rate": 1.3754078096160871e-05,
+      "loss": 0.2661,
+      "step": 6896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7596028889004636,
+      "learning_rate": 1.3752353226200359e-05,
+      "loss": 0.4223,
+      "step": 6897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3579206665965782,
+      "learning_rate": 1.3750628226296012e-05,
+      "loss": 0.2608,
+      "step": 6898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42991083239765604,
+      "learning_rate": 1.3748903096507576e-05,
+      "loss": 0.2457,
+      "step": 6899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46276695668103407,
+      "learning_rate": 1.3747177836894783e-05,
+      "loss": 0.4311,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33938255692060043,
+      "learning_rate": 1.3745452447517384e-05,
+      "loss": 0.253,
+      "step": 6901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33146128208950326,
+      "learning_rate": 1.3743726928435129e-05,
+      "loss": 0.1271,
+      "step": 6902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4516192355561343,
+      "learning_rate": 1.3742001279707771e-05,
+      "loss": 0.3194,
+      "step": 6903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3831757519966381,
+      "learning_rate": 1.3740275501395068e-05,
+      "loss": 0.2835,
+      "step": 6904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7703354652013342,
+      "learning_rate": 1.3738549593556787e-05,
+      "loss": 0.255,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6825758957008587,
+      "learning_rate": 1.3736823556252694e-05,
+      "loss": 0.5165,
+      "step": 6906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3866563201141378,
+      "learning_rate": 1.373509738954256e-05,
+      "loss": 0.3017,
+      "step": 6907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2428939493291775,
+      "learning_rate": 1.3733371093486168e-05,
+      "loss": 0.2285,
+      "step": 6908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6721461211190297,
+      "learning_rate": 1.3731644668143291e-05,
+      "loss": 0.4576,
+      "step": 6909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3443745352273169,
+      "learning_rate": 1.3729918113573723e-05,
+      "loss": 0.2499,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3699209177126759,
+      "learning_rate": 1.3728191429837247e-05,
+      "loss": 0.3297,
+      "step": 6911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9218896511237294,
+      "learning_rate": 1.3726464616993667e-05,
+      "loss": 0.5562,
+      "step": 6912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36061835491190963,
+      "learning_rate": 1.3724737675102774e-05,
+      "loss": 0.251,
+      "step": 6913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6697499130698259,
+      "learning_rate": 1.3723010604224381e-05,
+      "loss": 0.4177,
+      "step": 6914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22123181145234125,
+      "learning_rate": 1.3721283404418283e-05,
+      "loss": 0.1735,
+      "step": 6915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31925352597321077,
+      "learning_rate": 1.3719556075744306e-05,
+      "loss": 0.2455,
+      "step": 6916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9239303856324638,
+      "learning_rate": 1.3717828618262261e-05,
+      "loss": 0.5894,
+      "step": 6917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5162709979789786,
+      "learning_rate": 1.3716101032031972e-05,
+      "loss": 0.3261,
+      "step": 6918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29582580195856606,
+      "learning_rate": 1.371437331711326e-05,
+      "loss": 0.2552,
+      "step": 6919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8585276824390256,
+      "learning_rate": 1.3712645473565964e-05,
+      "loss": 0.4624,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3036535597617637,
+      "learning_rate": 1.3710917501449911e-05,
+      "loss": 0.1622,
+      "step": 6921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.375439624708647,
+      "learning_rate": 1.3709189400824948e-05,
+      "loss": 0.2807,
+      "step": 6922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38903684691791973,
+      "learning_rate": 1.3707461171750916e-05,
+      "loss": 0.2963,
+      "step": 6923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3030138881612745,
+      "learning_rate": 1.370573281428766e-05,
+      "loss": 0.8033,
+      "step": 6924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3366170348258912,
+      "learning_rate": 1.3704004328495037e-05,
+      "loss": 0.2032,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8395341835724895,
+      "learning_rate": 1.3702275714432905e-05,
+      "loss": 0.4634,
+      "step": 6926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22712153378717959,
+      "learning_rate": 1.3700546972161121e-05,
+      "loss": 0.2303,
+      "step": 6927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3062410578990784,
+      "learning_rate": 1.3698818101739554e-05,
+      "loss": 0.1977,
+      "step": 6928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0418163317880769,
+      "learning_rate": 1.3697089103228081e-05,
+      "loss": 0.7044,
+      "step": 6929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7665050260008015,
+      "learning_rate": 1.3695359976686568e-05,
+      "loss": 0.52,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2718208821347409,
+      "learning_rate": 1.3693630722174898e-05,
+      "loss": 0.225,
+      "step": 6931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5994959857163412,
+      "learning_rate": 1.3691901339752955e-05,
+      "loss": 0.48,
+      "step": 6932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42763359444722393,
+      "learning_rate": 1.3690171829480628e-05,
+      "loss": 0.279,
+      "step": 6933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2819399462682956,
+      "learning_rate": 1.3688442191417805e-05,
+      "loss": 0.169,
+      "step": 6934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3855701660647398,
+      "learning_rate": 1.3686712425624393e-05,
+      "loss": 0.3225,
+      "step": 6935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9462468602813544,
+      "learning_rate": 1.3684982532160285e-05,
+      "loss": 0.5966,
+      "step": 6936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34783075864199253,
+      "learning_rate": 1.3683252511085391e-05,
+      "loss": 0.2796,
+      "step": 6937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.702993674672384,
+      "learning_rate": 1.3681522362459623e-05,
+      "loss": 0.3554,
+      "step": 6938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2971323853404743,
+      "learning_rate": 1.3679792086342892e-05,
+      "loss": 0.2779,
+      "step": 6939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2884143111686183,
+      "learning_rate": 1.367806168279512e-05,
+      "loss": 0.2304,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5212153825592891,
+      "learning_rate": 1.3676331151876227e-05,
+      "loss": 0.2851,
+      "step": 6941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4431896490235908,
+      "learning_rate": 1.3674600493646146e-05,
+      "loss": 0.3398,
+      "step": 6942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4475954113556971,
+      "learning_rate": 1.367286970816481e-05,
+      "loss": 0.2763,
+      "step": 6943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.669305965541524,
+      "learning_rate": 1.3671138795492155e-05,
+      "loss": 0.3324,
+      "step": 6944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3080889359061354,
+      "learning_rate": 1.3669407755688117e-05,
+      "loss": 0.1614,
+      "step": 6945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40932396074492605,
+      "learning_rate": 1.366767658881265e-05,
+      "loss": 0.2977,
+      "step": 6946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32701274366526195,
+      "learning_rate": 1.36659452949257e-05,
+      "loss": 0.2876,
+      "step": 6947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7057822249995881,
+      "learning_rate": 1.3664213874087223e-05,
+      "loss": 0.3661,
+      "step": 6948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3797999918281963,
+      "learning_rate": 1.3662482326357172e-05,
+      "loss": 0.28,
+      "step": 6949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5054705360107299,
+      "learning_rate": 1.366075065179552e-05,
+      "loss": 0.373,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4181068108631131,
+      "learning_rate": 1.3659018850462226e-05,
+      "loss": 0.2382,
+      "step": 6951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2658521324344886,
+      "learning_rate": 1.3657286922417272e-05,
+      "loss": 0.1952,
+      "step": 6952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33355700539346833,
+      "learning_rate": 1.3655554867720623e-05,
+      "loss": 0.2598,
+      "step": 6953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7434397980880477,
+      "learning_rate": 1.3653822686432271e-05,
+      "loss": 0.3192,
+      "step": 6954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30312485770426095,
+      "learning_rate": 1.3652090378612198e-05,
+      "loss": 0.2706,
+      "step": 6955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0086654469423801,
+      "learning_rate": 1.3650357944320387e-05,
+      "loss": 0.6798,
+      "step": 6956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2880087205351191,
+      "learning_rate": 1.3648625383616841e-05,
+      "loss": 0.1437,
+      "step": 6957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29183807559357594,
+      "learning_rate": 1.3646892696561554e-05,
+      "loss": 0.2153,
+      "step": 6958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36485493480482983,
+      "learning_rate": 1.3645159883214528e-05,
+      "loss": 0.3043,
+      "step": 6959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6233292098008584,
+      "learning_rate": 1.3643426943635774e-05,
+      "loss": 0.4441,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35202902887011656,
+      "learning_rate": 1.36416938778853e-05,
+      "loss": 0.1655,
+      "step": 6961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4126226172553717,
+      "learning_rate": 1.3639960686023126e-05,
+      "loss": 0.3344,
+      "step": 6962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37428085092700025,
+      "learning_rate": 1.3638227368109268e-05,
+      "loss": 0.3312,
+      "step": 6963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30610032926280945,
+      "learning_rate": 1.3636493924203756e-05,
+      "loss": 0.1934,
+      "step": 6964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31909149409640514,
+      "learning_rate": 1.3634760354366612e-05,
+      "loss": 0.2953,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.816373087029119,
+      "learning_rate": 1.3633026658657872e-05,
+      "loss": 0.6193,
+      "step": 6966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2950274786523385,
+      "learning_rate": 1.3631292837137577e-05,
+      "loss": 0.2384,
+      "step": 6967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34463380268308075,
+      "learning_rate": 1.3629558889865768e-05,
+      "loss": 0.2693,
+      "step": 6968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.445343970358979,
+      "learning_rate": 1.3627824816902494e-05,
+      "loss": 0.7602,
+      "step": 6969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2910854511165072,
+      "learning_rate": 1.3626090618307796e-05,
+      "loss": 0.2366,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3727574615944739,
+      "learning_rate": 1.3624356294141738e-05,
+      "loss": 0.347,
+      "step": 6971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4251921909257409,
+      "learning_rate": 1.3622621844464379e-05,
+      "loss": 0.2888,
+      "step": 6972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3357843692589245,
+      "learning_rate": 1.362088726933578e-05,
+      "loss": 0.2384,
+      "step": 6973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36993298536045216,
+      "learning_rate": 1.361915256881601e-05,
+      "loss": 0.2396,
+      "step": 6974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3614041412393671,
+      "learning_rate": 1.3617417742965144e-05,
+      "loss": 0.3096,
+      "step": 6975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37952388985458196,
+      "learning_rate": 1.3615682791843257e-05,
+      "loss": 0.3231,
+      "step": 6976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3901412474468786,
+      "learning_rate": 1.3613947715510429e-05,
+      "loss": 0.3072,
+      "step": 6977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36218398221854425,
+      "learning_rate": 1.3612212514026746e-05,
+      "loss": 0.3417,
+      "step": 6978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8055590911647899,
+      "learning_rate": 1.3610477187452303e-05,
+      "loss": 0.4594,
+      "step": 6979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2289632841507798,
+      "learning_rate": 1.3608741735847186e-05,
+      "loss": 0.1503,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6391920773085977,
+      "learning_rate": 1.3607006159271503e-05,
+      "loss": 0.4503,
+      "step": 6981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3301091566046304,
+      "learning_rate": 1.3605270457785346e-05,
+      "loss": 0.3016,
+      "step": 6982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3309863876921998,
+      "learning_rate": 1.3603534631448831e-05,
+      "loss": 0.2646,
+      "step": 6983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2898443402936977,
+      "learning_rate": 1.3601798680322068e-05,
+      "loss": 0.7883,
+      "step": 6984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34812557437044317,
+      "learning_rate": 1.3600062604465168e-05,
+      "loss": 0.2385,
+      "step": 6985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.277167302217143,
+      "learning_rate": 1.3598326403938255e-05,
+      "loss": 0.2632,
+      "step": 6986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5174657148756132,
+      "learning_rate": 1.3596590078801458e-05,
+      "loss": 0.3016,
+      "step": 6987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4792618762975345,
+      "learning_rate": 1.3594853629114896e-05,
+      "loss": 0.3499,
+      "step": 6988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.41025013782946984,
+      "learning_rate": 1.359311705493871e-05,
+      "loss": 0.312,
+      "step": 6989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3738632934541031,
+      "learning_rate": 1.3591380356333038e-05,
+      "loss": 0.2848,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5999221057686619,
+      "learning_rate": 1.3589643533358013e-05,
+      "loss": 0.3353,
+      "step": 6991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32699926243927424,
+      "learning_rate": 1.358790658607379e-05,
+      "loss": 0.2355,
+      "step": 6992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43293676610358867,
+      "learning_rate": 1.3586169514540512e-05,
+      "loss": 0.2488,
+      "step": 6993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32886181991925645,
+      "learning_rate": 1.3584432318818344e-05,
+      "loss": 0.3064,
+      "step": 6994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35377350978253397,
+      "learning_rate": 1.3582694998967434e-05,
+      "loss": 0.316,
+      "step": 6995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.685901668579314,
+      "learning_rate": 1.3580957555047953e-05,
+      "loss": 0.4758,
+      "step": 6996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6877362501402533,
+      "learning_rate": 1.3579219987120065e-05,
+      "loss": 0.4549,
+      "step": 6997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27094743803612736,
+      "learning_rate": 1.3577482295243944e-05,
+      "loss": 0.2349,
+      "step": 6998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3247502461542125,
+      "learning_rate": 1.3575744479479764e-05,
+      "loss": 0.2856,
+      "step": 6999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6147448714871042,
+      "learning_rate": 1.3574006539887707e-05,
+      "loss": 0.2349,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38562569593393564,
+      "learning_rate": 1.3572268476527954e-05,
+      "loss": 0.2908,
+      "step": 7001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3642951593240954,
+      "learning_rate": 1.3570530289460701e-05,
+      "loss": 0.3264,
+      "step": 7002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5235074663626574,
+      "learning_rate": 1.3568791978746137e-05,
+      "loss": 0.2494,
+      "step": 7003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2952802191250271,
+      "learning_rate": 1.356705354444446e-05,
+      "loss": 0.2524,
+      "step": 7004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.946358029879268,
+      "learning_rate": 1.3565314986615871e-05,
+      "loss": 0.4903,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23563139165705427,
+      "learning_rate": 1.3563576305320579e-05,
+      "loss": 0.1791,
+      "step": 7006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40333290157473234,
+      "learning_rate": 1.356183750061879e-05,
+      "loss": 0.3251,
+      "step": 7007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2051917632433689,
+      "learning_rate": 1.3560098572570725e-05,
+      "loss": 0.821,
+      "step": 7008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37974187511583635,
+      "learning_rate": 1.35583595212366e-05,
+      "loss": 0.1655,
+      "step": 7009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3522486679741345,
+      "learning_rate": 1.3556620346676633e-05,
+      "loss": 0.2931,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6771158949431482,
+      "learning_rate": 1.355488104895106e-05,
+      "loss": 0.3763,
+      "step": 7011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2299304342135175,
+      "learning_rate": 1.3553141628120107e-05,
+      "loss": 0.1646,
+      "step": 7012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33815698261225796,
+      "learning_rate": 1.3551402084244014e-05,
+      "loss": 0.2277,
+      "step": 7013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3550244974142543,
+      "learning_rate": 1.3549662417383018e-05,
+      "loss": 0.3468,
+      "step": 7014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.499478939324147,
+      "learning_rate": 1.354792262759737e-05,
+      "loss": 0.7654,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33123396540469313,
+      "learning_rate": 1.3546182714947309e-05,
+      "loss": 0.2452,
+      "step": 7016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30626982106831946,
+      "learning_rate": 1.3544442679493095e-05,
+      "loss": 0.2432,
+      "step": 7017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36381971525818,
+      "learning_rate": 1.3542702521294981e-05,
+      "loss": 0.3203,
+      "step": 7018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2968935067872526,
+      "learning_rate": 1.3540962240413233e-05,
+      "loss": 0.2144,
+      "step": 7019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9758630593144416,
+      "learning_rate": 1.3539221836908113e-05,
+      "loss": 0.6957,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47204156166840916,
+      "learning_rate": 1.3537481310839897e-05,
+      "loss": 0.351,
+      "step": 7021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31181724975707065,
+      "learning_rate": 1.353574066226885e-05,
+      "loss": 0.2339,
+      "step": 7022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8749376420778082,
+      "learning_rate": 1.353399989125526e-05,
+      "loss": 0.4849,
+      "step": 7023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2859353718730429,
+      "learning_rate": 1.3532258997859404e-05,
+      "loss": 0.197,
+      "step": 7024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3325129256222981,
+      "learning_rate": 1.3530517982141574e-05,
+      "loss": 0.2746,
+      "step": 7025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3507133854637492,
+      "learning_rate": 1.3528776844162052e-05,
+      "loss": 0.277,
+      "step": 7026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5776205911314057,
+      "learning_rate": 1.3527035583981145e-05,
+      "loss": 0.4077,
+      "step": 7027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3720017793660261,
+      "learning_rate": 1.3525294201659145e-05,
+      "loss": 0.2941,
+      "step": 7028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6266844281338557,
+      "learning_rate": 1.3523552697256359e-05,
+      "loss": 0.3631,
+      "step": 7029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3606261379793363,
+      "learning_rate": 1.3521811070833095e-05,
+      "loss": 0.2194,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37657437220486956,
+      "learning_rate": 1.3520069322449663e-05,
+      "loss": 0.2835,
+      "step": 7031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8620238092878418,
+      "learning_rate": 1.3518327452166385e-05,
+      "loss": 0.4009,
+      "step": 7032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.654820668093201,
+      "learning_rate": 1.3516585460043576e-05,
+      "loss": 0.4591,
+      "step": 7033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3013622890908199,
+      "learning_rate": 1.3514843346141566e-05,
+      "loss": 0.2671,
+      "step": 7034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4458870442813115,
+      "learning_rate": 1.3513101110520678e-05,
+      "loss": 0.3693,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19906848799136267,
+      "learning_rate": 1.3511358753241254e-05,
+      "loss": 0.1152,
+      "step": 7036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32092656053178614,
+      "learning_rate": 1.3509616274363623e-05,
+      "loss": 0.2576,
+      "step": 7037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3677600497999778,
+      "learning_rate": 1.3507873673948137e-05,
+      "loss": 0.3478,
+      "step": 7038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9170207065182442,
+      "learning_rate": 1.3506130952055132e-05,
+      "loss": 0.4109,
+      "step": 7039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35086745820449605,
+      "learning_rate": 1.3504388108744966e-05,
+      "loss": 0.3033,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.099592946999764,
+      "learning_rate": 1.3502645144077987e-05,
+      "loss": 0.5843,
+      "step": 7041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.212123843237808,
+      "learning_rate": 1.350090205811456e-05,
+      "loss": 0.1806,
+      "step": 7042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33900519569136367,
+      "learning_rate": 1.3499158850915044e-05,
+      "loss": 0.2861,
+      "step": 7043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8621314063592519,
+      "learning_rate": 1.3497415522539807e-05,
+      "loss": 0.6942,
+      "step": 7044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3746493330911956,
+      "learning_rate": 1.3495672073049221e-05,
+      "loss": 0.3091,
+      "step": 7045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3969087626583301,
+      "learning_rate": 1.3493928502503664e-05,
+      "loss": 0.2964,
+      "step": 7046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48381748985794387,
+      "learning_rate": 1.3492184810963512e-05,
+      "loss": 0.3625,
+      "step": 7047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3345842253885811,
+      "learning_rate": 1.349044099848915e-05,
+      "loss": 0.2105,
+      "step": 7048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30381301666422506,
+      "learning_rate": 1.3488697065140964e-05,
+      "loss": 0.1968,
+      "step": 7049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41334840598392936,
+      "learning_rate": 1.348695301097935e-05,
+      "loss": 0.3586,
+      "step": 7050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7832888653278328,
+      "learning_rate": 1.3485208836064705e-05,
+      "loss": 0.4857,
+      "step": 7051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33241008219443263,
+      "learning_rate": 1.3483464540457428e-05,
+      "loss": 0.2076,
+      "step": 7052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49777611048133946,
+      "learning_rate": 1.348172012421792e-05,
+      "loss": 0.3706,
+      "step": 7053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5163034090447539,
+      "learning_rate": 1.3479975587406595e-05,
+      "loss": 0.3745,
+      "step": 7054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22940568788016039,
+      "learning_rate": 1.3478230930083868e-05,
+      "loss": 0.1524,
+      "step": 7055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9077703091017053,
+      "learning_rate": 1.3476486152310152e-05,
+      "loss": 0.444,
+      "step": 7056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5325747150064066,
+      "learning_rate": 1.3474741254145868e-05,
+      "loss": 0.397,
+      "step": 7057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2938576127412327,
+      "learning_rate": 1.3472996235651446e-05,
+      "loss": 0.227,
+      "step": 7058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2376508865292748,
+      "learning_rate": 1.3471251096887312e-05,
+      "loss": 0.6284,
+      "step": 7059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.400031588718824,
+      "learning_rate": 1.3469505837913903e-05,
+      "loss": 0.2609,
+      "step": 7060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3362257964775687,
+      "learning_rate": 1.3467760458791656e-05,
+      "loss": 0.2845,
+      "step": 7061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4483199711195374,
+      "learning_rate": 1.3466014959581013e-05,
+      "loss": 0.2888,
+      "step": 7062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0500941292938073,
+      "learning_rate": 1.3464269340342422e-05,
+      "loss": 0.6551,
+      "step": 7063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35210336824936767,
+      "learning_rate": 1.346252360113633e-05,
+      "loss": 0.2358,
+      "step": 7064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3657147809379971,
+      "learning_rate": 1.3460777742023202e-05,
+      "loss": 0.2663,
+      "step": 7065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.503373542099447,
+      "learning_rate": 1.3459031763063482e-05,
+      "loss": 0.365,
+      "step": 7066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42739082768414866,
+      "learning_rate": 1.3457285664317645e-05,
+      "loss": 0.3174,
+      "step": 7067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.358482347181096,
+      "learning_rate": 1.3455539445846151e-05,
+      "loss": 0.2484,
+      "step": 7068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3661152740979383,
+      "learning_rate": 1.3453793107709476e-05,
+      "loss": 0.3302,
+      "step": 7069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3814990470872192,
+      "learning_rate": 1.3452046649968091e-05,
+      "loss": 0.2572,
+      "step": 7070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3260559821888633,
+      "learning_rate": 1.3450300072682485e-05,
+      "loss": 0.1877,
+      "step": 7071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.823857743699866,
+      "learning_rate": 1.3448553375913132e-05,
+      "loss": 0.4651,
+      "step": 7072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35011088470733437,
+      "learning_rate": 1.3446806559720525e-05,
+      "loss": 0.3223,
+      "step": 7073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35841773511333275,
+      "learning_rate": 1.3445059624165156e-05,
+      "loss": 0.3149,
+      "step": 7074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9612427340225015,
+      "learning_rate": 1.3443312569307517e-05,
+      "loss": 0.5688,
+      "step": 7075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25064310434852977,
+      "learning_rate": 1.3441565395208114e-05,
+      "loss": 0.1798,
+      "step": 7076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5838118579171294,
+      "learning_rate": 1.343981810192745e-05,
+      "loss": 0.3781,
+      "step": 7077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40143952342120387,
+      "learning_rate": 1.3438070689526033e-05,
+      "loss": 0.2896,
+      "step": 7078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36313108466885224,
+      "learning_rate": 1.3436323158064373e-05,
+      "loss": 0.2885,
+      "step": 7079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7995956838032602,
+      "learning_rate": 1.3434575507602991e-05,
+      "loss": 0.5193,
+      "step": 7080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4270076800549399,
+      "learning_rate": 1.3432827738202407e-05,
+      "loss": 0.2954,
+      "step": 7081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3050621499863082,
+      "learning_rate": 1.3431079849923153e-05,
+      "loss": 0.1848,
+      "step": 7082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3510960749297572,
+      "learning_rate": 1.3429331842825742e-05,
+      "loss": 0.2506,
+      "step": 7083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6715341783485381,
+      "learning_rate": 1.342758371697072e-05,
+      "loss": 0.3628,
+      "step": 7084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4085613591910861,
+      "learning_rate": 1.342583547241862e-05,
+      "loss": 0.2762,
+      "step": 7085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32885154493809926,
+      "learning_rate": 1.3424087109229986e-05,
+      "loss": 0.3057,
+      "step": 7086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1248597962403253,
+      "learning_rate": 1.3422338627465364e-05,
+      "loss": 0.7771,
+      "step": 7087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2362658079430318,
+      "learning_rate": 1.3420590027185301e-05,
+      "loss": 0.1333,
+      "step": 7088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27743124046471523,
+      "learning_rate": 1.3418841308450353e-05,
+      "loss": 0.2448,
+      "step": 7089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0089479383803033,
+      "learning_rate": 1.3417092471321076e-05,
+      "loss": 0.439,
+      "step": 7090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.34819709868420223,
+      "learning_rate": 1.3415343515858035e-05,
+      "loss": 0.2192,
+      "step": 7091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5070973840014404,
+      "learning_rate": 1.3413594442121796e-05,
+      "loss": 0.3956,
+      "step": 7092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3532321667025554,
+      "learning_rate": 1.3411845250172928e-05,
+      "loss": 0.2936,
+      "step": 7093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3256958107973817,
+      "learning_rate": 1.3410095940072004e-05,
+      "loss": 0.1915,
+      "step": 7094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26218402620283915,
+      "learning_rate": 1.3408346511879604e-05,
+      "loss": 0.1904,
+      "step": 7095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9728861537090818,
+      "learning_rate": 1.340659696565631e-05,
+      "loss": 0.461,
+      "step": 7096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29866630802365746,
+      "learning_rate": 1.3404847301462713e-05,
+      "loss": 0.2247,
+      "step": 7097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5127063422711399,
+      "learning_rate": 1.3403097519359397e-05,
+      "loss": 0.3778,
+      "step": 7098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0938596974308756,
+      "learning_rate": 1.3401347619406966e-05,
+      "loss": 0.8696,
+      "step": 7099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3440710027431961,
+      "learning_rate": 1.3399597601666008e-05,
+      "loss": 0.214,
+      "step": 7100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33249458020431705,
+      "learning_rate": 1.3397847466197133e-05,
+      "loss": 0.2699,
+      "step": 7101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3150673143037075,
+      "learning_rate": 1.3396097213060943e-05,
+      "loss": 0.3037,
+      "step": 7102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9226442021708258,
+      "learning_rate": 1.3394346842318058e-05,
+      "loss": 0.5105,
+      "step": 7103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3788209793940605,
+      "learning_rate": 1.3392596354029084e-05,
+      "loss": 0.2471,
+      "step": 7104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36634771446717296,
+      "learning_rate": 1.3390845748254645e-05,
+      "loss": 0.3481,
+      "step": 7105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6158781133726381,
+      "learning_rate": 1.3389095025055363e-05,
+      "loss": 0.4241,
+      "step": 7106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3631589753231446,
+      "learning_rate": 1.3387344184491869e-05,
+      "loss": 0.2813,
+      "step": 7107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27955123400877324,
+      "learning_rate": 1.3385593226624787e-05,
+      "loss": 0.1621,
+      "step": 7108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3783695100880318,
+      "learning_rate": 1.338384215151476e-05,
+      "loss": 0.2813,
+      "step": 7109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6041099103124075,
+      "learning_rate": 1.3382090959222425e-05,
+      "loss": 0.3282,
+      "step": 7110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1035000418322451,
+      "learning_rate": 1.3380339649808425e-05,
+      "loss": 0.4968,
+      "step": 7111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4884716979614059,
+      "learning_rate": 1.337858822333341e-05,
+      "loss": 0.3715,
+      "step": 7112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3107210491119011,
+      "learning_rate": 1.3376836679858026e-05,
+      "loss": 0.2974,
+      "step": 7113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22911849054150063,
+      "learning_rate": 1.3375085019442937e-05,
+      "loss": 0.1643,
+      "step": 7114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5691552869077963,
+      "learning_rate": 1.3373333242148796e-05,
+      "loss": 0.3392,
+      "step": 7115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49816342959997756,
+      "learning_rate": 1.337158134803627e-05,
+      "loss": 0.3428,
+      "step": 7116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4375022231297627,
+      "learning_rate": 1.3369829337166031e-05,
+      "loss": 0.3051,
+      "step": 7117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5267758958092926,
+      "learning_rate": 1.3368077209598744e-05,
+      "loss": 0.3613,
+      "step": 7118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3807706588481771,
+      "learning_rate": 1.3366324965395088e-05,
+      "loss": 0.3324,
+      "step": 7119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23055407507354644,
+      "learning_rate": 1.3364572604615744e-05,
+      "loss": 0.156,
+      "step": 7120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47250351493960807,
+      "learning_rate": 1.3362820127321391e-05,
+      "loss": 0.3263,
+      "step": 7121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40387974042821045,
+      "learning_rate": 1.3361067533572726e-05,
+      "loss": 0.2879,
+      "step": 7122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6065089402375893,
+      "learning_rate": 1.3359314823430436e-05,
+      "loss": 0.5049,
+      "step": 7123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48384521210853454,
+      "learning_rate": 1.335756199695522e-05,
+      "loss": 0.3052,
+      "step": 7124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30084930743731136,
+      "learning_rate": 1.3355809054207774e-05,
+      "loss": 0.2649,
+      "step": 7125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9024344911983133,
+      "learning_rate": 1.3354055995248805e-05,
+      "loss": 0.6037,
+      "step": 7126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19286218457362345,
+      "learning_rate": 1.335230282013902e-05,
+      "loss": 0.1028,
+      "step": 7127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3553929943930889,
+      "learning_rate": 1.3350549528939135e-05,
+      "loss": 0.3059,
+      "step": 7128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3556388786278789,
+      "learning_rate": 1.3348796121709862e-05,
+      "loss": 0.3285,
+      "step": 7129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6547730919119816,
+      "learning_rate": 1.3347042598511926e-05,
+      "loss": 0.2819,
+      "step": 7130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36796600659289685,
+      "learning_rate": 1.3345288959406045e-05,
+      "loss": 0.2986,
+      "step": 7131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37729407521418595,
+      "learning_rate": 1.3343535204452953e-05,
+      "loss": 0.224,
+      "step": 7132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2750629669350354,
+      "learning_rate": 1.3341781333713381e-05,
+      "loss": 0.1908,
+      "step": 7133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44690214839091336,
+      "learning_rate": 1.3340027347248068e-05,
+      "loss": 0.3296,
+      "step": 7134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1545043346285377,
+      "learning_rate": 1.3338273245117745e-05,
+      "loss": 0.4603,
+      "step": 7135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40566944235161556,
+      "learning_rate": 1.3336519027383168e-05,
+      "loss": 0.3405,
+      "step": 7136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31670328181132457,
+      "learning_rate": 1.3334764694105079e-05,
+      "loss": 0.2335,
+      "step": 7137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8579819449562148,
+      "learning_rate": 1.3333010245344232e-05,
+      "loss": 0.5027,
+      "step": 7138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30341933853948494,
+      "learning_rate": 1.3331255681161386e-05,
+      "loss": 0.17,
+      "step": 7139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.38527026955972754,
+      "learning_rate": 1.3329501001617294e-05,
+      "loss": 0.2346,
+      "step": 7140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5092949576398142,
+      "learning_rate": 1.332774620677273e-05,
+      "loss": 0.3329,
+      "step": 7141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7811634793100883,
+      "learning_rate": 1.3325991296688455e-05,
+      "loss": 0.7817,
+      "step": 7142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3345933775300051,
+      "learning_rate": 1.3324236271425245e-05,
+      "loss": 0.2055,
+      "step": 7143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1826649808614909,
+      "learning_rate": 1.3322481131043876e-05,
+      "loss": 0.6385,
+      "step": 7144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25436360208374936,
+      "learning_rate": 1.332072587560513e-05,
+      "loss": 0.2162,
+      "step": 7145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3245351746881674,
+      "learning_rate": 1.3318970505169786e-05,
+      "loss": 0.2445,
+      "step": 7146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7222227073899474,
+      "learning_rate": 1.3317215019798639e-05,
+      "loss": 0.497,
+      "step": 7147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3898074555572452,
+      "learning_rate": 1.3315459419552477e-05,
+      "loss": 0.3174,
+      "step": 7148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.357549111885812,
+      "learning_rate": 1.33137037044921e-05,
+      "loss": 0.2622,
+      "step": 7149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8507678942147663,
+      "learning_rate": 1.3311947874678306e-05,
+      "loss": 0.2878,
+      "step": 7150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33806956641060104,
+      "learning_rate": 1.3310191930171898e-05,
+      "loss": 0.2648,
+      "step": 7151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41027889521502914,
+      "learning_rate": 1.3308435871033687e-05,
+      "loss": 0.2778,
+      "step": 7152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3561677104384658,
+      "learning_rate": 1.3306679697324485e-05,
+      "loss": 0.3,
+      "step": 7153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4096178462824716,
+      "learning_rate": 1.3304923409105104e-05,
+      "loss": 0.2723,
+      "step": 7154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.38934669547306056,
+      "learning_rate": 1.3303167006436371e-05,
+      "loss": 0.278,
+      "step": 7155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37412796658975134,
+      "learning_rate": 1.3301410489379103e-05,
+      "loss": 0.2778,
+      "step": 7156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7818448980631783,
+      "learning_rate": 1.3299653857994135e-05,
+      "loss": 0.4927,
+      "step": 7157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3268093806978687,
+      "learning_rate": 1.3297897112342294e-05,
+      "loss": 0.2912,
+      "step": 7158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.686057582809523,
+      "learning_rate": 1.3296140252484417e-05,
+      "loss": 0.419,
+      "step": 7159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2794205927146474,
+      "learning_rate": 1.3294383278481346e-05,
+      "loss": 0.2731,
+      "step": 7160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.306089163013698,
+      "learning_rate": 1.3292626190393923e-05,
+      "loss": 0.2232,
+      "step": 7161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0595231500511526,
+      "learning_rate": 1.3290868988282999e-05,
+      "loss": 0.5485,
+      "step": 7162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7733042625073977,
+      "learning_rate": 1.328911167220942e-05,
+      "loss": 0.3675,
+      "step": 7163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3460305444373183,
+      "learning_rate": 1.3287354242234047e-05,
+      "loss": 0.2629,
+      "step": 7164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3792220485449821,
+      "learning_rate": 1.3285596698417738e-05,
+      "loss": 0.3372,
+      "step": 7165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2128910770084968,
+      "learning_rate": 1.3283839040821355e-05,
+      "loss": 0.1166,
+      "step": 7166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35463706912168896,
+      "learning_rate": 1.3282081269505771e-05,
+      "loss": 0.2832,
+      "step": 7167
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5153840336512678,
+      "learning_rate": 1.3280323384531852e-05,
+      "loss": 0.3804,
+      "step": 7168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5089755863319722,
+      "learning_rate": 1.3278565385960476e-05,
+      "loss": 0.3391,
+      "step": 7169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4031836019221641,
+      "learning_rate": 1.3276807273852522e-05,
+      "loss": 0.2773,
+      "step": 7170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48049808170008806,
+      "learning_rate": 1.3275049048268869e-05,
+      "loss": 0.4095,
+      "step": 7171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2558281539455274,
+      "learning_rate": 1.327329070927041e-05,
+      "loss": 0.1998,
+      "step": 7172
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.38630900592037487,
+      "learning_rate": 1.3271532256918036e-05,
+      "loss": 0.295,
+      "step": 7173
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39008618040812376,
+      "learning_rate": 1.326977369127264e-05,
+      "loss": 0.2581,
+      "step": 7174
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6205933584816866,
+      "learning_rate": 1.326801501239512e-05,
+      "loss": 0.4402,
+      "step": 7175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31882272613528434,
+      "learning_rate": 1.3266256220346383e-05,
+      "loss": 0.1977,
+      "step": 7176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4407270029368251,
+      "learning_rate": 1.3264497315187334e-05,
+      "loss": 0.3436,
+      "step": 7177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.38673637353069634,
+      "learning_rate": 1.326273829697888e-05,
+      "loss": 0.244,
+      "step": 7178
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27113858546133,
+      "learning_rate": 1.3260979165781942e-05,
+      "loss": 0.1703,
+      "step": 7179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4419932070737161,
+      "learning_rate": 1.3259219921657436e-05,
+      "loss": 0.3672,
+      "step": 7180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45245896039209665,
+      "learning_rate": 1.3257460564666283e-05,
+      "loss": 0.3697,
+      "step": 7181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32808870623561165,
+      "learning_rate": 1.3255701094869408e-05,
+      "loss": 0.1927,
+      "step": 7182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1927858978015542,
+      "learning_rate": 1.325394151232775e-05,
+      "loss": 0.7927,
+      "step": 7183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4014879560091093,
+      "learning_rate": 1.3252181817102235e-05,
+      "loss": 0.3279,
+      "step": 7184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23244466150471954,
+      "learning_rate": 1.3250422009253802e-05,
+      "loss": 0.1628,
+      "step": 7185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4222245281834194,
+      "learning_rate": 1.3248662088843395e-05,
+      "loss": 0.2875,
+      "step": 7186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6467322169281083,
+      "learning_rate": 1.3246902055931961e-05,
+      "loss": 0.4428,
+      "step": 7187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.38186519736687974,
+      "learning_rate": 1.3245141910580446e-05,
+      "loss": 0.3217,
+      "step": 7188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3362029065075712,
+      "learning_rate": 1.324338165284981e-05,
+      "loss": 0.2656,
+      "step": 7189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40943953574665665,
+      "learning_rate": 1.3241621282801002e-05,
+      "loss": 0.267,
+      "step": 7190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4200553653950379,
+      "learning_rate": 1.3239860800494993e-05,
+      "loss": 0.3042,
+      "step": 7191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4031697492165725,
+      "learning_rate": 1.3238100205992739e-05,
+      "loss": 0.286,
+      "step": 7192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4600727603852913,
+      "learning_rate": 1.3236339499355217e-05,
+      "loss": 0.3152,
+      "step": 7193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3986539572582897,
+      "learning_rate": 1.3234578680643394e-05,
+      "loss": 0.3215,
+      "step": 7194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4169919148480683,
+      "learning_rate": 1.3232817749918256e-05,
+      "loss": 0.3092,
+      "step": 7195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3654175654707768,
+      "learning_rate": 1.3231056707240775e-05,
+      "loss": 0.3141,
+      "step": 7196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36406119272639237,
+      "learning_rate": 1.322929555267194e-05,
+      "loss": 0.3191,
+      "step": 7197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31941887730802226,
+      "learning_rate": 1.3227534286272741e-05,
+      "loss": 0.2926,
+      "step": 7198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2620408151493418,
+      "learning_rate": 1.3225772908104165e-05,
+      "loss": 0.0687,
+      "step": 7199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30135243734827055,
+      "learning_rate": 1.3224011418227215e-05,
+      "loss": 0.268,
+      "step": 7200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5366585174326642,
+      "learning_rate": 1.3222249816702885e-05,
+      "loss": 0.3695,
+      "step": 7201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.590033821849479,
+      "learning_rate": 1.3220488103592184e-05,
+      "loss": 0.381,
+      "step": 7202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3385399492527422,
+      "learning_rate": 1.3218726278956117e-05,
+      "loss": 0.2911,
+      "step": 7203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39860811495487514,
+      "learning_rate": 1.32169643428557e-05,
+      "loss": 0.393,
+      "step": 7204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25190821405668745,
+      "learning_rate": 1.3215202295351946e-05,
+      "loss": 0.1631,
+      "step": 7205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5825783438242356,
+      "learning_rate": 1.3213440136505872e-05,
+      "loss": 0.3373,
+      "step": 7206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46201854347371046,
+      "learning_rate": 1.3211677866378505e-05,
+      "loss": 0.3525,
+      "step": 7207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33494447559061813,
+      "learning_rate": 1.3209915485030872e-05,
+      "loss": 0.2768,
+      "step": 7208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5209714849109807,
+      "learning_rate": 1.3208152992524004e-05,
+      "loss": 0.334,
+      "step": 7209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37421682429920466,
+      "learning_rate": 1.3206390388918937e-05,
+      "loss": 0.3097,
+      "step": 7210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30114301634692187,
+      "learning_rate": 1.3204627674276706e-05,
+      "loss": 0.2041,
+      "step": 7211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30929089763537454,
+      "learning_rate": 1.320286484865836e-05,
+      "loss": 0.2373,
+      "step": 7212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5349211331103826,
+      "learning_rate": 1.3201101912124938e-05,
+      "loss": 0.4244,
+      "step": 7213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7905658849528135,
+      "learning_rate": 1.31993388647375e-05,
+      "loss": 0.4311,
+      "step": 7214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36725135920394,
+      "learning_rate": 1.3197575706557089e-05,
+      "loss": 0.2345,
+      "step": 7215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32845897394315215,
+      "learning_rate": 1.3195812437644771e-05,
+      "loss": 0.3105,
+      "step": 7216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29800837631567817,
+      "learning_rate": 1.3194049058061606e-05,
+      "loss": 0.1937,
+      "step": 7217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3439065437611227,
+      "learning_rate": 1.3192285567868662e-05,
+      "loss": 0.2248,
+      "step": 7218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5481495321388145,
+      "learning_rate": 1.3190521967127e-05,
+      "loss": 0.425,
+      "step": 7219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5796551641203623,
+      "learning_rate": 1.3188758255897705e-05,
+      "loss": 0.3267,
+      "step": 7220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5415045151729231,
+      "learning_rate": 1.3186994434241845e-05,
+      "loss": 0.2433,
+      "step": 7221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5321969246900784,
+      "learning_rate": 1.3185230502220508e-05,
+      "loss": 0.3406,
+      "step": 7222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2422636781051529,
+      "learning_rate": 1.3183466459894774e-05,
+      "loss": 0.2162,
+      "step": 7223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4012142713831807,
+      "learning_rate": 1.3181702307325732e-05,
+      "loss": 0.3343,
+      "step": 7224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4794094927724806,
+      "learning_rate": 1.3179938044574478e-05,
+      "loss": 0.2723,
+      "step": 7225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9303568882746546,
+      "learning_rate": 1.3178173671702106e-05,
+      "loss": 0.4344,
+      "step": 7226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6859661533697369,
+      "learning_rate": 1.3176409188769715e-05,
+      "loss": 0.4222,
+      "step": 7227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27217942492278485,
+      "learning_rate": 1.3174644595838411e-05,
+      "loss": 0.2338,
+      "step": 7228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3033095269716523,
+      "learning_rate": 1.3172879892969302e-05,
+      "loss": 0.1795,
+      "step": 7229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7835735876374561,
+      "learning_rate": 1.3171115080223498e-05,
+      "loss": 0.4627,
+      "step": 7230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6595555477893355,
+      "learning_rate": 1.3169350157662115e-05,
+      "loss": 0.2827,
+      "step": 7231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45468179409236015,
+      "learning_rate": 1.3167585125346271e-05,
+      "loss": 0.3494,
+      "step": 7232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6287105703435774,
+      "learning_rate": 1.3165819983337093e-05,
+      "loss": 0.3983,
+      "step": 7233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4100430193967899,
+      "learning_rate": 1.3164054731695706e-05,
+      "loss": 0.2633,
+      "step": 7234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3711852345127576,
+      "learning_rate": 1.3162289370483239e-05,
+      "loss": 0.2647,
+      "step": 7235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37139285226115515,
+      "learning_rate": 1.3160523899760824e-05,
+      "loss": 0.3054,
+      "step": 7236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4224536498097396,
+      "learning_rate": 1.3158758319589604e-05,
+      "loss": 0.3383,
+      "step": 7237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5765797692400455,
+      "learning_rate": 1.3156992630030719e-05,
+      "loss": 0.2837,
+      "step": 7238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38557712218290724,
+      "learning_rate": 1.3155226831145316e-05,
+      "loss": 0.2873,
+      "step": 7239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31621584525691937,
+      "learning_rate": 1.3153460922994543e-05,
+      "loss": 0.2885,
+      "step": 7240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29877105273955934,
+      "learning_rate": 1.3151694905639553e-05,
+      "loss": 0.164,
+      "step": 7241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5334188626320423,
+      "learning_rate": 1.3149928779141506e-05,
+      "loss": 0.4122,
+      "step": 7242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3828319018088226,
+      "learning_rate": 1.3148162543561557e-05,
+      "loss": 0.3108,
+      "step": 7243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27405350590100247,
+      "learning_rate": 1.3146396198960881e-05,
+      "loss": 0.2322,
+      "step": 7244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7653235271564569,
+      "learning_rate": 1.3144629745400632e-05,
+      "loss": 0.3892,
+      "step": 7245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40187932943275734,
+      "learning_rate": 1.3142863182941996e-05,
+      "loss": 0.2975,
+      "step": 7246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3764866541385056,
+      "learning_rate": 1.3141096511646141e-05,
+      "loss": 0.256,
+      "step": 7247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5491287247419514,
+      "learning_rate": 1.3139329731574248e-05,
+      "loss": 0.402,
+      "step": 7248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33817869550264207,
+      "learning_rate": 1.3137562842787502e-05,
+      "loss": 0.314,
+      "step": 7249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1455881396931549,
+      "learning_rate": 1.3135795845347091e-05,
+      "loss": 0.786,
+      "step": 7250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.260305077900469,
+      "learning_rate": 1.3134028739314204e-05,
+      "loss": 0.1759,
+      "step": 7251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3841956692601113,
+      "learning_rate": 1.3132261524750038e-05,
+      "loss": 0.2806,
+      "step": 7252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0728971327464343,
+      "learning_rate": 1.3130494201715786e-05,
+      "loss": 0.5737,
+      "step": 7253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4996657504738954,
+      "learning_rate": 1.312872677027266e-05,
+      "loss": 0.2811,
+      "step": 7254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4259459115097978,
+      "learning_rate": 1.3126959230481855e-05,
+      "loss": 0.3429,
+      "step": 7255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3836148834285065,
+      "learning_rate": 1.312519158240459e-05,
+      "loss": 0.3439,
+      "step": 7256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.195603144277449,
+      "learning_rate": 1.3123423826102074e-05,
+      "loss": 0.1038,
+      "step": 7257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3946116029677441,
+      "learning_rate": 1.3121655961635523e-05,
+      "loss": 0.3305,
+      "step": 7258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.429020225728886,
+      "learning_rate": 1.311988798906616e-05,
+      "loss": 0.3417,
+      "step": 7259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43981050593951315,
+      "learning_rate": 1.3118119908455214e-05,
+      "loss": 0.2962,
+      "step": 7260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36302632134687873,
+      "learning_rate": 1.3116351719863906e-05,
+      "loss": 0.297,
+      "step": 7261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44066367931176303,
+      "learning_rate": 1.3114583423353476e-05,
+      "loss": 0.3472,
+      "step": 7262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.301064575674817,
+      "learning_rate": 1.3112815018985154e-05,
+      "loss": 0.2503,
+      "step": 7263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31585777990855474,
+      "learning_rate": 1.311104650682018e-05,
+      "loss": 0.2119,
+      "step": 7264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7170533675281239,
+      "learning_rate": 1.3109277886919802e-05,
+      "loss": 0.4184,
+      "step": 7265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7914186842307775,
+      "learning_rate": 1.3107509159345262e-05,
+      "loss": 0.5291,
+      "step": 7266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.267323184964963,
+      "learning_rate": 1.3105740324157817e-05,
+      "loss": 0.2167,
+      "step": 7267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4627191275190162,
+      "learning_rate": 1.3103971381418713e-05,
+      "loss": 0.3843,
+      "step": 7268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2892535892426708,
+      "learning_rate": 1.310220233118922e-05,
+      "loss": 0.1798,
+      "step": 7269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42638022005244247,
+      "learning_rate": 1.3100433173530589e-05,
+      "loss": 0.222,
+      "step": 7270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6288156366173488,
+      "learning_rate": 1.3098663908504091e-05,
+      "loss": 0.3793,
+      "step": 7271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5008266918389817,
+      "learning_rate": 1.3096894536170994e-05,
+      "loss": 0.3658,
+      "step": 7272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34497131150322446,
+      "learning_rate": 1.3095125056592575e-05,
+      "loss": 0.2732,
+      "step": 7273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9667430455753272,
+      "learning_rate": 1.3093355469830107e-05,
+      "loss": 0.5544,
+      "step": 7274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2635365400899789,
+      "learning_rate": 1.3091585775944873e-05,
+      "loss": 0.2102,
+      "step": 7275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3788616946807546,
+      "learning_rate": 1.3089815974998154e-05,
+      "loss": 0.275,
+      "step": 7276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.897185868184525,
+      "learning_rate": 1.3088046067051243e-05,
+      "loss": 0.3707,
+      "step": 7277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9592855330664415,
+      "learning_rate": 1.308627605216543e-05,
+      "loss": 0.624,
+      "step": 7278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.347671499106493,
+      "learning_rate": 1.308450593040201e-05,
+      "loss": 0.2543,
+      "step": 7279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4070033059546702,
+      "learning_rate": 1.3082735701822281e-05,
+      "loss": 0.2806,
+      "step": 7280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30304605216105446,
+      "learning_rate": 1.3080965366487548e-05,
+      "loss": 0.1902,
+      "step": 7281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45872537263094,
+      "learning_rate": 1.3079194924459118e-05,
+      "loss": 0.2992,
+      "step": 7282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5951235637057674,
+      "learning_rate": 1.3077424375798295e-05,
+      "loss": 0.2988,
+      "step": 7283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.182568248414101,
+      "learning_rate": 1.3075653720566404e-05,
+      "loss": 0.7021,
+      "step": 7284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34373368560518014,
+      "learning_rate": 1.3073882958824755e-05,
+      "loss": 0.2923,
+      "step": 7285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.157375136004901,
+      "learning_rate": 1.307211209063467e-05,
+      "loss": 0.7641,
+      "step": 7286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2623129816600877,
+      "learning_rate": 1.3070341116057476e-05,
+      "loss": 0.2098,
+      "step": 7287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3824276930333361,
+      "learning_rate": 1.3068570035154503e-05,
+      "loss": 0.2981,
+      "step": 7288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4109022486432033,
+      "learning_rate": 1.306679884798708e-05,
+      "loss": 0.2912,
+      "step": 7289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5582932916423002,
+      "learning_rate": 1.3065027554616547e-05,
+      "loss": 0.302,
+      "step": 7290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34401297473123493,
+      "learning_rate": 1.3063256155104239e-05,
+      "loss": 0.2613,
+      "step": 7291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5353972926544106,
+      "learning_rate": 1.3061484649511503e-05,
+      "loss": 0.4175,
+      "step": 7292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5389049209563237,
+      "learning_rate": 1.3059713037899683e-05,
+      "loss": 0.3414,
+      "step": 7293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40682939769048926,
+      "learning_rate": 1.3057941320330134e-05,
+      "loss": 0.2576,
+      "step": 7294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.270271026585561,
+      "learning_rate": 1.3056169496864208e-05,
+      "loss": 0.2742,
+      "step": 7295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28083438824338586,
+      "learning_rate": 1.3054397567563266e-05,
+      "loss": 0.134,
+      "step": 7296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3797228091663637,
+      "learning_rate": 1.3052625532488663e-05,
+      "loss": 0.2747,
+      "step": 7297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8666369485609857,
+      "learning_rate": 1.3050853391701774e-05,
+      "loss": 0.6257,
+      "step": 7298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3821053180321841,
+      "learning_rate": 1.304908114526396e-05,
+      "loss": 0.3131,
+      "step": 7299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.327705833281437,
+      "learning_rate": 1.3047308793236599e-05,
+      "loss": 0.2423,
+      "step": 7300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32553038184470945,
+      "learning_rate": 1.3045536335681064e-05,
+      "loss": 0.2019,
+      "step": 7301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0963164215457855,
+      "learning_rate": 1.3043763772658739e-05,
+      "loss": 0.7163,
+      "step": 7302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2779405820004573,
+      "learning_rate": 1.3041991104231004e-05,
+      "loss": 0.2391,
+      "step": 7303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1690442501518032,
+      "learning_rate": 1.3040218330459249e-05,
+      "loss": 0.795,
+      "step": 7304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6931092185091348,
+      "learning_rate": 1.3038445451404862e-05,
+      "loss": 0.4273,
+      "step": 7305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33132146745020336,
+      "learning_rate": 1.3036672467129241e-05,
+      "loss": 0.2127,
+      "step": 7306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3629514950779212,
+      "learning_rate": 1.3034899377693782e-05,
+      "loss": 0.313,
+      "step": 7307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.297556760270983,
+      "learning_rate": 1.3033126183159887e-05,
+      "loss": 0.2063,
+      "step": 7308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2941708532876371,
+      "learning_rate": 1.3031352883588965e-05,
+      "loss": 0.2085,
+      "step": 7309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7852519937723851,
+      "learning_rate": 1.3029579479042423e-05,
+      "loss": 0.5261,
+      "step": 7310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4725984968530518,
+      "learning_rate": 1.3027805969581674e-05,
+      "loss": 0.327,
+      "step": 7311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.758050405985813,
+      "learning_rate": 1.3026032355268132e-05,
+      "loss": 0.3737,
+      "step": 7312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27727090117219044,
+      "learning_rate": 1.3024258636163221e-05,
+      "loss": 0.2148,
+      "step": 7313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27413178777534775,
+      "learning_rate": 1.3022484812328365e-05,
+      "loss": 0.2706,
+      "step": 7314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5412617681591296,
+      "learning_rate": 1.3020710883824987e-05,
+      "loss": 0.3523,
+      "step": 7315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45275149176644774,
+      "learning_rate": 1.3018936850714524e-05,
+      "loss": 0.2881,
+      "step": 7316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7923988739792034,
+      "learning_rate": 1.3017162713058404e-05,
+      "loss": 0.4677,
+      "step": 7317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39895495917725043,
+      "learning_rate": 1.3015388470918072e-05,
+      "loss": 0.2999,
+      "step": 7318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23533187055257196,
+      "learning_rate": 1.3013614124354969e-05,
+      "loss": 0.201,
+      "step": 7319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41061255560760335,
+      "learning_rate": 1.3011839673430536e-05,
+      "loss": 0.2942,
+      "step": 7320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3836948488403221,
+      "learning_rate": 1.3010065118206223e-05,
+      "loss": 0.3036,
+      "step": 7321
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8039043999491806,
+      "learning_rate": 1.3008290458743486e-05,
+      "loss": 0.3215,
+      "step": 7322
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3526267061237842,
+      "learning_rate": 1.3006515695103779e-05,
+      "loss": 0.3402,
+      "step": 7323
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3874565419841915,
+      "learning_rate": 1.3004740827348563e-05,
+      "loss": 0.2971,
+      "step": 7324
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4916981175451748,
+      "learning_rate": 1.3002965855539303e-05,
+      "loss": 0.3836,
+      "step": 7325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1963129197249394,
+      "learning_rate": 1.300119077973746e-05,
+      "loss": 0.1841,
+      "step": 7326
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3374193053149697,
+      "learning_rate": 1.2999415600004515e-05,
+      "loss": 0.2758,
+      "step": 7327
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9488092760590442,
+      "learning_rate": 1.2997640316401934e-05,
+      "loss": 0.4981,
+      "step": 7328
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6829150533818018,
+      "learning_rate": 1.2995864928991198e-05,
+      "loss": 0.3648,
+      "step": 7329
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5376235231735698,
+      "learning_rate": 1.2994089437833788e-05,
+      "loss": 0.3335,
+      "step": 7330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30612015196427483,
+      "learning_rate": 1.2992313842991189e-05,
+      "loss": 0.2881,
+      "step": 7331
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35369907883631363,
+      "learning_rate": 1.2990538144524894e-05,
+      "loss": 0.2176,
+      "step": 7332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6509986938076566,
+      "learning_rate": 1.2988762342496386e-05,
+      "loss": 0.366,
+      "step": 7333
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4988119608715147,
+      "learning_rate": 1.298698643696717e-05,
+      "loss": 0.3416,
+      "step": 7334
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2772638591631809,
+      "learning_rate": 1.2985210427998743e-05,
+      "loss": 0.2226,
+      "step": 7335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3951269297273973,
+      "learning_rate": 1.2983434315652606e-05,
+      "loss": 0.2974,
+      "step": 7336
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5102487142923199,
+      "learning_rate": 1.2981658099990266e-05,
+      "loss": 0.3945,
+      "step": 7337
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4414157270716968,
+      "learning_rate": 1.2979881781073235e-05,
+      "loss": 0.3039,
+      "step": 7338
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2909371552105041,
+      "learning_rate": 1.2978105358963026e-05,
+      "loss": 0.2469,
+      "step": 7339
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5649634546078088,
+      "learning_rate": 1.2976328833721157e-05,
+      "loss": 0.3579,
+      "step": 7340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3848501314342395,
+      "learning_rate": 1.2974552205409147e-05,
+      "loss": 0.2769,
+      "step": 7341
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3049961532656385,
+      "learning_rate": 1.2972775474088524e-05,
+      "loss": 0.2088,
+      "step": 7342
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34617722088590974,
+      "learning_rate": 1.297099863982081e-05,
+      "loss": 0.3243,
+      "step": 7343
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7259079907052219,
+      "learning_rate": 1.2969221702667547e-05,
+      "loss": 0.5144,
+      "step": 7344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38372525596694496,
+      "learning_rate": 1.2967444662690261e-05,
+      "loss": 0.2622,
+      "step": 7345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5209288626419487,
+      "learning_rate": 1.2965667519950494e-05,
+      "loss": 0.3831,
+      "step": 7346
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23884674114338472,
+      "learning_rate": 1.2963890274509789e-05,
+      "loss": 0.2358,
+      "step": 7347
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38743989121030176,
+      "learning_rate": 1.2962112926429691e-05,
+      "loss": 0.1331,
+      "step": 7348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43142029339243576,
+      "learning_rate": 1.2960335475771748e-05,
+      "loss": 0.3201,
+      "step": 7349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3556038132379808,
+      "learning_rate": 1.2958557922597516e-05,
+      "loss": 0.34,
+      "step": 7350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5629136553826217,
+      "learning_rate": 1.2956780266968552e-05,
+      "loss": 0.409,
+      "step": 7351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3375499747099566,
+      "learning_rate": 1.2955002508946413e-05,
+      "loss": 0.2461,
+      "step": 7352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26063028491401785,
+      "learning_rate": 1.2953224648592664e-05,
+      "loss": 0.1594,
+      "step": 7353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3286104317910206,
+      "learning_rate": 1.2951446685968874e-05,
+      "loss": 0.2815,
+      "step": 7354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33363688183935714,
+      "learning_rate": 1.294966862113661e-05,
+      "loss": 0.2217,
+      "step": 7355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6185356603885251,
+      "learning_rate": 1.2947890454157448e-05,
+      "loss": 0.4454,
+      "step": 7356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5052412376086751,
+      "learning_rate": 1.294611218509297e-05,
+      "loss": 0.3758,
+      "step": 7357
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31425887157261073,
+      "learning_rate": 1.2944333814004748e-05,
+      "loss": 0.2567,
+      "step": 7358
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3024510934501898,
+      "learning_rate": 1.2942555340954377e-05,
+      "loss": 0.2408,
+      "step": 7359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2721480122115268,
+      "learning_rate": 1.294077676600344e-05,
+      "loss": 0.1685,
+      "step": 7360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4334281254358977,
+      "learning_rate": 1.293899808921353e-05,
+      "loss": 0.3266,
+      "step": 7361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33199319188044935,
+      "learning_rate": 1.2937219310646242e-05,
+      "loss": 0.2896,
+      "step": 7362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5113244424261909,
+      "learning_rate": 1.2935440430363177e-05,
+      "loss": 0.3779,
+      "step": 7363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4415669571896982,
+      "learning_rate": 1.2933661448425933e-05,
+      "loss": 0.3187,
+      "step": 7364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.21104509788054693,
+      "learning_rate": 1.2931882364896125e-05,
+      "loss": 0.1266,
+      "step": 7365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39429657285673275,
+      "learning_rate": 1.2930103179835352e-05,
+      "loss": 0.3014,
+      "step": 7366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34096887677103693,
+      "learning_rate": 1.2928323893305233e-05,
+      "loss": 0.2779,
+      "step": 7367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5398849722107039,
+      "learning_rate": 1.2926544505367384e-05,
+      "loss": 0.3547,
+      "step": 7368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7583480793502161,
+      "learning_rate": 1.2924765016083427e-05,
+      "loss": 0.5297,
+      "step": 7369
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30692372439687576,
+      "learning_rate": 1.2922985425514977e-05,
+      "loss": 0.279,
+      "step": 7370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.270890793342731,
+      "learning_rate": 1.2921205733723672e-05,
+      "loss": 0.1777,
+      "step": 7371
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33009884553113095,
+      "learning_rate": 1.2919425940771138e-05,
+      "loss": 0.2584,
+      "step": 7372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3529582845492279,
+      "learning_rate": 1.2917646046719007e-05,
+      "loss": 0.304,
+      "step": 7373
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4422830853360155,
+      "learning_rate": 1.2915866051628923e-05,
+      "loss": 0.3514,
+      "step": 7374
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3936738090536721,
+      "learning_rate": 1.291408595556252e-05,
+      "loss": 0.258,
+      "step": 7375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3645622895731119,
+      "learning_rate": 1.2912305758581444e-05,
+      "loss": 0.3064,
+      "step": 7376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8760977353616266,
+      "learning_rate": 1.2910525460747346e-05,
+      "loss": 0.4878,
+      "step": 7377
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1923060008162484,
+      "learning_rate": 1.290874506212188e-05,
+      "loss": 0.1669,
+      "step": 7378
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.402353938670631,
+      "learning_rate": 1.2906964562766691e-05,
+      "loss": 0.3496,
+      "step": 7379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7286714606166612,
+      "learning_rate": 1.290518396274345e-05,
+      "loss": 0.437,
+      "step": 7380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35226470412228356,
+      "learning_rate": 1.290340326211381e-05,
+      "loss": 0.2338,
+      "step": 7381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3621217534989928,
+      "learning_rate": 1.290162246093944e-05,
+      "loss": 0.2912,
+      "step": 7382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4861495391034015,
+      "learning_rate": 1.289984155928201e-05,
+      "loss": 0.3546,
+      "step": 7383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.64343988125234,
+      "learning_rate": 1.289806055720319e-05,
+      "loss": 0.2483,
+      "step": 7384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3655898241983234,
+      "learning_rate": 1.2896279454764659e-05,
+      "loss": 0.2909,
+      "step": 7385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29398028827370154,
+      "learning_rate": 1.28944982520281e-05,
+      "loss": 0.2718,
+      "step": 7386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1800139635163578,
+      "learning_rate": 1.2892716949055184e-05,
+      "loss": 0.7545,
+      "step": 7387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3023709880759031,
+      "learning_rate": 1.2890935545907608e-05,
+      "loss": 0.2228,
+      "step": 7388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6965321600545177,
+      "learning_rate": 1.2889154042647056e-05,
+      "loss": 0.4642,
+      "step": 7389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35911214934551505,
+      "learning_rate": 1.2887372439335224e-05,
+      "loss": 0.3357,
+      "step": 7390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29662533744407216,
+      "learning_rate": 1.2885590736033808e-05,
+      "loss": 0.2187,
+      "step": 7391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27718955489176644,
+      "learning_rate": 1.2883808932804512e-05,
+      "loss": 0.1913,
+      "step": 7392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4086851848096156,
+      "learning_rate": 1.2882027029709034e-05,
+      "loss": 0.3497,
+      "step": 7393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3038815146416857,
+      "learning_rate": 1.2880245026809085e-05,
+      "loss": 0.2307,
+      "step": 7394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.500569565176652,
+      "learning_rate": 1.2878462924166374e-05,
+      "loss": 0.661,
+      "step": 7395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6963263635658477,
+      "learning_rate": 1.2876680721842616e-05,
+      "loss": 0.4307,
+      "step": 7396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.333813168785862,
+      "learning_rate": 1.2874898419899528e-05,
+      "loss": 0.2022,
+      "step": 7397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24787304670295343,
+      "learning_rate": 1.287311601839883e-05,
+      "loss": 0.2155,
+      "step": 7398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7433576356848729,
+      "learning_rate": 1.2871333517402251e-05,
+      "loss": 0.481,
+      "step": 7399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3939320064161549,
+      "learning_rate": 1.2869550916971512e-05,
+      "loss": 0.2881,
+      "step": 7400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8308971616875137,
+      "learning_rate": 1.2867768217168353e-05,
+      "loss": 0.3781,
+      "step": 7401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3497647575246814,
+      "learning_rate": 1.28659854180545e-05,
+      "loss": 0.3359,
+      "step": 7402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36233416373040706,
+      "learning_rate": 1.2864202519691698e-05,
+      "loss": 0.2455,
+      "step": 7403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21250382304746632,
+      "learning_rate": 1.2862419522141684e-05,
+      "loss": 0.1056,
+      "step": 7404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48055525319680503,
+      "learning_rate": 1.2860636425466207e-05,
+      "loss": 0.353,
+      "step": 7405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2929548307001326,
+      "learning_rate": 1.285885322972701e-05,
+      "loss": 0.267,
+      "step": 7406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9742707277009065,
+      "learning_rate": 1.2857069934985851e-05,
+      "loss": 0.3784,
+      "step": 7407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7845705775706029,
+      "learning_rate": 1.2855286541304481e-05,
+      "loss": 0.4938,
+      "step": 7408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3324160422402967,
+      "learning_rate": 1.2853503048744664e-05,
+      "loss": 0.2625,
+      "step": 7409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2891167340341053,
+      "learning_rate": 1.2851719457368157e-05,
+      "loss": 0.2068,
+      "step": 7410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42210199453028185,
+      "learning_rate": 1.2849935767236729e-05,
+      "loss": 0.3184,
+      "step": 7411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36308547586520445,
+      "learning_rate": 1.284815197841215e-05,
+      "loss": 0.3046,
+      "step": 7412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1643204041845043,
+      "learning_rate": 1.2846368090956185e-05,
+      "loss": 0.4728,
+      "step": 7413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31753775412680657,
+      "learning_rate": 1.284458410493062e-05,
+      "loss": 0.2662,
+      "step": 7414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38438890919644864,
+      "learning_rate": 1.2842800020397226e-05,
+      "loss": 0.2863,
+      "step": 7415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23075903001714582,
+      "learning_rate": 1.2841015837417792e-05,
+      "loss": 0.1904,
+      "step": 7416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3437999763100494,
+      "learning_rate": 1.2839231556054101e-05,
+      "loss": 0.2819,
+      "step": 7417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3856629904474758,
+      "learning_rate": 1.2837447176367944e-05,
+      "loss": 0.2964,
+      "step": 7418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6560835963344139,
+      "learning_rate": 1.2835662698421112e-05,
+      "loss": 0.3934,
+      "step": 7419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7348691624408498,
+      "learning_rate": 1.2833878122275407e-05,
+      "loss": 0.2447,
+      "step": 7420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34732153973656654,
+      "learning_rate": 1.283209344799262e-05,
+      "loss": 0.2776,
+      "step": 7421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29962271123904893,
+      "learning_rate": 1.283030867563456e-05,
+      "loss": 0.2691,
+      "step": 7422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7215084346836231,
+      "learning_rate": 1.282852380526303e-05,
+      "loss": 0.4093,
+      "step": 7423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3187780573476562,
+      "learning_rate": 1.2826738836939844e-05,
+      "loss": 0.2666,
+      "step": 7424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3051997295007057,
+      "learning_rate": 1.2824953770726813e-05,
+      "loss": 0.2971,
+      "step": 7425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5136628961032486,
+      "learning_rate": 1.2823168606685756e-05,
+      "loss": 0.3615,
+      "step": 7426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3726773739401942,
+      "learning_rate": 1.2821383344878491e-05,
+      "loss": 0.2014,
+      "step": 7427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1260492763383276,
+      "learning_rate": 1.2819597985366843e-05,
+      "loss": 0.5823,
+      "step": 7428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49804903503850456,
+      "learning_rate": 1.2817812528212635e-05,
+      "loss": 0.3953,
+      "step": 7429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.278617679213449,
+      "learning_rate": 1.2816026973477702e-05,
+      "loss": 0.2334,
+      "step": 7430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8077408746603993,
+      "learning_rate": 1.2814241321223876e-05,
+      "loss": 0.5714,
+      "step": 7431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2938940132824784,
+      "learning_rate": 1.2812455571512996e-05,
+      "loss": 0.178,
+      "step": 7432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31693402945068383,
+      "learning_rate": 1.28106697244069e-05,
+      "loss": 0.1713,
+      "step": 7433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39056029624267147,
+      "learning_rate": 1.2808883779967429e-05,
+      "loss": 0.2973,
+      "step": 7434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8784300285705219,
+      "learning_rate": 1.2807097738256436e-05,
+      "loss": 0.4679,
+      "step": 7435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6096405649192743,
+      "learning_rate": 1.2805311599335768e-05,
+      "loss": 0.3627,
+      "step": 7436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.280902071047559,
+      "learning_rate": 1.2803525363267281e-05,
+      "loss": 0.2223,
+      "step": 7437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33005696907181864,
+      "learning_rate": 1.2801739030112833e-05,
+      "loss": 0.2775,
+      "step": 7438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5559933672937305,
+      "learning_rate": 1.279995259993428e-05,
+      "loss": 0.3064,
+      "step": 7439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39420878524487984,
+      "learning_rate": 1.279816607279349e-05,
+      "loss": 0.2677,
+      "step": 7440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5102506758920938,
+      "learning_rate": 1.279637944875233e-05,
+      "loss": 0.3898,
+      "step": 7441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3100867254059578,
+      "learning_rate": 1.2794592727872665e-05,
+      "loss": 0.2904,
+      "step": 7442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2778272955079186,
+      "learning_rate": 1.279280591021638e-05,
+      "loss": 0.1926,
+      "step": 7443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3652081080576812,
+      "learning_rate": 1.2791018995845343e-05,
+      "loss": 0.2066,
+      "step": 7444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4003295351901793,
+      "learning_rate": 1.278923198482144e-05,
+      "loss": 0.2906,
+      "step": 7445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31584749571079734,
+      "learning_rate": 1.2787444877206552e-05,
+      "loss": 0.2526,
+      "step": 7446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6723729764280729,
+      "learning_rate": 1.2785657673062567e-05,
+      "loss": 0.4328,
+      "step": 7447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4069163041642835,
+      "learning_rate": 1.2783870372451377e-05,
+      "loss": 0.3099,
+      "step": 7448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5471645897491344,
+      "learning_rate": 1.278208297543488e-05,
+      "loss": 0.3764,
+      "step": 7449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22582651123557201,
+      "learning_rate": 1.2780295482074965e-05,
+      "loss": 0.1677,
+      "step": 7450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38504551814048305,
+      "learning_rate": 1.2778507892433538e-05,
+      "loss": 0.2699,
+      "step": 7451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5300421426970018,
+      "learning_rate": 1.2776720206572502e-05,
+      "loss": 0.3693,
+      "step": 7452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4316438122749434,
+      "learning_rate": 1.2774932424553763e-05,
+      "loss": 0.3068,
+      "step": 7453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6039224321850404,
+      "learning_rate": 1.2773144546439235e-05,
+      "loss": 0.3833,
+      "step": 7454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41397690083601735,
+      "learning_rate": 1.2771356572290834e-05,
+      "loss": 0.3263,
+      "step": 7455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22875703408665155,
+      "learning_rate": 1.2769568502170469e-05,
+      "loss": 0.1126,
+      "step": 7456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38290759108779704,
+      "learning_rate": 1.276778033614007e-05,
+      "loss": 0.2816,
+      "step": 7457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3645825413303027,
+      "learning_rate": 1.2765992074261555e-05,
+      "loss": 0.3286,
+      "step": 7458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7066287348903324,
+      "learning_rate": 1.276420371659685e-05,
+      "loss": 0.3722,
+      "step": 7459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6244111465506637,
+      "learning_rate": 1.2762415263207895e-05,
+      "loss": 0.3676,
+      "step": 7460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2867319890884178,
+      "learning_rate": 1.2760626714156614e-05,
+      "loss": 0.2733,
+      "step": 7461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37856860799168346,
+      "learning_rate": 1.2758838069504952e-05,
+      "loss": 0.2388,
+      "step": 7462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27983356612817223,
+      "learning_rate": 1.2757049329314842e-05,
+      "loss": 0.1823,
+      "step": 7463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.675249528883432,
+      "learning_rate": 1.2755260493648235e-05,
+      "loss": 0.3907,
+      "step": 7464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3990296921748976,
+      "learning_rate": 1.2753471562567074e-05,
+      "loss": 0.3226,
+      "step": 7465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3446096915891851,
+      "learning_rate": 1.2751682536133313e-05,
+      "loss": 0.2236,
+      "step": 7466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5313072118472417,
+      "learning_rate": 1.2749893414408903e-05,
+      "loss": 0.3751,
+      "step": 7467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.265255518955803,
+      "learning_rate": 1.2748104197455804e-05,
+      "loss": 0.1889,
+      "step": 7468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3086628494538964,
+      "learning_rate": 1.274631488533597e-05,
+      "loss": 0.235,
+      "step": 7469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3955694497758381,
+      "learning_rate": 1.2744525478111379e-05,
+      "loss": 0.3382,
+      "step": 7470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7593369964854633,
+      "learning_rate": 1.274273597584398e-05,
+      "loss": 0.5741,
+      "step": 7471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38222913352863014,
+      "learning_rate": 1.2740946378595758e-05,
+      "loss": 0.1657,
+      "step": 7472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3301246287937124,
+      "learning_rate": 1.2739156686428675e-05,
+      "loss": 0.2843,
+      "step": 7473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47079369177307934,
+      "learning_rate": 1.2737366899404718e-05,
+      "loss": 0.309,
+      "step": 7474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6375823621608645,
+      "learning_rate": 1.273557701758586e-05,
+      "loss": 0.4102,
+      "step": 7475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3015373234296537,
+      "learning_rate": 1.2733787041034092e-05,
+      "loss": 0.2052,
+      "step": 7476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3550705089146975,
+      "learning_rate": 1.2731996969811393e-05,
+      "loss": 0.2863,
+      "step": 7477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7378788393900801,
+      "learning_rate": 1.2730206803979754e-05,
+      "loss": 0.5696,
+      "step": 7478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3263003907992609,
+      "learning_rate": 1.2728416543601177e-05,
+      "loss": 0.2251,
+      "step": 7479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.775695398769665,
+      "learning_rate": 1.2726626188737647e-05,
+      "loss": 0.4716,
+      "step": 7480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33048671823166975,
+      "learning_rate": 1.2724835739451172e-05,
+      "loss": 0.2793,
+      "step": 7481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2443173066345539,
+      "learning_rate": 1.272304519580375e-05,
+      "loss": 0.1975,
+      "step": 7482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37993338584071,
+      "learning_rate": 1.2721254557857394e-05,
+      "loss": 0.255,
+      "step": 7483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5994907663851344,
+      "learning_rate": 1.2719463825674105e-05,
+      "loss": 0.3759,
+      "step": 7484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35639175499384124,
+      "learning_rate": 1.2717672999315904e-05,
+      "loss": 0.2441,
+      "step": 7485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5549223599700548,
+      "learning_rate": 1.2715882078844804e-05,
+      "loss": 0.3907,
+      "step": 7486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5760563913885721,
+      "learning_rate": 1.2714091064322824e-05,
+      "loss": 0.3488,
+      "step": 7487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30595159376311326,
+      "learning_rate": 1.2712299955811987e-05,
+      "loss": 0.2508,
+      "step": 7488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2638652449536984,
+      "learning_rate": 1.2710508753374317e-05,
+      "loss": 0.2126,
+      "step": 7489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.237358245192641,
+      "learning_rate": 1.2708717457071849e-05,
+      "loss": 0.7199,
+      "step": 7490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36818599307685185,
+      "learning_rate": 1.270692606696661e-05,
+      "loss": 0.2631,
+      "step": 7491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.471986003217928,
+      "learning_rate": 1.2705134583120638e-05,
+      "loss": 0.3152,
+      "step": 7492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42344872331296374,
+      "learning_rate": 1.2703343005595972e-05,
+      "loss": 0.3291,
+      "step": 7493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36850684943447043,
+      "learning_rate": 1.2701551334454652e-05,
+      "loss": 0.2939,
+      "step": 7494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21811476150078793,
+      "learning_rate": 1.2699759569758727e-05,
+      "loss": 0.1028,
+      "step": 7495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6464980054789515,
+      "learning_rate": 1.2697967711570243e-05,
+      "loss": 0.4877,
+      "step": 7496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28044990754809157,
+      "learning_rate": 1.2696175759951254e-05,
+      "loss": 0.2786,
+      "step": 7497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6001584093412066,
+      "learning_rate": 1.2694383714963818e-05,
+      "loss": 0.3558,
+      "step": 7498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6795937624798115,
+      "learning_rate": 1.2692591576669983e-05,
+      "loss": 0.443,
+      "step": 7499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2836693554883349,
+      "learning_rate": 1.2690799345131824e-05,
+      "loss": 0.2405,
+      "step": 7500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3201123731313662,
+      "learning_rate": 1.2689007020411394e-05,
+      "loss": 0.2822,
+      "step": 7501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36880094607275504,
+      "learning_rate": 1.268721460257077e-05,
+      "loss": 0.2344,
+      "step": 7502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42415078679604923,
+      "learning_rate": 1.2685422091672017e-05,
+      "loss": 0.3176,
+      "step": 7503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0388651965272275,
+      "learning_rate": 1.2683629487777219e-05,
+      "loss": 0.4227,
+      "step": 7504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26138461355474946,
+      "learning_rate": 1.268183679094844e-05,
+      "loss": 0.2142,
+      "step": 7505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38880693292751656,
+      "learning_rate": 1.2680044001247774e-05,
+      "loss": 0.3062,
+      "step": 7506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28854957691060595,
+      "learning_rate": 1.2678251118737293e-05,
+      "loss": 0.2059,
+      "step": 7507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6409188766262047,
+      "learning_rate": 1.2676458143479095e-05,
+      "loss": 0.306,
+      "step": 7508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3106921649626693,
+      "learning_rate": 1.2674665075535264e-05,
+      "loss": 0.2657,
+      "step": 7509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9982593919162711,
+      "learning_rate": 1.26728719149679e-05,
+      "loss": 0.4393,
+      "step": 7510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0040110349618188,
+      "learning_rate": 1.2671078661839093e-05,
+      "loss": 0.2504,
+      "step": 7511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27760287035235526,
+      "learning_rate": 1.2669285316210948e-05,
+      "loss": 0.2207,
+      "step": 7512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33738457628735213,
+      "learning_rate": 1.2667491878145568e-05,
+      "loss": 0.2674,
+      "step": 7513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8728551700618518,
+      "learning_rate": 1.266569834770506e-05,
+      "loss": 0.4935,
+      "step": 7514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5237880393460295,
+      "learning_rate": 1.266390472495153e-05,
+      "loss": 0.2501,
+      "step": 7515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2321177960808543,
+      "learning_rate": 1.2662111009947096e-05,
+      "loss": 0.7564,
+      "step": 7516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3852958785860179,
+      "learning_rate": 1.2660317202753873e-05,
+      "loss": 0.308,
+      "step": 7517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2878776513060847,
+      "learning_rate": 1.2658523303433979e-05,
+      "loss": 0.2114,
+      "step": 7518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3866565714290425,
+      "learning_rate": 1.2656729312049536e-05,
+      "loss": 0.2787,
+      "step": 7519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44929076242454746,
+      "learning_rate": 1.265493522866267e-05,
+      "loss": 0.3211,
+      "step": 7520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31019924286079437,
+      "learning_rate": 1.2653141053335513e-05,
+      "loss": 0.2021,
+      "step": 7521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5257400746389835,
+      "learning_rate": 1.2651346786130199e-05,
+      "loss": 0.3338,
+      "step": 7522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7700473026585529,
+      "learning_rate": 1.2649552427108856e-05,
+      "loss": 0.4551,
+      "step": 7523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37836722990428473,
+      "learning_rate": 1.2647757976333626e-05,
+      "loss": 0.2697,
+      "step": 7524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40109303353556514,
+      "learning_rate": 1.2645963433866653e-05,
+      "loss": 0.2891,
+      "step": 7525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5801053269128108,
+      "learning_rate": 1.2644168799770075e-05,
+      "loss": 0.4479,
+      "step": 7526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3038245233952448,
+      "learning_rate": 1.2642374074106052e-05,
+      "loss": 0.2673,
+      "step": 7527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23955365444396562,
+      "learning_rate": 1.2640579256936723e-05,
+      "loss": 0.195,
+      "step": 7528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9894905671520695,
+      "learning_rate": 1.263878434832425e-05,
+      "loss": 0.5351,
+      "step": 7529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33128274197373925,
+      "learning_rate": 1.2636989348330791e-05,
+      "loss": 0.2585,
+      "step": 7530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7002081987136007,
+      "learning_rate": 1.2635194257018501e-05,
+      "loss": 0.3638,
+      "step": 7531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3775323790192974,
+      "learning_rate": 1.2633399074449548e-05,
+      "loss": 0.3146,
+      "step": 7532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33698150077075945,
+      "learning_rate": 1.2631603800686099e-05,
+      "loss": 0.279,
+      "step": 7533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18218076324549323,
+      "learning_rate": 1.2629808435790322e-05,
+      "loss": 0.1041,
+      "step": 7534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.185980333284646,
+      "learning_rate": 1.2628012979824394e-05,
+      "loss": 0.7281,
+      "step": 7535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34323992321440966,
+      "learning_rate": 1.262621743285049e-05,
+      "loss": 0.262,
+      "step": 7536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37375107807679175,
+      "learning_rate": 1.2624421794930785e-05,
+      "loss": 0.3399,
+      "step": 7537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9873288871171293,
+      "learning_rate": 1.262262606612747e-05,
+      "loss": 0.3846,
+      "step": 7538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4163728309097746,
+      "learning_rate": 1.2620830246502729e-05,
+      "loss": 0.2865,
+      "step": 7539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29402576393144825,
+      "learning_rate": 1.2619034336118742e-05,
+      "loss": 0.2692,
+      "step": 7540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3024954021730123,
+      "learning_rate": 1.2617238335037717e-05,
+      "loss": 0.222,
+      "step": 7541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4399043946526552,
+      "learning_rate": 1.2615442243321837e-05,
+      "loss": 0.2605,
+      "step": 7542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6267856143786402,
+      "learning_rate": 1.2613646061033303e-05,
+      "loss": 0.3875,
+      "step": 7543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.383120587325467,
+      "learning_rate": 1.261184978823432e-05,
+      "loss": 0.2875,
+      "step": 7544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3409458484018371,
+      "learning_rate": 1.261005342498709e-05,
+      "loss": 0.277,
+      "step": 7545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25057745497441336,
+      "learning_rate": 1.2608256971353825e-05,
+      "loss": 0.1904,
+      "step": 7546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9892133398630651,
+      "learning_rate": 1.2606460427396729e-05,
+      "loss": 0.0705,
+      "step": 7547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3220890480663572,
+      "learning_rate": 1.2604663793178025e-05,
+      "loss": 0.2779,
+      "step": 7548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3346067280468399,
+      "learning_rate": 1.2602867068759921e-05,
+      "loss": 0.3242,
+      "step": 7549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9802497675330882,
+      "learning_rate": 1.2601070254204645e-05,
+      "loss": 0.7108,
+      "step": 7550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3310026474654711,
+      "learning_rate": 1.2599273349574417e-05,
+      "loss": 0.2248,
+      "step": 7551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4692480509388019,
+      "learning_rate": 1.2597476354931466e-05,
+      "loss": 0.3678,
+      "step": 7552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26325901123873297,
+      "learning_rate": 1.2595679270338019e-05,
+      "loss": 0.2001,
+      "step": 7553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3018764077905086,
+      "learning_rate": 1.2593882095856314e-05,
+      "loss": 0.2367,
+      "step": 7554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6035545300935379,
+      "learning_rate": 1.259208483154858e-05,
+      "loss": 0.4614,
+      "step": 7555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3817271648715765,
+      "learning_rate": 1.2590287477477063e-05,
+      "loss": 0.3635,
+      "step": 7556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3606104677787802,
+      "learning_rate": 1.2588490033703999e-05,
+      "loss": 0.1954,
+      "step": 7557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33357966342409084,
+      "learning_rate": 1.2586692500291639e-05,
+      "loss": 0.2763,
+      "step": 7558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4054777299779856,
+      "learning_rate": 1.2584894877302224e-05,
+      "loss": 0.3124,
+      "step": 7559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2770112585553754,
+      "learning_rate": 1.2583097164798015e-05,
+      "loss": 0.2086,
+      "step": 7560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33701165254254206,
+      "learning_rate": 1.2581299362841262e-05,
+      "loss": 0.3406,
+      "step": 7561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5593734746120151,
+      "learning_rate": 1.2579501471494222e-05,
+      "loss": 0.7475,
+      "step": 7562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6377691899908655,
+      "learning_rate": 1.2577703490819157e-05,
+      "loss": 0.3114,
+      "step": 7563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27445686268636504,
+      "learning_rate": 1.2575905420878332e-05,
+      "loss": 0.2575,
+      "step": 7564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26622696284658587,
+      "learning_rate": 1.2574107261734012e-05,
+      "loss": 0.2002,
+      "step": 7565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37561943751948673,
+      "learning_rate": 1.2572309013448467e-05,
+      "loss": 0.2893,
+      "step": 7566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4144161359543641,
+      "learning_rate": 1.2570510676083974e-05,
+      "loss": 0.2954,
+      "step": 7567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36156343975898747,
+      "learning_rate": 1.2568712249702806e-05,
+      "loss": 0.3277,
+      "step": 7568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41364106807789164,
+      "learning_rate": 1.2566913734367244e-05,
+      "loss": 0.2982,
+      "step": 7569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4642819832894536,
+      "learning_rate": 1.2565115130139565e-05,
+      "loss": 0.2253,
+      "step": 7570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5140464591085009,
+      "learning_rate": 1.2563316437082065e-05,
+      "loss": 0.4326,
+      "step": 7571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26369296742189074,
+      "learning_rate": 1.2561517655257025e-05,
+      "loss": 0.2555,
+      "step": 7572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25556862333416436,
+      "learning_rate": 1.255971878472674e-05,
+      "loss": 0.1509,
+      "step": 7573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0276481868351715,
+      "learning_rate": 1.2557919825553502e-05,
+      "loss": 0.6818,
+      "step": 7574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6329669628544323,
+      "learning_rate": 1.255612077779961e-05,
+      "loss": 0.3306,
+      "step": 7575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32794636640693425,
+      "learning_rate": 1.2554321641527367e-05,
+      "loss": 0.2998,
+      "step": 7576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5110473328508144,
+      "learning_rate": 1.2552522416799073e-05,
+      "loss": 0.3151,
+      "step": 7577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41058569117740423,
+      "learning_rate": 1.2550723103677039e-05,
+      "loss": 0.2154,
+      "step": 7578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32708584999844575,
+      "learning_rate": 1.2548923702223574e-05,
+      "loss": 0.251,
+      "step": 7579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3670176202785912,
+      "learning_rate": 1.254712421250099e-05,
+      "loss": 0.2829,
+      "step": 7580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5745557867723556,
+      "learning_rate": 1.2545324634571604e-05,
+      "loss": 0.3194,
+      "step": 7581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40854419253185176,
+      "learning_rate": 1.2543524968497738e-05,
+      "loss": 0.3315,
+      "step": 7582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6312594387495574,
+      "learning_rate": 1.2541725214341708e-05,
+      "loss": 0.3503,
+      "step": 7583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3346389065280756,
+      "learning_rate": 1.2539925372165847e-05,
+      "loss": 0.2932,
+      "step": 7584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23633536868367974,
+      "learning_rate": 1.2538125442032474e-05,
+      "loss": 0.2035,
+      "step": 7585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9129265657545105,
+      "learning_rate": 1.2536325424003932e-05,
+      "loss": 0.5529,
+      "step": 7586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.38335161706444687,
+      "learning_rate": 1.2534525318142546e-05,
+      "loss": 0.3163,
+      "step": 7587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37815106189321185,
+      "learning_rate": 1.2532725124510658e-05,
+      "loss": 0.3199,
+      "step": 7588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45945715714105756,
+      "learning_rate": 1.2530924843170609e-05,
+      "loss": 0.3619,
+      "step": 7589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25719376284057816,
+      "learning_rate": 1.2529124474184742e-05,
+      "loss": 0.154,
+      "step": 7590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.352646974082044,
+      "learning_rate": 1.25273240176154e-05,
+      "loss": 0.2434,
+      "step": 7591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40108702245255196,
+      "learning_rate": 1.2525523473524938e-05,
+      "loss": 0.3153,
+      "step": 7592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5606921794556875,
+      "learning_rate": 1.2523722841975707e-05,
+      "loss": 0.153,
+      "step": 7593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35881739246190164,
+      "learning_rate": 1.2521922123030061e-05,
+      "loss": 0.3164,
+      "step": 7594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4775366714403366,
+      "learning_rate": 1.252012131675036e-05,
+      "loss": 0.3781,
+      "step": 7595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3417720512306077,
+      "learning_rate": 1.2518320423198968e-05,
+      "loss": 0.2006,
+      "step": 7596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2217803683706333,
+      "learning_rate": 1.251651944243825e-05,
+      "loss": 0.207,
+      "step": 7597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9289910609233664,
+      "learning_rate": 1.2514718374530568e-05,
+      "loss": 0.6753,
+      "step": 7598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5061395116029092,
+      "learning_rate": 1.2512917219538297e-05,
+      "loss": 0.2249,
+      "step": 7599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3169813552029995,
+      "learning_rate": 1.2511115977523813e-05,
+      "loss": 0.3035,
+      "step": 7600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7382839722814659,
+      "learning_rate": 1.2509314648549491e-05,
+      "loss": 0.5035,
+      "step": 7601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7547998463825323,
+      "learning_rate": 1.2507513232677707e-05,
+      "loss": 0.397,
+      "step": 7602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2598908072442766,
+      "learning_rate": 1.250571172997085e-05,
+      "loss": 0.1986,
+      "step": 7603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40470480560147853,
+      "learning_rate": 1.2503910140491305e-05,
+      "loss": 0.3349,
+      "step": 7604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5543636608248991,
+      "learning_rate": 1.2502108464301456e-05,
+      "loss": 0.3531,
+      "step": 7605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4062889895351509,
+      "learning_rate": 1.2500306701463702e-05,
+      "loss": 0.2834,
+      "step": 7606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3447667781654027,
+      "learning_rate": 1.2498504852040433e-05,
+      "loss": 0.2808,
+      "step": 7607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.363996062685284,
+      "learning_rate": 1.2496702916094048e-05,
+      "loss": 0.2969,
+      "step": 7608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26788663410836405,
+      "learning_rate": 1.2494900893686949e-05,
+      "loss": 0.1817,
+      "step": 7609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9228599503181674,
+      "learning_rate": 1.2493098784881539e-05,
+      "loss": 0.5611,
+      "step": 7610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3804507345579722,
+      "learning_rate": 1.2491296589740224e-05,
+      "loss": 0.3399,
+      "step": 7611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2948051354691912,
+      "learning_rate": 1.2489494308325415e-05,
+      "loss": 0.2742,
+      "step": 7612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3363350955735608,
+      "learning_rate": 1.2487691940699529e-05,
+      "loss": 0.2011,
+      "step": 7613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1140463461241292,
+      "learning_rate": 1.2485889486924974e-05,
+      "loss": 0.6561,
+      "step": 7614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3030326221824844,
+      "learning_rate": 1.2484086947064175e-05,
+      "loss": 0.2679,
+      "step": 7615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36992495041043094,
+      "learning_rate": 1.2482284321179552e-05,
+      "loss": 0.2854,
+      "step": 7616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7049813104276725,
+      "learning_rate": 1.248048160933353e-05,
+      "loss": 0.3893,
+      "step": 7617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2756630402796699,
+      "learning_rate": 1.2478678811588535e-05,
+      "loss": 0.2303,
+      "step": 7618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26760176350237963,
+      "learning_rate": 1.2476875928006999e-05,
+      "loss": 0.1611,
+      "step": 7619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41829783407336724,
+      "learning_rate": 1.2475072958651358e-05,
+      "loss": 0.3304,
+      "step": 7620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35645273997227866,
+      "learning_rate": 1.2473269903584046e-05,
+      "loss": 0.2964,
+      "step": 7621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8604531292814376,
+      "learning_rate": 1.2471466762867506e-05,
+      "loss": 0.3573,
+      "step": 7622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3153244327743291,
+      "learning_rate": 1.2469663536564177e-05,
+      "loss": 0.3042,
+      "step": 7623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3374073034408949,
+      "learning_rate": 1.2467860224736501e-05,
+      "loss": 0.2689,
+      "step": 7624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2715934504776164,
+      "learning_rate": 1.2466056827446937e-05,
+      "loss": 0.1519,
+      "step": 7625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4920404791571963,
+      "learning_rate": 1.2464253344757929e-05,
+      "loss": 0.1767,
+      "step": 7626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40765441951874637,
+      "learning_rate": 1.2462449776731935e-05,
+      "loss": 0.3151,
+      "step": 7627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4929507662219086,
+      "learning_rate": 1.246064612343141e-05,
+      "loss": 0.3614,
+      "step": 7628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8465290944998883,
+      "learning_rate": 1.2458842384918815e-05,
+      "loss": 0.3429,
+      "step": 7629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37771971215198696,
+      "learning_rate": 1.2457038561256616e-05,
+      "loss": 0.2933,
+      "step": 7630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25817417503857,
+      "learning_rate": 1.2455234652507276e-05,
+      "loss": 0.2179,
+      "step": 7631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.003852559033286,
+      "learning_rate": 1.2453430658733265e-05,
+      "loss": 0.2929,
+      "step": 7632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33167390172321803,
+      "learning_rate": 1.2451626579997056e-05,
+      "loss": 0.2602,
+      "step": 7633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8273407599930446,
+      "learning_rate": 1.2449822416361123e-05,
+      "loss": 0.4349,
+      "step": 7634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3353811089623452,
+      "learning_rate": 1.2448018167887947e-05,
+      "loss": 0.264,
+      "step": 7635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34648324200161335,
+      "learning_rate": 1.2446213834640007e-05,
+      "loss": 0.2492,
+      "step": 7636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30946096151194946,
+      "learning_rate": 1.2444409416679786e-05,
+      "loss": 0.2142,
+      "step": 7637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.813049180954248,
+      "learning_rate": 1.2442604914069773e-05,
+      "loss": 0.4697,
+      "step": 7638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3757559066901632,
+      "learning_rate": 1.2440800326872457e-05,
+      "loss": 0.2301,
+      "step": 7639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4310357553833105,
+      "learning_rate": 1.2438995655150332e-05,
+      "loss": 0.3578,
+      "step": 7640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0897373835543203,
+      "learning_rate": 1.243719089896589e-05,
+      "loss": 0.7668,
+      "step": 7641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2947300212873115,
+      "learning_rate": 1.2435386058381634e-05,
+      "loss": 0.2155,
+      "step": 7642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2607968228356934,
+      "learning_rate": 1.243358113346006e-05,
+      "loss": 0.2048,
+      "step": 7643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.444883989342041,
+      "learning_rate": 1.243177612426368e-05,
+      "loss": 0.368,
+      "step": 7644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3069080743699087,
+      "learning_rate": 1.2429971030854993e-05,
+      "loss": 0.1943,
+      "step": 7645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8069208624859093,
+      "learning_rate": 1.2428165853296517e-05,
+      "loss": 0.4731,
+      "step": 7646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3425899314113828,
+      "learning_rate": 1.2426360591650761e-05,
+      "loss": 0.2992,
+      "step": 7647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3365953563747308,
+      "learning_rate": 1.2424555245980241e-05,
+      "loss": 0.2127,
+      "step": 7648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.281929789738832,
+      "learning_rate": 1.2422749816347479e-05,
+      "loss": 0.1837,
+      "step": 7649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8104273831878697,
+      "learning_rate": 1.2420944302814992e-05,
+      "loss": 0.4816,
+      "step": 7650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34569478851091506,
+      "learning_rate": 1.2419138705445314e-05,
+      "loss": 0.2962,
+      "step": 7651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3634484166137741,
+      "learning_rate": 1.241733302430096e-05,
+      "loss": 0.2862,
+      "step": 7652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0261468850569297,
+      "learning_rate": 1.2415527259444471e-05,
+      "loss": 0.6782,
+      "step": 7653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3215611242030219,
+      "learning_rate": 1.2413721410938373e-05,
+      "loss": 0.2798,
+      "step": 7654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2449067046131677,
+      "learning_rate": 1.2411915478845211e-05,
+      "loss": 0.1634,
+      "step": 7655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4588366813097026,
+      "learning_rate": 1.2410109463227519e-05,
+      "loss": 0.3376,
+      "step": 7656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3169675064491585,
+      "learning_rate": 1.240830336414784e-05,
+      "loss": 0.3055,
+      "step": 7657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8641165712233401,
+      "learning_rate": 1.2406497181668717e-05,
+      "loss": 0.4004,
+      "step": 7658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34837061522076307,
+      "learning_rate": 1.2404690915852701e-05,
+      "loss": 0.3421,
+      "step": 7659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3212739319107255,
+      "learning_rate": 1.2402884566762341e-05,
+      "loss": 0.2886,
+      "step": 7660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2776234930677816,
+      "learning_rate": 1.2401078134460194e-05,
+      "loss": 0.1348,
+      "step": 7661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41273755963460756,
+      "learning_rate": 1.2399271619008812e-05,
+      "loss": 0.3599,
+      "step": 7662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42377632401126736,
+      "learning_rate": 1.2397465020470757e-05,
+      "loss": 0.2785,
+      "step": 7663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3625811131545241,
+      "learning_rate": 1.2395658338908594e-05,
+      "loss": 0.3151,
+      "step": 7664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7998829300948974,
+      "learning_rate": 1.2393851574384886e-05,
+      "loss": 0.4532,
+      "step": 7665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40151280132662404,
+      "learning_rate": 1.23920447269622e-05,
+      "loss": 0.2937,
+      "step": 7666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3260653072198325,
+      "learning_rate": 1.2390237796703107e-05,
+      "loss": 0.3084,
+      "step": 7667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5939761261701787,
+      "learning_rate": 1.238843078367018e-05,
+      "loss": 0.3467,
+      "step": 7668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2672933275218854,
+      "learning_rate": 1.2386623687926001e-05,
+      "loss": 0.2261,
+      "step": 7669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4603784285403028,
+      "learning_rate": 1.2384816509533145e-05,
+      "loss": 0.3053,
+      "step": 7670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32124928699513156,
+      "learning_rate": 1.2383009248554195e-05,
+      "loss": 0.245,
+      "step": 7671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34815018711320994,
+      "learning_rate": 1.2381201905051738e-05,
+      "loss": 0.2903,
+      "step": 7672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.630884248029316,
+      "learning_rate": 1.2379394479088363e-05,
+      "loss": 0.4409,
+      "step": 7673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37651909244520654,
+      "learning_rate": 1.2377586970726658e-05,
+      "loss": 0.2451,
+      "step": 7674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.282839422926257,
+      "learning_rate": 1.2375779380029218e-05,
+      "loss": 0.2584,
+      "step": 7675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4063717632353682,
+      "learning_rate": 1.2373971707058643e-05,
+      "loss": 0.2808,
+      "step": 7676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9496003441841634,
+      "learning_rate": 1.2372163951877525e-05,
+      "loss": 0.6778,
+      "step": 7677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2875371706876149,
+      "learning_rate": 1.2370356114548476e-05,
+      "loss": 0.2067,
+      "step": 7678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3650829049658475,
+      "learning_rate": 1.2368548195134094e-05,
+      "loss": 0.3043,
+      "step": 7679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7485902503018145,
+      "learning_rate": 1.2366740193696991e-05,
+      "loss": 0.4163,
+      "step": 7680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23644972909627532,
+      "learning_rate": 1.2364932110299775e-05,
+      "loss": 0.1315,
+      "step": 7681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3908523069948069,
+      "learning_rate": 1.2363123945005064e-05,
+      "loss": 0.2984,
+      "step": 7682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36885901856842995,
+      "learning_rate": 1.2361315697875469e-05,
+      "loss": 0.3257,
+      "step": 7683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3138719168813281,
+      "learning_rate": 1.2359507368973618e-05,
+      "loss": 0.1951,
+      "step": 7684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6338918277920743,
+      "learning_rate": 1.2357698958362123e-05,
+      "loss": 0.3682,
+      "step": 7685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.075166134031894,
+      "learning_rate": 1.2355890466103619e-05,
+      "loss": 0.6696,
+      "step": 7686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2538342119211804,
+      "learning_rate": 1.2354081892260728e-05,
+      "loss": 0.2455,
+      "step": 7687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2697614085126694,
+      "learning_rate": 1.2352273236896082e-05,
+      "loss": 0.2123,
+      "step": 7688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.115602045075998,
+      "learning_rate": 1.2350464500072314e-05,
+      "loss": 0.5415,
+      "step": 7689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40017360263197127,
+      "learning_rate": 1.2348655681852064e-05,
+      "loss": 0.299,
+      "step": 7690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27748444108314924,
+      "learning_rate": 1.2346846782297966e-05,
+      "loss": 0.2675,
+      "step": 7691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.024308978329414,
+      "learning_rate": 1.2345037801472669e-05,
+      "loss": 0.5259,
+      "step": 7692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2658736169881998,
+      "learning_rate": 1.2343228739438811e-05,
+      "loss": 0.2082,
+      "step": 7693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5155626562375746,
+      "learning_rate": 1.2341419596259038e-05,
+      "loss": 0.295,
+      "step": 7694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26905222497297376,
+      "learning_rate": 1.2339610371996012e-05,
+      "loss": 0.2577,
+      "step": 7695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3606434958043496,
+      "learning_rate": 1.2337801066712376e-05,
+      "loss": 0.3027,
+      "step": 7696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49890333651717567,
+      "learning_rate": 1.2335991680470792e-05,
+      "loss": 0.2738,
+      "step": 7697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39467269562385565,
+      "learning_rate": 1.2334182213333912e-05,
+      "loss": 0.3395,
+      "step": 7698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5731151803650315,
+      "learning_rate": 1.2332372665364406e-05,
+      "loss": 0.3394,
+      "step": 7699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4065624798269097,
+      "learning_rate": 1.2330563036624934e-05,
+      "loss": 0.3214,
+      "step": 7700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36053450120448083,
+      "learning_rate": 1.2328753327178164e-05,
+      "loss": 0.2515,
+      "step": 7701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5689075010121288,
+      "learning_rate": 1.2326943537086766e-05,
+      "loss": 0.3741,
+      "step": 7702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2783654959597047,
+      "learning_rate": 1.2325133666413414e-05,
+      "loss": 0.2591,
+      "step": 7703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3908455924860992,
+      "learning_rate": 1.2323323715220783e-05,
+      "loss": 0.2327,
+      "step": 7704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47826736293204375,
+      "learning_rate": 1.2321513683571553e-05,
+      "loss": 0.3268,
+      "step": 7705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4121896236281999,
+      "learning_rate": 1.2319703571528403e-05,
+      "loss": 0.3567,
+      "step": 7706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35279691238740696,
+      "learning_rate": 1.231789337915402e-05,
+      "loss": 0.2763,
+      "step": 7707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37557769110490014,
+      "learning_rate": 1.2316083106511085e-05,
+      "loss": 0.2277,
+      "step": 7708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3360999778532407,
+      "learning_rate": 1.2314272753662295e-05,
+      "loss": 0.2438,
+      "step": 7709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46270153126000013,
+      "learning_rate": 1.2312462320670335e-05,
+      "loss": 0.2576,
+      "step": 7710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3572732447396155,
+      "learning_rate": 1.2310651807597909e-05,
+      "loss": 0.233,
+      "step": 7711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4912302945710772,
+      "learning_rate": 1.2308841214507708e-05,
+      "loss": 0.3823,
+      "step": 7712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5918930747562702,
+      "learning_rate": 1.2307030541462435e-05,
+      "loss": 0.4092,
+      "step": 7713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27316585495061246,
+      "learning_rate": 1.2305219788524794e-05,
+      "loss": 0.224,
+      "step": 7714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2925029836392248,
+      "learning_rate": 1.230340895575749e-05,
+      "loss": 0.248,
+      "step": 7715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5948366275202696,
+      "learning_rate": 1.2301598043223233e-05,
+      "loss": 0.2956,
+      "step": 7716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5585403674560185,
+      "learning_rate": 1.2299787050984736e-05,
+      "loss": 0.2098,
+      "step": 7717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.357036950905096,
+      "learning_rate": 1.2297975979104711e-05,
+      "loss": 0.3173,
+      "step": 7718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3489053320187153,
+      "learning_rate": 1.2296164827645875e-05,
+      "loss": 0.3219,
+      "step": 7719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26655390566547604,
+      "learning_rate": 1.2294353596670954e-05,
+      "loss": 0.1146,
+      "step": 7720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22260566173399035,
+      "learning_rate": 1.2292542286242663e-05,
+      "loss": 0.1856,
+      "step": 7721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5210406943968808,
+      "learning_rate": 1.2290730896423733e-05,
+      "loss": 0.3649,
+      "step": 7722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30909289929601924,
+      "learning_rate": 1.2288919427276889e-05,
+      "loss": 0.2107,
+      "step": 7723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3394742788735173,
+      "learning_rate": 1.2287107878864868e-05,
+      "loss": 0.3283,
+      "step": 7724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6441934367210518,
+      "learning_rate": 1.2285296251250396e-05,
+      "loss": 0.4857,
+      "step": 7725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33547352374751016,
+      "learning_rate": 1.2283484544496214e-05,
+      "loss": 0.3212,
+      "step": 7726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22535303095029596,
+      "learning_rate": 1.2281672758665057e-05,
+      "loss": 0.1664,
+      "step": 7727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1198897473384355,
+      "learning_rate": 1.2279860893819677e-05,
+      "loss": 0.5423,
+      "step": 7728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5315956158941079,
+      "learning_rate": 1.2278048950022807e-05,
+      "loss": 0.4195,
+      "step": 7729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3316998991025966,
+      "learning_rate": 1.2276236927337201e-05,
+      "loss": 0.2396,
+      "step": 7730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3949812567558907,
+      "learning_rate": 1.2274424825825608e-05,
+      "loss": 0.3378,
+      "step": 7731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5902242463857145,
+      "learning_rate": 1.2272612645550783e-05,
+      "loss": 0.4106,
+      "step": 7732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27268811226890377,
+      "learning_rate": 1.2270800386575479e-05,
+      "loss": 0.182,
+      "step": 7733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30230494310639255,
+      "learning_rate": 1.2268988048962454e-05,
+      "loss": 0.2506,
+      "step": 7734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7007259407472975,
+      "learning_rate": 1.2267175632774472e-05,
+      "loss": 0.4328,
+      "step": 7735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3078098222416652,
+      "learning_rate": 1.2265363138074294e-05,
+      "loss": 0.2572,
+      "step": 7736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7233622730142071,
+      "learning_rate": 1.226355056492469e-05,
+      "loss": 0.5583,
+      "step": 7737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3706652593809387,
+      "learning_rate": 1.2261737913388424e-05,
+      "loss": 0.3142,
+      "step": 7738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3325676696521458,
+      "learning_rate": 1.2259925183528275e-05,
+      "loss": 0.2802,
+      "step": 7739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.285868913478341,
+      "learning_rate": 1.225811237540701e-05,
+      "loss": 0.1341,
+      "step": 7740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6615062653955808,
+      "learning_rate": 1.2256299489087416e-05,
+      "loss": 0.3796,
+      "step": 7741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29847265124901856,
+      "learning_rate": 1.2254486524632263e-05,
+      "loss": 0.2744,
+      "step": 7742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39855067333182725,
+      "learning_rate": 1.225267348210434e-05,
+      "loss": 0.2633,
+      "step": 7743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6883013273514436,
+      "learning_rate": 1.225086036156643e-05,
+      "loss": 0.475,
+      "step": 7744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3010695239911448,
+      "learning_rate": 1.2249047163081325e-05,
+      "loss": 0.2446,
+      "step": 7745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.265481799290773,
+      "learning_rate": 1.2247233886711811e-05,
+      "loss": 0.1961,
+      "step": 7746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6405863195516796,
+      "learning_rate": 1.2245420532520687e-05,
+      "loss": 0.3769,
+      "step": 7747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3782124832205112,
+      "learning_rate": 1.2243607100570743e-05,
+      "loss": 0.3033,
+      "step": 7748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8096337104484459,
+      "learning_rate": 1.2241793590924785e-05,
+      "loss": 0.4681,
+      "step": 7749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2962372025335575,
+      "learning_rate": 1.2239980003645606e-05,
+      "loss": 0.2728,
+      "step": 7750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4048492625045128,
+      "learning_rate": 1.2238166338796021e-05,
+      "loss": 0.3013,
+      "step": 7751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.307624574114885,
+      "learning_rate": 1.2236352596438832e-05,
+      "loss": 0.1997,
+      "step": 7752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1069574333300105,
+      "learning_rate": 1.2234538776636846e-05,
+      "loss": 0.3514,
+      "step": 7753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31546488061455996,
+      "learning_rate": 1.223272487945288e-05,
+      "loss": 0.2837,
+      "step": 7754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4835867800375805,
+      "learning_rate": 1.2230910904949747e-05,
+      "loss": 0.3979,
+      "step": 7755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.959949705436744,
+      "learning_rate": 1.2229096853190265e-05,
+      "loss": 0.3485,
+      "step": 7756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3530160623767107,
+      "learning_rate": 1.2227282724237254e-05,
+      "loss": 0.2498,
+      "step": 7757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5273841705816225,
+      "learning_rate": 1.2225468518153543e-05,
+      "loss": 0.2861,
+      "step": 7758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9535026076828771,
+      "learning_rate": 1.2223654235001948e-05,
+      "loss": 0.2452,
+      "step": 7759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36487906220355076,
+      "learning_rate": 1.2221839874845303e-05,
+      "loss": 0.2854,
+      "step": 7760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1723647175488505,
+      "learning_rate": 1.2220025437746437e-05,
+      "loss": 0.8119,
+      "step": 7761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.395196663415754,
+      "learning_rate": 1.221821092376819e-05,
+      "loss": 0.2566,
+      "step": 7762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43346883361783417,
+      "learning_rate": 1.2216396332973391e-05,
+      "loss": 0.3076,
+      "step": 7763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8082014848351015,
+      "learning_rate": 1.2214581665424883e-05,
+      "loss": 0.442,
+      "step": 7764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2674446403474515,
+      "learning_rate": 1.2212766921185506e-05,
+      "loss": 0.2374,
+      "step": 7765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28619470020545024,
+      "learning_rate": 1.2210952100318107e-05,
+      "loss": 0.1977,
+      "step": 7766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5161481861718609,
+      "learning_rate": 1.220913720288553e-05,
+      "loss": 0.4137,
+      "step": 7767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4577512973298372,
+      "learning_rate": 1.2207322228950628e-05,
+      "loss": 0.7202,
+      "step": 7768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3080743765473305,
+      "learning_rate": 1.2205507178576252e-05,
+      "loss": 0.1984,
+      "step": 7769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.439692529348634,
+      "learning_rate": 1.220369205182526e-05,
+      "loss": 0.3539,
+      "step": 7770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2810154705395384,
+      "learning_rate": 1.2201876848760505e-05,
+      "loss": 0.2121,
+      "step": 7771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34147306795497445,
+      "learning_rate": 1.2200061569444848e-05,
+      "loss": 0.2114,
+      "step": 7772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1078075037797748,
+      "learning_rate": 1.2198246213941156e-05,
+      "loss": 0.6738,
+      "step": 7773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43075098792032634,
+      "learning_rate": 1.2196430782312292e-05,
+      "loss": 0.3043,
+      "step": 7774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3794600600488902,
+      "learning_rate": 1.2194615274621122e-05,
+      "loss": 0.2685,
+      "step": 7775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5827109851181835,
+      "learning_rate": 1.2192799690930521e-05,
+      "loss": 0.3391,
+      "step": 7776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35902548591666494,
+      "learning_rate": 1.2190984031303361e-05,
+      "loss": 0.2472,
+      "step": 7777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3640745890902366,
+      "learning_rate": 1.2189168295802519e-05,
+      "loss": 0.2622,
+      "step": 7778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34088465488055364,
+      "learning_rate": 1.218735248449087e-05,
+      "loss": 0.1616,
+      "step": 7779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7991062158846808,
+      "learning_rate": 1.21855365974313e-05,
+      "loss": 0.4441,
+      "step": 7780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3664535149552145,
+      "learning_rate": 1.2183720634686693e-05,
+      "loss": 0.2753,
+      "step": 7781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37060343525461587,
+      "learning_rate": 1.2181904596319933e-05,
+      "loss": 0.2746,
+      "step": 7782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26177238030517913,
+      "learning_rate": 1.2180088482393911e-05,
+      "loss": 0.1789,
+      "step": 7783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3335241264892765,
+      "learning_rate": 1.2178272292971519e-05,
+      "loss": 0.2824,
+      "step": 7784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7427978808671635,
+      "learning_rate": 1.217645602811565e-05,
+      "loss": 0.484,
+      "step": 7785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4443305120052916,
+      "learning_rate": 1.2174639687889202e-05,
+      "loss": 0.3144,
+      "step": 7786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39557505313196073,
+      "learning_rate": 1.2172823272355077e-05,
+      "loss": 0.2631,
+      "step": 7787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6536016348918546,
+      "learning_rate": 1.2171006781576173e-05,
+      "loss": 0.3575,
+      "step": 7788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2895298768999007,
+      "learning_rate": 1.2169190215615401e-05,
+      "loss": 0.1717,
+      "step": 7789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3057102143177872,
+      "learning_rate": 1.2167373574535661e-05,
+      "loss": 0.258,
+      "step": 7790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5624184239293547,
+      "learning_rate": 1.2165556858399874e-05,
+      "loss": 0.443,
+      "step": 7791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4411597730654119,
+      "learning_rate": 1.2163740067270941e-05,
+      "loss": 0.2398,
+      "step": 7792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33775600014775914,
+      "learning_rate": 1.2161923201211783e-05,
+      "loss": 0.273,
+      "step": 7793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.354861755129926,
+      "learning_rate": 1.2160106260285316e-05,
+      "loss": 0.293,
+      "step": 7794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7224381640648487,
+      "learning_rate": 1.2158289244554467e-05,
+      "loss": 0.3158,
+      "step": 7795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34561253546241905,
+      "learning_rate": 1.2156472154082149e-05,
+      "loss": 0.255,
+      "step": 7796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1561107168749716,
+      "learning_rate": 1.2154654988931296e-05,
+      "loss": 0.7711,
+      "step": 7797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37902420717857044,
+      "learning_rate": 1.2152837749164834e-05,
+      "loss": 0.2881,
+      "step": 7798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3027625498266799,
+      "learning_rate": 1.215102043484569e-05,
+      "loss": 0.224,
+      "step": 7799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4722136725028722,
+      "learning_rate": 1.2149203046036803e-05,
+      "loss": 0.2767,
+      "step": 7800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47601793386394814,
+      "learning_rate": 1.2147385582801106e-05,
+      "loss": 0.3417,
+      "step": 7801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30416584956647924,
+      "learning_rate": 1.214556804520154e-05,
+      "loss": 0.2099,
+      "step": 7802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5322689707569284,
+      "learning_rate": 1.2143750433301043e-05,
+      "loss": 0.3731,
+      "step": 7803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48119429572557804,
+      "learning_rate": 1.2141932747162564e-05,
+      "loss": 0.3049,
+      "step": 7804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2579226364490064,
+      "learning_rate": 1.2140114986849043e-05,
+      "loss": 0.138,
+      "step": 7805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32117319040447845,
+      "learning_rate": 1.2138297152423432e-05,
+      "loss": 0.2956,
+      "step": 7806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9370941699181503,
+      "learning_rate": 1.2136479243948683e-05,
+      "loss": 0.6731,
+      "step": 7807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3129022751134737,
+      "learning_rate": 1.2134661261487752e-05,
+      "loss": 0.1771,
+      "step": 7808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3969024582867962,
+      "learning_rate": 1.2132843205103591e-05,
+      "loss": 0.322,
+      "step": 7809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4938649637259429,
+      "learning_rate": 1.2131025074859164e-05,
+      "loss": 0.3403,
+      "step": 7810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.21944148398620011,
+      "learning_rate": 1.2129206870817428e-05,
+      "loss": 0.1346,
+      "step": 7811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3246560530948743,
+      "learning_rate": 1.2127388593041348e-05,
+      "loss": 0.24,
+      "step": 7812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48767280290232373,
+      "learning_rate": 1.2125570241593894e-05,
+      "loss": 0.3602,
+      "step": 7813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33893538071778284,
+      "learning_rate": 1.2123751816538037e-05,
+      "loss": 0.2849,
+      "step": 7814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49358569403480806,
+      "learning_rate": 1.2121933317936741e-05,
+      "loss": 0.3024,
+      "step": 7815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6051859788689352,
+      "learning_rate": 1.2120114745852989e-05,
+      "loss": 0.4997,
+      "step": 7816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2764055452780025,
+      "learning_rate": 1.2118296100349753e-05,
+      "loss": 0.2308,
+      "step": 7817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25970750334520953,
+      "learning_rate": 1.2116477381490013e-05,
+      "loss": 0.1922,
+      "step": 7818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8639977301196575,
+      "learning_rate": 1.2114658589336754e-05,
+      "loss": 0.5905,
+      "step": 7819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4017731721362114,
+      "learning_rate": 1.2112839723952958e-05,
+      "loss": 0.2922,
+      "step": 7820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3604215580984441,
+      "learning_rate": 1.2111020785401615e-05,
+      "loss": 0.2731,
+      "step": 7821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6402061764469947,
+      "learning_rate": 1.210920177374571e-05,
+      "loss": 0.3718,
+      "step": 7822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6287622775741888,
+      "learning_rate": 1.210738268904824e-05,
+      "loss": 0.3071,
+      "step": 7823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23030235373629632,
+      "learning_rate": 1.21055635313722e-05,
+      "loss": 0.1565,
+      "step": 7824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36849606576333455,
+      "learning_rate": 1.2103744300780586e-05,
+      "loss": 0.3311,
+      "step": 7825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5539057972494684,
+      "learning_rate": 1.2101924997336393e-05,
+      "loss": 0.3156,
+      "step": 7826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39605459800732945,
+      "learning_rate": 1.2100105621102631e-05,
+      "loss": 0.3243,
+      "step": 7827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8298682413471936,
+      "learning_rate": 1.20982861721423e-05,
+      "loss": 0.462,
+      "step": 7828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38887716172679526,
+      "learning_rate": 1.209646665051841e-05,
+      "loss": 0.2785,
+      "step": 7829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22922781355910388,
+      "learning_rate": 1.2094647056293969e-05,
+      "loss": 0.2256,
+      "step": 7830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6853732992858654,
+      "learning_rate": 1.2092827389531992e-05,
+      "loss": 0.3065,
+      "step": 7831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5288262475482187,
+      "learning_rate": 1.209100765029549e-05,
+      "loss": 0.3367,
+      "step": 7832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3683993478374688,
+      "learning_rate": 1.2089187838647484e-05,
+      "loss": 0.2997,
+      "step": 7833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46618559440736795,
+      "learning_rate": 1.2087367954650992e-05,
+      "loss": 0.3075,
+      "step": 7834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3954069825574937,
+      "learning_rate": 1.208554799836904e-05,
+      "loss": 0.2891,
+      "step": 7835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2476792511905212,
+      "learning_rate": 1.2083727969864652e-05,
+      "loss": 0.2011,
+      "step": 7836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4296079969449888,
+      "learning_rate": 1.208190786920085e-05,
+      "loss": 0.2942,
+      "step": 7837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6835585045500028,
+      "learning_rate": 1.208008769644067e-05,
+      "loss": 0.3775,
+      "step": 7838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.429431380987189,
+      "learning_rate": 1.2078267451647141e-05,
+      "loss": 0.319,
+      "step": 7839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9934331425210245,
+      "learning_rate": 1.20764471348833e-05,
+      "loss": 0.6671,
+      "step": 7840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35648383354079055,
+      "learning_rate": 1.2074626746212183e-05,
+      "loss": 0.2447,
+      "step": 7841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2711537520251157,
+      "learning_rate": 1.2072806285696836e-05,
+      "loss": 0.2366,
+      "step": 7842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5954785393208696,
+      "learning_rate": 1.2070985753400293e-05,
+      "loss": 0.3112,
+      "step": 7843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.506182932788288,
+      "learning_rate": 1.2069165149385603e-05,
+      "loss": 0.1937,
+      "step": 7844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3082872471946073,
+      "learning_rate": 1.2067344473715813e-05,
+      "loss": 0.2891,
+      "step": 7845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3423903304220317,
+      "learning_rate": 1.2065523726453971e-05,
+      "loss": 0.7992,
+      "step": 7846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4917154874110692,
+      "learning_rate": 1.206370290766313e-05,
+      "loss": 0.1749,
+      "step": 7847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.241763552867306,
+      "learning_rate": 1.206188201740635e-05,
+      "loss": 0.2102,
+      "step": 7848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3783400126203032,
+      "learning_rate": 1.206006105574668e-05,
+      "loss": 0.3344,
+      "step": 7849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.73194324677599,
+      "learning_rate": 1.205824002274719e-05,
+      "loss": 0.4001,
+      "step": 7850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31024938969861154,
+      "learning_rate": 1.2056418918470931e-05,
+      "loss": 0.2304,
+      "step": 7851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9587907228244877,
+      "learning_rate": 1.2054597742980975e-05,
+      "loss": 0.6535,
+      "step": 7852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.353162787883959,
+      "learning_rate": 1.2052776496340389e-05,
+      "loss": 0.2805,
+      "step": 7853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32028660740744386,
+      "learning_rate": 1.205095517861224e-05,
+      "loss": 0.2061,
+      "step": 7854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43666200782343506,
+      "learning_rate": 1.2049133789859602e-05,
+      "loss": 0.3048,
+      "step": 7855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.364837631487842,
+      "learning_rate": 1.2047312330145549e-05,
+      "loss": 0.2357,
+      "step": 7856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2910287527496673,
+      "learning_rate": 1.204549079953316e-05,
+      "loss": 0.253,
+      "step": 7857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8730968420317119,
+      "learning_rate": 1.2043669198085509e-05,
+      "loss": 0.6054,
+      "step": 7858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4669922543739415,
+      "learning_rate": 1.2041847525865681e-05,
+      "loss": 0.7951,
+      "step": 7859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3773913635551674,
+      "learning_rate": 1.2040025782936766e-05,
+      "loss": 0.2398,
+      "step": 7860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2937030728988133,
+      "learning_rate": 1.2038203969361841e-05,
+      "loss": 0.2758,
+      "step": 7861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39829030321756786,
+      "learning_rate": 1.2036382085204004e-05,
+      "loss": 0.2639,
+      "step": 7862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35718100232276506,
+      "learning_rate": 1.2034560130526341e-05,
+      "loss": 0.2594,
+      "step": 7863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8819076750989475,
+      "learning_rate": 1.2032738105391945e-05,
+      "loss": 0.4058,
+      "step": 7864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37511149307403135,
+      "learning_rate": 1.2030916009863921e-05,
+      "loss": 0.3146,
+      "step": 7865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3427182179012295,
+      "learning_rate": 1.2029093844005359e-05,
+      "loss": 0.2962,
+      "step": 7866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6080154357699522,
+      "learning_rate": 1.2027271607879368e-05,
+      "loss": 0.362,
+      "step": 7867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24903988506822783,
+      "learning_rate": 1.2025449301549046e-05,
+      "loss": 0.1698,
+      "step": 7868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3080660901489215,
+      "learning_rate": 1.2023626925077503e-05,
+      "loss": 0.272,
+      "step": 7869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1014977729702664,
+      "learning_rate": 1.2021804478527845e-05,
+      "loss": 0.401,
+      "step": 7870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6438849569428409,
+      "learning_rate": 1.2019981961963185e-05,
+      "loss": 0.4141,
+      "step": 7871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3717529729312076,
+      "learning_rate": 1.2018159375446636e-05,
+      "loss": 0.2579,
+      "step": 7872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3212381811770286,
+      "learning_rate": 1.2016336719041318e-05,
+      "loss": 0.2701,
+      "step": 7873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38203566172353803,
+      "learning_rate": 1.2014513992810344e-05,
+      "loss": 0.2637,
+      "step": 7874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3642964971611716,
+      "learning_rate": 1.2012691196816836e-05,
+      "loss": 0.275,
+      "step": 7875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37088618583675204,
+      "learning_rate": 1.2010868331123922e-05,
+      "loss": 0.2343,
+      "step": 7876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40984124709798897,
+      "learning_rate": 1.2009045395794723e-05,
+      "loss": 0.283,
+      "step": 7877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31850183923540143,
+      "learning_rate": 1.2007222390892365e-05,
+      "loss": 0.2847,
+      "step": 7878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6904602282111612,
+      "learning_rate": 1.2005399316479984e-05,
+      "loss": 0.4745,
+      "step": 7879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28231886869155365,
+      "learning_rate": 1.200357617262071e-05,
+      "loss": 0.1993,
+      "step": 7880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2864075909628802,
+      "learning_rate": 1.2001752959377681e-05,
+      "loss": 0.2689,
+      "step": 7881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5250714104982417,
+      "learning_rate": 1.1999929676814036e-05,
+      "loss": 0.2763,
+      "step": 7882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6084203947302514,
+      "learning_rate": 1.1998106324992906e-05,
+      "loss": 0.3492,
+      "step": 7883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3261697352200898,
+      "learning_rate": 1.1996282903977442e-05,
+      "loss": 0.3021,
+      "step": 7884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3600944615199071,
+      "learning_rate": 1.1994459413830785e-05,
+      "loss": 0.3117,
+      "step": 7885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43975626688178493,
+      "learning_rate": 1.1992635854616088e-05,
+      "loss": 0.2537,
+      "step": 7886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3502728085639639,
+      "learning_rate": 1.1990812226396496e-05,
+      "loss": 0.3062,
+      "step": 7887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43356883571119836,
+      "learning_rate": 1.1988988529235161e-05,
+      "loss": 0.305,
+      "step": 7888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29485545554777076,
+      "learning_rate": 1.198716476319524e-05,
+      "loss": 0.2916,
+      "step": 7889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31064543092327446,
+      "learning_rate": 1.198534092833989e-05,
+      "loss": 0.2028,
+      "step": 7890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8897180222566816,
+      "learning_rate": 1.1983517024732266e-05,
+      "loss": 0.59,
+      "step": 7891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3955907200246761,
+      "learning_rate": 1.1981693052435537e-05,
+      "loss": 0.3412,
+      "step": 7892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29039852394862853,
+      "learning_rate": 1.1979869011512859e-05,
+      "loss": 0.2017,
+      "step": 7893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39928201842865546,
+      "learning_rate": 1.1978044902027403e-05,
+      "loss": 0.3085,
+      "step": 7894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4353633841389933,
+      "learning_rate": 1.1976220724042336e-05,
+      "loss": 0.2926,
+      "step": 7895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32887704414837243,
+      "learning_rate": 1.1974396477620833e-05,
+      "loss": 0.1961,
+      "step": 7896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34286785641908873,
+      "learning_rate": 1.1972572162826061e-05,
+      "loss": 0.314,
+      "step": 7897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3682061627663276,
+      "learning_rate": 1.1970747779721203e-05,
+      "loss": 0.7447,
+      "step": 7898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29070012715612625,
+      "learning_rate": 1.1968923328369433e-05,
+      "loss": 0.219,
+      "step": 7899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4741431492632248,
+      "learning_rate": 1.1967098808833935e-05,
+      "loss": 0.3583,
+      "step": 7900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.298824920933953,
+      "learning_rate": 1.196527422117789e-05,
+      "loss": 0.2645,
+      "step": 7901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24902107345337846,
+      "learning_rate": 1.1963449565464478e-05,
+      "loss": 0.2042,
+      "step": 7902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0625977661145665,
+      "learning_rate": 1.1961624841756896e-05,
+      "loss": 0.3038,
+      "step": 7903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42317283623319885,
+      "learning_rate": 1.1959800050118328e-05,
+      "loss": 0.3523,
+      "step": 7904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3517066275803351,
+      "learning_rate": 1.195797519061197e-05,
+      "loss": 0.2577,
+      "step": 7905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7735868151817158,
+      "learning_rate": 1.1956150263301014e-05,
+      "loss": 0.3493,
+      "step": 7906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.310163938160124,
+      "learning_rate": 1.1954325268248662e-05,
+      "loss": 0.1994,
+      "step": 7907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.397983349336781,
+      "learning_rate": 1.1952500205518107e-05,
+      "loss": 0.241,
+      "step": 7908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3896181687123652,
+      "learning_rate": 1.1950675075172557e-05,
+      "loss": 0.2484,
+      "step": 7909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9368903933280383,
+      "learning_rate": 1.1948849877275209e-05,
+      "loss": 0.5781,
+      "step": 7910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39187076583274866,
+      "learning_rate": 1.1947024611889276e-05,
+      "loss": 0.2909,
+      "step": 7911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3288277529256562,
+      "learning_rate": 1.1945199279077962e-05,
+      "loss": 0.2785,
+      "step": 7912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5272823310712186,
+      "learning_rate": 1.1943373878904482e-05,
+      "loss": 0.2873,
+      "step": 7913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29573508428328504,
+      "learning_rate": 1.194154841143205e-05,
+      "loss": 0.1921,
+      "step": 7914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6374701531641034,
+      "learning_rate": 1.1939722876723878e-05,
+      "loss": 0.4224,
+      "step": 7915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3905831304108459,
+      "learning_rate": 1.1937897274843184e-05,
+      "loss": 0.2766,
+      "step": 7916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41233010441397044,
+      "learning_rate": 1.1936071605853195e-05,
+      "loss": 0.286,
+      "step": 7917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0236772511840988,
+      "learning_rate": 1.1934245869817127e-05,
+      "loss": 0.4651,
+      "step": 7918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27799086007368734,
+      "learning_rate": 1.193242006679821e-05,
+      "loss": 0.1277,
+      "step": 7919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2699895568941243,
+      "learning_rate": 1.1930594196859667e-05,
+      "loss": 0.2208,
+      "step": 7920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4589470961563092,
+      "learning_rate": 1.1928768260064732e-05,
+      "loss": 0.3091,
+      "step": 7921
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4691715710811628,
+      "learning_rate": 1.1926942256476632e-05,
+      "loss": 0.3769,
+      "step": 7922
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37149958982181935,
+      "learning_rate": 1.1925116186158605e-05,
+      "loss": 0.2761,
+      "step": 7923
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5020463855050367,
+      "learning_rate": 1.192329004917389e-05,
+      "loss": 0.3984,
+      "step": 7924
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31379760183411154,
+      "learning_rate": 1.1921463845585722e-05,
+      "loss": 0.2095,
+      "step": 7925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.304322085714914,
+      "learning_rate": 1.1919637575457343e-05,
+      "loss": 0.1815,
+      "step": 7926
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4280709827379945,
+      "learning_rate": 1.1917811238852e-05,
+      "loss": 0.3252,
+      "step": 7927
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5484844773978607,
+      "learning_rate": 1.1915984835832934e-05,
+      "loss": 0.3202,
+      "step": 7928
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49450017929156964,
+      "learning_rate": 1.1914158366463392e-05,
+      "loss": 0.2197,
+      "step": 7929
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6972758363104524,
+      "learning_rate": 1.1912331830806634e-05,
+      "loss": 0.4078,
+      "step": 7930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.163113774171106,
+      "learning_rate": 1.1910505228925903e-05,
+      "loss": 0.6485,
+      "step": 7931
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23145347746277942,
+      "learning_rate": 1.1908678560884462e-05,
+      "loss": 0.1715,
+      "step": 7932
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30815546200945876,
+      "learning_rate": 1.190685182674556e-05,
+      "loss": 0.2405,
+      "step": 7933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6610420498319607,
+      "learning_rate": 1.1905025026572466e-05,
+      "loss": 0.4497,
+      "step": 7934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3484729311489025,
+      "learning_rate": 1.1903198160428433e-05,
+      "loss": 0.2436,
+      "step": 7935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37479821138076064,
+      "learning_rate": 1.190137122837673e-05,
+      "loss": 0.3104,
+      "step": 7936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.903404747222253,
+      "learning_rate": 1.1899544230480623e-05,
+      "loss": 0.5389,
+      "step": 7937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3793020472891658,
+      "learning_rate": 1.1897717166803384e-05,
+      "loss": 0.2626,
+      "step": 7938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.21536537499616923,
+      "learning_rate": 1.189589003740828e-05,
+      "loss": 0.158,
+      "step": 7939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37013074965928505,
+      "learning_rate": 1.1894062842358585e-05,
+      "loss": 0.3346,
+      "step": 7940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4442486156485244,
+      "learning_rate": 1.1892235581717575e-05,
+      "loss": 0.2893,
+      "step": 7941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5788706585727145,
+      "learning_rate": 1.1890408255548526e-05,
+      "loss": 0.2968,
+      "step": 7942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.205983665744992,
+      "learning_rate": 1.1888580863914724e-05,
+      "loss": 0.613,
+      "step": 7943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3300548857885787,
+      "learning_rate": 1.188675340687945e-05,
+      "loss": 0.2639,
+      "step": 7944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23304407517691592,
+      "learning_rate": 1.188492588450598e-05,
+      "loss": 0.1743,
+      "step": 7945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6386065237872193,
+      "learning_rate": 1.1883098296857614e-05,
+      "loss": 0.444,
+      "step": 7946
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5258883988322739,
+      "learning_rate": 1.188127064399763e-05,
+      "loss": 0.2976,
+      "step": 7947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30515301220401475,
+      "learning_rate": 1.1879442925989327e-05,
+      "loss": 0.2492,
+      "step": 7948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9557336485183242,
+      "learning_rate": 1.1877615142895995e-05,
+      "loss": 0.5884,
+      "step": 7949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58178474444118,
+      "learning_rate": 1.1875787294780932e-05,
+      "loss": 0.324,
+      "step": 7950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4263233009317428,
+      "learning_rate": 1.1873959381707437e-05,
+      "loss": 0.323,
+      "step": 7951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2837105656508334,
+      "learning_rate": 1.1872131403738807e-05,
+      "loss": 0.2144,
+      "step": 7952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3091377428079506,
+      "learning_rate": 1.1870303360938349e-05,
+      "loss": 0.2136,
+      "step": 7953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38310899231449197,
+      "learning_rate": 1.1868475253369362e-05,
+      "loss": 0.313,
+      "step": 7954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9306600839745718,
+      "learning_rate": 1.1866647081095162e-05,
+      "loss": 0.2579,
+      "step": 7955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3050132731937429,
+      "learning_rate": 1.186481884417905e-05,
+      "loss": 0.2768,
+      "step": 7956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42166795211866515,
+      "learning_rate": 1.1862990542684345e-05,
+      "loss": 0.3322,
+      "step": 7957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2862765123153871,
+      "learning_rate": 1.1861162176674354e-05,
+      "loss": 0.1944,
+      "step": 7958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3983928765816261,
+      "learning_rate": 1.1859333746212403e-05,
+      "loss": 0.3066,
+      "step": 7959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37464310126796313,
+      "learning_rate": 1.1857505251361801e-05,
+      "loss": 0.3424,
+      "step": 7960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4032873786677103,
+      "learning_rate": 1.1855676692185872e-05,
+      "loss": 0.2819,
+      "step": 7961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4679357898041959,
+      "learning_rate": 1.1853848068747938e-05,
+      "loss": 0.3131,
+      "step": 7962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3412178021471056,
+      "learning_rate": 1.1852019381111326e-05,
+      "loss": 0.3276,
+      "step": 7963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2875821893013411,
+      "learning_rate": 1.185019062933936e-05,
+      "loss": 0.2642,
+      "step": 7964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4411680471167332,
+      "learning_rate": 1.1848361813495376e-05,
+      "loss": 0.2624,
+      "step": 7965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2880036099235008,
+      "learning_rate": 1.18465329336427e-05,
+      "loss": 0.2542,
+      "step": 7966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4021660522898735,
+      "learning_rate": 1.1844703989844666e-05,
+      "loss": 0.3471,
+      "step": 7967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33634227392857685,
+      "learning_rate": 1.1842874982164616e-05,
+      "loss": 0.2271,
+      "step": 7968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3303900188393209,
+      "learning_rate": 1.1841045910665881e-05,
+      "loss": 0.3177,
+      "step": 7969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9671616145587876,
+      "learning_rate": 1.1839216775411808e-05,
+      "loss": 0.609,
+      "step": 7970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.35308979972001275,
+      "learning_rate": 1.1837387576465735e-05,
+      "loss": 0.2297,
+      "step": 7971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24389530242129417,
+      "learning_rate": 1.183555831389101e-05,
+      "loss": 0.2464,
+      "step": 7972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4290048907771581,
+      "learning_rate": 1.183372898775098e-05,
+      "loss": 0.2936,
+      "step": 7973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4132601964727188,
+      "learning_rate": 1.1831899598108993e-05,
+      "loss": 0.2794,
+      "step": 7974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4167507268336926,
+      "learning_rate": 1.1830070145028404e-05,
+      "loss": 0.3419,
+      "step": 7975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4073307537683485,
+      "learning_rate": 1.1828240628572563e-05,
+      "loss": 0.3289,
+      "step": 7976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7557884514763259,
+      "learning_rate": 1.182641104880483e-05,
+      "loss": 0.4952,
+      "step": 7977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26140284114058343,
+      "learning_rate": 1.1824581405788558e-05,
+      "loss": 0.1731,
+      "step": 7978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39551359172857126,
+      "learning_rate": 1.1822751699587109e-05,
+      "loss": 0.2648,
+      "step": 7979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3736897265914453,
+      "learning_rate": 1.1820921930263851e-05,
+      "loss": 0.2989,
+      "step": 7980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38382012833460544,
+      "learning_rate": 1.1819092097882141e-05,
+      "loss": 0.231,
+      "step": 7981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.296301622420184,
+      "learning_rate": 1.1817262202505353e-05,
+      "loss": 0.4792,
+      "step": 7982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5060071372814816,
+      "learning_rate": 1.1815432244196849e-05,
+      "loss": 0.3495,
+      "step": 7983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.285491659038382,
+      "learning_rate": 1.181360222302001e-05,
+      "loss": 0.2228,
+      "step": 7984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43293902333173934,
+      "learning_rate": 1.18117721390382e-05,
+      "loss": 0.2953,
+      "step": 7985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42922801670662475,
+      "learning_rate": 1.1809941992314799e-05,
+      "loss": 0.2395,
+      "step": 7986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33588685351110065,
+      "learning_rate": 1.1808111782913188e-05,
+      "loss": 0.2126,
+      "step": 7987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40812800461735654,
+      "learning_rate": 1.180628151089674e-05,
+      "loss": 0.3165,
+      "step": 7988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6953194615167572,
+      "learning_rate": 1.1804451176328844e-05,
+      "loss": 0.524,
+      "step": 7989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38056672727143354,
+      "learning_rate": 1.1802620779272877e-05,
+      "loss": 0.3416,
+      "step": 7990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4374708121864055,
+      "learning_rate": 1.1800790319792234e-05,
+      "loss": 0.3152,
+      "step": 7991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24034260281344746,
+      "learning_rate": 1.1798959797950298e-05,
+      "loss": 0.1912,
+      "step": 7992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3737299347380316,
+      "learning_rate": 1.1797129213810462e-05,
+      "loss": 0.3117,
+      "step": 7993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8087123591015803,
+      "learning_rate": 1.179529856743612e-05,
+      "loss": 0.3435,
+      "step": 7994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3312821317125005,
+      "learning_rate": 1.1793467858890666e-05,
+      "loss": 0.2908,
+      "step": 7995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37260236653991463,
+      "learning_rate": 1.1791637088237493e-05,
+      "loss": 0.2907,
+      "step": 7996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5993123611192235,
+      "learning_rate": 1.1789806255540008e-05,
+      "loss": 0.3494,
+      "step": 7997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23486965767951468,
+      "learning_rate": 1.1787975360861607e-05,
+      "loss": 0.1612,
+      "step": 7998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.36941228758433226,
+      "learning_rate": 1.1786144404265701e-05,
+      "loss": 0.3084,
+      "step": 7999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38811622550655456,
+      "learning_rate": 1.1784313385815685e-05,
+      "loss": 0.2744,
+      "step": 8000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8777252991056873,
+      "learning_rate": 1.1782482305574976e-05,
+      "loss": 0.5783,
+      "step": 8001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33300708206449414,
+      "learning_rate": 1.178065116360698e-05,
+      "loss": 0.2683,
+      "step": 8002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3506847663910355,
+      "learning_rate": 1.1778819959975114e-05,
+      "loss": 0.3212,
+      "step": 8003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2039203231034237,
+      "learning_rate": 1.1776988694742786e-05,
+      "loss": 0.0902,
+      "step": 8004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32754667156190614,
+      "learning_rate": 1.1775157367973417e-05,
+      "loss": 0.2659,
+      "step": 8005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9840863478213498,
+      "learning_rate": 1.1773325979730428e-05,
+      "loss": 0.6268,
+      "step": 8006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4432248748342312,
+      "learning_rate": 1.1771494530077233e-05,
+      "loss": 0.3049,
+      "step": 8007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3207501681884594,
+      "learning_rate": 1.1769663019077258e-05,
+      "loss": 0.2882,
+      "step": 8008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8120058617367432,
+      "learning_rate": 1.176783144679393e-05,
+      "loss": 0.5619,
+      "step": 8009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.20825762019102417,
+      "learning_rate": 1.1765999813290674e-05,
+      "loss": 0.0947,
+      "step": 8010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4053812780070775,
+      "learning_rate": 1.1764168118630922e-05,
+      "loss": 0.277,
+      "step": 8011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3988311894219567,
+      "learning_rate": 1.1762336362878104e-05,
+      "loss": 0.3246,
+      "step": 8012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6500249623410269,
+      "learning_rate": 1.1760504546095653e-05,
+      "loss": 0.4227,
+      "step": 8013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29550813912741236,
+      "learning_rate": 1.1758672668347005e-05,
+      "loss": 0.19,
+      "step": 8014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34503398766813,
+      "learning_rate": 1.1756840729695598e-05,
+      "loss": 0.342,
+      "step": 8015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28919837746058225,
+      "learning_rate": 1.1755008730204873e-05,
+      "loss": 0.2108,
+      "step": 8016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31556082209556957,
+      "learning_rate": 1.1753176669938269e-05,
+      "loss": 0.1865,
+      "step": 8017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8571833591537861,
+      "learning_rate": 1.1751344548959233e-05,
+      "loss": 0.5349,
+      "step": 8018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3899301745997571,
+      "learning_rate": 1.174951236733121e-05,
+      "loss": 0.3314,
+      "step": 8019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2910575101476483,
+      "learning_rate": 1.1747680125117654e-05,
+      "loss": 0.2382,
+      "step": 8020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9408216363736895,
+      "learning_rate": 1.1745847822382004e-05,
+      "loss": 0.6722,
+      "step": 8021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3182946462400599,
+      "learning_rate": 1.174401545918772e-05,
+      "loss": 0.2214,
+      "step": 8022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2663466323044907,
+      "learning_rate": 1.1742183035598258e-05,
+      "loss": 0.2274,
+      "step": 8023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49731607824725593,
+      "learning_rate": 1.1740350551677073e-05,
+      "loss": 0.3698,
+      "step": 8024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9450188765195661,
+      "learning_rate": 1.1738518007487621e-05,
+      "loss": 0.4426,
+      "step": 8025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31989191955177276,
+      "learning_rate": 1.1736685403093367e-05,
+      "loss": 0.2602,
+      "step": 8026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4274492080687954,
+      "learning_rate": 1.1734852738557772e-05,
+      "loss": 0.2698,
+      "step": 8027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32934834987845174,
+      "learning_rate": 1.1733020013944301e-05,
+      "loss": 0.2708,
+      "step": 8028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2622235332469892,
+      "learning_rate": 1.1731187229316418e-05,
+      "loss": 0.2088,
+      "step": 8029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6319701577728358,
+      "learning_rate": 1.1729354384737602e-05,
+      "loss": 0.3641,
+      "step": 8030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39856243611000813,
+      "learning_rate": 1.1727521480271315e-05,
+      "loss": 0.3438,
+      "step": 8031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3740268094501319,
+      "learning_rate": 1.172568851598103e-05,
+      "loss": 0.2601,
+      "step": 8032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49836553108072196,
+      "learning_rate": 1.1723855491930232e-05,
+      "loss": 0.3013,
+      "step": 8033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3562893968298132,
+      "learning_rate": 1.1722022408182388e-05,
+      "loss": 0.288,
+      "step": 8034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3640195155365308,
+      "learning_rate": 1.1720189264800983e-05,
+      "loss": 0.2428,
+      "step": 8035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.706087521346157,
+      "learning_rate": 1.1718356061849496e-05,
+      "loss": 0.2559,
+      "step": 8036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2503475403720288,
+      "learning_rate": 1.1716522799391417e-05,
+      "loss": 0.5743,
+      "step": 8037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4138155307392456,
+      "learning_rate": 1.1714689477490224e-05,
+      "loss": 0.3201,
+      "step": 8038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3644446915489398,
+      "learning_rate": 1.1712856096209411e-05,
+      "loss": 0.292,
+      "step": 8039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7135645309500714,
+      "learning_rate": 1.1711022655612461e-05,
+      "loss": 0.2894,
+      "step": 8040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28773646997277114,
+      "learning_rate": 1.1709189155762872e-05,
+      "loss": 0.2276,
+      "step": 8041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8314554766196716,
+      "learning_rate": 1.1707355596724135e-05,
+      "loss": 0.4062,
+      "step": 8042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34693532859102244,
+      "learning_rate": 1.1705521978559748e-05,
+      "loss": 0.215,
+      "step": 8043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3694538811668248,
+      "learning_rate": 1.1703688301333211e-05,
+      "loss": 0.2752,
+      "step": 8044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.349991276656821,
+      "learning_rate": 1.1701854565108019e-05,
+      "loss": 0.7906,
+      "step": 8045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4942671800850725,
+      "learning_rate": 1.1700020769947675e-05,
+      "loss": 0.3041,
+      "step": 8046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28376124490302723,
+      "learning_rate": 1.1698186915915689e-05,
+      "loss": 0.258,
+      "step": 8047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5974320339927423,
+      "learning_rate": 1.1696353003075558e-05,
+      "loss": 0.4175,
+      "step": 8048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25196002764368464,
+      "learning_rate": 1.16945190314908e-05,
+      "loss": 0.1402,
+      "step": 8049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5777653906969942,
+      "learning_rate": 1.1692685001224918e-05,
+      "loss": 0.335,
+      "step": 8050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45541649563077186,
+      "learning_rate": 1.1690850912341427e-05,
+      "loss": 0.2842,
+      "step": 8051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9243817653395691,
+      "learning_rate": 1.1689016764903841e-05,
+      "loss": 0.4757,
+      "step": 8052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4541786271595083,
+      "learning_rate": 1.168718255897568e-05,
+      "loss": 0.2181,
+      "step": 8053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3405747715638682,
+      "learning_rate": 1.1685348294620457e-05,
+      "loss": 0.2851,
+      "step": 8054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31051270814141485,
+      "learning_rate": 1.1683513971901697e-05,
+      "loss": 0.2689,
+      "step": 8055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3512555459275891,
+      "learning_rate": 1.168167959088292e-05,
+      "loss": 0.2056,
+      "step": 8056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5055296842698704,
+      "learning_rate": 1.1679845151627648e-05,
+      "loss": 0.3748,
+      "step": 8057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7602648728429585,
+      "learning_rate": 1.1678010654199417e-05,
+      "loss": 0.4803,
+      "step": 8058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41065482545930426,
+      "learning_rate": 1.1676176098661742e-05,
+      "loss": 0.2424,
+      "step": 8059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47665228757374284,
+      "learning_rate": 1.1674341485078167e-05,
+      "loss": 0.3772,
+      "step": 8060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4010258063879982,
+      "learning_rate": 1.1672506813512217e-05,
+      "loss": 0.3169,
+      "step": 8061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3090499814345548,
+      "learning_rate": 1.1670672084027425e-05,
+      "loss": 0.2161,
+      "step": 8062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32229611953247095,
+      "learning_rate": 1.1668837296687332e-05,
+      "loss": 0.2474,
+      "step": 8063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.723866973557168,
+      "learning_rate": 1.1667002451555476e-05,
+      "loss": 0.4764,
+      "step": 8064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38413806263337497,
+      "learning_rate": 1.1665167548695395e-05,
+      "loss": 0.2978,
+      "step": 8065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.443860016995288,
+      "learning_rate": 1.1663332588170637e-05,
+      "loss": 0.2366,
+      "step": 8066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34763785680591497,
+      "learning_rate": 1.1661497570044737e-05,
+      "loss": 0.3424,
+      "step": 8067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8125107733973791,
+      "learning_rate": 1.1659662494381255e-05,
+      "loss": 0.4655,
+      "step": 8068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23319806165726573,
+      "learning_rate": 1.1657827361243725e-05,
+      "loss": 0.1526,
+      "step": 8069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38548295708245645,
+      "learning_rate": 1.1655992170695709e-05,
+      "loss": 0.3449,
+      "step": 8070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6652817103965797,
+      "learning_rate": 1.1654156922800757e-05,
+      "loss": 0.4387,
+      "step": 8071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40606341336922763,
+      "learning_rate": 1.1652321617622418e-05,
+      "loss": 0.2247,
+      "step": 8072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2211351878797165,
+      "learning_rate": 1.1650486255224254e-05,
+      "loss": 0.7863,
+      "step": 8073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2775457316895789,
+      "learning_rate": 1.1648650835669821e-05,
+      "loss": 0.1874,
+      "step": 8074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23958314554752064,
+      "learning_rate": 1.1646815359022683e-05,
+      "loss": 0.2295,
+      "step": 8075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5807486031966581,
+      "learning_rate": 1.1644979825346397e-05,
+      "loss": 0.4551,
+      "step": 8076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5713200590272652,
+      "learning_rate": 1.1643144234704531e-05,
+      "loss": 0.3726,
+      "step": 8077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4064945951489773,
+      "learning_rate": 1.1641308587160654e-05,
+      "loss": 0.3286,
+      "step": 8078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3521036176468557,
+      "learning_rate": 1.1639472882778328e-05,
+      "loss": 0.2556,
+      "step": 8079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.563728121201371,
+      "learning_rate": 1.1637637121621126e-05,
+      "loss": 0.3345,
+      "step": 8080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38309492806397877,
+      "learning_rate": 1.1635801303752622e-05,
+      "loss": 0.3351,
+      "step": 8081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3018626635818792,
+      "learning_rate": 1.1633965429236389e-05,
+      "loss": 0.2433,
+      "step": 8082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27316504593409413,
+      "learning_rate": 1.1632129498136005e-05,
+      "loss": 0.2163,
+      "step": 8083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5375141536539753,
+      "learning_rate": 1.1630293510515043e-05,
+      "loss": 0.3382,
+      "step": 8084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9905869281952622,
+      "learning_rate": 1.1628457466437091e-05,
+      "loss": 0.5418,
+      "step": 8085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43572613249512804,
+      "learning_rate": 1.1626621365965725e-05,
+      "loss": 0.2925,
+      "step": 8086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29459633182802186,
+      "learning_rate": 1.162478520916453e-05,
+      "loss": 0.267,
+      "step": 8087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.35473443242054237,
+      "learning_rate": 1.1622948996097095e-05,
+      "loss": 0.2073,
+      "step": 8088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6581069323859764,
+      "learning_rate": 1.1621112726827004e-05,
+      "loss": 0.4389,
+      "step": 8089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.36928184336908865,
+      "learning_rate": 1.161927640141785e-05,
+      "loss": 0.306,
+      "step": 8090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3721782809106861,
+      "learning_rate": 1.1617440019933226e-05,
+      "loss": 0.3317,
+      "step": 8091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4796932591821559,
+      "learning_rate": 1.1615603582436723e-05,
+      "loss": 0.2115,
+      "step": 8092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3904149073219585,
+      "learning_rate": 1.1613767088991935e-05,
+      "loss": 0.3234,
+      "step": 8093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.36772963613224796,
+      "learning_rate": 1.1611930539662463e-05,
+      "loss": 0.2398,
+      "step": 8094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23813307940661996,
+      "learning_rate": 1.1610093934511908e-05,
+      "loss": 0.1875,
+      "step": 8095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4083627297184952,
+      "learning_rate": 1.1608257273603864e-05,
+      "loss": 0.3244,
+      "step": 8096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3618253703159187,
+      "learning_rate": 1.1606420557001945e-05,
+      "loss": 0.7872,
+      "step": 8097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.282233577421707,
+      "learning_rate": 1.160458378476975e-05,
+      "loss": 0.2427,
+      "step": 8098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3926305634323061,
+      "learning_rate": 1.1602746956970886e-05,
+      "loss": 0.3108,
+      "step": 8099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8545494870759214,
+      "learning_rate": 1.1600910073668964e-05,
+      "loss": 0.5018,
+      "step": 8100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2593570986614807,
+      "learning_rate": 1.1599073134927597e-05,
+      "loss": 0.1651,
+      "step": 8101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34342933446892515,
+      "learning_rate": 1.1597236140810394e-05,
+      "loss": 0.217,
+      "step": 8102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3786898079471971,
+      "learning_rate": 1.1595399091380972e-05,
+      "loss": 0.3624,
+      "step": 8103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.661006336761479,
+      "learning_rate": 1.159356198670295e-05,
+      "loss": 0.4046,
+      "step": 8104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33321983617894935,
+      "learning_rate": 1.1591724826839943e-05,
+      "loss": 0.2469,
+      "step": 8105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3513617843380767,
+      "learning_rate": 1.1589887611855574e-05,
+      "loss": 0.3298,
+      "step": 8106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2512576604987261,
+      "learning_rate": 1.1588050341813466e-05,
+      "loss": 0.1484,
+      "step": 8107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2924953972160349,
+      "learning_rate": 1.1586213016777244e-05,
+      "loss": 0.2089,
+      "step": 8108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7746893833833915,
+      "learning_rate": 1.158437563681053e-05,
+      "loss": 0.4779,
+      "step": 8109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3601070577324971,
+      "learning_rate": 1.1582538201976958e-05,
+      "loss": 0.3198,
+      "step": 8110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27886876485206846,
+      "learning_rate": 1.1580700712340159e-05,
+      "loss": 0.2167,
+      "step": 8111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8927279441564796,
+      "learning_rate": 1.1578863167963761e-05,
+      "loss": 0.58,
+      "step": 8112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25440229767955796,
+      "learning_rate": 1.1577025568911395e-05,
+      "loss": 0.1733,
+      "step": 8113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28882003772125364,
+      "learning_rate": 1.1575187915246706e-05,
+      "loss": 0.262,
+      "step": 8114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47526159258141804,
+      "learning_rate": 1.1573350207033324e-05,
+      "loss": 0.3023,
+      "step": 8115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7238904758177274,
+      "learning_rate": 1.1571512444334894e-05,
+      "loss": 0.4253,
+      "step": 8116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3652372666944264,
+      "learning_rate": 1.1569674627215057e-05,
+      "loss": 0.2946,
+      "step": 8117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3398527200099904,
+      "learning_rate": 1.1567836755737452e-05,
+      "loss": 0.2535,
+      "step": 8118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25940567321868835,
+      "learning_rate": 1.156599882996573e-05,
+      "loss": 0.1895,
+      "step": 8119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41358852547742536,
+      "learning_rate": 1.1564160849963533e-05,
+      "loss": 0.2941,
+      "step": 8120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6890446909789003,
+      "learning_rate": 1.1562322815794516e-05,
+      "loss": 0.3646,
+      "step": 8121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39128113875580217,
+      "learning_rate": 1.1560484727522323e-05,
+      "loss": 0.3401,
+      "step": 8122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3501552171817226,
+      "learning_rate": 1.1558646585210615e-05,
+      "loss": 0.2678,
+      "step": 8123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8310288981112209,
+      "learning_rate": 1.1556808388923043e-05,
+      "loss": 0.3499,
+      "step": 8124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.32194066075690564,
+      "learning_rate": 1.155497013872326e-05,
+      "loss": 0.2148,
+      "step": 8125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3097208230540672,
+      "learning_rate": 1.1553131834674929e-05,
+      "loss": 0.2589,
+      "step": 8126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5418465886898102,
+      "learning_rate": 1.1551293476841712e-05,
+      "loss": 0.3533,
+      "step": 8127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9933329488460488,
+      "learning_rate": 1.1549455065287267e-05,
+      "loss": 0.5601,
+      "step": 8128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33251164007014095,
+      "learning_rate": 1.1547616600075262e-05,
+      "loss": 0.2702,
+      "step": 8129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48264580396110907,
+      "learning_rate": 1.1545778081269356e-05,
+      "loss": 0.3385,
+      "step": 8130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26561831151431287,
+      "learning_rate": 1.1543939508933226e-05,
+      "loss": 0.1731,
+      "step": 8131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38986731532730423,
+      "learning_rate": 1.1542100883130534e-05,
+      "loss": 0.2678,
+      "step": 8132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8541658768168554,
+      "learning_rate": 1.1540262203924957e-05,
+      "loss": 0.4704,
+      "step": 8133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37102575891994377,
+      "learning_rate": 1.1538423471380162e-05,
+      "loss": 0.2628,
+      "step": 8134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4084559161761699,
+      "learning_rate": 1.1536584685559833e-05,
+      "loss": 0.2971,
+      "step": 8135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.090623885447246,
+      "learning_rate": 1.1534745846527643e-05,
+      "loss": 0.5709,
+      "step": 8136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5691462758797022,
+      "learning_rate": 1.1532906954347265e-05,
+      "loss": 0.2653,
+      "step": 8137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3616131380469849,
+      "learning_rate": 1.1531068009082388e-05,
+      "loss": 0.2771,
+      "step": 8138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.35732383039321136,
+      "learning_rate": 1.1529229010796693e-05,
+      "loss": 0.2351,
+      "step": 8139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2853466987734752,
+      "learning_rate": 1.152738995955386e-05,
+      "loss": 0.8433,
+      "step": 8140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2917511597672303,
+      "learning_rate": 1.1525550855417579e-05,
+      "loss": 0.2001,
+      "step": 8141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41454996938168326,
+      "learning_rate": 1.152371169845154e-05,
+      "loss": 0.322,
+      "step": 8142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7889175233489341,
+      "learning_rate": 1.152187248871943e-05,
+      "loss": 0.4598,
+      "step": 8143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39376249011371567,
+      "learning_rate": 1.1520033226284942e-05,
+      "loss": 0.2423,
+      "step": 8144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42249860389750094,
+      "learning_rate": 1.1518193911211763e-05,
+      "loss": 0.3088,
+      "step": 8145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3832570482048342,
+      "learning_rate": 1.1516354543563603e-05,
+      "loss": 0.3023,
+      "step": 8146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27028430753402555,
+      "learning_rate": 1.1514515123404144e-05,
+      "loss": 0.1744,
+      "step": 8147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9814466672326471,
+      "learning_rate": 1.1512675650797093e-05,
+      "loss": 0.5945,
+      "step": 8148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5236065950219225,
+      "learning_rate": 1.1510836125806148e-05,
+      "loss": 0.3632,
+      "step": 8149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2567177557850539,
+      "learning_rate": 1.1508996548495015e-05,
+      "loss": 0.2272,
+      "step": 8150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48664804419956864,
+      "learning_rate": 1.1507156918927396e-05,
+      "loss": 0.2979,
+      "step": 8151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4403235305691835,
+      "learning_rate": 1.1505317237166997e-05,
+      "loss": 0.342,
+      "step": 8152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4455779532667353,
+      "learning_rate": 1.1503477503277526e-05,
+      "loss": 0.304,
+      "step": 8153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3233973459258135,
+      "learning_rate": 1.1501637717322695e-05,
+      "loss": 0.2459,
+      "step": 8154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6969251101798604,
+      "learning_rate": 1.1499797879366214e-05,
+      "loss": 0.4513,
+      "step": 8155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4039155950028945,
+      "learning_rate": 1.1497957989471798e-05,
+      "loss": 0.2988,
+      "step": 8156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5008602945483102,
+      "learning_rate": 1.1496118047703162e-05,
+      "loss": 0.2993,
+      "step": 8157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29040380927590353,
+      "learning_rate": 1.1494278054124019e-05,
+      "loss": 0.2655,
+      "step": 8158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39909885373138854,
+      "learning_rate": 1.1492438008798093e-05,
+      "loss": 0.264,
+      "step": 8159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4596488600608055,
+      "learning_rate": 1.1490597911789104e-05,
+      "loss": 0.2365,
+      "step": 8160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5358031775312709,
+      "learning_rate": 1.1488757763160771e-05,
+      "loss": 0.3619,
+      "step": 8161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.289051100704963,
+      "learning_rate": 1.148691756297682e-05,
+      "loss": 0.2643,
+      "step": 8162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46616362480908935,
+      "learning_rate": 1.1485077311300983e-05,
+      "loss": 0.2833,
+      "step": 8163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.43853018014503176,
+      "learning_rate": 1.1483237008196978e-05,
+      "loss": 0.3051,
+      "step": 8164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27216529737512757,
+      "learning_rate": 1.1481396653728542e-05,
+      "loss": 0.2282,
+      "step": 8165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3954700014749519,
+      "learning_rate": 1.14795562479594e-05,
+      "loss": 0.3425,
+      "step": 8166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.639378666232857,
+      "learning_rate": 1.1477715790953293e-05,
+      "loss": 0.3417,
+      "step": 8167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3230244326253486,
+      "learning_rate": 1.1475875282773948e-05,
+      "loss": 0.2642,
+      "step": 8168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0780614602536482,
+      "learning_rate": 1.1474034723485108e-05,
+      "loss": 0.6736,
+      "step": 8169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.299867058226962,
+      "learning_rate": 1.1472194113150507e-05,
+      "loss": 0.2598,
+      "step": 8170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2996405103989052,
+      "learning_rate": 1.1470353451833889e-05,
+      "loss": 0.2208,
+      "step": 8171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.62197502201416,
+      "learning_rate": 1.1468512739598991e-05,
+      "loss": 0.3585,
+      "step": 8172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28867732255449474,
+      "learning_rate": 1.1466671976509564e-05,
+      "loss": 0.2173,
+      "step": 8173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.613279875657845,
+      "learning_rate": 1.1464831162629346e-05,
+      "loss": 0.3476,
+      "step": 8174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3803647956262368,
+      "learning_rate": 1.146299029802209e-05,
+      "loss": 0.3359,
+      "step": 8175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0181850562146808,
+      "learning_rate": 1.1461149382751544e-05,
+      "loss": 0.6673,
+      "step": 8176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7603144529101236,
+      "learning_rate": 1.1459308416881454e-05,
+      "loss": 0.2102,
+      "step": 8177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30168260451446044,
+      "learning_rate": 1.145746740047558e-05,
+      "loss": 0.2995,
+      "step": 8178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3321460276508209,
+      "learning_rate": 1.1455626333597672e-05,
+      "loss": 0.2085,
+      "step": 8179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34455866728257245,
+      "learning_rate": 1.1453785216311484e-05,
+      "loss": 0.2043,
+      "step": 8180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3685518357521418,
+      "learning_rate": 1.1451944048680779e-05,
+      "loss": 0.3475,
+      "step": 8181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4831933139625984,
+      "learning_rate": 1.1450102830769314e-05,
+      "loss": 0.3685,
+      "step": 8182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3531600940862889,
+      "learning_rate": 1.1448261562640848e-05,
+      "loss": 0.2355,
+      "step": 8183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.439610405654752,
+      "learning_rate": 1.1446420244359148e-05,
+      "loss": 0.3821,
+      "step": 8184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26939221681386455,
+      "learning_rate": 1.1444578875987978e-05,
+      "loss": 0.2208,
+      "step": 8185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27650453372703404,
+      "learning_rate": 1.1442737457591102e-05,
+      "loss": 0.2099,
+      "step": 8186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3629929252444442,
+      "learning_rate": 1.144089598923229e-05,
+      "loss": 0.7209,
+      "step": 8187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6214268287048519,
+      "learning_rate": 1.1439054470975312e-05,
+      "loss": 0.506,
+      "step": 8188
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3119158539908204,
+      "learning_rate": 1.143721290288394e-05,
+      "loss": 0.2878,
+      "step": 8189
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3318970742581742,
+      "learning_rate": 1.1435371285021948e-05,
+      "loss": 0.262,
+      "step": 8190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2773017863670644,
+      "learning_rate": 1.1433529617453108e-05,
+      "loss": 0.1932,
+      "step": 8191
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8760997978912737,
+      "learning_rate": 1.1431687900241201e-05,
+      "loss": 0.4663,
+      "step": 8192
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3367584318928723,
+      "learning_rate": 1.142984613345e-05,
+      "loss": 0.2346,
+      "step": 8193
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4579062321186222,
+      "learning_rate": 1.1428004317143293e-05,
+      "loss": 0.3515,
+      "step": 8194
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6150169647018473,
+      "learning_rate": 1.1426162451384857e-05,
+      "loss": 0.3871,
+      "step": 8195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3216805849443447,
+      "learning_rate": 1.1424320536238478e-05,
+      "loss": 0.2268,
+      "step": 8196
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25155788365831927,
+      "learning_rate": 1.1422478571767937e-05,
+      "loss": 0.206,
+      "step": 8197
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34325115785592575,
+      "learning_rate": 1.1420636558037026e-05,
+      "loss": 0.2802,
+      "step": 8198
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3558996480535589,
+      "learning_rate": 1.1418794495109528e-05,
+      "loss": 0.2655,
+      "step": 8199
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7393643667231588,
+      "learning_rate": 1.1416952383049244e-05,
+      "loss": 0.4348,
+      "step": 8200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3308899478444616,
+      "learning_rate": 1.1415110221919958e-05,
+      "loss": 0.2869,
+      "step": 8201
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4330551569974486,
+      "learning_rate": 1.1413268011785463e-05,
+      "loss": 0.3299,
+      "step": 8202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16466774797088232,
+      "learning_rate": 1.1411425752709561e-05,
+      "loss": 0.0933,
+      "step": 8203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5658043381450948,
+      "learning_rate": 1.1409583444756043e-05,
+      "loss": 0.3813,
+      "step": 8204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39733186363612516,
+      "learning_rate": 1.1407741087988713e-05,
+      "loss": 0.311,
+      "step": 8205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5293473050356182,
+      "learning_rate": 1.1405898682471367e-05,
+      "loss": 0.3052,
+      "step": 8206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3863878784564842,
+      "learning_rate": 1.1404056228267813e-05,
+      "loss": 0.3073,
+      "step": 8207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5500023359847825,
+      "learning_rate": 1.140221372544185e-05,
+      "loss": 0.3302,
+      "step": 8208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26832299732179304,
+      "learning_rate": 1.1400371174057287e-05,
+      "loss": 0.2081,
+      "step": 8209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3489058273162678,
+      "learning_rate": 1.139852857417793e-05,
+      "loss": 0.2296,
+      "step": 8210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.35245208526279237,
+      "learning_rate": 1.139668592586759e-05,
+      "loss": 0.2888,
+      "step": 8211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6348054017588185,
+      "learning_rate": 1.1394843229190076e-05,
+      "loss": 0.3672,
+      "step": 8212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3565702444562063,
+      "learning_rate": 1.1393000484209202e-05,
+      "loss": 0.2839,
+      "step": 8213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.35892910995714983,
+      "learning_rate": 1.139115769098878e-05,
+      "loss": 0.2883,
+      "step": 8214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4652118043069549,
+      "learning_rate": 1.1389314849592626e-05,
+      "loss": 0.2623,
+      "step": 8215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.31534488534631117,
+      "learning_rate": 1.1387471960084557e-05,
+      "loss": 0.1603,
+      "step": 8216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3043071870765772,
+      "learning_rate": 1.1385629022528397e-05,
+      "loss": 0.2729,
+      "step": 8217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.791220490302,
+      "learning_rate": 1.1383786036987963e-05,
+      "loss": 0.433,
+      "step": 8218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3690113603730788,
+      "learning_rate": 1.1381943003527077e-05,
+      "loss": 0.1939,
+      "step": 8219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3844076124295093,
+      "learning_rate": 1.1380099922209564e-05,
+      "loss": 0.3177,
+      "step": 8220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3525552292700522,
+      "learning_rate": 1.1378256793099251e-05,
+      "loss": 0.3186,
+      "step": 8221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2606543369638635,
+      "learning_rate": 1.1376413616259965e-05,
+      "loss": 0.1299,
+      "step": 8222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45789735825217354,
+      "learning_rate": 1.1374570391755532e-05,
+      "loss": 0.3216,
+      "step": 8223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7984697434570868,
+      "learning_rate": 1.137272711964979e-05,
+      "loss": 0.4888,
+      "step": 8224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.31774601929632706,
+      "learning_rate": 1.1370883800006562e-05,
+      "loss": 0.2252,
+      "step": 8225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42364764216329703,
+      "learning_rate": 1.1369040432889691e-05,
+      "loss": 0.3239,
+      "step": 8226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.43654263319431763,
+      "learning_rate": 1.1367197018363005e-05,
+      "loss": 0.2615,
+      "step": 8227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5638770827996676,
+      "learning_rate": 1.1365353556490348e-05,
+      "loss": 0.3034,
+      "step": 8228
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26140699917356136,
+      "learning_rate": 1.1363510047335553e-05,
+      "loss": 0.2156,
+      "step": 8229
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8208422504947541,
+      "learning_rate": 1.1361666490962468e-05,
+      "loss": 0.4557,
+      "step": 8230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9822644350984603,
+      "learning_rate": 1.1359822887434927e-05,
+      "loss": 0.7397,
+      "step": 8231
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3314402613856236,
+      "learning_rate": 1.1357979236816781e-05,
+      "loss": 0.2076,
+      "step": 8232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4487027140708006,
+      "learning_rate": 1.135613553917187e-05,
+      "loss": 0.3122,
+      "step": 8233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9682417335085425,
+      "learning_rate": 1.1354291794564045e-05,
+      "loss": 0.4434,
+      "step": 8234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2746870381544497,
+      "learning_rate": 1.1352448003057153e-05,
+      "loss": 0.1845,
+      "step": 8235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3943326196154475,
+      "learning_rate": 1.1350604164715044e-05,
+      "loss": 0.237,
+      "step": 8236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3423516469836019,
+      "learning_rate": 1.1348760279601572e-05,
+      "loss": 0.3115,
+      "step": 8237
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3458819009956223,
+      "learning_rate": 1.134691634778059e-05,
+      "loss": 0.219,
+      "step": 8238
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8749680292286744,
+      "learning_rate": 1.1345072369315951e-05,
+      "loss": 0.4917,
+      "step": 8239
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.444178078876986,
+      "learning_rate": 1.1343228344271515e-05,
+      "loss": 0.352,
+      "step": 8240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7944684734767636,
+      "learning_rate": 1.1341384272711138e-05,
+      "loss": 0.2822,
+      "step": 8241
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28858447543694066,
+      "learning_rate": 1.1339540154698682e-05,
+      "loss": 0.1913,
+      "step": 8242
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33636234607029175,
+      "learning_rate": 1.133769599029801e-05,
+      "loss": 0.2435,
+      "step": 8243
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37425119175237026,
+      "learning_rate": 1.1335851779572979e-05,
+      "loss": 0.2879,
+      "step": 8244
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3420784443581969,
+      "learning_rate": 1.1334007522587462e-05,
+      "loss": 0.2764,
+      "step": 8245
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.626268923474981,
+      "learning_rate": 1.1332163219405318e-05,
+      "loss": 0.4184,
+      "step": 8246
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33327889942284983,
+      "learning_rate": 1.1330318870090427e-05,
+      "loss": 0.2542,
+      "step": 8247
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3209884215108911,
+      "learning_rate": 1.1328474474706643e-05,
+      "loss": 0.224,
+      "step": 8248
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38396143689121437,
+      "learning_rate": 1.132663003331785e-05,
+      "loss": 0.2768,
+      "step": 8249
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34285888950900517,
+      "learning_rate": 1.1324785545987911e-05,
+      "loss": 0.2586,
+      "step": 8250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7016183186742044,
+      "learning_rate": 1.1322941012780707e-05,
+      "loss": 0.3433,
+      "step": 8251
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4224569861018966,
+      "learning_rate": 1.1321096433760116e-05,
+      "loss": 0.3503,
+      "step": 8252
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2995614695059995,
+      "learning_rate": 1.1319251808990009e-05,
+      "loss": 0.2572,
+      "step": 8253
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2672865639780329,
+      "learning_rate": 1.1317407138534268e-05,
+      "loss": 0.714,
+      "step": 8254
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.20313439742018613,
+      "learning_rate": 1.1315562422456776e-05,
+      "loss": 0.1281,
+      "step": 8255
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37114924445860703,
+      "learning_rate": 1.1313717660821413e-05,
+      "loss": 0.2819,
+      "step": 8256
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4130806594357466,
+      "learning_rate": 1.1311872853692065e-05,
+      "loss": 0.3215,
+      "step": 8257
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7315074555515565,
+      "learning_rate": 1.1310028001132615e-05,
+      "loss": 0.2889,
+      "step": 8258
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3518162924006683,
+      "learning_rate": 1.1308183103206956e-05,
+      "loss": 0.2678,
+      "step": 8259
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.263822710183851,
+      "learning_rate": 1.1306338159978968e-05,
+      "loss": 0.8082,
+      "step": 8260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2290140963302349,
+      "learning_rate": 1.1304493171512548e-05,
+      "loss": 0.1771,
+      "step": 8261
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3609070041446921,
+      "learning_rate": 1.1302648137871584e-05,
+      "loss": 0.2591,
+      "step": 8262
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7500955160282625,
+      "learning_rate": 1.1300803059119969e-05,
+      "loss": 0.4461,
+      "step": 8263
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47247522252554397,
+      "learning_rate": 1.1298957935321604e-05,
+      "loss": 0.4008,
+      "step": 8264
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29187071697633193,
+      "learning_rate": 1.129711276654038e-05,
+      "loss": 0.2102,
+      "step": 8265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9689601512300138,
+      "learning_rate": 1.1295267552840198e-05,
+      "loss": 0.7327,
+      "step": 8266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39459373406513565,
+      "learning_rate": 1.1293422294284955e-05,
+      "loss": 0.255,
+      "step": 8267
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2841527542011416,
+      "learning_rate": 1.1291576990938556e-05,
+      "loss": 0.234,
+      "step": 8268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26846990348811833,
+      "learning_rate": 1.1289731642864896e-05,
+      "loss": 0.2712,
+      "step": 8269
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2160805062145421,
+      "learning_rate": 1.1287886250127888e-05,
+      "loss": 0.7619,
+      "step": 8270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3180662479564137,
+      "learning_rate": 1.1286040812791431e-05,
+      "loss": 0.2088,
+      "step": 8271
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2780419499498312,
+      "learning_rate": 1.1284195330919443e-05,
+      "loss": 0.763,
+      "step": 8272
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34751151421346865,
+      "learning_rate": 1.128234980457582e-05,
+      "loss": 0.3257,
+      "step": 8273
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29037446054088023,
+      "learning_rate": 1.1280504233824481e-05,
+      "loss": 0.2382,
+      "step": 8274
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.400642610476773,
+      "learning_rate": 1.1278658618729334e-05,
+      "loss": 0.2889,
+      "step": 8275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28690418079254904,
+      "learning_rate": 1.1276812959354295e-05,
+      "loss": 0.2466,
+      "step": 8276
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3976401918567556,
+      "learning_rate": 1.1274967255763278e-05,
+      "loss": 0.2747,
+      "step": 8277
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4454863521161837,
+      "learning_rate": 1.1273121508020202e-05,
+      "loss": 0.2982,
+      "step": 8278
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5729809893476822,
+      "learning_rate": 1.1271275716188978e-05,
+      "loss": 0.4256,
+      "step": 8279
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.410761763093016,
+      "learning_rate": 1.1269429880333533e-05,
+      "loss": 0.2653,
+      "step": 8280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2517747570353783,
+      "learning_rate": 1.1267584000517788e-05,
+      "loss": 0.2289,
+      "step": 8281
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4175701025964965,
+      "learning_rate": 1.1265738076805663e-05,
+      "loss": 0.2465,
+      "step": 8282
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4040133712318601,
+      "learning_rate": 1.1263892109261081e-05,
+      "loss": 0.2884,
+      "step": 8283
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3315685666860372,
+      "learning_rate": 1.126204609794797e-05,
+      "loss": 0.2601,
+      "step": 8284
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6738959505588366,
+      "learning_rate": 1.1260200042930257e-05,
+      "loss": 0.4439,
+      "step": 8285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3713996287944544,
+      "learning_rate": 1.125835394427187e-05,
+      "loss": 0.3126,
+      "step": 8286
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26116662030586907,
+      "learning_rate": 1.1256507802036742e-05,
+      "loss": 0.1798,
+      "step": 8287
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2923099580818043,
+      "learning_rate": 1.12546616162888e-05,
+      "loss": 0.241,
+      "step": 8288
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4591105534812718,
+      "learning_rate": 1.1252815387091984e-05,
+      "loss": 0.293,
+      "step": 8289
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5414758905754559,
+      "learning_rate": 1.1250969114510221e-05,
+      "loss": 0.4012,
+      "step": 8290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8942911584626979,
+      "learning_rate": 1.1249122798607454e-05,
+      "loss": 0.3676,
+      "step": 8291
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35940592339450533,
+      "learning_rate": 1.1247276439447616e-05,
+      "loss": 0.2682,
+      "step": 8292
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3883606131416445,
+      "learning_rate": 1.124543003709465e-05,
+      "loss": 0.3171,
+      "step": 8293
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2069420790155224,
+      "learning_rate": 1.1243583591612495e-05,
+      "loss": 0.1109,
+      "step": 8294
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3893463873924719,
+      "learning_rate": 1.1241737103065096e-05,
+      "loss": 0.2684,
+      "step": 8295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44638528446469994,
+      "learning_rate": 1.1239890571516389e-05,
+      "loss": 0.3573,
+      "step": 8296
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9358041698450383,
+      "learning_rate": 1.123804399703033e-05,
+      "loss": 0.3285,
+      "step": 8297
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6012185686365062,
+      "learning_rate": 1.1236197379670861e-05,
+      "loss": 0.3471,
+      "step": 8298
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36799061504790626,
+      "learning_rate": 1.1234350719501927e-05,
+      "loss": 0.3477,
+      "step": 8299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22412798408524393,
+      "learning_rate": 1.1232504016587482e-05,
+      "loss": 0.1747,
+      "step": 8300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4149487453115047,
+      "learning_rate": 1.1230657270991476e-05,
+      "loss": 0.304,
+      "step": 8301
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5459475622031438,
+      "learning_rate": 1.1228810482777859e-05,
+      "loss": 0.3991,
+      "step": 8302
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9562637175288264,
+      "learning_rate": 1.1226963652010592e-05,
+      "loss": 0.5005,
+      "step": 8303
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27131366109251476,
+      "learning_rate": 1.1225116778753622e-05,
+      "loss": 0.2124,
+      "step": 8304
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37477020048094534,
+      "learning_rate": 1.1223269863070913e-05,
+      "loss": 0.3101,
+      "step": 8305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37990887770967674,
+      "learning_rate": 1.1221422905026424e-05,
+      "loss": 0.1822,
+      "step": 8306
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4007847446197605,
+      "learning_rate": 1.1219575904684109e-05,
+      "loss": 0.2274,
+      "step": 8307
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3859141568843464,
+      "learning_rate": 1.1217728862107932e-05,
+      "loss": 0.3088,
+      "step": 8308
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5814554045581455,
+      "learning_rate": 1.1215881777361858e-05,
+      "loss": 0.3877,
+      "step": 8309
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31317402052160326,
+      "learning_rate": 1.1214034650509853e-05,
+      "loss": 0.2069,
+      "step": 8310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3864302814986327,
+      "learning_rate": 1.1212187481615875e-05,
+      "loss": 0.6831,
+      "step": 8311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24158373982842526,
+      "learning_rate": 1.1210340270743903e-05,
+      "loss": 0.2158,
+      "step": 8312
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38673341680171747,
+      "learning_rate": 1.1208493017957893e-05,
+      "loss": 0.2194,
+      "step": 8313
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3480815756689843,
+      "learning_rate": 1.1206645723321825e-05,
+      "loss": 0.3104,
+      "step": 8314
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.142989272685945,
+      "learning_rate": 1.1204798386899669e-05,
+      "loss": 0.8243,
+      "step": 8315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.334355108496965,
+      "learning_rate": 1.1202951008755395e-05,
+      "loss": 0.2723,
+      "step": 8316
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.342405895746166,
+      "learning_rate": 1.1201103588952979e-05,
+      "loss": 0.2292,
+      "step": 8317
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35075106621619223,
+      "learning_rate": 1.11992561275564e-05,
+      "loss": 0.2068,
+      "step": 8318
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5953512589583299,
+      "learning_rate": 1.1197408624629626e-05,
+      "loss": 0.4117,
+      "step": 8319
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27076852878493896,
+      "learning_rate": 1.119556108023665e-05,
+      "loss": 0.2509,
+      "step": 8320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2379250749740365,
+      "learning_rate": 1.119371349444144e-05,
+      "loss": 0.8403,
+      "step": 8321
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6147005884618181,
+      "learning_rate": 1.1191865867307987e-05,
+      "loss": 0.4054,
+      "step": 8322
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3673949040504266,
+      "learning_rate": 1.1190018198900267e-05,
+      "loss": 0.2359,
+      "step": 8323
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2579786089458528,
+      "learning_rate": 1.118817048928227e-05,
+      "loss": 0.2306,
+      "step": 8324
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5153506800417209,
+      "learning_rate": 1.1186322738517983e-05,
+      "loss": 0.3424,
+      "step": 8325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2880843628827029,
+      "learning_rate": 1.1184474946671384e-05,
+      "loss": 0.2012,
+      "step": 8326
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1330484123308262,
+      "learning_rate": 1.1182627113806475e-05,
+      "loss": 0.7822,
+      "step": 8327
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29700312651214944,
+      "learning_rate": 1.1180779239987233e-05,
+      "loss": 0.2688,
+      "step": 8328
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3926248997185353,
+      "learning_rate": 1.1178931325277662e-05,
+      "loss": 0.3018,
+      "step": 8329
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4880065387650843,
+      "learning_rate": 1.1177083369741749e-05,
+      "loss": 0.3457,
+      "step": 8330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4622472597385853,
+      "learning_rate": 1.117523537344349e-05,
+      "loss": 0.3208,
+      "step": 8331
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26773122651731995,
+      "learning_rate": 1.1173387336446879e-05,
+      "loss": 0.246,
+      "step": 8332
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27469813985436253,
+      "learning_rate": 1.1171539258815916e-05,
+      "loss": 0.2141,
+      "step": 8333
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6729771240210307,
+      "learning_rate": 1.1169691140614597e-05,
+      "loss": 0.4272,
+      "step": 8334
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40650217240748127,
+      "learning_rate": 1.1167842981906927e-05,
+      "loss": 0.3098,
+      "step": 8335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33788023906972464,
+      "learning_rate": 1.1165994782756902e-05,
+      "loss": 0.2863,
+      "step": 8336
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8965463163501143,
+      "learning_rate": 1.1164146543228529e-05,
+      "loss": 0.4183,
+      "step": 8337
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37608376944922794,
+      "learning_rate": 1.116229826338581e-05,
+      "loss": 0.2967,
+      "step": 8338
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1709144692691702,
+      "learning_rate": 1.1160449943292754e-05,
+      "loss": 0.7408,
+      "step": 8339
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20606643679960182,
+      "learning_rate": 1.1158601583013365e-05,
+      "loss": 0.1803,
+      "step": 8340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3825272882807248,
+      "learning_rate": 1.1156753182611655e-05,
+      "loss": 0.2924,
+      "step": 8341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.757439344922993,
+      "learning_rate": 1.1154904742151628e-05,
+      "loss": 0.408,
+      "step": 8342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47983847004271224,
+      "learning_rate": 1.1153056261697303e-05,
+      "loss": 0.2154,
+      "step": 8343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31408770009618836,
+      "learning_rate": 1.1151207741312688e-05,
+      "loss": 0.2766,
+      "step": 8344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1777256065853905,
+      "learning_rate": 1.11493591810618e-05,
+      "loss": 0.8841,
+      "step": 8345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.14795722092036293,
+      "learning_rate": 1.1147510581008654e-05,
+      "loss": 0.0727,
+      "step": 8346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44407525535508613,
+      "learning_rate": 1.114566194121726e-05,
+      "loss": 0.3331,
+      "step": 8347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4429171903471908,
+      "learning_rate": 1.1143813261751648e-05,
+      "loss": 0.3368,
+      "step": 8348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5326441778644128,
+      "learning_rate": 1.1141964542675831e-05,
+      "loss": 0.2415,
+      "step": 8349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35629373156998423,
+      "learning_rate": 1.1140115784053828e-05,
+      "loss": 0.2704,
+      "step": 8350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.473832740368251,
+      "learning_rate": 1.1138266985949668e-05,
+      "loss": 0.3667,
+      "step": 8351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24716345117047317,
+      "learning_rate": 1.113641814842737e-05,
+      "loss": 0.1892,
+      "step": 8352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37177419017363733,
+      "learning_rate": 1.1134569271550959e-05,
+      "loss": 0.2241,
+      "step": 8353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6795655987417271,
+      "learning_rate": 1.1132720355384466e-05,
+      "loss": 0.433,
+      "step": 8354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4546558880720396,
+      "learning_rate": 1.1130871399991912e-05,
+      "loss": 0.3476,
+      "step": 8355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2735908648714051,
+      "learning_rate": 1.1129022405437333e-05,
+      "loss": 0.2299,
+      "step": 8356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1504777351766975,
+      "learning_rate": 1.1127173371784755e-05,
+      "loss": 0.7199,
+      "step": 8357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.320853590136227,
+      "learning_rate": 1.112532429909821e-05,
+      "loss": 0.22,
+      "step": 8358
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2922420371940266,
+      "learning_rate": 1.1123475187441735e-05,
+      "loss": 0.2365,
+      "step": 8359
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4621379502333083,
+      "learning_rate": 1.1121626036879362e-05,
+      "loss": 0.3615,
+      "step": 8360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7091643725104544,
+      "learning_rate": 1.1119776847475128e-05,
+      "loss": 0.3798,
+      "step": 8361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32042337282411815,
+      "learning_rate": 1.1117927619293072e-05,
+      "loss": 0.2344,
+      "step": 8362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.46225350017426375,
+      "learning_rate": 1.1116078352397226e-05,
+      "loss": 0.3815,
+      "step": 8363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44794763153718986,
+      "learning_rate": 1.1114229046851639e-05,
+      "loss": 0.3283,
+      "step": 8364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35863928811423473,
+      "learning_rate": 1.1112379702720346e-05,
+      "loss": 0.2993,
+      "step": 8365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2997559968892406,
+      "learning_rate": 1.1110530320067395e-05,
+      "loss": 0.1826,
+      "step": 8366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31177442430328983,
+      "learning_rate": 1.110868089895682e-05,
+      "loss": 0.2818,
+      "step": 8367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34426675271911766,
+      "learning_rate": 1.1106831439452678e-05,
+      "loss": 0.2813,
+      "step": 8368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.159122972842692,
+      "learning_rate": 1.1104981941619008e-05,
+      "loss": 0.5404,
+      "step": 8369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6308084552291687,
+      "learning_rate": 1.1103132405519866e-05,
+      "loss": 0.4178,
+      "step": 8370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3508909818733613,
+      "learning_rate": 1.1101282831219292e-05,
+      "loss": 0.2844,
+      "step": 8371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24925963458866746,
+      "learning_rate": 1.1099433218781342e-05,
+      "loss": 0.1824,
+      "step": 8372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0146895770631508,
+      "learning_rate": 1.1097583568270068e-05,
+      "loss": 0.5618,
+      "step": 8373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3402111836733542,
+      "learning_rate": 1.1095733879749517e-05,
+      "loss": 0.2561,
+      "step": 8374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44394453768906555,
+      "learning_rate": 1.1093884153283755e-05,
+      "loss": 0.2666,
+      "step": 8375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5030724555789422,
+      "learning_rate": 1.1092034388936827e-05,
+      "loss": 0.3637,
+      "step": 8376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3053604077245766,
+      "learning_rate": 1.1090184586772798e-05,
+      "loss": 0.2541,
+      "step": 8377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26072680670778176,
+      "learning_rate": 1.1088334746855724e-05,
+      "loss": 0.1699,
+      "step": 8378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33213697117973734,
+      "learning_rate": 1.1086484869249664e-05,
+      "loss": 0.2501,
+      "step": 8379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3411697209711852,
+      "learning_rate": 1.1084634954018679e-05,
+      "loss": 0.2749,
+      "step": 8380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7937308936897097,
+      "learning_rate": 1.1082785001226833e-05,
+      "loss": 0.527,
+      "step": 8381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5427995751624755,
+      "learning_rate": 1.108093501093819e-05,
+      "loss": 0.2829,
+      "step": 8382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37423001686239854,
+      "learning_rate": 1.1079084983216812e-05,
+      "loss": 0.294,
+      "step": 8383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2610551272973318,
+      "learning_rate": 1.107723491812677e-05,
+      "loss": 0.2371,
+      "step": 8384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2848863442961672,
+      "learning_rate": 1.1075384815732126e-05,
+      "loss": 0.1655,
+      "step": 8385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36620040221470535,
+      "learning_rate": 1.1073534676096953e-05,
+      "loss": 0.3269,
+      "step": 8386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40895567597138205,
+      "learning_rate": 1.107168449928532e-05,
+      "loss": 0.3493,
+      "step": 8387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9694638436873771,
+      "learning_rate": 1.1069834285361299e-05,
+      "loss": 0.3581,
+      "step": 8388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3307306363709223,
+      "learning_rate": 1.1067984034388963e-05,
+      "loss": 0.2843,
+      "step": 8389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2628285063571431,
+      "learning_rate": 1.1066133746432388e-05,
+      "loss": 0.1729,
+      "step": 8390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39208782700167216,
+      "learning_rate": 1.1064283421555643e-05,
+      "loss": 0.2981,
+      "step": 8391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32337473433052144,
+      "learning_rate": 1.1062433059822813e-05,
+      "loss": 0.2115,
+      "step": 8392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9751414436431275,
+      "learning_rate": 1.106058266129797e-05,
+      "loss": 0.4711,
+      "step": 8393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.917279257161967,
+      "learning_rate": 1.1058732226045195e-05,
+      "loss": 0.504,
+      "step": 8394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27816877474586016,
+      "learning_rate": 1.1056881754128568e-05,
+      "loss": 0.2211,
+      "step": 8395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4602594483651638,
+      "learning_rate": 1.1055031245612172e-05,
+      "loss": 0.3818,
+      "step": 8396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2664228010575755,
+      "learning_rate": 1.1053180700560086e-05,
+      "loss": 0.1863,
+      "step": 8397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36439137476338745,
+      "learning_rate": 1.1051330119036404e-05,
+      "loss": 0.2241,
+      "step": 8398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.638885939720972,
+      "learning_rate": 1.1049479501105202e-05,
+      "loss": 0.3925,
+      "step": 8399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4978465949232517,
+      "learning_rate": 1.1047628846830571e-05,
+      "loss": 0.3931,
+      "step": 8400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33079743434334197,
+      "learning_rate": 1.1045778156276596e-05,
+      "loss": 0.1901,
+      "step": 8401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3101247786660534,
+      "learning_rate": 1.104392742950737e-05,
+      "loss": 0.2588,
+      "step": 8402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2805944667132338,
+      "learning_rate": 1.104207666658698e-05,
+      "loss": 0.2896,
+      "step": 8403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3730808898739765,
+      "learning_rate": 1.1040225867579522e-05,
+      "loss": 0.2604,
+      "step": 8404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5112786993131957,
+      "learning_rate": 1.1038375032549085e-05,
+      "loss": 0.33,
+      "step": 8405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9322563183427266,
+      "learning_rate": 1.1036524161559767e-05,
+      "loss": 0.5786,
+      "step": 8406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.355904824217904,
+      "learning_rate": 1.103467325467566e-05,
+      "loss": 0.2834,
+      "step": 8407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2743782333531795,
+      "learning_rate": 1.1032822311960866e-05,
+      "loss": 0.2112,
+      "step": 8408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4259052635545149,
+      "learning_rate": 1.1030971333479477e-05,
+      "loss": 0.3138,
+      "step": 8409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37998773225911436,
+      "learning_rate": 1.1029120319295597e-05,
+      "loss": 0.3443,
+      "step": 8410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34994216114806836,
+      "learning_rate": 1.1027269269473324e-05,
+      "loss": 0.2692,
+      "step": 8411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0286726329274982,
+      "learning_rate": 1.102541818407676e-05,
+      "loss": 0.6472,
+      "step": 8412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39969223773471535,
+      "learning_rate": 1.1023567063170008e-05,
+      "loss": 0.3095,
+      "step": 8413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4792364386637519,
+      "learning_rate": 1.1021715906817172e-05,
+      "loss": 0.2102,
+      "step": 8414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24998225105266594,
+      "learning_rate": 1.101986471508236e-05,
+      "loss": 0.2621,
+      "step": 8415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3296303617196106,
+      "learning_rate": 1.1018013488029675e-05,
+      "loss": 0.272,
+      "step": 8416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.831873250616347,
+      "learning_rate": 1.1016162225723227e-05,
+      "loss": 0.5823,
+      "step": 8417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3007402335020122,
+      "learning_rate": 1.1014310928227125e-05,
+      "loss": 0.2131,
+      "step": 8418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.364565455538296,
+      "learning_rate": 1.101245959560548e-05,
+      "loss": 0.2672,
+      "step": 8419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3912456702815622,
+      "learning_rate": 1.1010608227922401e-05,
+      "loss": 0.3275,
+      "step": 8420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4532113401173461,
+      "learning_rate": 1.1008756825242007e-05,
+      "loss": 0.2709,
+      "step": 8421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5767832010251636,
+      "learning_rate": 1.10069053876284e-05,
+      "loss": 0.3767,
+      "step": 8422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3190711161352703,
+      "learning_rate": 1.100505391514571e-05,
+      "loss": 0.309,
+      "step": 8423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2566852014475039,
+      "learning_rate": 1.1003202407858042e-05,
+      "loss": 0.1297,
+      "step": 8424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3887013888366776,
+      "learning_rate": 1.1001350865829519e-05,
+      "loss": 0.2731,
+      "step": 8425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49546719830955543,
+      "learning_rate": 1.0999499289124259e-05,
+      "loss": 0.371,
+      "step": 8426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3685020261407154,
+      "learning_rate": 1.0997647677806381e-05,
+      "loss": 0.318,
+      "step": 8427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.323905474118738,
+      "learning_rate": 1.0995796031940004e-05,
+      "loss": 0.2349,
+      "step": 8428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5429412404412678,
+      "learning_rate": 1.0993944351589257e-05,
+      "loss": 0.4474,
+      "step": 8429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2698732310913593,
+      "learning_rate": 1.0992092636818261e-05,
+      "loss": 0.1615,
+      "step": 8430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2698012591809379,
+      "learning_rate": 1.0990240887691135e-05,
+      "loss": 0.2257,
+      "step": 8431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5174891868486703,
+      "learning_rate": 1.0988389104272012e-05,
+      "loss": 0.3601,
+      "step": 8432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6007606961336437,
+      "learning_rate": 1.098653728662502e-05,
+      "loss": 0.4257,
+      "step": 8433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32474766627803614,
+      "learning_rate": 1.098468543481428e-05,
+      "loss": 0.2144,
+      "step": 8434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39732993885930923,
+      "learning_rate": 1.0982833548903926e-05,
+      "loss": 0.3183,
+      "step": 8435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26185582493114695,
+      "learning_rate": 1.0980981628958091e-05,
+      "loss": 0.2002,
+      "step": 8436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3630987442855988,
+      "learning_rate": 1.0979129675040902e-05,
+      "loss": 0.1708,
+      "step": 8437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4030675818976453,
+      "learning_rate": 1.0977277687216497e-05,
+      "loss": 0.3411,
+      "step": 8438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37339136570184556,
+      "learning_rate": 1.0975425665549005e-05,
+      "loss": 0.3173,
+      "step": 8439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8845059457123733,
+      "learning_rate": 1.0973573610102566e-05,
+      "loss": 0.4562,
+      "step": 8440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31660898433428125,
+      "learning_rate": 1.0971721520941312e-05,
+      "loss": 0.239,
+      "step": 8441
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23668371578828956,
+      "learning_rate": 1.0969869398129385e-05,
+      "loss": 0.1924,
+      "step": 8442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40438627675862854,
+      "learning_rate": 1.0968017241730922e-05,
+      "loss": 0.3176,
+      "step": 8443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4396742144328353,
+      "learning_rate": 1.0966165051810066e-05,
+      "loss": 0.2638,
+      "step": 8444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8565682090328727,
+      "learning_rate": 1.0964312828430952e-05,
+      "loss": 0.4941,
+      "step": 8445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.644386355055355,
+      "learning_rate": 1.096246057165773e-05,
+      "loss": 0.3571,
+      "step": 8446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2632623269668236,
+      "learning_rate": 1.0960608281554536e-05,
+      "loss": 0.258,
+      "step": 8447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3898501466449026,
+      "learning_rate": 1.0958755958185521e-05,
+      "loss": 0.2782,
+      "step": 8448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39076503964584275,
+      "learning_rate": 1.0956903601614827e-05,
+      "loss": 0.2519,
+      "step": 8449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34799939748395714,
+      "learning_rate": 1.0955051211906607e-05,
+      "loss": 0.2699,
+      "step": 8450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34692010363267356,
+      "learning_rate": 1.0953198789125e-05,
+      "loss": 0.3284,
+      "step": 8451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5508850692361121,
+      "learning_rate": 1.095134633333416e-05,
+      "loss": 0.3315,
+      "step": 8452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38972264224855135,
+      "learning_rate": 1.0949493844598237e-05,
+      "loss": 0.3421,
+      "step": 8453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2861316486348512,
+      "learning_rate": 1.0947641322981387e-05,
+      "loss": 0.2046,
+      "step": 8454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4901872853125752,
+      "learning_rate": 1.0945788768547754e-05,
+      "loss": 0.3408,
+      "step": 8455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34229442973648827,
+      "learning_rate": 1.0943936181361501e-05,
+      "loss": 0.3144,
+      "step": 8456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30739139755930395,
+      "learning_rate": 1.0942083561486775e-05,
+      "loss": 0.1327,
+      "step": 8457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3992928835707834,
+      "learning_rate": 1.0940230908987737e-05,
+      "loss": 0.3175,
+      "step": 8458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2975848893390521,
+      "learning_rate": 1.0938378223928545e-05,
+      "loss": 0.2855,
+      "step": 8459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6888931226475903,
+      "learning_rate": 1.0936525506373353e-05,
+      "loss": 0.3482,
+      "step": 8460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5397706730536465,
+      "learning_rate": 1.0934672756386324e-05,
+      "loss": 0.4013,
+      "step": 8461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2018669105288497,
+      "learning_rate": 1.0932819974031616e-05,
+      "loss": 0.1768,
+      "step": 8462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3994888465967652,
+      "learning_rate": 1.0930967159373393e-05,
+      "loss": 0.2612,
+      "step": 8463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7874372028522038,
+      "learning_rate": 1.0929114312475818e-05,
+      "loss": 0.4882,
+      "step": 8464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3514672262598415,
+      "learning_rate": 1.0927261433403055e-05,
+      "loss": 0.3112,
+      "step": 8465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5423505007021057,
+      "learning_rate": 1.092540852221927e-05,
+      "loss": 0.3962,
+      "step": 8466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.47049811026281263,
+      "learning_rate": 1.0923555578988624e-05,
+      "loss": 0.2611,
+      "step": 8467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39710438531386655,
+      "learning_rate": 1.0921702603775288e-05,
+      "loss": 0.2879,
+      "step": 8468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2792052805726842,
+      "learning_rate": 1.0919849596643434e-05,
+      "loss": 0.1829,
+      "step": 8469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3047677379880632,
+      "learning_rate": 1.0917996557657224e-05,
+      "loss": 0.2355,
+      "step": 8470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40934080266730394,
+      "learning_rate": 1.0916143486880836e-05,
+      "loss": 0.3056,
+      "step": 8471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8182633785891162,
+      "learning_rate": 1.0914290384378436e-05,
+      "loss": 0.4636,
+      "step": 8472
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5738428159353489,
+      "learning_rate": 1.09124372502142e-05,
+      "loss": 0.284,
+      "step": 8473
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2757029445511652,
+      "learning_rate": 1.09105840844523e-05,
+      "loss": 0.2102,
+      "step": 8474
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24592269658330476,
+      "learning_rate": 1.0908730887156915e-05,
+      "loss": 0.2472,
+      "step": 8475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4877436876419471,
+      "learning_rate": 1.090687765839222e-05,
+      "loss": 0.2014,
+      "step": 8476
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3949928498732002,
+      "learning_rate": 1.0905024398222386e-05,
+      "loss": 0.3212,
+      "step": 8477
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5094201877628184,
+      "learning_rate": 1.09031711067116e-05,
+      "loss": 0.3381,
+      "step": 8478
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9709515948124112,
+      "learning_rate": 1.0901317783924032e-05,
+      "loss": 0.4758,
+      "step": 8479
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3008208066982179,
+      "learning_rate": 1.0899464429923874e-05,
+      "loss": 0.2053,
+      "step": 8480
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2425174984408598,
+      "learning_rate": 1.0897611044775299e-05,
+      "loss": 0.1676,
+      "step": 8481
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3567217475863464,
+      "learning_rate": 1.0895757628542492e-05,
+      "loss": 0.3243,
+      "step": 8482
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3865743507662977,
+      "learning_rate": 1.0893904181289637e-05,
+      "loss": 0.2332,
+      "step": 8483
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9081131636400291,
+      "learning_rate": 1.0892050703080918e-05,
+      "loss": 0.4564,
+      "step": 8484
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.337498254558959,
+      "learning_rate": 1.0890197193980523e-05,
+      "loss": 0.8662,
+      "step": 8485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3029460382510471,
+      "learning_rate": 1.0888343654052636e-05,
+      "loss": 0.2082,
+      "step": 8486
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22595061674438283,
+      "learning_rate": 1.0886490083361445e-05,
+      "loss": 0.2207,
+      "step": 8487
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6513624736080968,
+      "learning_rate": 1.0884636481971145e-05,
+      "loss": 0.4554,
+      "step": 8488
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3546626256247301,
+      "learning_rate": 1.0882782849945917e-05,
+      "loss": 0.2338,
+      "step": 8489
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4004379725262802,
+      "learning_rate": 1.088092918734996e-05,
+      "loss": 0.3369,
+      "step": 8490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.06392676133309,
+      "learning_rate": 1.0879075494247459e-05,
+      "loss": 0.6164,
+      "step": 8491
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33643344970466654,
+      "learning_rate": 1.0877221770702618e-05,
+      "loss": 0.2818,
+      "step": 8492
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20948852930825718,
+      "learning_rate": 1.087536801677962e-05,
+      "loss": 0.1037,
+      "step": 8493
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3585673923696074,
+      "learning_rate": 1.0873514232542665e-05,
+      "loss": 0.3189,
+      "step": 8494
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35729435806128756,
+      "learning_rate": 1.0871660418055954e-05,
+      "loss": 0.2793,
+      "step": 8495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.616608207430578,
+      "learning_rate": 1.0869806573383675e-05,
+      "loss": 0.362,
+      "step": 8496
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3825893173323816,
+      "learning_rate": 1.0867952698590036e-05,
+      "loss": 0.6839,
+      "step": 8497
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27447229366512443,
+      "learning_rate": 1.0866098793739229e-05,
+      "loss": 0.2518,
+      "step": 8498
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21800662896904763,
+      "learning_rate": 1.0864244858895461e-05,
+      "loss": 0.1464,
+      "step": 8499
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5799123938435271,
+      "learning_rate": 1.086239089412293e-05,
+      "loss": 0.4081,
+      "step": 8500
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3538200919550498,
+      "learning_rate": 1.086053689948584e-05,
+      "loss": 0.2998,
+      "step": 8501
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.441401374507026,
+      "learning_rate": 1.085868287504839e-05,
+      "loss": 0.3884,
+      "step": 8502
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4445663999073777,
+      "learning_rate": 1.0856828820874794e-05,
+      "loss": 0.2894,
+      "step": 8503
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3221303501831563,
+      "learning_rate": 1.0854974737029248e-05,
+      "loss": 0.2564,
+      "step": 8504
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2526415990634436,
+      "learning_rate": 1.0853120623575968e-05,
+      "loss": 0.1755,
+      "step": 8505
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3396586056153518,
+      "learning_rate": 1.0851266480579155e-05,
+      "loss": 0.2631,
+      "step": 8506
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.32619397347087004,
+      "learning_rate": 1.0849412308103023e-05,
+      "loss": 0.2644,
+      "step": 8507
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8049881277903056,
+      "learning_rate": 1.0847558106211775e-05,
+      "loss": 0.6005,
+      "step": 8508
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39477084899592646,
+      "learning_rate": 1.0845703874969629e-05,
+      "loss": 0.2584,
+      "step": 8509
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3284641100010038,
+      "learning_rate": 1.0843849614440793e-05,
+      "loss": 0.2518,
+      "step": 8510
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3164534348135442,
+      "learning_rate": 1.0841995324689482e-05,
+      "loss": 0.2347,
+      "step": 8511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6746338727776406,
+      "learning_rate": 1.0840141005779907e-05,
+      "loss": 0.3469,
+      "step": 8512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3406299935814668,
+      "learning_rate": 1.0838286657776289e-05,
+      "loss": 0.2758,
+      "step": 8513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.368507523241613,
+      "learning_rate": 1.0836432280742837e-05,
+      "loss": 0.3614,
+      "step": 8514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4196989763584233,
+      "learning_rate": 1.0834577874743772e-05,
+      "loss": 0.2835,
+      "step": 8515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.286344189141085,
+      "learning_rate": 1.0832723439843313e-05,
+      "loss": 0.2135,
+      "step": 8516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8103735048771367,
+      "learning_rate": 1.0830868976105677e-05,
+      "loss": 0.4709,
+      "step": 8517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3806559083984567,
+      "learning_rate": 1.0829014483595081e-05,
+      "loss": 0.3363,
+      "step": 8518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22981873993975727,
+      "learning_rate": 1.0827159962375753e-05,
+      "loss": 0.1717,
+      "step": 8519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1201661013138515,
+      "learning_rate": 1.0825305412511906e-05,
+      "loss": 0.7243,
+      "step": 8520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29503616397144117,
+      "learning_rate": 1.0823450834067772e-05,
+      "loss": 0.2378,
+      "step": 8521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2551033595031021,
+      "learning_rate": 1.0821596227107572e-05,
+      "loss": 0.2024,
+      "step": 8522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7899047872546144,
+      "learning_rate": 1.0819741591695526e-05,
+      "loss": 0.4307,
+      "step": 8523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6356314127947628,
+      "learning_rate": 1.0817886927895866e-05,
+      "loss": 0.4684,
+      "step": 8524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.344138473561077,
+      "learning_rate": 1.0816032235772816e-05,
+      "loss": 0.194,
+      "step": 8525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3089788067804815,
+      "learning_rate": 1.0814177515390605e-05,
+      "loss": 0.297,
+      "step": 8526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.32417727591946566,
+      "learning_rate": 1.081232276681346e-05,
+      "loss": 0.2092,
+      "step": 8527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4153972292522417,
+      "learning_rate": 1.0810467990105617e-05,
+      "loss": 0.2856,
+      "step": 8528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.655625681624118,
+      "learning_rate": 1.0808613185331297e-05,
+      "loss": 0.3203,
+      "step": 8529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3785813411354443,
+      "learning_rate": 1.0806758352554743e-05,
+      "loss": 0.3061,
+      "step": 8530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37824841761056816,
+      "learning_rate": 1.0804903491840178e-05,
+      "loss": 0.2755,
+      "step": 8531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21373532506861836,
+      "learning_rate": 1.080304860325184e-05,
+      "loss": 0.1536,
+      "step": 8532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4860820363746777,
+      "learning_rate": 1.0801193686853964e-05,
+      "loss": 0.3688,
+      "step": 8533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2978394874490292,
+      "learning_rate": 1.0799338742710788e-05,
+      "loss": 0.24,
+      "step": 8534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48456413182199815,
+      "learning_rate": 1.0797483770886542e-05,
+      "loss": 0.2991,
+      "step": 8535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6425265613421929,
+      "learning_rate": 1.0795628771445467e-05,
+      "loss": 0.4898,
+      "step": 8536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3882124012170663,
+      "learning_rate": 1.0793773744451804e-05,
+      "loss": 0.2956,
+      "step": 8537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29591018567833727,
+      "learning_rate": 1.079191868996979e-05,
+      "loss": 0.2537,
+      "step": 8538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.275873528349889,
+      "learning_rate": 1.0790063608063664e-05,
+      "loss": 0.2111,
+      "step": 8539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.309022281433369,
+      "learning_rate": 1.078820849879767e-05,
+      "loss": 0.2702,
+      "step": 8540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7002771887809298,
+      "learning_rate": 1.0786353362236051e-05,
+      "loss": 0.4943,
+      "step": 8541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3367810286858073,
+      "learning_rate": 1.0784498198443048e-05,
+      "loss": 0.267,
+      "step": 8542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6251041919918563,
+      "learning_rate": 1.0782643007482908e-05,
+      "loss": 0.3424,
+      "step": 8543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4196247509404695,
+      "learning_rate": 1.0780787789419868e-05,
+      "loss": 0.3126,
+      "step": 8544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.222604196010212,
+      "learning_rate": 1.0778932544318185e-05,
+      "loss": 0.1972,
+      "step": 8545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5323880097633098,
+      "learning_rate": 1.0777077272242103e-05,
+      "loss": 0.3519,
+      "step": 8546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3522018391320144,
+      "learning_rate": 1.0775221973255866e-05,
+      "loss": 0.3304,
+      "step": 8547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7935920052225027,
+      "learning_rate": 1.0773366647423724e-05,
+      "loss": 0.3471,
+      "step": 8548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3614599034269359,
+      "learning_rate": 1.0771511294809933e-05,
+      "loss": 0.2682,
+      "step": 8549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30816488025273653,
+      "learning_rate": 1.0769655915478734e-05,
+      "loss": 0.2914,
+      "step": 8550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41552829457555635,
+      "learning_rate": 1.076780050949439e-05,
+      "loss": 0.2073,
+      "step": 8551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3957754702487912,
+      "learning_rate": 1.0765945076921143e-05,
+      "loss": 0.3081,
+      "step": 8552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3438857205809205,
+      "learning_rate": 1.0764089617823252e-05,
+      "loss": 0.2495,
+      "step": 8553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3916725126911459,
+      "learning_rate": 1.0762234132264969e-05,
+      "loss": 0.3022,
+      "step": 8554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3472478565503047,
+      "learning_rate": 1.0760378620310551e-05,
+      "loss": 0.2252,
+      "step": 8555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5140747501812202,
+      "learning_rate": 1.0758523082024255e-05,
+      "loss": 0.39,
+      "step": 8556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46369297394140513,
+      "learning_rate": 1.0756667517470337e-05,
+      "loss": 0.3532,
+      "step": 8557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2763335707004524,
+      "learning_rate": 1.0754811926713053e-05,
+      "loss": 0.233,
+      "step": 8558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2642442656817401,
+      "learning_rate": 1.075295630981667e-05,
+      "loss": 0.1955,
+      "step": 8559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8901340050690897,
+      "learning_rate": 1.0751100666845437e-05,
+      "loss": 0.5739,
+      "step": 8560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33780611056603577,
+      "learning_rate": 1.0749244997863624e-05,
+      "loss": 0.1916,
+      "step": 8561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.314626187389568,
+      "learning_rate": 1.0747389302935487e-05,
+      "loss": 0.2898,
+      "step": 8562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5661987660912312,
+      "learning_rate": 1.074553358212529e-05,
+      "loss": 0.4068,
+      "step": 8563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22613700731661518,
+      "learning_rate": 1.07436778354973e-05,
+      "loss": 0.1042,
+      "step": 8564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26126861145961344,
+      "learning_rate": 1.0741822063115774e-05,
+      "loss": 0.2493,
+      "step": 8565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48082663751471855,
+      "learning_rate": 1.0739966265044985e-05,
+      "loss": 0.3952,
+      "step": 8566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.776697669372256,
+      "learning_rate": 1.0738110441349194e-05,
+      "loss": 0.568,
+      "step": 8567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35344836698827203,
+      "learning_rate": 1.0736254592092674e-05,
+      "loss": 0.2445,
+      "step": 8568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41759160329707223,
+      "learning_rate": 1.0734398717339687e-05,
+      "loss": 0.3361,
+      "step": 8569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36406742647874546,
+      "learning_rate": 1.0732542817154505e-05,
+      "loss": 0.268,
+      "step": 8570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21027776814527477,
+      "learning_rate": 1.0730686891601394e-05,
+      "loss": 0.1567,
+      "step": 8571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0486613084863996,
+      "learning_rate": 1.072883094074463e-05,
+      "loss": 0.66,
+      "step": 8572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30409813856147094,
+      "learning_rate": 1.0726974964648478e-05,
+      "loss": 0.2823,
+      "step": 8573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3632538787528237,
+      "learning_rate": 1.072511896337722e-05,
+      "loss": 0.2477,
+      "step": 8574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6178813686029212,
+      "learning_rate": 1.0723262936995118e-05,
+      "loss": 0.4058,
+      "step": 8575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5408942484954933,
+      "learning_rate": 1.0721406885566455e-05,
+      "loss": 0.3104,
+      "step": 8576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22758907901569544,
+      "learning_rate": 1.07195508091555e-05,
+      "loss": 0.1969,
+      "step": 8577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29891880470997145,
+      "learning_rate": 1.0717694707826534e-05,
+      "loss": 0.2564,
+      "step": 8578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6764541746498948,
+      "learning_rate": 1.0715838581643829e-05,
+      "loss": 0.4602,
+      "step": 8579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3694126858365215,
+      "learning_rate": 1.0713982430671668e-05,
+      "loss": 0.3039,
+      "step": 8580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3388532266528095,
+      "learning_rate": 1.0712126254974325e-05,
+      "loss": 0.2766,
+      "step": 8581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8196371361869945,
+      "learning_rate": 1.0710270054616077e-05,
+      "loss": 0.3736,
+      "step": 8582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2575004355415229,
+      "learning_rate": 1.070841382966121e-05,
+      "loss": 0.2168,
+      "step": 8583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3064963483696533,
+      "learning_rate": 1.0706557580174002e-05,
+      "loss": 0.1623,
+      "step": 8584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30059761729633716,
+      "learning_rate": 1.0704701306218737e-05,
+      "loss": 0.2986,
+      "step": 8585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3834769871882269,
+      "learning_rate": 1.0702845007859697e-05,
+      "loss": 0.2874,
+      "step": 8586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7050746353337347,
+      "learning_rate": 1.0700988685161162e-05,
+      "loss": 0.3589,
+      "step": 8587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8554268457849118,
+      "learning_rate": 1.069913233818742e-05,
+      "loss": 0.5188,
+      "step": 8588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2600277824616293,
+      "learning_rate": 1.0697275967002754e-05,
+      "loss": 0.26,
+      "step": 8589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2661738871990709,
+      "learning_rate": 1.069541957167145e-05,
+      "loss": 0.1824,
+      "step": 8590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5728659494558793,
+      "learning_rate": 1.06935631522578e-05,
+      "loss": 0.2859,
+      "step": 8591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36724855285541735,
+      "learning_rate": 1.0691706708826084e-05,
+      "loss": 0.3184,
+      "step": 8592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36684883389905065,
+      "learning_rate": 1.0689850241440598e-05,
+      "loss": 0.3254,
+      "step": 8593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.448707775246444,
+      "learning_rate": 1.0687993750165623e-05,
+      "loss": 0.2209,
+      "step": 8594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40597125420971497,
+      "learning_rate": 1.0686137235065458e-05,
+      "loss": 0.3169,
+      "step": 8595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34023890211065355,
+      "learning_rate": 1.068428069620439e-05,
+      "loss": 0.2522,
+      "step": 8596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41372972524648555,
+      "learning_rate": 1.0682424133646712e-05,
+      "loss": 0.275,
+      "step": 8597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3490980753166338,
+      "learning_rate": 1.068056754745671e-05,
+      "loss": 0.2738,
+      "step": 8598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9919442403924039,
+      "learning_rate": 1.0678710937698689e-05,
+      "loss": 0.7263,
+      "step": 8599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.477348484888539,
+      "learning_rate": 1.0676854304436936e-05,
+      "loss": 0.1544,
+      "step": 8600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29238634288399545,
+      "learning_rate": 1.0674997647735745e-05,
+      "loss": 0.2659,
+      "step": 8601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4137595889493302,
+      "learning_rate": 1.0673140967659418e-05,
+      "loss": 0.2705,
+      "step": 8602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7976162801001311,
+      "learning_rate": 1.0671284264272249e-05,
+      "loss": 0.4373,
+      "step": 8603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35688291511303527,
+      "learning_rate": 1.066942753763853e-05,
+      "loss": 0.2226,
+      "step": 8604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28563980083841095,
+      "learning_rate": 1.0667570787822568e-05,
+      "loss": 0.2589,
+      "step": 8605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8815030483266307,
+      "learning_rate": 1.0665714014888657e-05,
+      "loss": 0.5098,
+      "step": 8606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3215761513411512,
+      "learning_rate": 1.0663857218901097e-05,
+      "loss": 0.2103,
+      "step": 8607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9842923959569092,
+      "learning_rate": 1.0662000399924193e-05,
+      "loss": 0.4303,
+      "step": 8608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34745505755938316,
+      "learning_rate": 1.066014355802224e-05,
+      "loss": 0.3268,
+      "step": 8609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.31255754616013104,
+      "learning_rate": 1.0658286693259544e-05,
+      "loss": 0.1985,
+      "step": 8610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.273327258396647,
+      "learning_rate": 1.065642980570041e-05,
+      "loss": 0.1981,
+      "step": 8611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4237158009498341,
+      "learning_rate": 1.0654572895409142e-05,
+      "loss": 0.3071,
+      "step": 8612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3323909110807824,
+      "learning_rate": 1.065271596245004e-05,
+      "loss": 0.2247,
+      "step": 8613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.197242401674588,
+      "learning_rate": 1.0650859006887412e-05,
+      "loss": 0.4439,
+      "step": 8614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1780449474286145,
+      "learning_rate": 1.0649002028785564e-05,
+      "loss": 0.7566,
+      "step": 8615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3025044117036119,
+      "learning_rate": 1.0647145028208808e-05,
+      "loss": 0.2614,
+      "step": 8616
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21862780541713553,
+      "learning_rate": 1.0645288005221443e-05,
+      "loss": 0.1848,
+      "step": 8617
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9973641936045772,
+      "learning_rate": 1.0643430959887786e-05,
+      "loss": 0.5432,
+      "step": 8618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3686939049542852,
+      "learning_rate": 1.064157389227214e-05,
+      "loss": 0.2805,
+      "step": 8619
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3729969071937,
+      "learning_rate": 1.063971680243882e-05,
+      "loss": 0.3485,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41373947900301006,
+      "learning_rate": 1.063785969045213e-05,
+      "loss": 0.3296,
+      "step": 8621
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3333660307427822,
+      "learning_rate": 1.063600255637639e-05,
+      "loss": 0.2753,
+      "step": 8622
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1846235788447922,
+      "learning_rate": 1.0634145400275906e-05,
+      "loss": 0.0857,
+      "step": 8623
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4198005388095872,
+      "learning_rate": 1.0632288222214998e-05,
+      "loss": 0.333,
+      "step": 8624
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4751118320166682,
+      "learning_rate": 1.0630431022257975e-05,
+      "loss": 0.2883,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5403740119919342,
+      "learning_rate": 1.062857380046915e-05,
+      "loss": 0.3654,
+      "step": 8626
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.906474785955322,
+      "learning_rate": 1.0626716556912845e-05,
+      "loss": 0.656,
+      "step": 8627
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40934538461942965,
+      "learning_rate": 1.062485929165337e-05,
+      "loss": 0.3045,
+      "step": 8628
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2483934265695432,
+      "learning_rate": 1.0623002004755045e-05,
+      "loss": 0.2398,
+      "step": 8629
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8469478281869853,
+      "learning_rate": 1.0621144696282187e-05,
+      "loss": 0.2506,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41460984148063096,
+      "learning_rate": 1.0619287366299116e-05,
+      "loss": 0.2685,
+      "step": 8631
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7057658416397729,
+      "learning_rate": 1.061743001487015e-05,
+      "loss": 0.3861,
+      "step": 8632
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44006599321635453,
+      "learning_rate": 1.0615572642059608e-05,
+      "loss": 0.2791,
+      "step": 8633
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35366409363198736,
+      "learning_rate": 1.0613715247931811e-05,
+      "loss": 0.2718,
+      "step": 8634
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28522502179300874,
+      "learning_rate": 1.0611857832551088e-05,
+      "loss": 0.1987,
+      "step": 8635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4444119647481306,
+      "learning_rate": 1.0610000395981748e-05,
+      "loss": 0.2836,
+      "step": 8636
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29795136052669713,
+      "learning_rate": 1.0608142938288122e-05,
+      "loss": 0.2719,
+      "step": 8637
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7921438236185548,
+      "learning_rate": 1.0606285459534531e-05,
+      "loss": 0.4997,
+      "step": 8638
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.113545353925166,
+      "learning_rate": 1.0604427959785305e-05,
+      "loss": 0.5553,
+      "step": 8639
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35533823163207906,
+      "learning_rate": 1.0602570439104758e-05,
+      "loss": 0.2753,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4131396122160858,
+      "learning_rate": 1.0600712897557229e-05,
+      "loss": 0.3235,
+      "step": 8641
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.578787262125609,
+      "learning_rate": 1.0598855335207032e-05,
+      "loss": 0.3251,
+      "step": 8642
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2622329097119207,
+      "learning_rate": 1.0596997752118505e-05,
+      "loss": 0.1882,
+      "step": 8643
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6319084606170982,
+      "learning_rate": 1.0595140148355971e-05,
+      "loss": 0.7732,
+      "step": 8644
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38998737359035557,
+      "learning_rate": 1.059328252398376e-05,
+      "loss": 0.329,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3543340025838131,
+      "learning_rate": 1.0591424879066199e-05,
+      "loss": 0.1859,
+      "step": 8646
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4751509222241394,
+      "learning_rate": 1.058956721366762e-05,
+      "loss": 0.4029,
+      "step": 8647
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40849080293214024,
+      "learning_rate": 1.0587709527852354e-05,
+      "loss": 0.3217,
+      "step": 8648
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2225141351938953,
+      "learning_rate": 1.0585851821684731e-05,
+      "loss": 0.1289,
+      "step": 8649
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35872861250269855,
+      "learning_rate": 1.0583994095229086e-05,
+      "loss": 0.2722,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3321392581467473,
+      "learning_rate": 1.0582136348549751e-05,
+      "loss": 0.7329,
+      "step": 8651
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2864610099756312,
+      "learning_rate": 1.0580278581711062e-05,
+      "loss": 0.1966,
+      "step": 8652
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36678325852272914,
+      "learning_rate": 1.0578420794777347e-05,
+      "loss": 0.3385,
+      "step": 8653
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6607280902420078,
+      "learning_rate": 1.0576562987812946e-05,
+      "loss": 0.4297,
+      "step": 8654
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26838225983595704,
+      "learning_rate": 1.057470516088219e-05,
+      "loss": 0.2015,
+      "step": 8655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.23625021150066397,
+      "learning_rate": 1.0572847314049424e-05,
+      "loss": 0.2171,
+      "step": 8656
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3904254222964338,
+      "learning_rate": 1.0570989447378977e-05,
+      "loss": 0.8084,
+      "step": 8657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3996383469966337,
+      "learning_rate": 1.056913156093519e-05,
+      "loss": 0.3012,
+      "step": 8658
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47744810436214125,
+      "learning_rate": 1.0567273654782402e-05,
+      "loss": 0.2646,
+      "step": 8659
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.37398280158270103,
+      "learning_rate": 1.0565415728984954e-05,
+      "loss": 0.3085,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2956477490756631,
+      "learning_rate": 1.0563557783607182e-05,
+      "loss": 0.173,
+      "step": 8661
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26479803730153023,
+      "learning_rate": 1.0561699818713427e-05,
+      "loss": 0.1764,
+      "step": 8662
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1463907876064328,
+      "learning_rate": 1.0559841834368032e-05,
+      "loss": 0.6209,
+      "step": 8663
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3129762774498092,
+      "learning_rate": 1.055798383063534e-05,
+      "loss": 0.2687,
+      "step": 8664
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3635218135502235,
+      "learning_rate": 1.0556125807579691e-05,
+      "loss": 0.3221,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.692553553878542,
+      "learning_rate": 1.0554267765265428e-05,
+      "loss": 0.3318,
+      "step": 8666
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3232964697528742,
+      "learning_rate": 1.0552409703756896e-05,
+      "loss": 0.2057,
+      "step": 8667
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2696835374164365,
+      "learning_rate": 1.0550551623118442e-05,
+      "loss": 0.2448,
+      "step": 8668
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4353801277831827,
+      "learning_rate": 1.0548693523414408e-05,
+      "loss": 0.3043,
+      "step": 8669
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6110278457480871,
+      "learning_rate": 1.0546835404709142e-05,
+      "loss": 0.3705,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4375873666361446,
+      "learning_rate": 1.0544977267066986e-05,
+      "loss": 0.3336,
+      "step": 8671
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3254332303228954,
+      "learning_rate": 1.0543119110552293e-05,
+      "loss": 0.2477,
+      "step": 8672
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4354120294911844,
+      "learning_rate": 1.054126093522941e-05,
+      "loss": 0.3037,
+      "step": 8673
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2439818838843636,
+      "learning_rate": 1.053940274116268e-05,
+      "loss": 0.1891,
+      "step": 8674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0463292921514271,
+      "learning_rate": 1.0537544528416462e-05,
+      "loss": 0.396,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2922783263343222,
+      "learning_rate": 1.0535686297055095e-05,
+      "loss": 0.2758,
+      "step": 8676
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33915434052209564,
+      "learning_rate": 1.0533828047142936e-05,
+      "loss": 0.313,
+      "step": 8677
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2238073735137502,
+      "learning_rate": 1.0531969778744333e-05,
+      "loss": 0.7319,
+      "step": 8678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24234025299640638,
+      "learning_rate": 1.0530111491923642e-05,
+      "loss": 0.1626,
+      "step": 8679
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2732911508199734,
+      "learning_rate": 1.0528253186745212e-05,
+      "loss": 0.2467,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5100215847515718,
+      "learning_rate": 1.05263948632734e-05,
+      "loss": 0.4121,
+      "step": 8681
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7570537100397893,
+      "learning_rate": 1.052453652157255e-05,
+      "loss": 0.2958,
+      "step": 8682
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40521850584754937,
+      "learning_rate": 1.0522678161707028e-05,
+      "loss": 0.3425,
+      "step": 8683
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35722274671332144,
+      "learning_rate": 1.0520819783741183e-05,
+      "loss": 0.3157,
+      "step": 8684
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.378465025509005,
+      "learning_rate": 1.0518961387739371e-05,
+      "loss": 0.1266,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28385498048625774,
+      "learning_rate": 1.0517102973765947e-05,
+      "loss": 0.2468,
+      "step": 8686
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6575638629333442,
+      "learning_rate": 1.0515244541885272e-05,
+      "loss": 0.4461,
+      "step": 8687
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2883829375610862,
+      "learning_rate": 1.0513386092161698e-05,
+      "loss": 0.2233,
+      "step": 8688
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2855510500547569,
+      "learning_rate": 1.0511527624659585e-05,
+      "loss": 0.2701,
+      "step": 8689
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9124086442455803,
+      "learning_rate": 1.0509669139443298e-05,
+      "loss": 0.5854,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3336901059066564,
+      "learning_rate": 1.0507810636577183e-05,
+      "loss": 0.1928,
+      "step": 8691
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.263736365371567,
+      "learning_rate": 1.0505952116125613e-05,
+      "loss": 0.2334,
+      "step": 8692
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.731716892313344,
+      "learning_rate": 1.0504093578152939e-05,
+      "loss": 0.4945,
+      "step": 8693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5263262316670798,
+      "learning_rate": 1.050223502272353e-05,
+      "loss": 0.3654,
+      "step": 8694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2641345245374191,
+      "learning_rate": 1.050037644990174e-05,
+      "loss": 0.1886,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36928392656217834,
+      "learning_rate": 1.0498517859751937e-05,
+      "loss": 0.3137,
+      "step": 8696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.609889087343759,
+      "learning_rate": 1.0496659252338481e-05,
+      "loss": 0.3196,
+      "step": 8697
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32459381957992633,
+      "learning_rate": 1.049480062772574e-05,
+      "loss": 0.2378,
+      "step": 8698
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7666747207357101,
+      "learning_rate": 1.0492941985978068e-05,
+      "loss": 0.4289,
+      "step": 8699
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3039731492970164,
+      "learning_rate": 1.049108332715984e-05,
+      "loss": 0.2918,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30935907905723153,
+      "learning_rate": 1.048922465133542e-05,
+      "loss": 0.1984,
+      "step": 8701
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3842240130182527,
+      "learning_rate": 1.0487365958569168e-05,
+      "loss": 0.2887,
+      "step": 8702
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4564138910898665,
+      "learning_rate": 1.0485507248925455e-05,
+      "loss": 0.7804,
+      "step": 8703
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2959826982997899,
+      "learning_rate": 1.0483648522468648e-05,
+      "loss": 0.2633,
+      "step": 8704
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6727655285294992,
+      "learning_rate": 1.0481789779263112e-05,
+      "loss": 0.3068,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8354526542235158,
+      "learning_rate": 1.0479931019373218e-05,
+      "loss": 0.5327,
+      "step": 8706
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28734045291129795,
+      "learning_rate": 1.0478072242863329e-05,
+      "loss": 0.2392,
+      "step": 8707
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2747471101459275,
+      "learning_rate": 1.0476213449797823e-05,
+      "loss": 0.2075,
+      "step": 8708
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7019374127703849,
+      "learning_rate": 1.0474354640241065e-05,
+      "loss": 0.4068,
+      "step": 8709
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36357077422014367,
+      "learning_rate": 1.0472495814257426e-05,
+      "loss": 0.294,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6829568268139392,
+      "learning_rate": 1.0470636971911277e-05,
+      "loss": 0.3449,
+      "step": 8711
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3321892388960226,
+      "learning_rate": 1.046877811326699e-05,
+      "loss": 0.3047,
+      "step": 8712
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3844356675302627,
+      "learning_rate": 1.0466919238388937e-05,
+      "loss": 0.2823,
+      "step": 8713
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2569912841925734,
+      "learning_rate": 1.046506034734149e-05,
+      "loss": 0.1119,
+      "step": 8714
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5010661824434921,
+      "learning_rate": 1.0463201440189026e-05,
+      "loss": 0.3222,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3171331861477249,
+      "learning_rate": 1.0461342516995911e-05,
+      "loss": 0.2934,
+      "step": 8716
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43958228142956374,
+      "learning_rate": 1.0459483577826531e-05,
+      "loss": 0.3712,
+      "step": 8717
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3325947509858215,
+      "learning_rate": 1.0457624622745249e-05,
+      "loss": 0.2327,
+      "step": 8718
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4079029698952522,
+      "learning_rate": 1.0455765651816447e-05,
+      "loss": 0.2972,
+      "step": 8719
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.23506748013692239,
+      "learning_rate": 1.0453906665104503e-05,
+      "loss": 0.2195,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5733380652274306,
+      "learning_rate": 1.045204766267379e-05,
+      "loss": 0.3302,
+      "step": 8721
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3160060820791429,
+      "learning_rate": 1.0450188644588684e-05,
+      "loss": 0.2767,
+      "step": 8722
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4478524014302271,
+      "learning_rate": 1.0448329610913566e-05,
+      "loss": 0.3633,
+      "step": 8723
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3882152266604375,
+      "learning_rate": 1.0446470561712811e-05,
+      "loss": 0.2507,
+      "step": 8724
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34054329567829944,
+      "learning_rate": 1.0444611497050802e-05,
+      "loss": 0.2896,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2992192854650521,
+      "learning_rate": 1.0442752416991912e-05,
+      "loss": 0.1979,
+      "step": 8726
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45370200964194424,
+      "learning_rate": 1.0440893321600529e-05,
+      "loss": 0.2756,
+      "step": 8727
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32295994917909365,
+      "learning_rate": 1.0439034210941029e-05,
+      "loss": 0.2951,
+      "step": 8728
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5271479543748887,
+      "learning_rate": 1.043717508507779e-05,
+      "loss": 0.6307,
+      "step": 8729
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2596569844559131,
+      "learning_rate": 1.0435315944075202e-05,
+      "loss": 0.8572,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3131215205253095,
+      "learning_rate": 1.0433456787997636e-05,
+      "loss": 0.198,
+      "step": 8731
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24968167949562506,
+      "learning_rate": 1.0431597616909483e-05,
+      "loss": 0.2025,
+      "step": 8732
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.48335571779497233,
+      "learning_rate": 1.0429738430875123e-05,
+      "loss": 0.4133,
+      "step": 8733
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31200844208398004,
+      "learning_rate": 1.042787922995894e-05,
+      "loss": 0.1918,
+      "step": 8734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45930456874730524,
+      "learning_rate": 1.0426020014225313e-05,
+      "loss": 0.3733,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40373907057363134,
+      "learning_rate": 1.0424160783738637e-05,
+      "loss": 0.342,
+      "step": 8736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3350587314333596,
+      "learning_rate": 1.042230153856329e-05,
+      "loss": 0.2144,
+      "step": 8737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3563876657519294,
+      "learning_rate": 1.0420442278763658e-05,
+      "loss": 0.2922,
+      "step": 8738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31504314484019597,
+      "learning_rate": 1.0418583004404128e-05,
+      "loss": 0.2687,
+      "step": 8739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30162193745736154,
+      "learning_rate": 1.0416723715549086e-05,
+      "loss": 0.2186,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3170050829277982,
+      "learning_rate": 1.041486441226292e-05,
+      "loss": 0.8277,
+      "step": 8741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3552064424071086,
+      "learning_rate": 1.0413005094610018e-05,
+      "loss": 0.8414,
+      "step": 8742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.380534543132684,
+      "learning_rate": 1.0411145762654767e-05,
+      "loss": 0.2638,
+      "step": 8743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36920878357651454,
+      "learning_rate": 1.0409286416461557e-05,
+      "loss": 0.2848,
+      "step": 8744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2959147152051227,
+      "learning_rate": 1.0407427056094772e-05,
+      "loss": 0.2261,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34193258538459975,
+      "learning_rate": 1.040556768161881e-05,
+      "loss": 0.2593,
+      "step": 8746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45975870258676144,
+      "learning_rate": 1.0403708293098054e-05,
+      "loss": 0.2501,
+      "step": 8747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4738679517364789,
+      "learning_rate": 1.04018488905969e-05,
+      "loss": 0.3933,
+      "step": 8748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3316696976128548,
+      "learning_rate": 1.0399989474179735e-05,
+      "loss": 0.2681,
+      "step": 8749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.567887887250414,
+      "learning_rate": 1.0398130043910949e-05,
+      "loss": 0.3258,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2892458714833695,
+      "learning_rate": 1.0396270599854939e-05,
+      "loss": 0.2585,
+      "step": 8751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4389296101630068,
+      "learning_rate": 1.0394411142076092e-05,
+      "loss": 0.2814,
+      "step": 8752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35415454086871245,
+      "learning_rate": 1.039255167063881e-05,
+      "loss": 0.2628,
+      "step": 8753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6841238312775954,
+      "learning_rate": 1.0390692185607479e-05,
+      "loss": 0.3674,
+      "step": 8754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41943936748101157,
+      "learning_rate": 1.0388832687046493e-05,
+      "loss": 0.3115,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3136131249500698,
+      "learning_rate": 1.0386973175020248e-05,
+      "loss": 0.3109,
+      "step": 8756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2241031296136957,
+      "learning_rate": 1.0385113649593137e-05,
+      "loss": 0.1017,
+      "step": 8757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25406672869397945,
+      "learning_rate": 1.0383254110829557e-05,
+      "loss": 0.2081,
+      "step": 8758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3567819115485671,
+      "learning_rate": 1.0381394558793907e-05,
+      "loss": 0.3352,
+      "step": 8759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8288749721657916,
+      "learning_rate": 1.0379534993550574e-05,
+      "loss": 0.4522,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35424045018318084,
+      "learning_rate": 1.0377675415163965e-05,
+      "loss": 0.2828,
+      "step": 8761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7610154560364215,
+      "learning_rate": 1.0375815823698471e-05,
+      "loss": 0.4219,
+      "step": 8762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3571386552082526,
+      "learning_rate": 1.0373956219218495e-05,
+      "loss": 0.2569,
+      "step": 8763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27089130035377157,
+      "learning_rate": 1.0372096601788426e-05,
+      "loss": 0.2144,
+      "step": 8764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4068884998575498,
+      "learning_rate": 1.0370236971472671e-05,
+      "loss": 0.251,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7412488012692768,
+      "learning_rate": 1.0368377328335623e-05,
+      "loss": 0.5212,
+      "step": 8766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27216313191409564,
+      "learning_rate": 1.0366517672441687e-05,
+      "loss": 0.2245,
+      "step": 8767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.429758312937244,
+      "learning_rate": 1.0364658003855256e-05,
+      "loss": 0.3467,
+      "step": 8768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4300319006050432,
+      "learning_rate": 1.0362798322640736e-05,
+      "loss": 0.2656,
+      "step": 8769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2356666590382897,
+      "learning_rate": 1.0360938628862527e-05,
+      "loss": 0.1391,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3233756808129512,
+      "learning_rate": 1.0359078922585029e-05,
+      "loss": 0.2775,
+      "step": 8771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8330387344459389,
+      "learning_rate": 1.0357219203872641e-05,
+      "loss": 0.4525,
+      "step": 8772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34761547025612005,
+      "learning_rate": 1.035535947278977e-05,
+      "loss": 0.2182,
+      "step": 8773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4854843573284205,
+      "learning_rate": 1.035349972940081e-05,
+      "loss": 0.3956,
+      "step": 8774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3489720860066471,
+      "learning_rate": 1.0351639973770175e-05,
+      "loss": 0.2953,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29659453460115837,
+      "learning_rate": 1.0349780205962264e-05,
+      "loss": 0.1995,
+      "step": 8776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2555721583072774,
+      "learning_rate": 1.0347920426041475e-05,
+      "loss": 0.188,
+      "step": 8777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1484304454055863,
+      "learning_rate": 1.034606063407222e-05,
+      "loss": 0.4487,
+      "step": 8778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3225354838444197,
+      "learning_rate": 1.0344200830118899e-05,
+      "loss": 0.2934,
+      "step": 8779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36739556613949326,
+      "learning_rate": 1.0342341014245918e-05,
+      "loss": 0.2392,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7747310744381001,
+      "learning_rate": 1.0340481186517678e-05,
+      "loss": 0.4761,
+      "step": 8781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27809320794294334,
+      "learning_rate": 1.0338621346998596e-05,
+      "loss": 0.2168,
+      "step": 8782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.399339632887001,
+      "learning_rate": 1.0336761495753067e-05,
+      "loss": 0.2646,
+      "step": 8783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7950306926717371,
+      "learning_rate": 1.0334901632845504e-05,
+      "loss": 0.3684,
+      "step": 8784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.37529210841211885,
+      "learning_rate": 1.0333041758340312e-05,
+      "loss": 0.2767,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2740845981263997,
+      "learning_rate": 1.0331181872301898e-05,
+      "loss": 0.1405,
+      "step": 8786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3420145569049249,
+      "learning_rate": 1.0329321974794671e-05,
+      "loss": 0.2961,
+      "step": 8787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5831138037987301,
+      "learning_rate": 1.0327462065883036e-05,
+      "loss": 0.3274,
+      "step": 8788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2896592046347293,
+      "learning_rate": 1.0325602145631403e-05,
+      "loss": 0.223,
+      "step": 8789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.43954585956748027,
+      "learning_rate": 1.0323742214104185e-05,
+      "loss": 0.3921,
+      "step": 8790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3625988463122244,
+      "learning_rate": 1.0321882271365786e-05,
+      "loss": 0.2806,
+      "step": 8791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3160804484447476,
+      "learning_rate": 1.0320022317480618e-05,
+      "loss": 0.2677,
+      "step": 8792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0366060736983413,
+      "learning_rate": 1.031816235251309e-05,
+      "loss": 0.3159,
+      "step": 8793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5889250828589246,
+      "learning_rate": 1.0316302376527616e-05,
+      "loss": 0.3359,
+      "step": 8794
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.270751161601078,
+      "learning_rate": 1.0314442389588603e-05,
+      "loss": 0.2913,
+      "step": 8795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6323219923443495,
+      "learning_rate": 1.0312582391760462e-05,
+      "loss": 0.3475,
+      "step": 8796
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4125990956262591,
+      "learning_rate": 1.0310722383107608e-05,
+      "loss": 0.2992,
+      "step": 8797
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23697489736616656,
+      "learning_rate": 1.030886236369445e-05,
+      "loss": 0.1779,
+      "step": 8798
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3704019970521971,
+      "learning_rate": 1.0307002333585404e-05,
+      "loss": 0.2598,
+      "step": 8799
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44374597809610156,
+      "learning_rate": 1.0305142292844876e-05,
+      "loss": 0.2887,
+      "step": 8800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4554383476417022,
+      "learning_rate": 1.0303282241537287e-05,
+      "loss": 0.3395,
+      "step": 8801
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3874203617171426,
+      "learning_rate": 1.0301422179727045e-05,
+      "loss": 0.3243,
+      "step": 8802
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3125693395295528,
+      "learning_rate": 1.0299562107478569e-05,
+      "loss": 0.2679,
+      "step": 8803
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25728668270074057,
+      "learning_rate": 1.0297702024856268e-05,
+      "loss": 0.1966,
+      "step": 8804
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0480113904388246,
+      "learning_rate": 1.0295841931924559e-05,
+      "loss": 0.5405,
+      "step": 8805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.663364129540765,
+      "learning_rate": 1.0293981828747857e-05,
+      "loss": 0.2988,
+      "step": 8806
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2966797163973741,
+      "learning_rate": 1.0292121715390576e-05,
+      "loss": 0.2727,
+      "step": 8807
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6970721071112972,
+      "learning_rate": 1.0290261591917132e-05,
+      "loss": 0.5367,
+      "step": 8808
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38124847024879227,
+      "learning_rate": 1.0288401458391943e-05,
+      "loss": 0.1685,
+      "step": 8809
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2504095296687366,
+      "learning_rate": 1.0286541314879424e-05,
+      "loss": 0.2031,
+      "step": 8810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3809688844646517,
+      "learning_rate": 1.028468116144399e-05,
+      "loss": 0.3369,
+      "step": 8811
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4955121367154073,
+      "learning_rate": 1.028282099815006e-05,
+      "loss": 0.2312,
+      "step": 8812
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4093725872251935,
+      "learning_rate": 1.0280960825062054e-05,
+      "loss": 0.3284,
+      "step": 8813
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1636539511340667,
+      "learning_rate": 1.0279100642244382e-05,
+      "loss": 0.8014,
+      "step": 8814
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28726866422333247,
+      "learning_rate": 1.027724044976147e-05,
+      "loss": 0.2295,
+      "step": 8815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24419854470114155,
+      "learning_rate": 1.0275380247677733e-05,
+      "loss": 0.1984,
+      "step": 8816
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9213705414253189,
+      "learning_rate": 1.0273520036057587e-05,
+      "loss": 0.517,
+      "step": 8817
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4880627076951752,
+      "learning_rate": 1.0271659814965457e-05,
+      "loss": 0.3437,
+      "step": 8818
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25429892494500267,
+      "learning_rate": 1.0269799584465758e-05,
+      "loss": 0.2382,
+      "step": 8819
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1443560930208179,
+      "learning_rate": 1.0267939344622912e-05,
+      "loss": 0.7465,
+      "step": 8820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47916608663820603,
+      "learning_rate": 1.0266079095501338e-05,
+      "loss": 0.3078,
+      "step": 8821
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22200518903738178,
+      "learning_rate": 1.0264218837165459e-05,
+      "loss": 0.1626,
+      "step": 8822
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38383767303997884,
+      "learning_rate": 1.0262358569679686e-05,
+      "loss": 0.3217,
+      "step": 8823
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6558207841443644,
+      "learning_rate": 1.0260498293108452e-05,
+      "loss": 0.4036,
+      "step": 8824
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3591505119739801,
+      "learning_rate": 1.025863800751617e-05,
+      "loss": 0.2283,
+      "step": 8825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36957746518241613,
+      "learning_rate": 1.025677771296727e-05,
+      "loss": 0.3354,
+      "step": 8826
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4188426062019395,
+      "learning_rate": 1.0254917409526163e-05,
+      "loss": 0.7654,
+      "step": 8827
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30075617474235133,
+      "learning_rate": 1.0253057097257281e-05,
+      "loss": 0.2479,
+      "step": 8828
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3521081179987658,
+      "learning_rate": 1.025119677622504e-05,
+      "loss": 0.2155,
+      "step": 8829
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4290538039923925,
+      "learning_rate": 1.0249336446493869e-05,
+      "loss": 0.3082,
+      "step": 8830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28813142501039557,
+      "learning_rate": 1.0247476108128183e-05,
+      "loss": 0.2763,
+      "step": 8831
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7845090507131877,
+      "learning_rate": 1.0245615761192414e-05,
+      "loss": 0.4415,
+      "step": 8832
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8072316896704939,
+      "learning_rate": 1.024375540575098e-05,
+      "loss": 0.4035,
+      "step": 8833
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2726303021214843,
+      "learning_rate": 1.0241895041868306e-05,
+      "loss": 0.2369,
+      "step": 8834
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36492778874634096,
+      "learning_rate": 1.024003466960882e-05,
+      "loss": 0.2715,
+      "step": 8835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.360313855746173,
+      "learning_rate": 1.0238174289036942e-05,
+      "loss": 0.2589,
+      "step": 8836
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36305815876103154,
+      "learning_rate": 1.0236313900217099e-05,
+      "loss": 0.2774,
+      "step": 8837
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35402837649284924,
+      "learning_rate": 1.0234453503213715e-05,
+      "loss": 0.286,
+      "step": 8838
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6168437114744936,
+      "learning_rate": 1.0232593098091215e-05,
+      "loss": 0.3838,
+      "step": 8839
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4155028386232177,
+      "learning_rate": 1.0230732684914029e-05,
+      "loss": 0.3072,
+      "step": 8840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5305176733751163,
+      "learning_rate": 1.022887226374658e-05,
+      "loss": 0.4033,
+      "step": 8841
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20583299188599177,
+      "learning_rate": 1.022701183465329e-05,
+      "loss": 0.1745,
+      "step": 8842
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3890161245544743,
+      "learning_rate": 1.0225151397698597e-05,
+      "loss": 0.3028,
+      "step": 8843
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9808698140562502,
+      "learning_rate": 1.0223290952946914e-05,
+      "loss": 0.6994,
+      "step": 8844
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6121490823974144,
+      "learning_rate": 1.0221430500462677e-05,
+      "loss": 0.3077,
+      "step": 8845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3066633720162846,
+      "learning_rate": 1.0219570040310312e-05,
+      "loss": 0.2655,
+      "step": 8846
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34538918095313825,
+      "learning_rate": 1.0217709572554247e-05,
+      "loss": 0.3273,
+      "step": 8847
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1859650911764527,
+      "learning_rate": 1.0215849097258905e-05,
+      "loss": 0.0884,
+      "step": 8848
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3703599398367777,
+      "learning_rate": 1.0213988614488721e-05,
+      "loss": 0.289,
+      "step": 8849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4604316756843436,
+      "learning_rate": 1.0212128124308121e-05,
+      "loss": 0.3816,
+      "step": 8850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38250000158150516,
+      "learning_rate": 1.0210267626781532e-05,
+      "loss": 0.2585,
+      "step": 8851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.350693722826386,
+      "learning_rate": 1.0208407121973383e-05,
+      "loss": 0.3005,
+      "step": 8852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7957509715262783,
+      "learning_rate": 1.0206546609948107e-05,
+      "loss": 0.4827,
+      "step": 8853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.235842723340732,
+      "learning_rate": 1.020468609077013e-05,
+      "loss": 0.216,
+      "step": 8854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30607668526376075,
+      "learning_rate": 1.0202825564503885e-05,
+      "loss": 0.2021,
+      "step": 8855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1314724300561396,
+      "learning_rate": 1.0200965031213795e-05,
+      "loss": 0.7834,
+      "step": 8856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7721586850426669,
+      "learning_rate": 1.0199104490964296e-05,
+      "loss": 0.4185,
+      "step": 8857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3153541338301299,
+      "learning_rate": 1.0197243943819816e-05,
+      "loss": 0.2,
+      "step": 8858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3830402764051056,
+      "learning_rate": 1.0195383389844789e-05,
+      "loss": 0.3396,
+      "step": 8859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2716412665480148,
+      "learning_rate": 1.0193522829103643e-05,
+      "loss": 0.183,
+      "step": 8860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3956289122519081,
+      "learning_rate": 1.0191662261660809e-05,
+      "loss": 0.2267,
+      "step": 8861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3719950768589626,
+      "learning_rate": 1.018980168758072e-05,
+      "loss": 0.3198,
+      "step": 8862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.073061393618599,
+      "learning_rate": 1.0187941106927803e-05,
+      "loss": 0.4103,
+      "step": 8863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3268498389256042,
+      "learning_rate": 1.0186080519766499e-05,
+      "loss": 0.2171,
+      "step": 8864
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.932570603873562,
+      "learning_rate": 1.0184219926161229e-05,
+      "loss": 0.5862,
+      "step": 8865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2876069287491476,
+      "learning_rate": 1.0182359326176437e-05,
+      "loss": 0.2418,
+      "step": 8866
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2760938844417518,
+      "learning_rate": 1.0180498719876546e-05,
+      "loss": 0.2243,
+      "step": 8867
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9738896002035565,
+      "learning_rate": 1.0178638107325993e-05,
+      "loss": 0.5697,
+      "step": 8868
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9192592820798557,
+      "learning_rate": 1.0176777488589206e-05,
+      "loss": 0.4044,
+      "step": 8869
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2846380947105442,
+      "learning_rate": 1.0174916863730628e-05,
+      "loss": 0.244,
+      "step": 8870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4158621955491559,
+      "learning_rate": 1.0173056232814684e-05,
+      "loss": 0.2808,
+      "step": 8871
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4635268387198839,
+      "learning_rate": 1.0171195595905811e-05,
+      "loss": 0.2996,
+      "step": 8872
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.33203805455122837,
+      "learning_rate": 1.0169334953068442e-05,
+      "loss": 0.2639,
+      "step": 8873
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3697048151949442,
+      "learning_rate": 1.0167474304367011e-05,
+      "loss": 0.2824,
+      "step": 8874
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4844233725817165,
+      "learning_rate": 1.0165613649865951e-05,
+      "loss": 0.3137,
+      "step": 8875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3984242851912228,
+      "learning_rate": 1.0163752989629698e-05,
+      "loss": 0.2471,
+      "step": 8876
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47180498037132856,
+      "learning_rate": 1.0161892323722684e-05,
+      "loss": 0.274,
+      "step": 8877
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3669597020882024,
+      "learning_rate": 1.0160031652209348e-05,
+      "loss": 0.2947,
+      "step": 8878
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.386735586286777,
+      "learning_rate": 1.0158170975154121e-05,
+      "loss": 0.2804,
+      "step": 8879
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.472778223664258,
+      "learning_rate": 1.015631029262144e-05,
+      "loss": 0.3634,
+      "step": 8880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31498560183851326,
+      "learning_rate": 1.0154449604675745e-05,
+      "loss": 0.2056,
+      "step": 8881
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31106123240309985,
+      "learning_rate": 1.015258891138146e-05,
+      "loss": 0.2597,
+      "step": 8882
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32163140709549864,
+      "learning_rate": 1.0150728212803034e-05,
+      "loss": 0.248,
+      "step": 8883
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0163468760891674,
+      "learning_rate": 1.0148867509004892e-05,
+      "loss": 0.3417,
+      "step": 8884
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3535521516167228,
+      "learning_rate": 1.0147006800051475e-05,
+      "loss": 0.273,
+      "step": 8885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32974085632081673,
+      "learning_rate": 1.0145146086007219e-05,
+      "loss": 0.3063,
+      "step": 8886
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37848673394257054,
+      "learning_rate": 1.0143285366936562e-05,
+      "loss": 0.2336,
+      "step": 8887
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27078939239282207,
+      "learning_rate": 1.0141424642903936e-05,
+      "loss": 0.208,
+      "step": 8888
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0221622920030797,
+      "learning_rate": 1.0139563913973787e-05,
+      "loss": 0.5142,
+      "step": 8889
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34970550608599016,
+      "learning_rate": 1.0137703180210538e-05,
+      "loss": 0.2834,
+      "step": 8890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35189717797611997,
+      "learning_rate": 1.0135842441678639e-05,
+      "loss": 0.2754,
+      "step": 8891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6613352038736564,
+      "learning_rate": 1.0133981698442519e-05,
+      "loss": 0.411,
+      "step": 8892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49333120019293597,
+      "learning_rate": 1.013212095056662e-05,
+      "loss": 0.4239,
+      "step": 8893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2691281560347184,
+      "learning_rate": 1.0130260198115376e-05,
+      "loss": 0.2087,
+      "step": 8894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25133156816965974,
+      "learning_rate": 1.012839944115323e-05,
+      "loss": 0.1896,
+      "step": 8895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.644446483143422,
+      "learning_rate": 1.0126538679744615e-05,
+      "loss": 0.4346,
+      "step": 8896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31080535051286085,
+      "learning_rate": 1.0124677913953971e-05,
+      "loss": 0.2094,
+      "step": 8897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2925056576358622,
+      "learning_rate": 1.0122817143845736e-05,
+      "loss": 0.29,
+      "step": 8898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1700861136940273,
+      "learning_rate": 1.0120956369484352e-05,
+      "loss": 0.8421,
+      "step": 8899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18646929013358124,
+      "learning_rate": 1.011909559093425e-05,
+      "loss": 0.1259,
+      "step": 8900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5031315658177309,
+      "learning_rate": 1.0117234808259875e-05,
+      "loss": 0.3329,
+      "step": 8901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3837696227726515,
+      "learning_rate": 1.0115374021525664e-05,
+      "loss": 0.2908,
+      "step": 8902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3224777264779916,
+      "learning_rate": 1.0113513230796052e-05,
+      "loss": 0.2378,
+      "step": 8903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7486941413631881,
+      "learning_rate": 1.0111652436135486e-05,
+      "loss": 0.4879,
+      "step": 8904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4577781436154072,
+      "learning_rate": 1.01097916376084e-05,
+      "loss": 0.4012,
+      "step": 8905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3227090641491666,
+      "learning_rate": 1.0107930835279234e-05,
+      "loss": 0.2838,
+      "step": 8906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25374936217959587,
+      "learning_rate": 1.0106070029212424e-05,
+      "loss": 0.1262,
+      "step": 8907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6524870029201337,
+      "learning_rate": 1.0104209219472418e-05,
+      "loss": 0.3958,
+      "step": 8908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37586082968520385,
+      "learning_rate": 1.010234840612365e-05,
+      "loss": 0.2897,
+      "step": 8909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3307571873747576,
+      "learning_rate": 1.010048758923056e-05,
+      "loss": 0.2648,
+      "step": 8910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8419980738165997,
+      "learning_rate": 1.0098626768857591e-05,
+      "loss": 0.6101,
+      "step": 8911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6204223015862753,
+      "learning_rate": 1.009676594506918e-05,
+      "loss": 0.3407,
+      "step": 8912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28043986557667766,
+      "learning_rate": 1.0094905117929767e-05,
+      "loss": 0.1755,
+      "step": 8913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35706107090418515,
+      "learning_rate": 1.0093044287503797e-05,
+      "loss": 0.3173,
+      "step": 8914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.564795786408195,
+      "learning_rate": 1.0091183453855706e-05,
+      "loss": 0.2933,
+      "step": 8915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37689587316033274,
+      "learning_rate": 1.0089322617049936e-05,
+      "loss": 0.3088,
+      "step": 8916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4513654962243863,
+      "learning_rate": 1.0087461777150926e-05,
+      "loss": 0.2704,
+      "step": 8917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3927024710302205,
+      "learning_rate": 1.0085600934223121e-05,
+      "loss": 0.2996,
+      "step": 8918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26729943984490195,
+      "learning_rate": 1.008374008833096e-05,
+      "loss": 0.2113,
+      "step": 8919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6478463082866031,
+      "learning_rate": 1.0081879239538881e-05,
+      "loss": 0.332,
+      "step": 8920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3996713800799195,
+      "learning_rate": 1.0080018387911328e-05,
+      "loss": 0.2707,
+      "step": 8921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3221994879121444,
+      "learning_rate": 1.0078157533512742e-05,
+      "loss": 0.3001,
+      "step": 8922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8538514540552568,
+      "learning_rate": 1.0076296676407565e-05,
+      "loss": 0.404,
+      "step": 8923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41161279723696287,
+      "learning_rate": 1.0074435816660235e-05,
+      "loss": 0.2951,
+      "step": 8924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35715732608291045,
+      "learning_rate": 1.00725749543352e-05,
+      "loss": 0.2687,
+      "step": 8925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3368658460212253,
+      "learning_rate": 1.0070714089496891e-05,
+      "loss": 0.2479,
+      "step": 8926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44449844022617024,
+      "learning_rate": 1.006885322220976e-05,
+      "loss": 0.3089,
+      "step": 8927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5794718664509232,
+      "learning_rate": 1.0066992352538245e-05,
+      "loss": 0.3989,
+      "step": 8928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40439067942242224,
+      "learning_rate": 1.0065131480546788e-05,
+      "loss": 0.327,
+      "step": 8929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7094782461959981,
+      "learning_rate": 1.006327060629983e-05,
+      "loss": 0.2071,
+      "step": 8930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32681962706815443,
+      "learning_rate": 1.0061409729861814e-05,
+      "loss": 0.3041,
+      "step": 8931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29148296068996754,
+      "learning_rate": 1.0059548851297178e-05,
+      "loss": 0.2023,
+      "step": 8932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3429142652730741,
+      "learning_rate": 1.0057687970670372e-05,
+      "loss": 0.2153,
+      "step": 8933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34103484654619365,
+      "learning_rate": 1.005582708804583e-05,
+      "loss": 0.2975,
+      "step": 8934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0197996496950272,
+      "learning_rate": 1.0053966203488003e-05,
+      "loss": 0.677,
+      "step": 8935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36527832406679867,
+      "learning_rate": 1.0052105317061327e-05,
+      "loss": 0.182,
+      "step": 8936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29033341390121514,
+      "learning_rate": 1.0050244428830246e-05,
+      "loss": 0.2933,
+      "step": 8937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42647799922546814,
+      "learning_rate": 1.0048383538859202e-05,
+      "loss": 0.354,
+      "step": 8938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1574000968867463,
+      "learning_rate": 1.0046522647212642e-05,
+      "loss": 0.087,
+      "step": 8939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41188216839380715,
+      "learning_rate": 1.0044661753955001e-05,
+      "loss": 0.3561,
+      "step": 8940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4381041332600563,
+      "learning_rate": 1.0042800859150726e-05,
+      "loss": 0.3384,
+      "step": 8941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3689628610234497,
+      "learning_rate": 1.0040939962864258e-05,
+      "loss": 0.311,
+      "step": 8942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3731679048957679,
+      "learning_rate": 1.0039079065160042e-05,
+      "loss": 0.2616,
+      "step": 8943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41815692396401,
+      "learning_rate": 1.0037218166102518e-05,
+      "loss": 0.2991,
+      "step": 8944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2680388741950821,
+      "learning_rate": 1.0035357265756134e-05,
+      "loss": 0.2294,
+      "step": 8945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3967633759634121,
+      "learning_rate": 1.003349636418533e-05,
+      "loss": 0.2559,
+      "step": 8946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7089779370160822,
+      "learning_rate": 1.0031635461454544e-05,
+      "loss": 0.5292,
+      "step": 8947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8142086961246247,
+      "learning_rate": 1.0029774557628224e-05,
+      "loss": 0.3636,
+      "step": 8948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3104295700200426,
+      "learning_rate": 1.0027913652770813e-05,
+      "loss": 0.2351,
+      "step": 8949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3667967386304453,
+      "learning_rate": 1.0026052746946756e-05,
+      "loss": 0.2733,
+      "step": 8950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.45760292541987996,
+      "learning_rate": 1.002419184022049e-05,
+      "loss": 0.282,
+      "step": 8951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2975732477040278,
+      "learning_rate": 1.0022330932656463e-05,
+      "loss": 0.2097,
+      "step": 8952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36123061495278136,
+      "learning_rate": 1.0020470024319115e-05,
+      "loss": 0.2913,
+      "step": 8953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6903017856807934,
+      "learning_rate": 1.0018609115272896e-05,
+      "loss": 0.3646,
+      "step": 8954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34900373266645707,
+      "learning_rate": 1.0016748205582238e-05,
+      "loss": 0.2853,
+      "step": 8955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6639895995416654,
+      "learning_rate": 1.0014887295311595e-05,
+      "loss": 0.3339,
+      "step": 8956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21779948373374317,
+      "learning_rate": 1.0013026384525404e-05,
+      "loss": 0.2023,
+      "step": 8957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3989690662409945,
+      "learning_rate": 1.0011165473288108e-05,
+      "loss": 0.3183,
+      "step": 8958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9027589397240701,
+      "learning_rate": 1.0009304561664154e-05,
+      "loss": 0.359,
+      "step": 8959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6025413935849674,
+      "learning_rate": 1.0007443649717985e-05,
+      "loss": 0.3867,
+      "step": 8960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3535480063313833,
+      "learning_rate": 1.0005582737514039e-05,
+      "loss": 0.2922,
+      "step": 8961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34940809513436644,
+      "learning_rate": 1.0003721825116766e-05,
+      "loss": 0.2565,
+      "step": 8962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23572550353217794,
+      "learning_rate": 1.0001860912590604e-05,
+      "loss": 0.1681,
+      "step": 8963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4371986239915237,
+      "learning_rate": 1e-05,
+      "loss": 0.3286,
+      "step": 8964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5882681268821435,
+      "learning_rate": 9.998139087409399e-06,
+      "loss": 0.3003,
+      "step": 8965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.998562898170738,
+      "learning_rate": 9.996278174883236e-06,
+      "loss": 0.4324,
+      "step": 8966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32535718394732693,
+      "learning_rate": 9.994417262485963e-06,
+      "loss": 0.2808,
+      "step": 8967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8845442812310932,
+      "learning_rate": 9.992556350282018e-06,
+      "loss": 0.5168,
+      "step": 8968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2605945618913011,
+      "learning_rate": 9.990695438335847e-06,
+      "loss": 0.2,
+      "step": 8969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3912375752348602,
+      "learning_rate": 9.988834526711893e-06,
+      "loss": 0.267,
+      "step": 8970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8202549742402676,
+      "learning_rate": 9.9869736154746e-06,
+      "loss": 0.4242,
+      "step": 8971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2838554077787522,
+      "learning_rate": 9.985112704688406e-06,
+      "loss": 0.1891,
+      "step": 8972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2958047596139349,
+      "learning_rate": 9.983251794417763e-06,
+      "loss": 0.2796,
+      "step": 8973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.146356366991916,
+      "learning_rate": 9.981390884727106e-06,
+      "loss": 0.6581,
+      "step": 8974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4452802708748933,
+      "learning_rate": 9.979529975680885e-06,
+      "loss": 0.2432,
+      "step": 8975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38232368072779344,
+      "learning_rate": 9.977669067343537e-06,
+      "loss": 0.3037,
+      "step": 8976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3709480557645404,
+      "learning_rate": 9.975808159779512e-06,
+      "loss": 0.2857,
+      "step": 8977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2848080628979088,
+      "learning_rate": 9.973947253053248e-06,
+      "loss": 0.1488,
+      "step": 8978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36194839241988114,
+      "learning_rate": 9.972086347229187e-06,
+      "loss": 0.3028,
+      "step": 8979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9172902205994092,
+      "learning_rate": 9.970225442371778e-06,
+      "loss": 0.5909,
+      "step": 8980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3406603301566798,
+      "learning_rate": 9.968364538545461e-06,
+      "loss": 0.3361,
+      "step": 8981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3044827668809311,
+      "learning_rate": 9.966503635814677e-06,
+      "loss": 0.2007,
+      "step": 8982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5498909403562776,
+      "learning_rate": 9.96464273424387e-06,
+      "loss": 0.4395,
+      "step": 8983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2707344192696009,
+      "learning_rate": 9.962781833897484e-06,
+      "loss": 0.2111,
+      "step": 8984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27766021070239455,
+      "learning_rate": 9.960920934839963e-06,
+      "loss": 0.2124,
+      "step": 8985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0646904612166812,
+      "learning_rate": 9.959060037135745e-06,
+      "loss": 0.5268,
+      "step": 8986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7117116899008378,
+      "learning_rate": 9.95719914084928e-06,
+      "loss": 0.4817,
+      "step": 8987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2762098719031825,
+      "learning_rate": 9.955338246045004e-06,
+      "loss": 0.2282,
+      "step": 8988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3723392795742885,
+      "learning_rate": 9.953477352787363e-06,
+      "loss": 0.3297,
+      "step": 8989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4044795205958676,
+      "learning_rate": 9.9516164611408e-06,
+      "loss": 0.2681,
+      "step": 8990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2391132673895818,
+      "learning_rate": 9.949755571169757e-06,
+      "loss": 0.2128,
+      "step": 8991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9771484815283424,
+      "learning_rate": 9.947894682938676e-06,
+      "loss": 0.2187,
+      "step": 8992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3742803679785533,
+      "learning_rate": 9.946033796511999e-06,
+      "loss": 0.3154,
+      "step": 8993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36770528833525723,
+      "learning_rate": 9.944172911954173e-06,
+      "loss": 0.2913,
+      "step": 8994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7614209857492386,
+      "learning_rate": 9.942312029329631e-06,
+      "loss": 0.4018,
+      "step": 8995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27484815842745103,
+      "learning_rate": 9.940451148702826e-06,
+      "loss": 0.2612,
+      "step": 8996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31502147368921235,
+      "learning_rate": 9.938590270138191e-06,
+      "loss": 0.2551,
+      "step": 8997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43320342033214754,
+      "learning_rate": 9.936729393700176e-06,
+      "loss": 0.1657,
+      "step": 8998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7310020498893697,
+      "learning_rate": 9.934868519453215e-06,
+      "loss": 0.4141,
+      "step": 8999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3868343055255914,
+      "learning_rate": 9.933007647461758e-06,
+      "loss": 0.2496,
+      "step": 9000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.304909472398119,
+      "learning_rate": 9.931146777790241e-06,
+      "loss": 0.2584,
+      "step": 9001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43745718846557247,
+      "learning_rate": 9.929285910503112e-06,
+      "loss": 0.2569,
+      "step": 9002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2749356993094986,
+      "learning_rate": 9.927425045664804e-06,
+      "loss": 0.202,
+      "step": 9003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44467210424965653,
+      "learning_rate": 9.925564183339768e-06,
+      "loss": 0.3516,
+      "step": 9004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3609803021729023,
+      "learning_rate": 9.92370332359244e-06,
+      "loss": 0.2799,
+      "step": 9005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32493117789295467,
+      "learning_rate": 9.92184246648726e-06,
+      "loss": 0.2776,
+      "step": 9006
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0347469550216204,
+      "learning_rate": 9.919981612088676e-06,
+      "loss": 0.7681,
+      "step": 9007
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33748772354766315,
+      "learning_rate": 9.91812076046112e-06,
+      "loss": 0.2435,
+      "step": 9008
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26513883317931297,
+      "learning_rate": 9.916259911669044e-06,
+      "loss": 0.2091,
+      "step": 9009
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.46014096183328146,
+      "learning_rate": 9.914399065776879e-06,
+      "loss": 0.2877,
+      "step": 9010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6082292573499609,
+      "learning_rate": 9.912538222849074e-06,
+      "loss": 0.3486,
+      "step": 9011
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34178859785901966,
+      "learning_rate": 9.910677382950064e-06,
+      "loss": 0.2791,
+      "step": 9012
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3524107406208595,
+      "learning_rate": 9.908816546144296e-06,
+      "loss": 0.3128,
+      "step": 9013
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2773592503100069,
+      "learning_rate": 9.906955712496203e-06,
+      "loss": 0.1512,
+      "step": 9014
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3265634728938896,
+      "learning_rate": 9.905094882070234e-06,
+      "loss": 0.2592,
+      "step": 9015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40937969659931533,
+      "learning_rate": 9.903234054930824e-06,
+      "loss": 0.3167,
+      "step": 9016
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33132514739426283,
+      "learning_rate": 9.901373231142416e-06,
+      "loss": 0.3214,
+      "step": 9017
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35046526567785413,
+      "learning_rate": 9.899512410769443e-06,
+      "loss": 0.1871,
+      "step": 9018
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4758272095091834,
+      "learning_rate": 9.897651593876356e-06,
+      "loss": 0.4102,
+      "step": 9019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32645946398502895,
+      "learning_rate": 9.895790780527585e-06,
+      "loss": 0.2765,
+      "step": 9020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33879706747924987,
+      "learning_rate": 9.89392997078758e-06,
+      "loss": 0.1954,
+      "step": 9021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.420195059981251,
+      "learning_rate": 9.892069164720771e-06,
+      "loss": 0.3489,
+      "step": 9022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27953102565373356,
+      "learning_rate": 9.890208362391606e-06,
+      "loss": 0.1685,
+      "step": 9023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33324986313949784,
+      "learning_rate": 9.888347563864517e-06,
+      "loss": 0.2261,
+      "step": 9024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4158775305414562,
+      "learning_rate": 9.886486769203951e-06,
+      "loss": 0.3704,
+      "step": 9025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1118021729267935,
+      "learning_rate": 9.884625978474341e-06,
+      "loss": 0.6197,
+      "step": 9026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2940205641682392,
+      "learning_rate": 9.88276519174013e-06,
+      "loss": 0.2177,
+      "step": 9027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.632974645298959,
+      "learning_rate": 9.880904409065753e-06,
+      "loss": 0.4462,
+      "step": 9028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21338348316802783,
+      "learning_rate": 9.879043630515651e-06,
+      "loss": 0.2196,
+      "step": 9029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4146681584696092,
+      "learning_rate": 9.877182856154267e-06,
+      "loss": 0.3047,
+      "step": 9030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4489759474178878,
+      "learning_rate": 9.87532208604603e-06,
+      "loss": 0.2851,
+      "step": 9031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36796592998808814,
+      "learning_rate": 9.873461320255388e-06,
+      "loss": 0.3119,
+      "step": 9032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3727674532757231,
+      "learning_rate": 9.871600558846772e-06,
+      "loss": 0.2888,
+      "step": 9033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4707617476263641,
+      "learning_rate": 9.869739801884627e-06,
+      "loss": 0.2876,
+      "step": 9034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2443750988765055,
+      "learning_rate": 9.867879049433383e-06,
+      "loss": 0.1874,
+      "step": 9035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39422198057692154,
+      "learning_rate": 9.866018301557484e-06,
+      "loss": 0.2668,
+      "step": 9036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29803699521115307,
+      "learning_rate": 9.864157558321364e-06,
+      "loss": 0.2592,
+      "step": 9037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4831922205500076,
+      "learning_rate": 9.862296819789464e-06,
+      "loss": 0.3992,
+      "step": 9038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5305541289799832,
+      "learning_rate": 9.860436086026218e-06,
+      "loss": 0.3633,
+      "step": 9039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2684294559842977,
+      "learning_rate": 9.858575357096064e-06,
+      "loss": 0.2333,
+      "step": 9040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25245525380867007,
+      "learning_rate": 9.85671463306344e-06,
+      "loss": 0.1646,
+      "step": 9041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5212038708909201,
+      "learning_rate": 9.854853913992783e-06,
+      "loss": 0.3353,
+      "step": 9042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3956254025587982,
+      "learning_rate": 9.852993199948527e-06,
+      "loss": 0.3331,
+      "step": 9043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3427049716407958,
+      "learning_rate": 9.85113249099511e-06,
+      "loss": 0.276,
+      "step": 9044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.46855115446728846,
+      "learning_rate": 9.849271787196971e-06,
+      "loss": 0.3368,
+      "step": 9045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3783074851487932,
+      "learning_rate": 9.847411088618539e-06,
+      "loss": 0.3037,
+      "step": 9046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18974821301931535,
+      "learning_rate": 9.845550395324259e-06,
+      "loss": 0.0865,
+      "step": 9047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2976246395263267,
+      "learning_rate": 9.843689707378558e-06,
+      "loss": 0.2641,
+      "step": 9048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3856321462242667,
+      "learning_rate": 9.841829024845882e-06,
+      "loss": 0.3264,
+      "step": 9049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8229565405098527,
+      "learning_rate": 9.839968347790657e-06,
+      "loss": 0.3381,
+      "step": 9050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5464022009127502,
+      "learning_rate": 9.83810767627732e-06,
+      "loss": 0.3216,
+      "step": 9051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3146404356039652,
+      "learning_rate": 9.836247010370308e-06,
+      "loss": 0.2941,
+      "step": 9052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29591631958317766,
+      "learning_rate": 9.834386350134052e-06,
+      "loss": 0.1869,
+      "step": 9053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7751683065263755,
+      "learning_rate": 9.832525695632994e-06,
+      "loss": 0.4217,
+      "step": 9054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4249413501512791,
+      "learning_rate": 9.830665046931563e-06,
+      "loss": 0.2905,
+      "step": 9055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5985386409545421,
+      "learning_rate": 9.828804404094192e-06,
+      "loss": 0.3324,
+      "step": 9056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20868793648929493,
+      "learning_rate": 9.82694376718532e-06,
+      "loss": 0.0741,
+      "step": 9057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3726465468468453,
+      "learning_rate": 9.825083136269375e-06,
+      "loss": 0.3055,
+      "step": 9058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4228458238529731,
+      "learning_rate": 9.823222511410795e-06,
+      "loss": 0.2778,
+      "step": 9059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29394146644376107,
+      "learning_rate": 9.82136189267401e-06,
+      "loss": 0.2364,
+      "step": 9060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43066936972759234,
+      "learning_rate": 9.819501280123458e-06,
+      "loss": 0.3554,
+      "step": 9061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6720219973395819,
+      "learning_rate": 9.817640673823566e-06,
+      "loss": 0.4414,
+      "step": 9062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22057283222630394,
+      "learning_rate": 9.815780073838773e-06,
+      "loss": 0.0968,
+      "step": 9063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27013861542171497,
+      "learning_rate": 9.813919480233503e-06,
+      "loss": 0.283,
+      "step": 9064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7716994426637789,
+      "learning_rate": 9.812058893072199e-06,
+      "loss": 0.5575,
+      "step": 9065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34090727687569566,
+      "learning_rate": 9.810198312419284e-06,
+      "loss": 0.2088,
+      "step": 9066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4536499925740262,
+      "learning_rate": 9.808337738339194e-06,
+      "loss": 0.355,
+      "step": 9067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3564000526722853,
+      "learning_rate": 9.80647717089636e-06,
+      "loss": 0.3234,
+      "step": 9068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23213456762043275,
+      "learning_rate": 9.804616610155215e-06,
+      "loss": 0.1778,
+      "step": 9069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3640646952779975,
+      "learning_rate": 9.802756056180187e-06,
+      "loss": 0.2301,
+      "step": 9070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2155181820109497,
+      "learning_rate": 9.800895509035708e-06,
+      "loss": 0.8132,
+      "step": 9071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3652622733588144,
+      "learning_rate": 9.799034968786209e-06,
+      "loss": 0.2981,
+      "step": 9072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3160744756567268,
+      "learning_rate": 9.797174435496119e-06,
+      "loss": 0.2346,
+      "step": 9073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3198722182667888,
+      "learning_rate": 9.795313909229872e-06,
+      "loss": 0.2408,
+      "step": 9074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43873354265547654,
+      "learning_rate": 9.793453390051894e-06,
+      "loss": 0.3458,
+      "step": 9075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27745929926602364,
+      "learning_rate": 9.791592878026617e-06,
+      "loss": 0.2203,
+      "step": 9076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.962358758398474,
+      "learning_rate": 9.789732373218468e-06,
+      "loss": 0.5281,
+      "step": 9077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5943049810697584,
+      "learning_rate": 9.78787187569188e-06,
+      "loss": 0.3826,
+      "step": 9078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3278974230441489,
+      "learning_rate": 9.786011385511279e-06,
+      "loss": 0.2963,
+      "step": 9079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3273067224989936,
+      "learning_rate": 9.784150902741095e-06,
+      "loss": 0.2385,
+      "step": 9080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27902999532971506,
+      "learning_rate": 9.782290427445755e-06,
+      "loss": 0.1677,
+      "step": 9081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3192097580526225,
+      "learning_rate": 9.78042995968969e-06,
+      "loss": 0.2732,
+      "step": 9082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6567345097224239,
+      "learning_rate": 9.778569499537327e-06,
+      "loss": 0.3247,
+      "step": 9083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30540297396342236,
+      "learning_rate": 9.77670904705309e-06,
+      "loss": 0.2962,
+      "step": 9084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.368732504569253,
+      "learning_rate": 9.77484860230141e-06,
+      "loss": 0.292,
+      "step": 9085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8805145259459297,
+      "learning_rate": 9.772988165346715e-06,
+      "loss": 0.4137,
+      "step": 9086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22124704588757194,
+      "learning_rate": 9.771127736253426e-06,
+      "loss": 0.1537,
+      "step": 9087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3510303095693387,
+      "learning_rate": 9.769267315085976e-06,
+      "loss": 0.2932,
+      "step": 9088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4614132599295984,
+      "learning_rate": 9.767406901908787e-06,
+      "loss": 0.2882,
+      "step": 9089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8664585977799852,
+      "learning_rate": 9.76554649678629e-06,
+      "loss": 0.4554,
+      "step": 9090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3287018378571944,
+      "learning_rate": 9.763686099782905e-06,
+      "loss": 0.2582,
+      "step": 9091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31568488663477967,
+      "learning_rate": 9.761825710963063e-06,
+      "loss": 0.2942,
+      "step": 9092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20206609002791392,
+      "learning_rate": 9.759965330391182e-06,
+      "loss": 0.0898,
+      "step": 9093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3151578729799421,
+      "learning_rate": 9.758104958131696e-06,
+      "loss": 0.2367,
+      "step": 9094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.741103555846677,
+      "learning_rate": 9.756244594249024e-06,
+      "loss": 0.4884,
+      "step": 9095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.368566194853679,
+      "learning_rate": 9.754384238807589e-06,
+      "loss": 0.2917,
+      "step": 9096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32908075448316854,
+      "learning_rate": 9.752523891871819e-06,
+      "loss": 0.2796,
+      "step": 9097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9739561754616147,
+      "learning_rate": 9.750663553506134e-06,
+      "loss": 0.6682,
+      "step": 9098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19140722525879253,
+      "learning_rate": 9.748803223774962e-06,
+      "loss": 0.1553,
+      "step": 9099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29655050143646866,
+      "learning_rate": 9.746942902742722e-06,
+      "loss": 0.2873,
+      "step": 9100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8291151161681248,
+      "learning_rate": 9.745082590473839e-06,
+      "loss": 0.4544,
+      "step": 9101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6396141518635055,
+      "learning_rate": 9.743222287032734e-06,
+      "loss": 0.3046,
+      "step": 9102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3562190167610645,
+      "learning_rate": 9.741361992483832e-06,
+      "loss": 0.2791,
+      "step": 9103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3584912974204283,
+      "learning_rate": 9.739501706891551e-06,
+      "loss": 0.3235,
+      "step": 9104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28189812469588293,
+      "learning_rate": 9.737641430320315e-06,
+      "loss": 0.1743,
+      "step": 9105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30844703586005295,
+      "learning_rate": 9.735781162834546e-06,
+      "loss": 0.1798,
+      "step": 9106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8857786094208216,
+      "learning_rate": 9.733920904498664e-06,
+      "loss": 0.3613,
+      "step": 9107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4731519913769104,
+      "learning_rate": 9.73206065537709e-06,
+      "loss": 0.3541,
+      "step": 9108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.291075703344724,
+      "learning_rate": 9.730200415534242e-06,
+      "loss": 0.1994,
+      "step": 9109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2261999869951392,
+      "learning_rate": 9.728340185034545e-06,
+      "loss": 0.7143,
+      "step": 9110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31050079628050553,
+      "learning_rate": 9.726479963942412e-06,
+      "loss": 0.2554,
+      "step": 9111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26466322118910507,
+      "learning_rate": 9.72461975232227e-06,
+      "loss": 0.1923,
+      "step": 9112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6613580619617312,
+      "learning_rate": 9.72275955023853e-06,
+      "loss": 0.4344,
+      "step": 9113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3485587659150748,
+      "learning_rate": 9.720899357755618e-06,
+      "loss": 0.8047,
+      "step": 9114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29397938026410353,
+      "learning_rate": 9.719039174937948e-06,
+      "loss": 0.2338,
+      "step": 9115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4540959149910563,
+      "learning_rate": 9.717179001849942e-06,
+      "loss": 0.3446,
+      "step": 9116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7506127708683249,
+      "learning_rate": 9.715318838556014e-06,
+      "loss": 0.4407,
+      "step": 9117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3762547929412095,
+      "learning_rate": 9.71345868512058e-06,
+      "loss": 0.2911,
+      "step": 9118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32214352086262227,
+      "learning_rate": 9.711598541608062e-06,
+      "loss": 0.2218,
+      "step": 9119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3327082599332806,
+      "learning_rate": 9.709738408082873e-06,
+      "loss": 0.2906,
+      "step": 9120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3958060769664337,
+      "learning_rate": 9.707878284609429e-06,
+      "loss": 0.2693,
+      "step": 9121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4592079526507075,
+      "learning_rate": 9.706018171252148e-06,
+      "loss": 0.2672,
+      "step": 9122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34486135538857204,
+      "learning_rate": 9.704158068075445e-06,
+      "loss": 0.314,
+      "step": 9123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3742732158714571,
+      "learning_rate": 9.702297975143737e-06,
+      "loss": 0.2434,
+      "step": 9124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2477396220146489,
+      "learning_rate": 9.700437892521434e-06,
+      "loss": 0.1747,
+      "step": 9125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2186332134145217,
+      "learning_rate": 9.698577820272958e-06,
+      "loss": 0.81,
+      "step": 9126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4225976525036132,
+      "learning_rate": 9.696717758462716e-06,
+      "loss": 0.2788,
+      "step": 9127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2970628514194815,
+      "learning_rate": 9.694857707155126e-06,
+      "loss": 0.2496,
+      "step": 9128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8264539934908051,
+      "learning_rate": 9.6929976664146e-06,
+      "loss": 0.4638,
+      "step": 9129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33276779372939486,
+      "learning_rate": 9.691137636305554e-06,
+      "loss": 0.2866,
+      "step": 9130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3552299923460887,
+      "learning_rate": 9.689277616892396e-06,
+      "loss": 0.3332,
+      "step": 9131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21249879155980383,
+      "learning_rate": 9.687417608239541e-06,
+      "loss": 0.1038,
+      "step": 9132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3190251853080398,
+      "learning_rate": 9.6855576104114e-06,
+      "loss": 0.2811,
+      "step": 9133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1275659444191837,
+      "learning_rate": 9.683697623472387e-06,
+      "loss": 0.6108,
+      "step": 9134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4722264674934552,
+      "learning_rate": 9.681837647486912e-06,
+      "loss": 0.2835,
+      "step": 9135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3269246466042898,
+      "learning_rate": 9.679977682519385e-06,
+      "loss": 0.2699,
+      "step": 9136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6250715059284526,
+      "learning_rate": 9.678117728634217e-06,
+      "loss": 0.6513,
+      "step": 9137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23143067024166109,
+      "learning_rate": 9.676257785895817e-06,
+      "loss": 0.134,
+      "step": 9138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39965635338532934,
+      "learning_rate": 9.674397854368598e-06,
+      "loss": 0.2694,
+      "step": 9139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3096227242107919,
+      "learning_rate": 9.672537934116966e-06,
+      "loss": 0.3045,
+      "step": 9140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5551639828567845,
+      "learning_rate": 9.670678025205332e-06,
+      "loss": 0.3214,
+      "step": 9141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3793778776155998,
+      "learning_rate": 9.668818127698103e-06,
+      "loss": 0.2593,
+      "step": 9142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3052842446630093,
+      "learning_rate": 9.66695824165969e-06,
+      "loss": 0.2919,
+      "step": 9143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3974048517058976,
+      "learning_rate": 9.665098367154496e-06,
+      "loss": 0.2662,
+      "step": 9144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30419717373936883,
+      "learning_rate": 9.663238504246933e-06,
+      "loss": 0.2007,
+      "step": 9145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3827763477642118,
+      "learning_rate": 9.661378653001404e-06,
+      "loss": 0.2769,
+      "step": 9146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3473256096785234,
+      "learning_rate": 9.65951881348232e-06,
+      "loss": 0.3079,
+      "step": 9147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3339712098319444,
+      "learning_rate": 9.657658985754085e-06,
+      "loss": 0.162,
+      "step": 9148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.507041220843939,
+      "learning_rate": 9.655799169881103e-06,
+      "loss": 0.3754,
+      "step": 9149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3073406626204165,
+      "learning_rate": 9.653939365927785e-06,
+      "loss": 0.7838,
+      "step": 9150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28633539371312033,
+      "learning_rate": 9.652079573958529e-06,
+      "loss": 0.2202,
+      "step": 9151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3691162201113786,
+      "learning_rate": 9.650219794037741e-06,
+      "loss": 0.238,
+      "step": 9152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41623336946895095,
+      "learning_rate": 9.648360026229828e-06,
+      "loss": 0.3067,
+      "step": 9153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35646845738442257,
+      "learning_rate": 9.646500270599191e-06,
+      "loss": 0.2647,
+      "step": 9154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28283345328135584,
+      "learning_rate": 9.644640527210235e-06,
+      "loss": 0.2443,
+      "step": 9155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1947336670116657,
+      "learning_rate": 9.642780796127362e-06,
+      "loss": 0.6928,
+      "step": 9156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5466204845370135,
+      "learning_rate": 9.640921077414975e-06,
+      "loss": 0.3285,
+      "step": 9157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35102520321136116,
+      "learning_rate": 9.639061371137475e-06,
+      "loss": 0.2716,
+      "step": 9158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24634222677994205,
+      "learning_rate": 9.637201677359266e-06,
+      "loss": 0.2251,
+      "step": 9159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5712291239043696,
+      "learning_rate": 9.635341996144747e-06,
+      "loss": 0.3271,
+      "step": 9160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34496339742721027,
+      "learning_rate": 9.633482327558316e-06,
+      "loss": 0.2331,
+      "step": 9161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8177903308392893,
+      "learning_rate": 9.63162267166438e-06,
+      "loss": 0.6087,
+      "step": 9162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3266062325787849,
+      "learning_rate": 9.629763028527332e-06,
+      "loss": 0.2651,
+      "step": 9163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.42567311427029086,
+      "learning_rate": 9.627903398211577e-06,
+      "loss": 0.2801,
+      "step": 9164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3037908672972681,
+      "learning_rate": 9.626043780781508e-06,
+      "loss": 0.1905,
+      "step": 9165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.623542999401843,
+      "learning_rate": 9.62418417630153e-06,
+      "loss": 0.3826,
+      "step": 9166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31378311285481736,
+      "learning_rate": 9.622324584836036e-06,
+      "loss": 0.2816,
+      "step": 9167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.04329856978667,
+      "learning_rate": 9.620465006449427e-06,
+      "loss": 0.4929,
+      "step": 9168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6656212816613781,
+      "learning_rate": 9.618605441206098e-06,
+      "loss": 0.3594,
+      "step": 9169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3644528569994132,
+      "learning_rate": 9.616745889170446e-06,
+      "loss": 0.3054,
+      "step": 9170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1996589631527687,
+      "learning_rate": 9.614886350406865e-06,
+      "loss": 0.1666,
+      "step": 9171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3726302590229509,
+      "learning_rate": 9.613026824979757e-06,
+      "loss": 0.2799,
+      "step": 9172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5874746934630642,
+      "learning_rate": 9.61116731295351e-06,
+      "loss": 0.3598,
+      "step": 9173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4754181343382663,
+      "learning_rate": 9.609307814392525e-06,
+      "loss": 0.3157,
+      "step": 9174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3358430754062846,
+      "learning_rate": 9.607448329361193e-06,
+      "loss": 0.2765,
+      "step": 9175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.368547482274709,
+      "learning_rate": 9.605588857923906e-06,
+      "loss": 0.3226,
+      "step": 9176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24855846818078145,
+      "learning_rate": 9.603729400145063e-06,
+      "loss": 0.1304,
+      "step": 9177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8221257736419765,
+      "learning_rate": 9.601869956089051e-06,
+      "loss": 0.3965,
+      "step": 9178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2844103771443083,
+      "learning_rate": 9.60001052582027e-06,
+      "loss": 0.2756,
+      "step": 9179
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5660900591073189,
+      "learning_rate": 9.598151109403102e-06,
+      "loss": 0.4276,
+      "step": 9180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4704157923383228,
+      "learning_rate": 9.596291706901946e-06,
+      "loss": 0.2179,
+      "step": 9181
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4820397687697669,
+      "learning_rate": 9.59443231838119e-06,
+      "loss": 0.3491,
+      "step": 9182
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38983163889496963,
+      "learning_rate": 9.59257294390523e-06,
+      "loss": 0.3285,
+      "step": 9183
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2131714527848812,
+      "learning_rate": 9.59071358353845e-06,
+      "loss": 0.1013,
+      "step": 9184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3957734107129791,
+      "learning_rate": 9.588854237345238e-06,
+      "loss": 0.3476,
+      "step": 9185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9320135709989978,
+      "learning_rate": 9.586994905389985e-06,
+      "loss": 0.4693,
+      "step": 9186
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3124984898779434,
+      "learning_rate": 9.585135587737085e-06,
+      "loss": 0.2266,
+      "step": 9187
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45270701652538803,
+      "learning_rate": 9.583276284450917e-06,
+      "loss": 0.3355,
+      "step": 9188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44931577222450564,
+      "learning_rate": 9.581416995595877e-06,
+      "loss": 0.2828,
+      "step": 9189
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23532191378371217,
+      "learning_rate": 9.579557721236345e-06,
+      "loss": 0.1269,
+      "step": 9190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34700214444362804,
+      "learning_rate": 9.577698461436715e-06,
+      "loss": 0.2821,
+      "step": 9191
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1069957482466108,
+      "learning_rate": 9.575839216261366e-06,
+      "loss": 0.4473,
+      "step": 9192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.539025147105325,
+      "learning_rate": 9.573979985774689e-06,
+      "loss": 0.3448,
+      "step": 9193
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4325392202671178,
+      "learning_rate": 9.572120770041065e-06,
+      "loss": 0.2551,
+      "step": 9194
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37278555312821754,
+      "learning_rate": 9.570261569124882e-06,
+      "loss": 0.3115,
+      "step": 9195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31715591922240705,
+      "learning_rate": 9.568402383090519e-06,
+      "loss": 0.1692,
+      "step": 9196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3931850717706773,
+      "learning_rate": 9.566543212002365e-06,
+      "loss": 0.2381,
+      "step": 9197
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4006785161291414,
+      "learning_rate": 9.564684055924801e-06,
+      "loss": 0.3082,
+      "step": 9198
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9563471812307568,
+      "learning_rate": 9.562824914922211e-06,
+      "loss": 0.6652,
+      "step": 9199
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29294573478695535,
+      "learning_rate": 9.560965789058975e-06,
+      "loss": 0.2166,
+      "step": 9200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.056246726061743,
+      "learning_rate": 9.559106678399473e-06,
+      "loss": 0.5694,
+      "step": 9201
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27936225502157397,
+      "learning_rate": 9.55724758300809e-06,
+      "loss": 0.1959,
+      "step": 9202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3162580378410517,
+      "learning_rate": 9.555388502949201e-06,
+      "loss": 0.2596,
+      "step": 9203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6572106044140896,
+      "learning_rate": 9.553529438287192e-06,
+      "loss": 0.4197,
+      "step": 9204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8055700596054751,
+      "learning_rate": 9.551670389086438e-06,
+      "loss": 0.5843,
+      "step": 9205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.372228470556821,
+      "learning_rate": 9.54981135541132e-06,
+      "loss": 0.2797,
+      "step": 9206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3561420751933189,
+      "learning_rate": 9.547952337326214e-06,
+      "loss": 0.2531,
+      "step": 9207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37148428650642595,
+      "learning_rate": 9.546093334895498e-06,
+      "loss": 0.2567,
+      "step": 9208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36363796098136053,
+      "learning_rate": 9.544234348183553e-06,
+      "loss": 0.279,
+      "step": 9209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2959650585604881,
+      "learning_rate": 9.542375377254753e-06,
+      "loss": 0.238,
+      "step": 9210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6060688718356312,
+      "learning_rate": 9.54051642217347e-06,
+      "loss": 0.399,
+      "step": 9211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3963007962854568,
+      "learning_rate": 9.538657483004088e-06,
+      "loss": 0.3126,
+      "step": 9212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7595275160558781,
+      "learning_rate": 9.536798559810978e-06,
+      "loss": 0.3205,
+      "step": 9213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3413631598251066,
+      "learning_rate": 9.53493965265851e-06,
+      "loss": 0.3021,
+      "step": 9214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.327046762074509,
+      "learning_rate": 9.533080761611066e-06,
+      "loss": 0.2814,
+      "step": 9215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21901892281945962,
+      "learning_rate": 9.53122188673301e-06,
+      "loss": 0.1508,
+      "step": 9216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0974836030044757,
+      "learning_rate": 9.529363028088725e-06,
+      "loss": 0.7183,
+      "step": 9217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34926035459810145,
+      "learning_rate": 9.52750418574258e-06,
+      "loss": 0.265,
+      "step": 9218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3801483295500776,
+      "learning_rate": 9.525645359758939e-06,
+      "loss": 0.3289,
+      "step": 9219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1846574314137546,
+      "learning_rate": 9.523786550202182e-06,
+      "loss": 0.3465,
+      "step": 9220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35928822346783335,
+      "learning_rate": 9.521927757136673e-06,
+      "loss": 0.2593,
+      "step": 9221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25623013450271037,
+      "learning_rate": 9.520068980626789e-06,
+      "loss": 0.1978,
+      "step": 9222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37505431562660235,
+      "learning_rate": 9.518210220736892e-06,
+      "loss": 0.2664,
+      "step": 9223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37891600500710415,
+      "learning_rate": 9.516351477531357e-06,
+      "loss": 0.2994,
+      "step": 9224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9228496278274424,
+      "learning_rate": 9.51449275107455e-06,
+      "loss": 0.4677,
+      "step": 9225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3824480382733829,
+      "learning_rate": 9.512634041430835e-06,
+      "loss": 0.2635,
+      "step": 9226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41175544519260576,
+      "learning_rate": 9.510775348664584e-06,
+      "loss": 0.3034,
+      "step": 9227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26939742805578254,
+      "learning_rate": 9.508916672840161e-06,
+      "loss": 0.1751,
+      "step": 9228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.172159570979836,
+      "learning_rate": 9.507058014021933e-06,
+      "loss": 0.5465,
+      "step": 9229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35814475799141565,
+      "learning_rate": 9.505199372274264e-06,
+      "loss": 0.27,
+      "step": 9230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43551068922587277,
+      "learning_rate": 9.50334074766152e-06,
+      "loss": 0.3128,
+      "step": 9231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7961712293430698,
+      "learning_rate": 9.501482140248064e-06,
+      "loss": 0.4733,
+      "step": 9232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3121898207774114,
+      "learning_rate": 9.499623550098262e-06,
+      "loss": 0.1957,
+      "step": 9233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2227896224663921,
+      "learning_rate": 9.497764977276473e-06,
+      "loss": 0.2213,
+      "step": 9234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5631757758484857,
+      "learning_rate": 9.495906421847063e-06,
+      "loss": 0.792,
+      "step": 9235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3107139886169333,
+      "learning_rate": 9.49404788387439e-06,
+      "loss": 0.2195,
+      "step": 9236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7244261519601902,
+      "learning_rate": 9.492189363422819e-06,
+      "loss": 0.4441,
+      "step": 9237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3605582089400721,
+      "learning_rate": 9.490330860556707e-06,
+      "loss": 0.3125,
+      "step": 9238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2873919445850608,
+      "learning_rate": 9.488472375340417e-06,
+      "loss": 0.2106,
+      "step": 9239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39888390668387463,
+      "learning_rate": 9.486613907838306e-06,
+      "loss": 0.2769,
+      "step": 9240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4867865905958677,
+      "learning_rate": 9.484755458114732e-06,
+      "loss": 0.3455,
+      "step": 9241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.310940342891282,
+      "learning_rate": 9.482897026234056e-06,
+      "loss": 0.2665,
+      "step": 9242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.501526397698931,
+      "learning_rate": 9.48103861226063e-06,
+      "loss": 0.3035,
+      "step": 9243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7521124857164626,
+      "learning_rate": 9.47918021625882e-06,
+      "loss": 0.5031,
+      "step": 9244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.393541576127973,
+      "learning_rate": 9.477321838292972e-06,
+      "loss": 0.279,
+      "step": 9245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3212033534001573,
+      "learning_rate": 9.475463478427451e-06,
+      "loss": 0.2417,
+      "step": 9246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7874596505559587,
+      "learning_rate": 9.473605136726602e-06,
+      "loss": 0.6043,
+      "step": 9247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3961777779705941,
+      "learning_rate": 9.471746813254788e-06,
+      "loss": 0.2912,
+      "step": 9248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27725278619850396,
+      "learning_rate": 9.469888508076357e-06,
+      "loss": 0.1882,
+      "step": 9249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3362006708506701,
+      "learning_rate": 9.468030221255667e-06,
+      "loss": 0.3169,
+      "step": 9250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38625457503857924,
+      "learning_rate": 9.46617195285707e-06,
+      "loss": 0.2824,
+      "step": 9251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4559670161202051,
+      "learning_rate": 9.464313702944912e-06,
+      "loss": 0.272,
+      "step": 9252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8027237790991724,
+      "learning_rate": 9.462455471583545e-06,
+      "loss": 0.5867,
+      "step": 9253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2779288614551335,
+      "learning_rate": 9.460597258837325e-06,
+      "loss": 0.2458,
+      "step": 9254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43085408967342126,
+      "learning_rate": 9.458739064770595e-06,
+      "loss": 0.351,
+      "step": 9255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.22237328741189238,
+      "learning_rate": 9.456880889447712e-06,
+      "loss": 0.093,
+      "step": 9256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3228147343712859,
+      "learning_rate": 9.455022732933017e-06,
+      "loss": 0.2727,
+      "step": 9257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4869726882315915,
+      "learning_rate": 9.453164595290865e-06,
+      "loss": 0.3442,
+      "step": 9258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4765999794871249,
+      "learning_rate": 9.451306476585595e-06,
+      "loss": 0.3115,
+      "step": 9259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37056859041713597,
+      "learning_rate": 9.449448376881563e-06,
+      "loss": 0.2723,
+      "step": 9260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9036552153671678,
+      "learning_rate": 9.447590296243106e-06,
+      "loss": 0.4958,
+      "step": 9261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21923658796039047,
+      "learning_rate": 9.445732234734576e-06,
+      "loss": 0.1774,
+      "step": 9262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36316978936914124,
+      "learning_rate": 9.443874192420312e-06,
+      "loss": 0.2585,
+      "step": 9263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5077816681609485,
+      "learning_rate": 9.442016169364664e-06,
+      "loss": 0.3649,
+      "step": 9264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33319325267717365,
+      "learning_rate": 9.440158165631972e-06,
+      "loss": 0.2898,
+      "step": 9265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4992737656729488,
+      "learning_rate": 9.438300181286576e-06,
+      "loss": 0.284,
+      "step": 9266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4975012780414002,
+      "learning_rate": 9.436442216392823e-06,
+      "loss": 0.4145,
+      "step": 9267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2694784425314511,
+      "learning_rate": 9.43458427101505e-06,
+      "loss": 0.1441,
+      "step": 9268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3240022249400918,
+      "learning_rate": 9.4327263452176e-06,
+      "loss": 0.1771,
+      "step": 9269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3343235297742924,
+      "learning_rate": 9.430868439064813e-06,
+      "loss": 0.3089,
+      "step": 9270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6910139457529847,
+      "learning_rate": 9.429010552621027e-06,
+      "loss": 0.4781,
+      "step": 9271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35862362126432507,
+      "learning_rate": 9.42715268595058e-06,
+      "loss": 0.243,
+      "step": 9272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49851971643361387,
+      "learning_rate": 9.425294839117812e-06,
+      "loss": 0.391,
+      "step": 9273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23422338164763837,
+      "learning_rate": 9.423437012187057e-06,
+      "loss": 0.2075,
+      "step": 9274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3086926262711173,
+      "learning_rate": 9.421579205222657e-06,
+      "loss": 0.2067,
+      "step": 9275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.310073728398527,
+      "learning_rate": 9.41972141828894e-06,
+      "loss": 0.6654,
+      "step": 9276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5841986899196692,
+      "learning_rate": 9.41786365145025e-06,
+      "loss": 0.3337,
+      "step": 9277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2536795044961282,
+      "learning_rate": 9.416005904770916e-06,
+      "loss": 0.2111,
+      "step": 9278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2215967277636537,
+      "learning_rate": 9.414148178315268e-06,
+      "loss": 0.8136,
+      "step": 9279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3058690669119812,
+      "learning_rate": 9.412290472147648e-06,
+      "loss": 0.1969,
+      "step": 9280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7589007743382404,
+      "learning_rate": 9.41043278633238e-06,
+      "loss": 0.409,
+      "step": 9281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3057591447527336,
+      "learning_rate": 9.408575120933804e-06,
+      "loss": 0.2498,
+      "step": 9282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6995045368657311,
+      "learning_rate": 9.406717476016242e-06,
+      "loss": 0.4688,
+      "step": 9283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5984000260809333,
+      "learning_rate": 9.40485985164403e-06,
+      "loss": 0.3472,
+      "step": 9284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.375960869641853,
+      "learning_rate": 9.403002247881499e-06,
+      "loss": 0.2511,
+      "step": 9285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2200191456962238,
+      "learning_rate": 9.40114466479297e-06,
+      "loss": 0.209,
+      "step": 9286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5939482634092973,
+      "learning_rate": 9.399287102442776e-06,
+      "loss": 0.333,
+      "step": 9287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3786700230040074,
+      "learning_rate": 9.397429560895243e-06,
+      "loss": 0.2799,
+      "step": 9288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.47843379765938265,
+      "learning_rate": 9.395572040214702e-06,
+      "loss": 0.3526,
+      "step": 9289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3725937972585066,
+      "learning_rate": 9.393714540465474e-06,
+      "loss": 0.2929,
+      "step": 9290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34822543851312543,
+      "learning_rate": 9.391857061711883e-06,
+      "loss": 0.2587,
+      "step": 9291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2866067925550406,
+      "learning_rate": 9.389999604018258e-06,
+      "loss": 0.1637,
+      "step": 9292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3738816323207938,
+      "learning_rate": 9.388142167448917e-06,
+      "loss": 0.3084,
+      "step": 9293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29667286372725454,
+      "learning_rate": 9.38628475206819e-06,
+      "loss": 0.2788,
+      "step": 9294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6007497819778793,
+      "learning_rate": 9.384427357940394e-06,
+      "loss": 0.3506,
+      "step": 9295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41917867119031543,
+      "learning_rate": 9.382569985129854e-06,
+      "loss": 0.3215,
+      "step": 9296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5345523972264065,
+      "learning_rate": 9.380712633700887e-06,
+      "loss": 0.37,
+      "step": 9297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20961728479650352,
+      "learning_rate": 9.378855303717817e-06,
+      "loss": 0.1741,
+      "step": 9298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5007677585180572,
+      "learning_rate": 9.376997995244957e-06,
+      "loss": 0.3596,
+      "step": 9299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30960347623118034,
+      "learning_rate": 9.375140708346634e-06,
+      "loss": 0.2864,
+      "step": 9300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3443777763500569,
+      "learning_rate": 9.373283443087159e-06,
+      "loss": 0.2645,
+      "step": 9301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7207968905840805,
+      "learning_rate": 9.371426199530853e-06,
+      "loss": 0.4504,
+      "step": 9302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3588460744152287,
+      "learning_rate": 9.369568977742028e-06,
+      "loss": 0.3213,
+      "step": 9303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6060702782939552,
+      "learning_rate": 9.367711777785004e-06,
+      "loss": 0.324,
+      "step": 9304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3480442813272276,
+      "learning_rate": 9.365854599724096e-06,
+      "loss": 0.2939,
+      "step": 9305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23986228711116675,
+      "learning_rate": 9.363997443623612e-06,
+      "loss": 0.253,
+      "step": 9306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4254434470268579,
+      "learning_rate": 9.362140309547873e-06,
+      "loss": 0.2891,
+      "step": 9307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4305528719251428,
+      "learning_rate": 9.360283197561185e-06,
+      "loss": 0.1963,
+      "step": 9308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.32013724856043474,
+      "learning_rate": 9.358426107727862e-06,
+      "loss": 0.2812,
+      "step": 9309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49958189678242726,
+      "learning_rate": 9.356569040112216e-06,
+      "loss": 0.351,
+      "step": 9310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4889885463858359,
+      "learning_rate": 9.354711994778558e-06,
+      "loss": 0.2716,
+      "step": 9311
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2855203325007413,
+      "learning_rate": 9.352854971791192e-06,
+      "loss": 0.2248,
+      "step": 9312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27880863601709716,
+      "learning_rate": 9.350997971214434e-06,
+      "loss": 0.2507,
+      "step": 9313
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3792599482042661,
+      "learning_rate": 9.349140993112588e-06,
+      "loss": 0.2169,
+      "step": 9314
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38684532344974176,
+      "learning_rate": 9.347284037549962e-06,
+      "loss": 0.3065,
+      "step": 9315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8003084968892227,
+      "learning_rate": 9.34542710459086e-06,
+      "loss": 0.4174,
+      "step": 9316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3193010117283392,
+      "learning_rate": 9.343570194299591e-06,
+      "loss": 0.2854,
+      "step": 9317
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32379332391684124,
+      "learning_rate": 9.341713306740457e-06,
+      "loss": 0.2349,
+      "step": 9318
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30827057912219974,
+      "learning_rate": 9.339856441977767e-06,
+      "loss": 0.2344,
+      "step": 9319
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1331119563905843,
+      "learning_rate": 9.337999600075814e-06,
+      "loss": 0.6408,
+      "step": 9320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2831832360300413,
+      "learning_rate": 9.336142781098908e-06,
+      "loss": 0.2151,
+      "step": 9321
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5405839352824342,
+      "learning_rate": 9.33428598511135e-06,
+      "loss": 0.3459,
+      "step": 9322
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9654227292116724,
+      "learning_rate": 9.332429212177438e-06,
+      "loss": 0.4725,
+      "step": 9323
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23482472356001713,
+      "learning_rate": 9.330572462361474e-06,
+      "loss": 0.1729,
+      "step": 9324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4684689393595989,
+      "learning_rate": 9.328715735727758e-06,
+      "loss": 0.3859,
+      "step": 9325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.297606052083825,
+      "learning_rate": 9.326859032340585e-06,
+      "loss": 0.2425,
+      "step": 9326
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4108345649916392,
+      "learning_rate": 9.325002352264257e-06,
+      "loss": 0.2403,
+      "step": 9327
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9941058792175076,
+      "learning_rate": 9.323145695563067e-06,
+      "loss": 0.4418,
+      "step": 9328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32359946359922775,
+      "learning_rate": 9.321289062301313e-06,
+      "loss": 0.2928,
+      "step": 9329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3409617393200946,
+      "learning_rate": 9.319432452543292e-06,
+      "loss": 0.2698,
+      "step": 9330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24943573775102337,
+      "learning_rate": 9.317575866353293e-06,
+      "loss": 0.151,
+      "step": 9331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47506315731861976,
+      "learning_rate": 9.315719303795614e-06,
+      "loss": 0.2959,
+      "step": 9332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36402547396764845,
+      "learning_rate": 9.313862764934543e-06,
+      "loss": 0.2759,
+      "step": 9333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5439363586977312,
+      "learning_rate": 9.312006249834378e-06,
+      "loss": 0.2909,
+      "step": 9334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7236335627092366,
+      "learning_rate": 9.310149758559405e-06,
+      "loss": 0.3895,
+      "step": 9335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36068005871087383,
+      "learning_rate": 9.30829329117392e-06,
+      "loss": 0.2747,
+      "step": 9336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3165484712620228,
+      "learning_rate": 9.306436847742203e-06,
+      "loss": 0.2673,
+      "step": 9337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38534131183340886,
+      "learning_rate": 9.304580428328552e-06,
+      "loss": 0.2414,
+      "step": 9338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3115474788513869,
+      "learning_rate": 9.30272403299725e-06,
+      "loss": 0.2647,
+      "step": 9339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3285219921891418,
+      "learning_rate": 9.300867661812585e-06,
+      "loss": 0.1915,
+      "step": 9340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40017987691315415,
+      "learning_rate": 9.29901131483884e-06,
+      "loss": 0.3231,
+      "step": 9341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34327243727753615,
+      "learning_rate": 9.297154992140307e-06,
+      "loss": 0.2567,
+      "step": 9342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2467554100617364,
+      "learning_rate": 9.295298693781267e-06,
+      "loss": 0.7486,
+      "step": 9343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3513490054263849,
+      "learning_rate": 9.293442419825998e-06,
+      "loss": 0.2287,
+      "step": 9344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29446477079644806,
+      "learning_rate": 9.291586170338793e-06,
+      "loss": 0.2604,
+      "step": 9345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3906785495034913,
+      "learning_rate": 9.289729945383924e-06,
+      "loss": 0.2681,
+      "step": 9346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.693760446359131,
+      "learning_rate": 9.28787374502568e-06,
+      "loss": 0.2316,
+      "step": 9347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34998798570881384,
+      "learning_rate": 9.286017569328334e-06,
+      "loss": 0.2665,
+      "step": 9348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3792652287586318,
+      "learning_rate": 9.284161418356171e-06,
+      "loss": 0.3214,
+      "step": 9349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6013268840193028,
+      "learning_rate": 9.282305292173467e-06,
+      "loss": 0.3168,
+      "step": 9350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40681134523342405,
+      "learning_rate": 9.280449190844501e-06,
+      "loss": 0.2882,
+      "step": 9351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32759946772814447,
+      "learning_rate": 9.278593114433547e-06,
+      "loss": 0.2228,
+      "step": 9352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38246028250584957,
+      "learning_rate": 9.276737063004884e-06,
+      "loss": 0.2469,
+      "step": 9353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3570238331886624,
+      "learning_rate": 9.274881036622785e-06,
+      "loss": 0.2766,
+      "step": 9354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8536521626653547,
+      "learning_rate": 9.273025035351526e-06,
+      "loss": 0.5547,
+      "step": 9355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6642206879063963,
+      "learning_rate": 9.271169059255376e-06,
+      "loss": 0.4194,
+      "step": 9356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2544235207525964,
+      "learning_rate": 9.269313108398611e-06,
+      "loss": 0.2227,
+      "step": 9357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27520770254521076,
+      "learning_rate": 9.2674571828455e-06,
+      "loss": 0.1867,
+      "step": 9358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9026816205634214,
+      "learning_rate": 9.265601282660318e-06,
+      "loss": 0.5432,
+      "step": 9359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3427639614610134,
+      "learning_rate": 9.263745407907329e-06,
+      "loss": 0.2059,
+      "step": 9360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4234555347458187,
+      "learning_rate": 9.261889558650809e-06,
+      "loss": 0.3324,
+      "step": 9361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0330872273748664,
+      "learning_rate": 9.260033734955018e-06,
+      "loss": 0.421,
+      "step": 9362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3320606941363923,
+      "learning_rate": 9.25817793688423e-06,
+      "loss": 0.2124,
+      "step": 9363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32162030886140774,
+      "learning_rate": 9.256322164502704e-06,
+      "loss": 0.1961,
+      "step": 9364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3775542765593719,
+      "learning_rate": 9.254466417874714e-06,
+      "loss": 0.3093,
+      "step": 9365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3245176121995681,
+      "learning_rate": 9.252610697064516e-06,
+      "loss": 0.1544,
+      "step": 9366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6532298265771281,
+      "learning_rate": 9.25075500213638e-06,
+      "loss": 0.3814,
+      "step": 9367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39797220000426015,
+      "learning_rate": 9.248899333154565e-06,
+      "loss": 0.3394,
+      "step": 9368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3901046536766521,
+      "learning_rate": 9.247043690183334e-06,
+      "loss": 0.2712,
+      "step": 9369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2983187384453381,
+      "learning_rate": 9.245188073286949e-06,
+      "loss": 0.1997,
+      "step": 9370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4306936639740022,
+      "learning_rate": 9.243332482529665e-06,
+      "loss": 0.3333,
+      "step": 9371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4124403701710802,
+      "learning_rate": 9.241476917975748e-06,
+      "loss": 0.2959,
+      "step": 9372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3857979000578478,
+      "learning_rate": 9.239621379689452e-06,
+      "loss": 0.2773,
+      "step": 9373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7877835736477719,
+      "learning_rate": 9.237765867735035e-06,
+      "loss": 0.5544,
+      "step": 9374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3805108940267185,
+      "learning_rate": 9.235910382176751e-06,
+      "loss": 0.2811,
+      "step": 9375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2183372803331831,
+      "learning_rate": 9.234054923078862e-06,
+      "loss": 0.1758,
+      "step": 9376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.44649442283170376,
+      "learning_rate": 9.232199490505613e-06,
+      "loss": 0.3603,
+      "step": 9377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3408016275041556,
+      "learning_rate": 9.230344084521266e-06,
+      "loss": 0.2609,
+      "step": 9378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8404217815307761,
+      "learning_rate": 9.228488705190069e-06,
+      "loss": 0.3583,
+      "step": 9379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34473996252645167,
+      "learning_rate": 9.226633352576276e-06,
+      "loss": 0.3451,
+      "step": 9380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3742015521440866,
+      "learning_rate": 9.224778026744135e-06,
+      "loss": 0.2612,
+      "step": 9381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9236809927403062,
+      "learning_rate": 9.222922727757899e-06,
+      "loss": 0.5384,
+      "step": 9382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2849500682429431,
+      "learning_rate": 9.221067455681817e-06,
+      "loss": 0.1571,
+      "step": 9383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4049939669342111,
+      "learning_rate": 9.219212210580132e-06,
+      "loss": 0.318,
+      "step": 9384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29913254424780833,
+      "learning_rate": 9.217356992517097e-06,
+      "loss": 0.2886,
+      "step": 9385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0583760086865348,
+      "learning_rate": 9.215501801556954e-06,
+      "loss": 0.4293,
+      "step": 9386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.61751445391918,
+      "learning_rate": 9.213646637763954e-06,
+      "loss": 0.3031,
+      "step": 9387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4187489970847992,
+      "learning_rate": 9.211791501202333e-06,
+      "loss": 0.3314,
+      "step": 9388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2328530107323132,
+      "learning_rate": 9.209936391936339e-06,
+      "loss": 0.2078,
+      "step": 9389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42827604942936515,
+      "learning_rate": 9.208081310030216e-06,
+      "loss": 0.273,
+      "step": 9390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5354159507363013,
+      "learning_rate": 9.2062262555482e-06,
+      "loss": 0.3933,
+      "step": 9391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4463089490948037,
+      "learning_rate": 9.204371228554538e-06,
+      "loss": 0.2974,
+      "step": 9392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3010100233617082,
+      "learning_rate": 9.202516229113462e-06,
+      "loss": 0.2662,
+      "step": 9393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.508972715684767,
+      "learning_rate": 9.200661257289217e-06,
+      "loss": 0.3506,
+      "step": 9394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3217573305101514,
+      "learning_rate": 9.19880631314604e-06,
+      "loss": 0.2201,
+      "step": 9395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29336545819378834,
+      "learning_rate": 9.196951396748164e-06,
+      "loss": 0.2371,
+      "step": 9396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4756760206599183,
+      "learning_rate": 9.195096508159826e-06,
+      "loss": 0.323,
+      "step": 9397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0804796916074983,
+      "learning_rate": 9.193241647445262e-06,
+      "loss": 0.77,
+      "step": 9398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3241278568868132,
+      "learning_rate": 9.191386814668704e-06,
+      "loss": 0.201,
+      "step": 9399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8403532514129441,
+      "learning_rate": 9.189532009894387e-06,
+      "loss": 0.7033,
+      "step": 9400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32997349588142116,
+      "learning_rate": 9.187677233186541e-06,
+      "loss": 0.318,
+      "step": 9401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37909775087026976,
+      "learning_rate": 9.185822484609397e-06,
+      "loss": 0.2387,
+      "step": 9402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26048756876240237,
+      "learning_rate": 9.183967764227188e-06,
+      "loss": 0.2072,
+      "step": 9403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3638139283443146,
+      "learning_rate": 9.182113072104137e-06,
+      "loss": 0.3321,
+      "step": 9404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7221700317212014,
+      "learning_rate": 9.180258408304478e-06,
+      "loss": 0.3916,
+      "step": 9405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3277302216404148,
+      "learning_rate": 9.178403772892433e-06,
+      "loss": 0.2227,
+      "step": 9406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6393616769102479,
+      "learning_rate": 9.176549165932231e-06,
+      "loss": 0.3885,
+      "step": 9407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3951954935016409,
+      "learning_rate": 9.174694587488097e-06,
+      "loss": 0.2902,
+      "step": 9408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.22636610131498916,
+      "learning_rate": 9.17284003762425e-06,
+      "loss": 0.1794,
+      "step": 9409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2675540511328809,
+      "learning_rate": 9.170985516404922e-06,
+      "loss": 0.7407,
+      "step": 9410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59893272808341,
+      "learning_rate": 9.169131023894325e-06,
+      "loss": 0.2856,
+      "step": 9411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28230999589030675,
+      "learning_rate": 9.16727656015669e-06,
+      "loss": 0.2509,
+      "step": 9412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8586813691539498,
+      "learning_rate": 9.165422125256228e-06,
+      "loss": 0.4289,
+      "step": 9413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5314017571523516,
+      "learning_rate": 9.163567719257164e-06,
+      "loss": 0.3422,
+      "step": 9414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31946600380526713,
+      "learning_rate": 9.161713342223711e-06,
+      "loss": 0.2,
+      "step": 9415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3639625388147878,
+      "learning_rate": 9.159858994220092e-06,
+      "loss": 0.3099,
+      "step": 9416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2937050738295473,
+      "learning_rate": 9.15800467531052e-06,
+      "loss": 0.1719,
+      "step": 9417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42166472914570036,
+      "learning_rate": 9.156150385559208e-06,
+      "loss": 0.3034,
+      "step": 9418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4713264661584862,
+      "learning_rate": 9.154296125030371e-06,
+      "loss": 0.3491,
+      "step": 9419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28848784892017665,
+      "learning_rate": 9.15244189378823e-06,
+      "loss": 0.2629,
+      "step": 9420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5160241916906537,
+      "learning_rate": 9.150587691896984e-06,
+      "loss": 0.344,
+      "step": 9421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.183426975433477,
+      "learning_rate": 9.14873351942085e-06,
+      "loss": 0.1326,
+      "step": 9422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7740940117037703,
+      "learning_rate": 9.146879376424037e-06,
+      "loss": 0.3699,
+      "step": 9423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36265406185090515,
+      "learning_rate": 9.145025262970757e-06,
+      "loss": 0.2904,
+      "step": 9424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7197643231905059,
+      "learning_rate": 9.143171179125212e-06,
+      "loss": 0.2937,
+      "step": 9425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6953661847377259,
+      "learning_rate": 9.141317124951613e-06,
+      "loss": 0.349,
+      "step": 9426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32798387026370746,
+      "learning_rate": 9.139463100514166e-06,
+      "loss": 0.2851,
+      "step": 9427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3583724552430915,
+      "learning_rate": 9.137609105877075e-06,
+      "loss": 0.2518,
+      "step": 9428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21343072518400563,
+      "learning_rate": 9.135755141104544e-06,
+      "loss": 0.1528,
+      "step": 9429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4644618331947836,
+      "learning_rate": 9.133901206260773e-06,
+      "loss": 0.3436,
+      "step": 9430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8170445704875847,
+      "learning_rate": 9.132047301409968e-06,
+      "loss": 0.4853,
+      "step": 9431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27928105311252444,
+      "learning_rate": 9.130193426616327e-06,
+      "loss": 0.2082,
+      "step": 9432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41359155985409485,
+      "learning_rate": 9.12833958194405e-06,
+      "loss": 0.31,
+      "step": 9433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1546078883310094,
+      "learning_rate": 9.126485767457336e-06,
+      "loss": 0.7811,
+      "step": 9434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3413134026821165,
+      "learning_rate": 9.124631983220384e-06,
+      "loss": 0.1997,
+      "step": 9435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3295985332780704,
+      "learning_rate": 9.122778229297387e-06,
+      "loss": 0.2877,
+      "step": 9436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3604904526688763,
+      "learning_rate": 9.120924505752543e-06,
+      "loss": 0.2817,
+      "step": 9437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45457565674240885,
+      "learning_rate": 9.119070812650044e-06,
+      "loss": 0.155,
+      "step": 9438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4349575002134113,
+      "learning_rate": 9.117217150054087e-06,
+      "loss": 0.3293,
+      "step": 9439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40328109153502684,
+      "learning_rate": 9.115363518028858e-06,
+      "loss": 0.3267,
+      "step": 9440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7613078562304955,
+      "learning_rate": 9.113509916638557e-06,
+      "loss": 0.2908,
+      "step": 9441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27332115307343685,
+      "learning_rate": 9.111656345947367e-06,
+      "loss": 0.2487,
+      "step": 9442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4121158185394958,
+      "learning_rate": 9.10980280601948e-06,
+      "loss": 0.2577,
+      "step": 9443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34272848135119255,
+      "learning_rate": 9.107949296919084e-06,
+      "loss": 0.3144,
+      "step": 9444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3178424358331697,
+      "learning_rate": 9.106095818710367e-06,
+      "loss": 0.2214,
+      "step": 9445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8274851482459838,
+      "learning_rate": 9.10424237145751e-06,
+      "loss": 0.4761,
+      "step": 9446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4605336475712775,
+      "learning_rate": 9.102388955224703e-06,
+      "loss": 0.3781,
+      "step": 9447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2601899726482571,
+      "learning_rate": 9.10053557007613e-06,
+      "loss": 0.2162,
+      "step": 9448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2737288914687979,
+      "learning_rate": 9.098682216075968e-06,
+      "loss": 0.1548,
+      "step": 9449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7389906969176646,
+      "learning_rate": 9.096828893288404e-06,
+      "loss": 0.3916,
+      "step": 9450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35483877986484735,
+      "learning_rate": 9.094975601777615e-06,
+      "loss": 0.231,
+      "step": 9451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38272983487780293,
+      "learning_rate": 9.093122341607782e-06,
+      "loss": 0.3315,
+      "step": 9452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7223595246219131,
+      "learning_rate": 9.091269112843084e-06,
+      "loss": 0.4254,
+      "step": 9453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34796801665638943,
+      "learning_rate": 9.089415915547702e-06,
+      "loss": 0.2188,
+      "step": 9454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30394563246761563,
+      "learning_rate": 9.087562749785805e-06,
+      "loss": 0.1796,
+      "step": 9455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3333343056858731,
+      "learning_rate": 9.085709615621567e-06,
+      "loss": 0.2878,
+      "step": 9456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3613631132845819,
+      "learning_rate": 9.083856513119169e-06,
+      "loss": 0.2873,
+      "step": 9457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6852597899585396,
+      "learning_rate": 9.082003442342779e-06,
+      "loss": 0.3644,
+      "step": 9458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4615449689732116,
+      "learning_rate": 9.080150403356571e-06,
+      "loss": 0.3572,
+      "step": 9459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27980817096660626,
+      "learning_rate": 9.078297396224716e-06,
+      "loss": 0.2563,
+      "step": 9460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25145383400299104,
+      "learning_rate": 9.07644442101138e-06,
+      "loss": 0.1177,
+      "step": 9461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4001380538324593,
+      "learning_rate": 9.074591477780736e-06,
+      "loss": 0.7624,
+      "step": 9462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.322723513333794,
+      "learning_rate": 9.072738566596948e-06,
+      "loss": 0.2826,
+      "step": 9463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3481210569513449,
+      "learning_rate": 9.070885687524184e-06,
+      "loss": 0.2795,
+      "step": 9464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.013606524113796,
+      "learning_rate": 9.069032840626608e-06,
+      "loss": 0.6801,
+      "step": 9465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3272263147205583,
+      "learning_rate": 9.067180025968387e-06,
+      "loss": 0.2762,
+      "step": 9466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20755738352285572,
+      "learning_rate": 9.065327243613679e-06,
+      "loss": 0.0841,
+      "step": 9467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35855047633435044,
+      "learning_rate": 9.06347449362665e-06,
+      "loss": 0.3212,
+      "step": 9468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3261481520927009,
+      "learning_rate": 9.061621776071458e-06,
+      "loss": 0.2756,
+      "step": 9469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6254025257211152,
+      "learning_rate": 9.059769091012265e-06,
+      "loss": 0.3968,
+      "step": 9470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3530171497321099,
+      "learning_rate": 9.057916438513226e-06,
+      "loss": 0.2633,
+      "step": 9471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36080066395287175,
+      "learning_rate": 9.056063818638502e-06,
+      "loss": 0.3188,
+      "step": 9472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2664798556440693,
+      "learning_rate": 9.054211231452248e-06,
+      "loss": 0.2171,
+      "step": 9473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5965216255280918,
+      "learning_rate": 9.052358677018615e-06,
+      "loss": 0.2632,
+      "step": 9474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.376392912726472,
+      "learning_rate": 9.050506155401764e-06,
+      "loss": 0.2729,
+      "step": 9475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3404934744042724,
+      "learning_rate": 9.048653666665841e-06,
+      "loss": 0.3248,
+      "step": 9476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.040043861261599,
+      "learning_rate": 9.046801210875002e-06,
+      "loss": 0.4977,
+      "step": 9477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3204967432613155,
+      "learning_rate": 9.044948788093396e-06,
+      "loss": 0.2685,
+      "step": 9478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.22589665188179028,
+      "learning_rate": 9.043096398385174e-06,
+      "loss": 0.208,
+      "step": 9479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4867779438407952,
+      "learning_rate": 9.041244041814479e-06,
+      "loss": 0.3523,
+      "step": 9480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3340647249545658,
+      "learning_rate": 9.039391718445466e-06,
+      "loss": 0.2391,
+      "step": 9481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8060812933217341,
+      "learning_rate": 9.03753942834227e-06,
+      "loss": 0.5286,
+      "step": 9482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.518908014472298,
+      "learning_rate": 9.03568717156905e-06,
+      "loss": 0.3698,
+      "step": 9483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28165729999453204,
+      "learning_rate": 9.033834948189936e-06,
+      "loss": 0.2253,
+      "step": 9484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.099977564464965,
+      "learning_rate": 9.031982758269078e-06,
+      "loss": 0.626,
+      "step": 9485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40518128341349524,
+      "learning_rate": 9.030130601870615e-06,
+      "loss": 0.2398,
+      "step": 9486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27175186242628224,
+      "learning_rate": 9.02827847905869e-06,
+      "loss": 0.2198,
+      "step": 9487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47784160961695693,
+      "learning_rate": 9.02642638989744e-06,
+      "loss": 0.3479,
+      "step": 9488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1214209878128412,
+      "learning_rate": 9.024574334451002e-06,
+      "loss": 0.7194,
+      "step": 9489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32215965041453914,
+      "learning_rate": 9.02272231278351e-06,
+      "loss": 0.1616,
+      "step": 9490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28186208190846546,
+      "learning_rate": 9.020870324959103e-06,
+      "loss": 0.2528,
+      "step": 9491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4286246139050462,
+      "learning_rate": 9.019018371041914e-06,
+      "loss": 0.3669,
+      "step": 9492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.433043538846795,
+      "learning_rate": 9.017166451096077e-06,
+      "loss": 0.2856,
+      "step": 9493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.252036692216471,
+      "learning_rate": 9.015314565185724e-06,
+      "loss": 0.1777,
+      "step": 9494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3567792880638821,
+      "learning_rate": 9.013462713374986e-06,
+      "loss": 0.3031,
+      "step": 9495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4525301068269673,
+      "learning_rate": 9.01161089572799e-06,
+      "loss": 0.3151,
+      "step": 9496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4602195249201141,
+      "learning_rate": 9.009759112308867e-06,
+      "loss": 0.2832,
+      "step": 9497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6256387629393889,
+      "learning_rate": 9.007907363181742e-06,
+      "loss": 0.4464,
+      "step": 9498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2897790404691268,
+      "learning_rate": 9.006055648410745e-06,
+      "loss": 0.2242,
+      "step": 9499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2947409062044821,
+      "learning_rate": 9.004203968059997e-06,
+      "loss": 0.203,
+      "step": 9500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8779701484144261,
+      "learning_rate": 9.002352322193622e-06,
+      "loss": 0.5903,
+      "step": 9501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31908124321909886,
+      "learning_rate": 9.000500710875746e-06,
+      "loss": 0.2542,
+      "step": 9502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.450366955751652,
+      "learning_rate": 8.998649134170484e-06,
+      "loss": 0.2701,
+      "step": 9503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5082350399387127,
+      "learning_rate": 8.996797592141962e-06,
+      "loss": 0.337,
+      "step": 9504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2484863804487214,
+      "learning_rate": 8.994946084854294e-06,
+      "loss": 0.2122,
+      "step": 9505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3488933962122594,
+      "learning_rate": 8.9930946123716e-06,
+      "loss": 0.2352,
+      "step": 9506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3152421095593874,
+      "learning_rate": 8.991243174757997e-06,
+      "loss": 0.2676,
+      "step": 9507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4266176579389591,
+      "learning_rate": 8.9893917720776e-06,
+      "loss": 0.3114,
+      "step": 9508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4938255856927612,
+      "learning_rate": 8.987540404394521e-06,
+      "loss": 0.3598,
+      "step": 9509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3394737874270989,
+      "learning_rate": 8.985689071772877e-06,
+      "loss": 0.2153,
+      "step": 9510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37853025059595635,
+      "learning_rate": 8.983837774276774e-06,
+      "loss": 0.2724,
+      "step": 9511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2888364011937344,
+      "learning_rate": 8.981986511970327e-06,
+      "loss": 0.2864,
+      "step": 9512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28390144484902724,
+      "learning_rate": 8.980135284917644e-06,
+      "loss": 0.1234,
+      "step": 9513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35464412918329336,
+      "learning_rate": 8.97828409318283e-06,
+      "loss": 0.2683,
+      "step": 9514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.286350130274255,
+      "learning_rate": 8.976432936829995e-06,
+      "loss": 0.2999,
+      "step": 9515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6943883642781716,
+      "learning_rate": 8.974581815923242e-06,
+      "loss": 0.3359,
+      "step": 9516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40022381769441084,
+      "learning_rate": 8.972730730526679e-06,
+      "loss": 0.243,
+      "step": 9517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5276954077075993,
+      "learning_rate": 8.970879680704404e-06,
+      "loss": 0.4043,
+      "step": 9518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28500783585719586,
+      "learning_rate": 8.969028666520524e-06,
+      "loss": 0.2542,
+      "step": 9519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24384593736610496,
+      "learning_rate": 8.967177688039135e-06,
+      "loss": 0.1619,
+      "step": 9520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48283497815577264,
+      "learning_rate": 8.96532674532434e-06,
+      "loss": 0.3448,
+      "step": 9521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8189859030284328,
+      "learning_rate": 8.963475838440237e-06,
+      "loss": 0.4574,
+      "step": 9522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2617952258223295,
+      "learning_rate": 8.961624967450917e-06,
+      "loss": 0.2188,
+      "step": 9523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48019246869526944,
+      "learning_rate": 8.959774132420481e-06,
+      "loss": 0.3712,
+      "step": 9524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29000128064258673,
+      "learning_rate": 8.957923333413024e-06,
+      "loss": 0.2129,
+      "step": 9525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.526363917290331,
+      "learning_rate": 8.956072570492635e-06,
+      "loss": 0.2352,
+      "step": 9526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2704870327838035,
+      "learning_rate": 8.954221843723409e-06,
+      "loss": 0.2755,
+      "step": 9527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0257404142025073,
+      "learning_rate": 8.952371153169435e-06,
+      "loss": 0.4987,
+      "step": 9528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2940447139600711,
+      "learning_rate": 8.950520498894803e-06,
+      "loss": 0.1177,
+      "step": 9529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38604536877936113,
+      "learning_rate": 8.9486698809636e-06,
+      "loss": 0.3443,
+      "step": 9530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33418039366402746,
+      "learning_rate": 8.946819299439915e-06,
+      "loss": 0.3424,
+      "step": 9531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5842996172246837,
+      "learning_rate": 8.944968754387832e-06,
+      "loss": 0.3918,
+      "step": 9532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20671343129546188,
+      "learning_rate": 8.943118245871437e-06,
+      "loss": 0.1549,
+      "step": 9533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8778299994060808,
+      "learning_rate": 8.941267773954809e-06,
+      "loss": 0.475,
+      "step": 9534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3016216513617832,
+      "learning_rate": 8.939417338702034e-06,
+      "loss": 0.285,
+      "step": 9535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3695515460906184,
+      "learning_rate": 8.93756694017719e-06,
+      "loss": 0.2673,
+      "step": 9536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7614491699387651,
+      "learning_rate": 8.935716578444358e-06,
+      "loss": 0.4398,
+      "step": 9537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.341722836767795,
+      "learning_rate": 8.933866253567615e-06,
+      "loss": 0.2352,
+      "step": 9538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25011501068791103,
+      "learning_rate": 8.932015965611039e-06,
+      "loss": 0.2378,
+      "step": 9539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39852031588177406,
+      "learning_rate": 8.930165714638705e-06,
+      "loss": 0.1926,
+      "step": 9540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5183265685725699,
+      "learning_rate": 8.928315500714682e-06,
+      "loss": 0.3365,
+      "step": 9541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3523020413766487,
+      "learning_rate": 8.92646532390305e-06,
+      "loss": 0.2559,
+      "step": 9542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4905252780482375,
+      "learning_rate": 8.924615184267876e-06,
+      "loss": 0.3216,
+      "step": 9543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5270482680361739,
+      "learning_rate": 8.922765081873235e-06,
+      "loss": 0.3081,
+      "step": 9544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25970806818561254,
+      "learning_rate": 8.92091501678319e-06,
+      "loss": 0.2239,
+      "step": 9545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3026984576541297,
+      "learning_rate": 8.919064989061813e-06,
+      "loss": 0.194,
+      "step": 9546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4164454229618692,
+      "learning_rate": 8.917214998773169e-06,
+      "loss": 0.3015,
+      "step": 9547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32582574801627273,
+      "learning_rate": 8.915365045981323e-06,
+      "loss": 0.285,
+      "step": 9548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7217319400615111,
+      "learning_rate": 8.913515130750336e-06,
+      "loss": 0.3338,
+      "step": 9549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7820623728911443,
+      "learning_rate": 8.911665253144277e-06,
+      "loss": 0.4671,
+      "step": 9550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24558093961398544,
+      "learning_rate": 8.9098154132272e-06,
+      "loss": 0.2646,
+      "step": 9551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19719586673338096,
+      "learning_rate": 8.907965611063173e-06,
+      "loss": 0.073,
+      "step": 9552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5547086944121592,
+      "learning_rate": 8.906115846716247e-06,
+      "loss": 0.3461,
+      "step": 9553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3694099603084277,
+      "learning_rate": 8.904266120250483e-06,
+      "loss": 0.3258,
+      "step": 9554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4425574975466733,
+      "learning_rate": 8.902416431729939e-06,
+      "loss": 0.2769,
+      "step": 9555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5256141648470348,
+      "learning_rate": 8.900566781218665e-06,
+      "loss": 0.3479,
+      "step": 9556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3022772910244815,
+      "learning_rate": 8.898717168780713e-06,
+      "loss": 0.2375,
+      "step": 9557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.409281807309569,
+      "learning_rate": 8.896867594480141e-06,
+      "loss": 0.2499,
+      "step": 9558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3148261067696356,
+      "learning_rate": 8.895018058380995e-06,
+      "loss": 0.2314,
+      "step": 9559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3935927137427299,
+      "learning_rate": 8.893168560547327e-06,
+      "loss": 0.3369,
+      "step": 9560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6398746892434829,
+      "learning_rate": 8.891319101043181e-06,
+      "loss": 0.4199,
+      "step": 9561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39721004210862404,
+      "learning_rate": 8.889469679932612e-06,
+      "loss": 0.2113,
+      "step": 9562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28574456478066207,
+      "learning_rate": 8.887620297279656e-06,
+      "loss": 0.2957,
+      "step": 9563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26193897143332595,
+      "learning_rate": 8.885770953148364e-06,
+      "loss": 0.1744,
+      "step": 9564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6740258025951248,
+      "learning_rate": 8.883921647602777e-06,
+      "loss": 0.2764,
+      "step": 9565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3760220070940269,
+      "learning_rate": 8.882072380706931e-06,
+      "loss": 0.3045,
+      "step": 9566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3810517214167689,
+      "learning_rate": 8.880223152524875e-06,
+      "loss": 0.3385,
+      "step": 9567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9536008985955754,
+      "learning_rate": 8.87837396312064e-06,
+      "loss": 0.5612,
+      "step": 9568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22293813348609617,
+      "learning_rate": 8.876524812558269e-06,
+      "loss": 0.1594,
+      "step": 9569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3957805630650527,
+      "learning_rate": 8.874675700901791e-06,
+      "loss": 0.2501,
+      "step": 9570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38770512159826404,
+      "learning_rate": 8.87282662821525e-06,
+      "loss": 0.3198,
+      "step": 9571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3391420199028762,
+      "learning_rate": 8.87097759456267e-06,
+      "loss": 0.2365,
+      "step": 9572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.202499855422108,
+      "learning_rate": 8.869128600008092e-06,
+      "loss": 0.818,
+      "step": 9573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.45885122466342415,
+      "learning_rate": 8.867279644615537e-06,
+      "loss": 0.3281,
+      "step": 9574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31119385570500957,
+      "learning_rate": 8.865430728449043e-06,
+      "loss": 0.2231,
+      "step": 9575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2595652472913027,
+      "learning_rate": 8.863581851572633e-06,
+      "loss": 0.1539,
+      "step": 9576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4554085077648678,
+      "learning_rate": 8.861733014050334e-06,
+      "loss": 0.3657,
+      "step": 9577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3500096854930678,
+      "learning_rate": 8.859884215946174e-06,
+      "loss": 0.2237,
+      "step": 9578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3774828819594551,
+      "learning_rate": 8.858035457324172e-06,
+      "loss": 0.3163,
+      "step": 9579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7309407493660592,
+      "learning_rate": 8.856186738248355e-06,
+      "loss": 0.4135,
+      "step": 9580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3706488154985834,
+      "learning_rate": 8.85433805878274e-06,
+      "loss": 0.288,
+      "step": 9581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.21088115991413472,
+      "learning_rate": 8.85248941899135e-06,
+      "loss": 0.1959,
+      "step": 9582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.850973894592171,
+      "learning_rate": 8.850640818938202e-06,
+      "loss": 0.4281,
+      "step": 9583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31743512024875603,
+      "learning_rate": 8.848792258687312e-06,
+      "loss": 0.2678,
+      "step": 9584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.773444547415871,
+      "learning_rate": 8.846943738302697e-06,
+      "loss": 0.5385,
+      "step": 9585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40078285238849626,
+      "learning_rate": 8.845095257848372e-06,
+      "loss": 0.3058,
+      "step": 9586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.364521054432038,
+      "learning_rate": 8.843246817388345e-06,
+      "loss": 0.2559,
+      "step": 9587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3184618070839194,
+      "learning_rate": 8.841398416986635e-06,
+      "loss": 0.3191,
+      "step": 9588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3073635659674038,
+      "learning_rate": 8.83955005670725e-06,
+      "loss": 0.2267,
+      "step": 9589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2940086234938814,
+      "learning_rate": 8.837701736614194e-06,
+      "loss": 0.2482,
+      "step": 9590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29719764315455904,
+      "learning_rate": 8.835853456771476e-06,
+      "loss": 0.2412,
+      "step": 9591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9759795203201609,
+      "learning_rate": 8.834005217243103e-06,
+      "loss": 0.593,
+      "step": 9592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3481287823907185,
+      "learning_rate": 8.832157018093078e-06,
+      "loss": 0.2459,
+      "step": 9593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5531869767322063,
+      "learning_rate": 8.830308859385408e-06,
+      "loss": 0.3371,
+      "step": 9594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34635142391335155,
+      "learning_rate": 8.828460741184089e-06,
+      "loss": 0.2135,
+      "step": 9595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4122181775243568,
+      "learning_rate": 8.826612663553126e-06,
+      "loss": 0.2867,
+      "step": 9596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4546733222643512,
+      "learning_rate": 8.824764626556514e-06,
+      "loss": 0.3519,
+      "step": 9597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30491457595218185,
+      "learning_rate": 8.822916630258255e-06,
+      "loss": 0.2441,
+      "step": 9598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35806364278988395,
+      "learning_rate": 8.82106867472234e-06,
+      "loss": 0.2882,
+      "step": 9599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6824198286469831,
+      "learning_rate": 8.819220760012768e-06,
+      "loss": 0.4241,
+      "step": 9600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2923506374089054,
+      "learning_rate": 8.81737288619353e-06,
+      "loss": 0.1254,
+      "step": 9601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3244416849157942,
+      "learning_rate": 8.815525053328617e-06,
+      "loss": 0.2596,
+      "step": 9602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28496715222960844,
+      "learning_rate": 8.81367726148202e-06,
+      "loss": 0.2617,
+      "step": 9603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7812979157651535,
+      "learning_rate": 8.811829510717731e-06,
+      "loss": 0.3444,
+      "step": 9604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31375336364159706,
+      "learning_rate": 8.809981801099735e-06,
+      "loss": 0.2695,
+      "step": 9605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.392233640141547,
+      "learning_rate": 8.808134132692015e-06,
+      "loss": 0.3176,
+      "step": 9606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3762085157918527,
+      "learning_rate": 8.806286505558563e-06,
+      "loss": 0.6777,
+      "step": 9607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3082493565251616,
+      "learning_rate": 8.804438919763352e-06,
+      "loss": 0.219,
+      "step": 9608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4539330693613155,
+      "learning_rate": 8.802591375370375e-06,
+      "loss": 0.2814,
+      "step": 9609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30749822562633533,
+      "learning_rate": 8.800743872443605e-06,
+      "loss": 0.2762,
+      "step": 9610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3004335057158179,
+      "learning_rate": 8.798896411047024e-06,
+      "loss": 0.2093,
+      "step": 9611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.651005317404619,
+      "learning_rate": 8.797048991244606e-06,
+      "loss": 0.4288,
+      "step": 9612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9232587293760873,
+      "learning_rate": 8.795201613100334e-06,
+      "loss": 0.627,
+      "step": 9613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2907613834716408,
+      "learning_rate": 8.793354276678176e-06,
+      "loss": 0.1997,
+      "step": 9614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3186326974408788,
+      "learning_rate": 8.791506982042107e-06,
+      "loss": 0.2948,
+      "step": 9615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5159427141535858,
+      "learning_rate": 8.789659729256099e-06,
+      "loss": 0.2764,
+      "step": 9616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3412013433538205,
+      "learning_rate": 8.787812518384125e-06,
+      "loss": 0.1942,
+      "step": 9617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35941120379477665,
+      "learning_rate": 8.78596534949015e-06,
+      "loss": 0.283,
+      "step": 9618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0007832706509505,
+      "learning_rate": 8.784118222638142e-06,
+      "loss": 0.6069,
+      "step": 9619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4190917204278424,
+      "learning_rate": 8.78227113789207e-06,
+      "loss": 0.3474,
+      "step": 9620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25232330543617304,
+      "learning_rate": 8.780424095315893e-06,
+      "loss": 0.2104,
+      "step": 9621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3665822398797408,
+      "learning_rate": 8.778577094973579e-06,
+      "loss": 0.3238,
+      "step": 9622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41883098987828316,
+      "learning_rate": 8.77673013692909e-06,
+      "loss": 0.3045,
+      "step": 9623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3007795637019158,
+      "learning_rate": 8.77488322124638e-06,
+      "loss": 0.1753,
+      "step": 9624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.344525945362686,
+      "learning_rate": 8.773036347989413e-06,
+      "loss": 0.7523,
+      "step": 9625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26911175627901446,
+      "learning_rate": 8.771189517222143e-06,
+      "loss": 0.265,
+      "step": 9626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42709579508130063,
+      "learning_rate": 8.769342729008529e-06,
+      "loss": 0.2923,
+      "step": 9627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6199825882275222,
+      "learning_rate": 8.767495983412521e-06,
+      "loss": 0.3981,
+      "step": 9628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25178532363408446,
+      "learning_rate": 8.765649280498076e-06,
+      "loss": 0.197,
+      "step": 9629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2845849191707805,
+      "learning_rate": 8.763802620329146e-06,
+      "loss": 0.2088,
+      "step": 9630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0896261704932444,
+      "learning_rate": 8.761956002969672e-06,
+      "loss": 0.7473,
+      "step": 9631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5818223951584665,
+      "learning_rate": 8.760109428483613e-06,
+      "loss": 0.3236,
+      "step": 9632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41990213718374075,
+      "learning_rate": 8.758262896934909e-06,
+      "loss": 0.3098,
+      "step": 9633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.399969974428762,
+      "learning_rate": 8.756416408387507e-06,
+      "loss": 0.2806,
+      "step": 9634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3299745775632892,
+      "learning_rate": 8.754569962905351e-06,
+      "loss": 0.2081,
+      "step": 9635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3008769243224494,
+      "learning_rate": 8.752723560552386e-06,
+      "loss": 0.2482,
+      "step": 9636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9013635666683019,
+      "learning_rate": 8.750877201392547e-06,
+      "loss": 0.3913,
+      "step": 9637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28342819302573247,
+      "learning_rate": 8.749030885489782e-06,
+      "loss": 0.2537,
+      "step": 9638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5810023181066697,
+      "learning_rate": 8.747184612908019e-06,
+      "loss": 0.3466,
+      "step": 9639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6679312576871723,
+      "learning_rate": 8.745338383711202e-06,
+      "loss": 0.3194,
+      "step": 9640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29402336429450904,
+      "learning_rate": 8.74349219796326e-06,
+      "loss": 0.2023,
+      "step": 9641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26832438600589964,
+      "learning_rate": 8.741646055728133e-06,
+      "loss": 0.2494,
+      "step": 9642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.104583917262513,
+      "learning_rate": 8.739799957069747e-06,
+      "loss": 0.681,
+      "step": 9643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35403449470833537,
+      "learning_rate": 8.737953902052031e-06,
+      "loss": 0.2109,
+      "step": 9644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6163331604318542,
+      "learning_rate": 8.736107890738922e-06,
+      "loss": 0.3423,
+      "step": 9645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.44648031150704204,
+      "learning_rate": 8.73426192319434e-06,
+      "loss": 0.3083,
+      "step": 9646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32973426498978503,
+      "learning_rate": 8.732415999482214e-06,
+      "loss": 0.2018,
+      "step": 9647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29871322069317574,
+      "learning_rate": 8.730570119666465e-06,
+      "loss": 0.18,
+      "step": 9648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5177502319297812,
+      "learning_rate": 8.728724283811024e-06,
+      "loss": 0.3988,
+      "step": 9649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3241226683385959,
+      "learning_rate": 8.7268784919798e-06,
+      "loss": 0.2099,
+      "step": 9650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.45021331398071146,
+      "learning_rate": 8.725032744236723e-06,
+      "loss": 0.3137,
+      "step": 9651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7969625967830009,
+      "learning_rate": 8.723187040645704e-06,
+      "loss": 0.5277,
+      "step": 9652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27981093470443125,
+      "learning_rate": 8.721341381270668e-06,
+      "loss": 0.0984,
+      "step": 9653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2597047101805497,
+      "learning_rate": 8.719495766175519e-06,
+      "loss": 0.2336,
+      "step": 9654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1135473319190738,
+      "learning_rate": 8.717650195424182e-06,
+      "loss": 0.6095,
+      "step": 9655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6191392989069763,
+      "learning_rate": 8.715804669080559e-06,
+      "loss": 0.3826,
+      "step": 9656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3043730092485221,
+      "learning_rate": 8.713959187208572e-06,
+      "loss": 0.2419,
+      "step": 9657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4603338423923025,
+      "learning_rate": 8.712113749872117e-06,
+      "loss": 0.3772,
+      "step": 9658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5274180041754517,
+      "learning_rate": 8.710268357135109e-06,
+      "loss": 0.3372,
+      "step": 9659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20931527382171677,
+      "learning_rate": 8.70842300906145e-06,
+      "loss": 0.1499,
+      "step": 9660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45032579061938405,
+      "learning_rate": 8.70657770571505e-06,
+      "loss": 0.3645,
+      "step": 9661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4065200216698055,
+      "learning_rate": 8.704732447159807e-06,
+      "loss": 0.2926,
+      "step": 9662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41762152764098875,
+      "learning_rate": 8.702887233459625e-06,
+      "loss": 0.2756,
+      "step": 9663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3246494443880283,
+      "learning_rate": 8.7010420646784e-06,
+      "loss": 0.8345,
+      "step": 9664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34809237419193545,
+      "learning_rate": 8.699196940880032e-06,
+      "loss": 0.2729,
+      "step": 9665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2492864391185059,
+      "learning_rate": 8.69735186212842e-06,
+      "loss": 0.2019,
+      "step": 9666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5684745090978655,
+      "learning_rate": 8.695506828487457e-06,
+      "loss": 0.285,
+      "step": 9667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4835198325712554,
+      "learning_rate": 8.693661840021035e-06,
+      "loss": 0.3404,
+      "step": 9668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41002756921375183,
+      "learning_rate": 8.691816896793049e-06,
+      "loss": 0.3262,
+      "step": 9669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3301324479001951,
+      "learning_rate": 8.689971998867386e-06,
+      "loss": 0.271,
+      "step": 9670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8899810603812406,
+      "learning_rate": 8.688127146307938e-06,
+      "loss": 0.4589,
+      "step": 9671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26078813677723967,
+      "learning_rate": 8.68628233917859e-06,
+      "loss": 0.2169,
+      "step": 9672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2721884189437094,
+      "learning_rate": 8.684437577543227e-06,
+      "loss": 0.2406,
+      "step": 9673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5468489073374028,
+      "learning_rate": 8.682592861465735e-06,
+      "loss": 0.3952,
+      "step": 9674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3647467307258367,
+      "learning_rate": 8.680748191009995e-06,
+      "loss": 0.295,
+      "step": 9675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7706378980657368,
+      "learning_rate": 8.67890356623989e-06,
+      "loss": 0.3905,
+      "step": 9676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3748336812202746,
+      "learning_rate": 8.677058987219294e-06,
+      "loss": 0.3049,
+      "step": 9677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2454214308007065,
+      "learning_rate": 8.675214454012092e-06,
+      "loss": 0.233,
+      "step": 9678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45322202546424467,
+      "learning_rate": 8.673369966682154e-06,
+      "loss": 0.2341,
+      "step": 9679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5365199502435255,
+      "learning_rate": 8.67152552529336e-06,
+      "loss": 0.3428,
+      "step": 9680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4117158142211994,
+      "learning_rate": 8.669681129909578e-06,
+      "loss": 0.3221,
+      "step": 9681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34998858461194654,
+      "learning_rate": 8.667836780594682e-06,
+      "loss": 0.336,
+      "step": 9682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3850112115715607,
+      "learning_rate": 8.665992477412541e-06,
+      "loss": 0.1246,
+      "step": 9683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3888304992876215,
+      "learning_rate": 8.664148220427023e-06,
+      "loss": 0.3117,
+      "step": 9684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38035497224060644,
+      "learning_rate": 8.662304009701994e-06,
+      "loss": 0.3066,
+      "step": 9685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2546173089481329,
+      "learning_rate": 8.66045984530132e-06,
+      "loss": 0.1335,
+      "step": 9686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35621827719168836,
+      "learning_rate": 8.658615727288863e-06,
+      "loss": 0.2854,
+      "step": 9687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0338373334263178,
+      "learning_rate": 8.656771655728487e-06,
+      "loss": 0.6266,
+      "step": 9688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3064225676834638,
+      "learning_rate": 8.65492763068405e-06,
+      "loss": 0.2283,
+      "step": 9689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33605010056283907,
+      "learning_rate": 8.653083652219417e-06,
+      "loss": 0.302,
+      "step": 9690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6969521748952202,
+      "learning_rate": 8.651239720398433e-06,
+      "loss": 0.4025,
+      "step": 9691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22137321403803517,
+      "learning_rate": 8.64939583528496e-06,
+      "loss": 0.114,
+      "step": 9692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31067351486742123,
+      "learning_rate": 8.647551996942852e-06,
+      "loss": 0.2666,
+      "step": 9693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3017237884993278,
+      "learning_rate": 8.645708205435959e-06,
+      "loss": 0.237,
+      "step": 9694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7947918989546467,
+      "learning_rate": 8.643864460828135e-06,
+      "loss": 0.4867,
+      "step": 9695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30815653925994546,
+      "learning_rate": 8.642020763183224e-06,
+      "loss": 0.2316,
+      "step": 9696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4541266781213096,
+      "learning_rate": 8.640177112565078e-06,
+      "loss": 0.3681,
+      "step": 9697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2920545878428287,
+      "learning_rate": 8.638333509037537e-06,
+      "loss": 0.2506,
+      "step": 9698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3173736734585809,
+      "learning_rate": 8.63648995266445e-06,
+      "loss": 0.211,
+      "step": 9699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38440130858759364,
+      "learning_rate": 8.634646443509656e-06,
+      "loss": 0.2679,
+      "step": 9700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32061793907154607,
+      "learning_rate": 8.632802981636998e-06,
+      "loss": 0.3013,
+      "step": 9701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31661079459501085,
+      "learning_rate": 8.630959567110314e-06,
+      "loss": 0.2308,
+      "step": 9702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.839441115520653,
+      "learning_rate": 8.629116199993441e-06,
+      "loss": 0.4207,
+      "step": 9703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1032675562865573,
+      "learning_rate": 8.627272880350214e-06,
+      "loss": 0.7347,
+      "step": 9704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26791299634997684,
+      "learning_rate": 8.62542960824447e-06,
+      "loss": 0.212,
+      "step": 9705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2568316260810858,
+      "learning_rate": 8.623586383740037e-06,
+      "loss": 0.2,
+      "step": 9706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0746648875443938,
+      "learning_rate": 8.621743206900752e-06,
+      "loss": 0.407,
+      "step": 9707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3060183253023324,
+      "learning_rate": 8.619900077790439e-06,
+      "loss": 0.2603,
+      "step": 9708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3354086747479937,
+      "learning_rate": 8.618056996472925e-06,
+      "loss": 0.2808,
+      "step": 9709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.25565098457294,
+      "learning_rate": 8.616213963012042e-06,
+      "loss": 0.7637,
+      "step": 9710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34005069327044857,
+      "learning_rate": 8.614370977471604e-06,
+      "loss": 0.27,
+      "step": 9711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18000557563902286,
+      "learning_rate": 8.612528039915444e-06,
+      "loss": 0.0875,
+      "step": 9712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5228609692059021,
+      "learning_rate": 8.610685150407376e-06,
+      "loss": 0.3247,
+      "step": 9713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3874828243771713,
+      "learning_rate": 8.608842309011224e-06,
+      "loss": 0.2822,
+      "step": 9714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9765194619225318,
+      "learning_rate": 8.606999515790801e-06,
+      "loss": 0.2146,
+      "step": 9715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4766873680274542,
+      "learning_rate": 8.605156770809926e-06,
+      "loss": 0.3851,
+      "step": 9716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33428061403293535,
+      "learning_rate": 8.603314074132411e-06,
+      "loss": 0.2666,
+      "step": 9717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29411656533246155,
+      "learning_rate": 8.60147142582207e-06,
+      "loss": 0.2137,
+      "step": 9718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45824528014285415,
+      "learning_rate": 8.599628825942713e-06,
+      "loss": 0.2705,
+      "step": 9719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37961707152946494,
+      "learning_rate": 8.597786274558152e-06,
+      "loss": 0.2672,
+      "step": 9720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.375183535077221,
+      "learning_rate": 8.595943771732187e-06,
+      "loss": 0.3041,
+      "step": 9721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2665256641960105,
+      "learning_rate": 8.594101317528634e-06,
+      "loss": 0.4257,
+      "step": 9722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3198305090286384,
+      "learning_rate": 8.59225891201129e-06,
+      "loss": 0.227,
+      "step": 9723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5603332070743713,
+      "learning_rate": 8.590416555243962e-06,
+      "loss": 0.4177,
+      "step": 9724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34251976719886146,
+      "learning_rate": 8.588574247290444e-06,
+      "loss": 0.2623,
+      "step": 9725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26791062951199285,
+      "learning_rate": 8.586731988214542e-06,
+      "loss": 0.21,
+      "step": 9726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43863608264963566,
+      "learning_rate": 8.584889778080049e-06,
+      "loss": 0.2461,
+      "step": 9727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4854419789559304,
+      "learning_rate": 8.583047616950761e-06,
+      "loss": 0.3134,
+      "step": 9728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32119730040485317,
+      "learning_rate": 8.581205504890474e-06,
+      "loss": 0.2772,
+      "step": 9729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8707909117456779,
+      "learning_rate": 8.57936344196298e-06,
+      "loss": 0.4635,
+      "step": 9730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49345414113964153,
+      "learning_rate": 8.577521428232067e-06,
+      "loss": 0.2946,
+      "step": 9731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23250652886101789,
+      "learning_rate": 8.575679463761527e-06,
+      "loss": 0.1565,
+      "step": 9732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3752357585942652,
+      "learning_rate": 8.573837548615144e-06,
+      "loss": 0.3268,
+      "step": 9733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1076609438672396,
+      "learning_rate": 8.57199568285671e-06,
+      "loss": 0.6822,
+      "step": 9734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31491087231924286,
+      "learning_rate": 8.570153866550002e-06,
+      "loss": 0.2089,
+      "step": 9735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3889797767245986,
+      "learning_rate": 8.568312099758802e-06,
+      "loss": 0.4492,
+      "step": 9736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36769358781100087,
+      "learning_rate": 8.566470382546896e-06,
+      "loss": 0.3141,
+      "step": 9737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.21757133357345376,
+      "learning_rate": 8.564628714978055e-06,
+      "loss": 0.1293,
+      "step": 9738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34371264630242915,
+      "learning_rate": 8.562787097116063e-06,
+      "loss": 0.2301,
+      "step": 9739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36254630132426113,
+      "learning_rate": 8.56094552902469e-06,
+      "loss": 0.354,
+      "step": 9740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.321207211418575,
+      "learning_rate": 8.559104010767713e-06,
+      "loss": 0.1934,
+      "step": 9741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.018284315811994,
+      "learning_rate": 8.5572625424089e-06,
+      "loss": 0.4464,
+      "step": 9742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.050265768211607,
+      "learning_rate": 8.555421124012026e-06,
+      "loss": 0.5515,
+      "step": 9743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23753029210938723,
+      "learning_rate": 8.553579755640853e-06,
+      "loss": 0.1735,
+      "step": 9744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3188849467880673,
+      "learning_rate": 8.551738437359154e-06,
+      "loss": 0.2574,
+      "step": 9745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7259221869316915,
+      "learning_rate": 8.549897169230689e-06,
+      "loss": 0.5058,
+      "step": 9746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4847398425340026,
+      "learning_rate": 8.548055951319223e-06,
+      "loss": 0.3058,
+      "step": 9747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4390574813917582,
+      "learning_rate": 8.546214783688518e-06,
+      "loss": 0.2942,
+      "step": 9748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3452402290361021,
+      "learning_rate": 8.544373666402331e-06,
+      "loss": 0.3128,
+      "step": 9749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3458191816422412,
+      "learning_rate": 8.542532599524422e-06,
+      "loss": 0.2582,
+      "step": 9750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18158561259795064,
+      "learning_rate": 8.540691583118545e-06,
+      "loss": 0.0899,
+      "step": 9751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4406492459517275,
+      "learning_rate": 8.53885061724846e-06,
+      "loss": 0.3035,
+      "step": 9752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3702396463379801,
+      "learning_rate": 8.537009701977909e-06,
+      "loss": 0.276,
+      "step": 9753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8471022600906709,
+      "learning_rate": 8.535168837370656e-06,
+      "loss": 0.357,
+      "step": 9754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8915584935820783,
+      "learning_rate": 8.533328023490438e-06,
+      "loss": 0.5318,
+      "step": 9755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3460314169075071,
+      "learning_rate": 8.531487260401009e-06,
+      "loss": 0.2315,
+      "step": 9756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2540429650530294,
+      "learning_rate": 8.529646548166113e-06,
+      "loss": 0.2456,
+      "step": 9757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8162276591758156,
+      "learning_rate": 8.527805886849496e-06,
+      "loss": 0.3404,
+      "step": 9758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6224711881741437,
+      "learning_rate": 8.525965276514897e-06,
+      "loss": 0.4223,
+      "step": 9759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41310833028155763,
+      "learning_rate": 8.524124717226057e-06,
+      "loss": 0.3289,
+      "step": 9760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3572582038454073,
+      "learning_rate": 8.522284209046713e-06,
+      "loss": 0.2494,
+      "step": 9761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3742903886740477,
+      "learning_rate": 8.520443752040604e-06,
+      "loss": 0.2831,
+      "step": 9762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2706597483671145,
+      "learning_rate": 8.518603346271463e-06,
+      "loss": 0.2184,
+      "step": 9763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38218243499476895,
+      "learning_rate": 8.516762991803027e-06,
+      "loss": 0.2738,
+      "step": 9764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4708346428419067,
+      "learning_rate": 8.51492268869902e-06,
+      "loss": 0.2916,
+      "step": 9765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5177037495364069,
+      "learning_rate": 8.513082437023182e-06,
+      "loss": 0.4067,
+      "step": 9766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8805166112100425,
+      "learning_rate": 8.511242236839232e-06,
+      "loss": 0.2842,
+      "step": 9767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32887364777368694,
+      "learning_rate": 8.509402088210901e-06,
+      "loss": 0.2746,
+      "step": 9768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26738679639436275,
+      "learning_rate": 8.507561991201908e-06,
+      "loss": 0.2453,
+      "step": 9769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41725219260940183,
+      "learning_rate": 8.505721945875985e-06,
+      "loss": 0.3137,
+      "step": 9770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3031029677592879,
+      "learning_rate": 8.503881952296842e-06,
+      "loss": 0.2186,
+      "step": 9771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.254047894329755,
+      "learning_rate": 8.502042010528205e-06,
+      "loss": 0.7768,
+      "step": 9772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3559282773742175,
+      "learning_rate": 8.50020212063379e-06,
+      "loss": 0.3021,
+      "step": 9773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3623989422829272,
+      "learning_rate": 8.498362282677308e-06,
+      "loss": 0.1237,
+      "step": 9774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33483177445849416,
+      "learning_rate": 8.496522496722476e-06,
+      "loss": 0.2634,
+      "step": 9775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32082119699761735,
+      "learning_rate": 8.494682762833004e-06,
+      "loss": 0.2741,
+      "step": 9776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3887285850178219,
+      "learning_rate": 8.492843081072609e-06,
+      "loss": 0.1544,
+      "step": 9777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4023507557929457,
+      "learning_rate": 8.491003451504987e-06,
+      "loss": 0.3336,
+      "step": 9778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8923515179116659,
+      "learning_rate": 8.489163874193854e-06,
+      "loss": 0.5183,
+      "step": 9779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32433786243686613,
+      "learning_rate": 8.487324349202909e-06,
+      "loss": 0.2238,
+      "step": 9780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3837847982567723,
+      "learning_rate": 8.485484876595859e-06,
+      "loss": 0.3084,
+      "step": 9781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45428916728106616,
+      "learning_rate": 8.4836454564364e-06,
+      "loss": 0.2587,
+      "step": 9782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5838701470386773,
+      "learning_rate": 8.481806088788235e-06,
+      "loss": 0.3887,
+      "step": 9783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23669142326985326,
+      "learning_rate": 8.47996677371506e-06,
+      "loss": 0.2022,
+      "step": 9784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4869737162516183,
+      "learning_rate": 8.478127511280571e-06,
+      "loss": 0.399,
+      "step": 9785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5174517432230594,
+      "learning_rate": 8.476288301548458e-06,
+      "loss": 0.3913,
+      "step": 9786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36426826902545983,
+      "learning_rate": 8.474449144582419e-06,
+      "loss": 0.2611,
+      "step": 9787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.339033501439424,
+      "learning_rate": 8.472610040446142e-06,
+      "loss": 0.3378,
+      "step": 9788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6482947546336639,
+      "learning_rate": 8.470770989203309e-06,
+      "loss": 0.4807,
+      "step": 9789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2700529166944222,
+      "learning_rate": 8.468931990917613e-06,
+      "loss": 0.1869,
+      "step": 9790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38726787939089174,
+      "learning_rate": 8.467093045652736e-06,
+      "loss": 0.2809,
+      "step": 9791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3210957803126272,
+      "learning_rate": 8.465254153472362e-06,
+      "loss": 0.2913,
+      "step": 9792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3114223586414909,
+      "learning_rate": 8.463415314440172e-06,
+      "loss": 0.2626,
+      "step": 9793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9172670409997502,
+      "learning_rate": 8.46157652861984e-06,
+      "loss": 0.6035,
+      "step": 9794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7608940459805851,
+      "learning_rate": 8.45973779607505e-06,
+      "loss": 0.4936,
+      "step": 9795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2258225225155695,
+      "learning_rate": 8.457899116869469e-06,
+      "loss": 0.2257,
+      "step": 9796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2627940231472568,
+      "learning_rate": 8.45606049106678e-06,
+      "loss": 0.1897,
+      "step": 9797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5623646815898292,
+      "learning_rate": 8.454221918730646e-06,
+      "loss": 0.3763,
+      "step": 9798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3464750026436955,
+      "learning_rate": 8.452383399924743e-06,
+      "loss": 0.2948,
+      "step": 9799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3381683606193286,
+      "learning_rate": 8.450544934712736e-06,
+      "loss": 0.2535,
+      "step": 9800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6811041565579721,
+      "learning_rate": 8.44870652315829e-06,
+      "loss": 0.4924,
+      "step": 9801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2585541466426452,
+      "learning_rate": 8.446868165325073e-06,
+      "loss": 0.2242,
+      "step": 9802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4408860747267185,
+      "learning_rate": 8.445029861276742e-06,
+      "loss": 0.1794,
+      "step": 9803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30415946757221135,
+      "learning_rate": 8.443191611076962e-06,
+      "loss": 0.2894,
+      "step": 9804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3788792995472617,
+      "learning_rate": 8.441353414789386e-06,
+      "loss": 0.2975,
+      "step": 9805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3016544580089682,
+      "learning_rate": 8.439515272477679e-06,
+      "loss": 0.6672,
+      "step": 9806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4190179098500511,
+      "learning_rate": 8.437677184205488e-06,
+      "loss": 0.2239,
+      "step": 9807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3089991674098943,
+      "learning_rate": 8.43583915003647e-06,
+      "loss": 0.2844,
+      "step": 9808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2639149079109884,
+      "learning_rate": 8.434001170034273e-06,
+      "loss": 0.2243,
+      "step": 9809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.556166511583863,
+      "learning_rate": 8.432163244262551e-06,
+      "loss": 0.3236,
+      "step": 9810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32418630784592684,
+      "learning_rate": 8.430325372784946e-06,
+      "loss": 0.2629,
+      "step": 9811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3574565696512591,
+      "learning_rate": 8.428487555665108e-06,
+      "loss": 0.3277,
+      "step": 9812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43378655974248065,
+      "learning_rate": 8.426649792966679e-06,
+      "loss": 0.2564,
+      "step": 9813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3671253402319884,
+      "learning_rate": 8.424812084753297e-06,
+      "loss": 0.2683,
+      "step": 9814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6749395956942922,
+      "learning_rate": 8.422974431088607e-06,
+      "loss": 0.3903,
+      "step": 9815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2000129236440116,
+      "learning_rate": 8.421136832036242e-06,
+      "loss": 0.1704,
+      "step": 9816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3516750238446244,
+      "learning_rate": 8.419299287659844e-06,
+      "loss": 0.2996,
+      "step": 9817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.055794532902847,
+      "learning_rate": 8.417461798023042e-06,
+      "loss": 0.6188,
+      "step": 9818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4547799461366989,
+      "learning_rate": 8.41562436318947e-06,
+      "loss": 0.3617,
+      "step": 9819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28796199443351,
+      "learning_rate": 8.413786983222758e-06,
+      "loss": 0.2079,
+      "step": 9820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.828814734820839,
+      "learning_rate": 8.411949658186536e-06,
+      "loss": 0.3814,
+      "step": 9821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29849571473835407,
+      "learning_rate": 8.410112388144426e-06,
+      "loss": 0.2056,
+      "step": 9822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3716093461533886,
+      "learning_rate": 8.408275173160059e-06,
+      "loss": 0.207,
+      "step": 9823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3931711629255428,
+      "learning_rate": 8.406438013297052e-06,
+      "loss": 0.3198,
+      "step": 9824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8020438520926569,
+      "learning_rate": 8.404600908619033e-06,
+      "loss": 0.5594,
+      "step": 9825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39734590436170697,
+      "learning_rate": 8.40276385918961e-06,
+      "loss": 0.2346,
+      "step": 9826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8506087767154734,
+      "learning_rate": 8.40092686507241e-06,
+      "loss": 0.4187,
+      "step": 9827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.255074633086988,
+      "learning_rate": 8.39908992633104e-06,
+      "loss": 0.2158,
+      "step": 9828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31639947374690586,
+      "learning_rate": 8.39725304302912e-06,
+      "loss": 0.1998,
+      "step": 9829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2054591386012041,
+      "learning_rate": 8.395416215230255e-06,
+      "loss": 0.7528,
+      "step": 9830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4489425221494689,
+      "learning_rate": 8.39357944299806e-06,
+      "loss": 0.3473,
+      "step": 9831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3019501530801851,
+      "learning_rate": 8.391742726396138e-06,
+      "loss": 0.2471,
+      "step": 9832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6973811792398468,
+      "learning_rate": 8.389906065488099e-06,
+      "loss": 0.3255,
+      "step": 9833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28463618303054367,
+      "learning_rate": 8.38806946033754e-06,
+      "loss": 0.1894,
+      "step": 9834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35151178079415824,
+      "learning_rate": 8.386232911008069e-06,
+      "loss": 0.2611,
+      "step": 9835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40041939562222845,
+      "learning_rate": 8.38439641756328e-06,
+      "loss": 0.2496,
+      "step": 9836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.876402219608769,
+      "learning_rate": 8.382559980066778e-06,
+      "loss": 0.4565,
+      "step": 9837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31643695032595764,
+      "learning_rate": 8.380723598582152e-06,
+      "loss": 0.2451,
+      "step": 9838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9533380440004469,
+      "learning_rate": 8.378887273172997e-06,
+      "loss": 0.3012,
+      "step": 9839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20603783465622966,
+      "learning_rate": 8.37705100390291e-06,
+      "loss": 0.2236,
+      "step": 9840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37732886480516004,
+      "learning_rate": 8.375214790835471e-06,
+      "loss": 0.2641,
+      "step": 9841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48964154358706147,
+      "learning_rate": 8.37337863403428e-06,
+      "loss": 0.2918,
+      "step": 9842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4250539613089976,
+      "learning_rate": 8.371542533562912e-06,
+      "loss": 0.3089,
+      "step": 9843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35552109627010425,
+      "learning_rate": 8.369706489484958e-06,
+      "loss": 0.2745,
+      "step": 9844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9900011631599458,
+      "learning_rate": 8.367870501863999e-06,
+      "loss": 0.4337,
+      "step": 9845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18633542192754032,
+      "learning_rate": 8.366034570763614e-06,
+      "loss": 0.0689,
+      "step": 9846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3414535184442479,
+      "learning_rate": 8.36419869624738e-06,
+      "loss": 0.2813,
+      "step": 9847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37358328219071146,
+      "learning_rate": 8.362362878378876e-06,
+      "loss": 0.329,
+      "step": 9848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8248357384042281,
+      "learning_rate": 8.360527117221675e-06,
+      "loss": 0.3291,
+      "step": 9849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36147984520054255,
+      "learning_rate": 8.358691412839351e-06,
+      "loss": 0.3034,
+      "step": 9850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5104748583363516,
+      "learning_rate": 8.35685576529547e-06,
+      "loss": 0.3727,
+      "step": 9851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40198951064850447,
+      "learning_rate": 8.355020174653605e-06,
+      "loss": 0.2165,
+      "step": 9852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24647184833766747,
+      "learning_rate": 8.35318464097732e-06,
+      "loss": 0.2172,
+      "step": 9853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3810295500607417,
+      "learning_rate": 8.35134916433018e-06,
+      "loss": 0.2782,
+      "step": 9854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3857340949697676,
+      "learning_rate": 8.349513744775748e-06,
+      "loss": 0.289,
+      "step": 9855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34685648534680974,
+      "learning_rate": 8.347678382377584e-06,
+      "loss": 0.2889,
+      "step": 9856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0213706600766048,
+      "learning_rate": 8.345843077199247e-06,
+      "loss": 0.604,
+      "step": 9857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.43460712379136257,
+      "learning_rate": 8.344007829304291e-06,
+      "loss": 0.2663,
+      "step": 9858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2396752936639507,
+      "learning_rate": 8.342172638756276e-06,
+      "loss": 0.1397,
+      "step": 9859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3169017785415661,
+      "learning_rate": 8.34033750561875e-06,
+      "loss": 0.29,
+      "step": 9860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9931863567969424,
+      "learning_rate": 8.338502429955264e-06,
+      "loss": 0.4568,
+      "step": 9861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3480195681104575,
+      "learning_rate": 8.33666741182937e-06,
+      "loss": 0.1948,
+      "step": 9862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37565505477248806,
+      "learning_rate": 8.334832451304607e-06,
+      "loss": 0.3106,
+      "step": 9863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.45130905555980977,
+      "learning_rate": 8.332997548444528e-06,
+      "loss": 0.3433,
+      "step": 9864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3033398640505855,
+      "learning_rate": 8.331162703312671e-06,
+      "loss": 0.1934,
+      "step": 9865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5363775633765079,
+      "learning_rate": 8.329327915972578e-06,
+      "loss": 0.2919,
+      "step": 9866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.747351620662827,
+      "learning_rate": 8.32749318648779e-06,
+      "loss": 0.3933,
+      "step": 9867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26553280669644175,
+      "learning_rate": 8.325658514921838e-06,
+      "loss": 0.211,
+      "step": 9868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4145040507342814,
+      "learning_rate": 8.32382390133826e-06,
+      "loss": 0.227,
+      "step": 9869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1950814437419346,
+      "learning_rate": 8.321989345800587e-06,
+      "loss": 0.6061,
+      "step": 9870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3002830485967934,
+      "learning_rate": 8.320154848372353e-06,
+      "loss": 0.2718,
+      "step": 9871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3226666488160495,
+      "learning_rate": 8.318320409117082e-06,
+      "loss": 0.2688,
+      "step": 9872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37089315357089786,
+      "learning_rate": 8.316486028098306e-06,
+      "loss": 0.2992,
+      "step": 9873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29002209727779943,
+      "learning_rate": 8.314651705379544e-06,
+      "loss": 0.2231,
+      "step": 9874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3455087689241484,
+      "learning_rate": 8.312817441024324e-06,
+      "loss": 0.2336,
+      "step": 9875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5069761719351095,
+      "learning_rate": 8.31098323509616e-06,
+      "loss": 0.3668,
+      "step": 9876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38171523506523025,
+      "learning_rate": 8.309149087658576e-06,
+      "loss": 0.278,
+      "step": 9877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37069169080183434,
+      "learning_rate": 8.307314998775087e-06,
+      "loss": 0.2917,
+      "step": 9878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3392341659426202,
+      "learning_rate": 8.305480968509204e-06,
+      "loss": 0.3168,
+      "step": 9879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5341042702280032,
+      "learning_rate": 8.303646996924445e-06,
+      "loss": 0.3165,
+      "step": 9880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2169102274872214,
+      "learning_rate": 8.301813084084315e-06,
+      "loss": 0.1475,
+      "step": 9881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5998300732121216,
+      "learning_rate": 8.299979230052327e-06,
+      "loss": 0.4361,
+      "step": 9882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31913217855811254,
+      "learning_rate": 8.298145434891983e-06,
+      "loss": 0.2759,
+      "step": 9883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3384502713432209,
+      "learning_rate": 8.296311698666792e-06,
+      "loss": 0.3219,
+      "step": 9884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7524840483991162,
+      "learning_rate": 8.29447802144025e-06,
+      "loss": 0.4665,
+      "step": 9885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2888469947900723,
+      "learning_rate": 8.292644403275865e-06,
+      "loss": 0.2242,
+      "step": 9886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2532786012933536,
+      "learning_rate": 8.290810844237128e-06,
+      "loss": 0.2485,
+      "step": 9887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7263770853364647,
+      "learning_rate": 8.28897734438754e-06,
+      "loss": 0.2903,
+      "step": 9888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3729535266775858,
+      "learning_rate": 8.28714390379059e-06,
+      "loss": 0.2678,
+      "step": 9889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4971089032472461,
+      "learning_rate": 8.285310522509777e-06,
+      "loss": 0.3529,
+      "step": 9890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32657011764466537,
+      "learning_rate": 8.283477200608585e-06,
+      "loss": 0.2734,
+      "step": 9891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4085045169813695,
+      "learning_rate": 8.281643938150504e-06,
+      "loss": 0.2973,
+      "step": 9892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.25379677627879427,
+      "learning_rate": 8.27981073519902e-06,
+      "loss": 0.1814,
+      "step": 9893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8729493209016004,
+      "learning_rate": 8.277977591817617e-06,
+      "loss": 0.3863,
+      "step": 9894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30599910372112143,
+      "learning_rate": 8.276144508069775e-06,
+      "loss": 0.2213,
+      "step": 9895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38495626365299385,
+      "learning_rate": 8.274311484018975e-06,
+      "loss": 0.3128,
+      "step": 9896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9544093017904444,
+      "learning_rate": 8.27247851972869e-06,
+      "loss": 0.6904,
+      "step": 9897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4332396370100671,
+      "learning_rate": 8.270645615262405e-06,
+      "loss": 0.1635,
+      "step": 9898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.21958238359771295,
+      "learning_rate": 8.268812770683583e-06,
+      "loss": 0.2311,
+      "step": 9899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7528076032824318,
+      "learning_rate": 8.266979986055704e-06,
+      "loss": 0.3034,
+      "step": 9900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2277184270609133,
+      "learning_rate": 8.265147261442232e-06,
+      "loss": 0.0727,
+      "step": 9901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3900048604849335,
+      "learning_rate": 8.263314596906636e-06,
+      "loss": 0.3282,
+      "step": 9902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3516000182623754,
+      "learning_rate": 8.261481992512382e-06,
+      "loss": 0.3266,
+      "step": 9903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41670733313075314,
+      "learning_rate": 8.25964944832293e-06,
+      "loss": 0.1948,
+      "step": 9904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.309400465234977,
+      "learning_rate": 8.257816964401745e-06,
+      "loss": 0.2583,
+      "step": 9905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4465142891914701,
+      "learning_rate": 8.255984540812281e-06,
+      "loss": 0.3008,
+      "step": 9906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28221141190811483,
+      "learning_rate": 8.254152177618e-06,
+      "loss": 0.2664,
+      "step": 9907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3505287455300988,
+      "learning_rate": 8.252319874882351e-06,
+      "loss": 0.2389,
+      "step": 9908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0994942283600821,
+      "learning_rate": 8.250487632668793e-06,
+      "loss": 0.7445,
+      "step": 9909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.522218227682,
+      "learning_rate": 8.248655451040768e-06,
+      "loss": 0.2764,
+      "step": 9910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2631576808456909,
+      "learning_rate": 8.246823330061734e-06,
+      "loss": 0.2347,
+      "step": 9911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40714845276580147,
+      "learning_rate": 8.24499126979513e-06,
+      "loss": 0.281,
+      "step": 9912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28989171280742454,
+      "learning_rate": 8.243159270304406e-06,
+      "loss": 0.1528,
+      "step": 9913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3758477950586018,
+      "learning_rate": 8.241327331652997e-06,
+      "loss": 0.2369,
+      "step": 9914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35371479814125156,
+      "learning_rate": 8.23949545390435e-06,
+      "loss": 0.3125,
+      "step": 9915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7429248777607208,
+      "learning_rate": 8.237663637121897e-06,
+      "loss": 0.4045,
+      "step": 9916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32476985877232545,
+      "learning_rate": 8.23583188136908e-06,
+      "loss": 0.2445,
+      "step": 9917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2535536568991639,
+      "learning_rate": 8.234000186709327e-06,
+      "loss": 0.2115,
+      "step": 9918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.424295206181053,
+      "learning_rate": 8.232168553206072e-06,
+      "loss": 0.3238,
+      "step": 9919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32505346775808885,
+      "learning_rate": 8.230336980922744e-06,
+      "loss": 0.2858,
+      "step": 9920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9237249641595617,
+      "learning_rate": 8.228505469922769e-06,
+      "loss": 0.4477,
+      "step": 9921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4630004104417319,
+      "learning_rate": 8.226674020269576e-06,
+      "loss": 0.3362,
+      "step": 9922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2790719952889474,
+      "learning_rate": 8.224842632026583e-06,
+      "loss": 0.2751,
+      "step": 9923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7399588708936765,
+      "learning_rate": 8.223011305257214e-06,
+      "loss": 0.2648,
+      "step": 9924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24700257914947743,
+      "learning_rate": 8.221180040024887e-06,
+      "loss": 0.1378,
+      "step": 9925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36109041110206347,
+      "learning_rate": 8.21934883639302e-06,
+      "loss": 0.2974,
+      "step": 9926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36150839242761174,
+      "learning_rate": 8.217517694425027e-06,
+      "loss": 0.2756,
+      "step": 9927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5338739079601947,
+      "learning_rate": 8.215686614184317e-06,
+      "loss": 0.3363,
+      "step": 9928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3856577519384225,
+      "learning_rate": 8.213855595734306e-06,
+      "loss": 0.3088,
+      "step": 9929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4388607395567644,
+      "learning_rate": 8.212024639138398e-06,
+      "loss": 0.2918,
+      "step": 9930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22893949698281893,
+      "learning_rate": 8.210193744459997e-06,
+      "loss": 0.1762,
+      "step": 9931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38411850632766775,
+      "learning_rate": 8.20836291176251e-06,
+      "loss": 0.3018,
+      "step": 9932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7225551464888974,
+      "learning_rate": 8.20653214110934e-06,
+      "loss": 0.5503,
+      "step": 9933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4470092332721713,
+      "learning_rate": 8.204701432563886e-06,
+      "loss": 0.2851,
+      "step": 9934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28828857751457004,
+      "learning_rate": 8.202870786189541e-06,
+      "loss": 0.2635,
+      "step": 9935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9999090217246076,
+      "learning_rate": 8.201040202049705e-06,
+      "loss": 0.6211,
+      "step": 9936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1682762823248837,
+      "learning_rate": 8.199209680207768e-06,
+      "loss": 0.0922,
+      "step": 9937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35653206640661295,
+      "learning_rate": 8.197379220727124e-06,
+      "loss": 0.2672,
+      "step": 9938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34611855850166906,
+      "learning_rate": 8.19554882367116e-06,
+      "loss": 0.3211,
+      "step": 9939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7223852385709048,
+      "learning_rate": 8.193718489103261e-06,
+      "loss": 0.3111,
+      "step": 9940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3878265662993998,
+      "learning_rate": 8.191888217086813e-06,
+      "loss": 0.3133,
+      "step": 9941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9687515986579172,
+      "learning_rate": 8.190058007685203e-06,
+      "loss": 0.5916,
+      "step": 9942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.21520660321755714,
+      "learning_rate": 8.188227860961804e-06,
+      "loss": 0.1676,
+      "step": 9943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.343814550581402,
+      "learning_rate": 8.186397776979992e-06,
+      "loss": 0.2711,
+      "step": 9944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9159296933741887,
+      "learning_rate": 8.184567755803153e-06,
+      "loss": 0.4973,
+      "step": 9945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38092338926104646,
+      "learning_rate": 8.18273779749465e-06,
+      "loss": 0.3144,
+      "step": 9946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29189600354117673,
+      "learning_rate": 8.180907902117862e-06,
+      "loss": 0.1929,
+      "step": 9947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1424900738622954,
+      "learning_rate": 8.179078069736152e-06,
+      "loss": 0.6561,
+      "step": 9948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22268099279160083,
+      "learning_rate": 8.177248300412893e-06,
+      "loss": 0.1552,
+      "step": 9949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3583785102766591,
+      "learning_rate": 8.175418594211445e-06,
+      "loss": 0.2161,
+      "step": 9950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5628579774742724,
+      "learning_rate": 8.173588951195175e-06,
+      "loss": 0.3217,
+      "step": 9951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9119904572848944,
+      "learning_rate": 8.171759371427439e-06,
+      "loss": 0.492,
+      "step": 9952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2712431380235841,
+      "learning_rate": 8.169929854971598e-06,
+      "loss": 0.2144,
+      "step": 9953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41873206115862327,
+      "learning_rate": 8.168100401891007e-06,
+      "loss": 0.3233,
+      "step": 9954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41447916444623933,
+      "learning_rate": 8.166271012249022e-06,
+      "loss": 0.26,
+      "step": 9955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2745882143555028,
+      "learning_rate": 8.164441686108991e-06,
+      "loss": 0.1833,
+      "step": 9956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8934275316001319,
+      "learning_rate": 8.162612423534266e-06,
+      "loss": 0.4585,
+      "step": 9957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37576440877617384,
+      "learning_rate": 8.160783224588196e-06,
+      "loss": 0.3174,
+      "step": 9958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34228438562146457,
+      "learning_rate": 8.15895408933412e-06,
+      "loss": 0.2592,
+      "step": 9959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0015700431321963,
+      "learning_rate": 8.157125017835389e-06,
+      "loss": 0.2788,
+      "step": 9960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3365784712512407,
+      "learning_rate": 8.15529601015534e-06,
+      "loss": 0.2207,
+      "step": 9961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3051365494564319,
+      "learning_rate": 8.153467066357305e-06,
+      "loss": 0.2439,
+      "step": 9962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4885135862243723,
+      "learning_rate": 8.15163818650463e-06,
+      "loss": 0.3013,
+      "step": 9963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38036239490366125,
+      "learning_rate": 8.149809370660643e-06,
+      "loss": 0.2774,
+      "step": 9964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.39821866202603395,
+      "learning_rate": 8.14798061888868e-06,
+      "loss": 0.2925,
+      "step": 9965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3459147936489671,
+      "learning_rate": 8.146151931252067e-06,
+      "loss": 0.2577,
+      "step": 9966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.590203539009197,
+      "learning_rate": 8.144323307814133e-06,
+      "loss": 0.3359,
+      "step": 9967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3644950630409327,
+      "learning_rate": 8.142494748638204e-06,
+      "loss": 0.2875,
+      "step": 9968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38221492079631614,
+      "learning_rate": 8.140666253787602e-06,
+      "loss": 0.2538,
+      "step": 9969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.399983798988496,
+      "learning_rate": 8.138837823325647e-06,
+      "loss": 0.3374,
+      "step": 9970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35302886052202936,
+      "learning_rate": 8.137009457315658e-06,
+      "loss": 0.2862,
+      "step": 9971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40507971925164943,
+      "learning_rate": 8.135181155820953e-06,
+      "loss": 0.2151,
+      "step": 9972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5666935537735343,
+      "learning_rate": 8.13335291890484e-06,
+      "loss": 0.3137,
+      "step": 9973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32067561902428615,
+      "learning_rate": 8.13152474663064e-06,
+      "loss": 0.2686,
+      "step": 9974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33814247044172163,
+      "learning_rate": 8.129696639061654e-06,
+      "loss": 0.286,
+      "step": 9975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3246915939454487,
+      "learning_rate": 8.127868596261198e-06,
+      "loss": 0.1764,
+      "step": 9976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26615674398341954,
+      "learning_rate": 8.126040618292566e-06,
+      "loss": 0.2069,
+      "step": 9977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4585736080937011,
+      "learning_rate": 8.124212705219071e-06,
+      "loss": 0.3394,
+      "step": 9978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4703515008785008,
+      "learning_rate": 8.122384857104006e-06,
+      "loss": 0.2791,
+      "step": 9979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3081159193556577,
+      "learning_rate": 8.120557074010677e-06,
+      "loss": 0.251,
+      "step": 9980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0902723825131568,
+      "learning_rate": 8.118729356002371e-06,
+      "loss": 0.6268,
+      "step": 9981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3609668953923149,
+      "learning_rate": 8.11690170314239e-06,
+      "loss": 0.3442,
+      "step": 9982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23808650024305908,
+      "learning_rate": 8.115074115494022e-06,
+      "loss": 0.1346,
+      "step": 9983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34721727640723965,
+      "learning_rate": 8.113246593120554e-06,
+      "loss": 0.2742,
+      "step": 9984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8939576818059674,
+      "learning_rate": 8.111419136085278e-06,
+      "loss": 0.4012,
+      "step": 9985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37324261637531364,
+      "learning_rate": 8.109591744451472e-06,
+      "loss": 0.2016,
+      "step": 9986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.407045014028079,
+      "learning_rate": 8.107764418282427e-06,
+      "loss": 0.3083,
+      "step": 9987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0490869448387106,
+      "learning_rate": 8.105937157641416e-06,
+      "loss": 0.6916,
+      "step": 9988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2294748663492301,
+      "learning_rate": 8.104109962591722e-06,
+      "loss": 0.1641,
+      "step": 9989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28437388048864876,
+      "learning_rate": 8.102282833196616e-06,
+      "loss": 0.2555,
+      "step": 9990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.774892354817154,
+      "learning_rate": 8.100455769519377e-06,
+      "loss": 0.4188,
+      "step": 9991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29569982792830685,
+      "learning_rate": 8.09862877162327e-06,
+      "loss": 0.1961,
+      "step": 9992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0072369091205406,
+      "learning_rate": 8.096801839571569e-06,
+      "loss": 0.6177,
+      "step": 9993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3855864777216197,
+      "learning_rate": 8.094974973427541e-06,
+      "loss": 0.3251,
+      "step": 9994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.39142238746136504,
+      "learning_rate": 8.093148173254445e-06,
+      "loss": 0.2776,
+      "step": 9995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23644514454316207,
+      "learning_rate": 8.091321439115543e-06,
+      "loss": 0.1532,
+      "step": 9996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7087803679347151,
+      "learning_rate": 8.089494771074102e-06,
+      "loss": 0.4177,
+      "step": 9997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2797651307049413,
+      "learning_rate": 8.08766816919337e-06,
+      "loss": 0.2555,
+      "step": 9998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4959220267266461,
+      "learning_rate": 8.085841633536611e-06,
+      "loss": 0.2582,
+      "step": 9999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0009644158738145,
+      "learning_rate": 8.084015164167071e-06,
+      "loss": 0.6752,
+      "step": 10000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32313335118258313,
+      "learning_rate": 8.082188761148007e-06,
+      "loss": 0.1926,
+      "step": 10001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2669066233121326,
+      "learning_rate": 8.08036242454266e-06,
+      "loss": 0.2554,
+      "step": 10002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4574088631849031,
+      "learning_rate": 8.078536154414283e-06,
+      "loss": 0.2827,
+      "step": 10003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.507204777496856,
+      "learning_rate": 8.076709950826113e-06,
+      "loss": 0.2304,
+      "step": 10004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38781786620956266,
+      "learning_rate": 8.074883813841397e-06,
+      "loss": 0.2686,
+      "step": 10005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36301702077513615,
+      "learning_rate": 8.073057743523371e-06,
+      "loss": 0.3049,
+      "step": 10006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4861354405825026,
+      "learning_rate": 8.071231739935272e-06,
+      "loss": 0.3207,
+      "step": 10007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35373086697907874,
+      "learning_rate": 8.069405803140338e-06,
+      "loss": 0.3251,
+      "step": 10008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2171049695260447,
+      "learning_rate": 8.067579933201793e-06,
+      "loss": 0.0853,
+      "step": 10009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26086832974147256,
+      "learning_rate": 8.065754130182876e-06,
+      "loss": 0.2698,
+      "step": 10010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5370388860704075,
+      "learning_rate": 8.063928394146806e-06,
+      "loss": 0.4125,
+      "step": 10011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7740071781472809,
+      "learning_rate": 8.062102725156818e-06,
+      "loss": 0.3091,
+      "step": 10012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36948235193836254,
+      "learning_rate": 8.060277123276125e-06,
+      "loss": 0.3018,
+      "step": 10013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40449012511592974,
+      "learning_rate": 8.058451588567954e-06,
+      "loss": 0.3211,
+      "step": 10014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23658106450923316,
+      "learning_rate": 8.05662612109552e-06,
+      "loss": 0.1578,
+      "step": 10015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3824604885675158,
+      "learning_rate": 8.05480072092204e-06,
+      "loss": 0.2864,
+      "step": 10016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5705382545976666,
+      "learning_rate": 8.052975388110726e-06,
+      "loss": 0.3922,
+      "step": 10017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3746308523465573,
+      "learning_rate": 8.051150122724793e-06,
+      "loss": 0.2776,
+      "step": 10018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6099067258468585,
+      "learning_rate": 8.049324924827447e-06,
+      "loss": 0.3158,
+      "step": 10019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38826376456495426,
+      "learning_rate": 8.047499794481894e-06,
+      "loss": 0.3084,
+      "step": 10020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24314199497277397,
+      "learning_rate": 8.045674731751338e-06,
+      "loss": 0.2092,
+      "step": 10021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31113245462098016,
+      "learning_rate": 8.043849736698986e-06,
+      "loss": 0.1859,
+      "step": 10022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37576660120750494,
+      "learning_rate": 8.04202480938803e-06,
+      "loss": 0.3239,
+      "step": 10023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7344082199207784,
+      "learning_rate": 8.040199949881672e-06,
+      "loss": 0.4354,
+      "step": 10024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45688911337935884,
+      "learning_rate": 8.038375158243108e-06,
+      "loss": 0.2071,
+      "step": 10025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2761160102991935,
+      "learning_rate": 8.036550434535522e-06,
+      "loss": 0.2761,
+      "step": 10026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.061528670972983,
+      "learning_rate": 8.034725778822114e-06,
+      "loss": 0.5102,
+      "step": 10027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.16482060521116781,
+      "learning_rate": 8.032901191166071e-06,
+      "loss": 0.0969,
+      "step": 10028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30696431779330907,
+      "learning_rate": 8.03107667163057e-06,
+      "loss": 0.2962,
+      "step": 10029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5561323310489525,
+      "learning_rate": 8.029252220278802e-06,
+      "loss": 0.3496,
+      "step": 10030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45238876227242075,
+      "learning_rate": 8.02742783717394e-06,
+      "loss": 0.2456,
+      "step": 10031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5396238256737506,
+      "learning_rate": 8.025603522379172e-06,
+      "loss": 0.3749,
+      "step": 10032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3195175238113092,
+      "learning_rate": 8.023779275957668e-06,
+      "loss": 0.2389,
+      "step": 10033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24479568402304927,
+      "learning_rate": 8.021955097972602e-06,
+      "loss": 0.2131,
+      "step": 10034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3826033703959546,
+      "learning_rate": 8.020130988487146e-06,
+      "loss": 0.2464,
+      "step": 10035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7693896830289918,
+      "learning_rate": 8.01830694756447e-06,
+      "loss": 0.4276,
+      "step": 10036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3963773983506641,
+      "learning_rate": 8.016482975267738e-06,
+      "loss": 0.2943,
+      "step": 10037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28363379615809237,
+      "learning_rate": 8.014659071660113e-06,
+      "loss": 0.2321,
+      "step": 10038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41409455127608996,
+      "learning_rate": 8.012835236804764e-06,
+      "loss": 0.2602,
+      "step": 10039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33374553909505866,
+      "learning_rate": 8.01101147076484e-06,
+      "loss": 0.2047,
+      "step": 10040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3514808777272365,
+      "learning_rate": 8.009187773603508e-06,
+      "loss": 0.246,
+      "step": 10041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3632666739681491,
+      "learning_rate": 8.007364145383914e-06,
+      "loss": 0.3166,
+      "step": 10042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.809247352433276,
+      "learning_rate": 8.005540586169216e-06,
+      "loss": 0.4513,
+      "step": 10043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31999804489355055,
+      "learning_rate": 8.003717096022561e-06,
+      "loss": 0.2042,
+      "step": 10044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45816885307178257,
+      "learning_rate": 8.001893675007098e-06,
+      "loss": 0.3312,
+      "step": 10045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.218345571007771,
+      "learning_rate": 8.00007032318597e-06,
+      "loss": 0.1845,
+      "step": 10046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.337060392878494,
+      "learning_rate": 7.99824704062232e-06,
+      "loss": 0.2972,
+      "step": 10047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7301838529307065,
+      "learning_rate": 7.996423827379292e-06,
+      "loss": 0.3609,
+      "step": 10048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32799631732078843,
+      "learning_rate": 7.994600683520018e-06,
+      "loss": 0.3072,
+      "step": 10049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4005912109187204,
+      "learning_rate": 7.992777609107638e-06,
+      "loss": 0.3005,
+      "step": 10050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8157284221590041,
+      "learning_rate": 7.99095460420528e-06,
+      "loss": 0.3047,
+      "step": 10051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3615354345631635,
+      "learning_rate": 7.989131668876081e-06,
+      "loss": 0.2347,
+      "step": 10052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3831280211895221,
+      "learning_rate": 7.987308803183164e-06,
+      "loss": 0.3028,
+      "step": 10053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3977051857632209,
+      "learning_rate": 7.985486007189658e-06,
+      "loss": 0.2917,
+      "step": 10054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3577210765059678,
+      "learning_rate": 7.983663280958682e-06,
+      "loss": 0.2307,
+      "step": 10055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37861640099556787,
+      "learning_rate": 7.981840624553364e-06,
+      "loss": 0.2822,
+      "step": 10056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3719306516421558,
+      "learning_rate": 7.980018038036815e-06,
+      "loss": 0.313,
+      "step": 10057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7053848501492725,
+      "learning_rate": 7.978195521472157e-06,
+      "loss": 0.3032,
+      "step": 10058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3060816409264136,
+      "learning_rate": 7.976373074922498e-06,
+      "loss": 0.2589,
+      "step": 10059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7008923801011527,
+      "learning_rate": 7.974550698450956e-06,
+      "loss": 0.5585,
+      "step": 10060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22270529904066977,
+      "learning_rate": 7.972728392120634e-06,
+      "loss": 0.1837,
+      "step": 10061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2558814298202647,
+      "learning_rate": 7.970906155994646e-06,
+      "loss": 0.2174,
+      "step": 10062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0344165023559295,
+      "learning_rate": 7.969083990136084e-06,
+      "loss": 0.5085,
+      "step": 10063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7397533513729492,
+      "learning_rate": 7.967261894608058e-06,
+      "loss": 0.3023,
+      "step": 10064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.298771462575836,
+      "learning_rate": 7.965439869473664e-06,
+      "loss": 0.2718,
+      "step": 10065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.492680151076164,
+      "learning_rate": 7.963617914796002e-06,
+      "loss": 0.3987,
+      "step": 10066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17351851611022887,
+      "learning_rate": 7.961796030638162e-06,
+      "loss": 0.096,
+      "step": 10067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34316673447226265,
+      "learning_rate": 7.95997421706324e-06,
+      "loss": 0.2893,
+      "step": 10068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3809433763662325,
+      "learning_rate": 7.95815247413432e-06,
+      "loss": 0.3244,
+      "step": 10069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8279519992417633,
+      "learning_rate": 7.956330801914495e-06,
+      "loss": 0.3853,
+      "step": 10070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3236230187941723,
+      "learning_rate": 7.954509200466845e-06,
+      "loss": 0.2006,
+      "step": 10071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9845789891659482,
+      "learning_rate": 7.952687669854453e-06,
+      "loss": 0.7532,
+      "step": 10072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2719432805518241,
+      "learning_rate": 7.950866210140401e-06,
+      "loss": 0.2321,
+      "step": 10073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25669260749844697,
+      "learning_rate": 7.949044821387761e-06,
+      "loss": 0.1774,
+      "step": 10074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8589351191506319,
+      "learning_rate": 7.947223503659613e-06,
+      "loss": 0.4398,
+      "step": 10075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8104627329921519,
+      "learning_rate": 7.945402257019026e-06,
+      "loss": 0.4334,
+      "step": 10076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28954849191593923,
+      "learning_rate": 7.943581081529072e-06,
+      "loss": 0.2135,
+      "step": 10077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5205399924008268,
+      "learning_rate": 7.941759977252815e-06,
+      "loss": 0.4581,
+      "step": 10078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28954303976259166,
+      "learning_rate": 7.939938944253321e-06,
+      "loss": 0.1576,
+      "step": 10079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3201876263622221,
+      "learning_rate": 7.938117982593653e-06,
+      "loss": 0.203,
+      "step": 10080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6596627329890012,
+      "learning_rate": 7.936297092336872e-06,
+      "loss": 0.3682,
+      "step": 10081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5906861427138822,
+      "learning_rate": 7.934476273546032e-06,
+      "loss": 0.3605,
+      "step": 10082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30195190043595455,
+      "learning_rate": 7.932655526284192e-06,
+      "loss": 0.2666,
+      "step": 10083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.992932228624963,
+      "learning_rate": 7.930834850614399e-06,
+      "loss": 0.5499,
+      "step": 10084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28475196579428813,
+      "learning_rate": 7.92901424659971e-06,
+      "loss": 0.2582,
+      "step": 10085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32554733501197314,
+      "learning_rate": 7.927193714303166e-06,
+      "loss": 0.2448,
+      "step": 10086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38443162661917063,
+      "learning_rate": 7.925373253787817e-06,
+      "loss": 0.2281,
+      "step": 10087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4907863733563514,
+      "learning_rate": 7.923552865116701e-06,
+      "loss": 0.3353,
+      "step": 10088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39379032559924637,
+      "learning_rate": 7.92173254835286e-06,
+      "loss": 0.2464,
+      "step": 10089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37205057590132345,
+      "learning_rate": 7.919912303559334e-06,
+      "loss": 0.2769,
+      "step": 10090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37416042075941736,
+      "learning_rate": 7.91809213079915e-06,
+      "loss": 0.2306,
+      "step": 10091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36138527919034713,
+      "learning_rate": 7.916272030135353e-06,
+      "loss": 0.2915,
+      "step": 10092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3876317520648567,
+      "learning_rate": 7.91445200163096e-06,
+      "loss": 0.2792,
+      "step": 10093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3951992348618812,
+      "learning_rate": 7.912632045349008e-06,
+      "loss": 0.2647,
+      "step": 10094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33494661203789694,
+      "learning_rate": 7.910812161352517e-06,
+      "loss": 0.2683,
+      "step": 10095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9173367452830046,
+      "learning_rate": 7.908992349704515e-06,
+      "loss": 0.7132,
+      "step": 10096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3487243075292584,
+      "learning_rate": 7.907172610468015e-06,
+      "loss": 0.2215,
+      "step": 10097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32608359551160127,
+      "learning_rate": 7.905352943706035e-06,
+      "loss": 0.2646,
+      "step": 10098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42778870571825733,
+      "learning_rate": 7.903533349481596e-06,
+      "loss": 0.2771,
+      "step": 10099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29251505997739846,
+      "learning_rate": 7.901713827857705e-06,
+      "loss": 0.1692,
+      "step": 10100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28561704547379524,
+      "learning_rate": 7.899894378897374e-06,
+      "loss": 0.275,
+      "step": 10101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1525092467625975,
+      "learning_rate": 7.898075002663612e-06,
+      "loss": 0.7976,
+      "step": 10102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6225456716829496,
+      "learning_rate": 7.89625569921942e-06,
+      "loss": 0.335,
+      "step": 10103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3281355566608314,
+      "learning_rate": 7.894436468627804e-06,
+      "loss": 0.2564,
+      "step": 10104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36381880629843233,
+      "learning_rate": 7.892617310951761e-06,
+      "loss": 0.3175,
+      "step": 10105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2159222699318665,
+      "learning_rate": 7.890798226254291e-06,
+      "loss": 0.1054,
+      "step": 10106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35946385596297437,
+      "learning_rate": 7.888979214598387e-06,
+      "loss": 0.2295,
+      "step": 10107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4936241577077283,
+      "learning_rate": 7.887160276047045e-06,
+      "loss": 0.376,
+      "step": 10108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33073585211025036,
+      "learning_rate": 7.885341410663248e-06,
+      "loss": 0.2954,
+      "step": 10109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34079160331467057,
+      "learning_rate": 7.883522618509989e-06,
+      "loss": 0.2056,
+      "step": 10110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.351615537794108,
+      "learning_rate": 7.881703899650249e-06,
+      "loss": 0.2591,
+      "step": 10111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39871038149815147,
+      "learning_rate": 7.879885254147014e-06,
+      "loss": 0.2566,
+      "step": 10112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24185991194383902,
+      "learning_rate": 7.878066682063262e-06,
+      "loss": 0.2185,
+      "step": 10113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.558532170715619,
+      "learning_rate": 7.876248183461967e-06,
+      "loss": 0.7215,
+      "step": 10114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6384019101854085,
+      "learning_rate": 7.874429758406108e-06,
+      "loss": 0.3869,
+      "step": 10115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3028402041371424,
+      "learning_rate": 7.872611406958653e-06,
+      "loss": 0.2154,
+      "step": 10116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3622538842476559,
+      "learning_rate": 7.870793129182577e-06,
+      "loss": 0.2996,
+      "step": 10117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27173806317093663,
+      "learning_rate": 7.86897492514084e-06,
+      "loss": 0.1832,
+      "step": 10118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30861485287327556,
+      "learning_rate": 7.86715679489641e-06,
+      "loss": 0.2098,
+      "step": 10119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2068126960800758,
+      "learning_rate": 7.86533873851225e-06,
+      "loss": 0.6609,
+      "step": 10120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36384977954229525,
+      "learning_rate": 7.863520756051317e-06,
+      "loss": 0.3154,
+      "step": 10121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4049015882904989,
+      "learning_rate": 7.861702847576568e-06,
+      "loss": 0.293,
+      "step": 10122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5157148425875564,
+      "learning_rate": 7.859885013150959e-06,
+      "loss": 0.2408,
+      "step": 10123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23292577160493733,
+      "learning_rate": 7.858067252837437e-06,
+      "loss": 0.2067,
+      "step": 10124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5576874400588665,
+      "learning_rate": 7.856249566698957e-06,
+      "loss": 0.3294,
+      "step": 10125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.331359061847375,
+      "learning_rate": 7.854431954798463e-06,
+      "loss": 0.2648,
+      "step": 10126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5790983417813569,
+      "learning_rate": 7.852614417198894e-06,
+      "loss": 0.3896,
+      "step": 10127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5886527390523273,
+      "learning_rate": 7.850796953963198e-06,
+      "loss": 0.3371,
+      "step": 10128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2781699195992228,
+      "learning_rate": 7.848979565154314e-06,
+      "loss": 0.2399,
+      "step": 10129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25233740774202973,
+      "learning_rate": 7.847162250835171e-06,
+      "loss": 0.1594,
+      "step": 10130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5636084765264904,
+      "learning_rate": 7.845345011068709e-06,
+      "loss": 0.4195,
+      "step": 10131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29482534737792404,
+      "learning_rate": 7.843527845917855e-06,
+      "loss": 0.2604,
+      "step": 10132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.44948437825709303,
+      "learning_rate": 7.84171075544554e-06,
+      "loss": 0.3688,
+      "step": 10133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.386973790913395,
+      "learning_rate": 7.839893739714686e-06,
+      "loss": 0.2615,
+      "step": 10134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4697085406586142,
+      "learning_rate": 7.83807679878822e-06,
+      "loss": 0.3673,
+      "step": 10135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2578535107262174,
+      "learning_rate": 7.836259932729062e-06,
+      "loss": 0.203,
+      "step": 10136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3996683546043715,
+      "learning_rate": 7.834443141600131e-06,
+      "loss": 0.3033,
+      "step": 10137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5165287449268846,
+      "learning_rate": 7.83262642546434e-06,
+      "loss": 0.3747,
+      "step": 10138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4618223329699041,
+      "learning_rate": 7.830809784384602e-06,
+      "loss": 0.2441,
+      "step": 10139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30623960016940566,
+      "learning_rate": 7.828993218423829e-06,
+      "loss": 0.2678,
+      "step": 10140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35976134707155416,
+      "learning_rate": 7.827176727644925e-06,
+      "loss": 0.3079,
+      "step": 10141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28324400036157693,
+      "learning_rate": 7.825360312110801e-06,
+      "loss": 0.1523,
+      "step": 10142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7363848108703631,
+      "learning_rate": 7.823543971884353e-06,
+      "loss": 0.3966,
+      "step": 10143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3260077080948101,
+      "learning_rate": 7.821727707028486e-06,
+      "loss": 0.2802,
+      "step": 10144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27632507657679467,
+      "learning_rate": 7.81991151760609e-06,
+      "loss": 0.2405,
+      "step": 10145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3488908220712134,
+      "learning_rate": 7.81809540368007e-06,
+      "loss": 0.1139,
+      "step": 10146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37529873675347386,
+      "learning_rate": 7.81627936531331e-06,
+      "loss": 0.2968,
+      "step": 10147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3596295347448901,
+      "learning_rate": 7.814463402568703e-06,
+      "loss": 0.3217,
+      "step": 10148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6919007612501898,
+      "learning_rate": 7.812647515509131e-06,
+      "loss": 0.245,
+      "step": 10149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36005726352566714,
+      "learning_rate": 7.810831704197486e-06,
+      "loss": 0.3193,
+      "step": 10150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3847714209623518,
+      "learning_rate": 7.80901596869664e-06,
+      "loss": 0.2601,
+      "step": 10151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.21052130026797297,
+      "learning_rate": 7.807200309069482e-06,
+      "loss": 0.1744,
+      "step": 10152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40305583102744774,
+      "learning_rate": 7.805384725378881e-06,
+      "loss": 0.3051,
+      "step": 10153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0107321479306213,
+      "learning_rate": 7.803569217687711e-06,
+      "loss": 0.5662,
+      "step": 10154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6122037063934481,
+      "learning_rate": 7.801753786058847e-06,
+      "loss": 0.2446,
+      "step": 10155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5214675627145766,
+      "learning_rate": 7.799938430555152e-06,
+      "loss": 0.3822,
+      "step": 10156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35482096271578345,
+      "learning_rate": 7.798123151239497e-06,
+      "loss": 0.3462,
+      "step": 10157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24059969586075008,
+      "learning_rate": 7.79630794817474e-06,
+      "loss": 0.1636,
+      "step": 10158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39992023315441017,
+      "learning_rate": 7.794492821423747e-06,
+      "loss": 0.2323,
+      "step": 10159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36592149517228917,
+      "learning_rate": 7.79267777104937e-06,
+      "loss": 0.3129,
+      "step": 10160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6464934747226485,
+      "learning_rate": 7.79086279711447e-06,
+      "loss": 0.3873,
+      "step": 10161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32063480202839684,
+      "learning_rate": 7.789047899681893e-06,
+      "loss": 0.2247,
+      "step": 10162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2778589463458745,
+      "learning_rate": 7.787233078814497e-06,
+      "loss": 0.2119,
+      "step": 10163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3474062922703908,
+      "learning_rate": 7.785418334575122e-06,
+      "loss": 0.3213,
+      "step": 10164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3227767768744291,
+      "learning_rate": 7.783603667026616e-06,
+      "loss": 0.2222,
+      "step": 10165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7112062417756925,
+      "learning_rate": 7.781789076231815e-06,
+      "loss": 0.387,
+      "step": 10166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5940613791092809,
+      "learning_rate": 7.779974562253568e-06,
+      "loss": 0.3781,
+      "step": 10167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24669557310007448,
+      "learning_rate": 7.778160125154702e-06,
+      "loss": 0.2217,
+      "step": 10168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3234491850879115,
+      "learning_rate": 7.776345764998059e-06,
+      "loss": 0.8241,
+      "step": 10169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2135022809142091,
+      "learning_rate": 7.774531481846464e-06,
+      "loss": 0.1409,
+      "step": 10170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33929896012752037,
+      "learning_rate": 7.77271727576275e-06,
+      "loss": 0.2906,
+      "step": 10171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.423664471917811,
+      "learning_rate": 7.770903146809738e-06,
+      "loss": 0.2725,
+      "step": 10172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6369728062982355,
+      "learning_rate": 7.769089095050258e-06,
+      "loss": 0.3813,
+      "step": 10173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3777810720627077,
+      "learning_rate": 7.767275120547123e-06,
+      "loss": 0.2917,
+      "step": 10174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24151161950766775,
+      "learning_rate": 7.765461223363158e-06,
+      "loss": 0.1624,
+      "step": 10175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26233751290851975,
+      "learning_rate": 7.76364740356117e-06,
+      "loss": 0.2287,
+      "step": 10176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4027707555176561,
+      "learning_rate": 7.76183366120398e-06,
+      "loss": 0.333,
+      "step": 10177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9316383400032427,
+      "learning_rate": 7.760019996354396e-06,
+      "loss": 0.3163,
+      "step": 10178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.900838231219768,
+      "learning_rate": 7.75820640907522e-06,
+      "loss": 0.6042,
+      "step": 10179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26204651119783984,
+      "learning_rate": 7.75639289942926e-06,
+      "loss": 0.2604,
+      "step": 10180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4479463747107403,
+      "learning_rate": 7.754579467479318e-06,
+      "loss": 0.2993,
+      "step": 10181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33894552794765487,
+      "learning_rate": 7.752766113288192e-06,
+      "loss": 0.1901,
+      "step": 10182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3762907960074976,
+      "learning_rate": 7.750952836918679e-06,
+      "loss": 0.2689,
+      "step": 10183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.414454215601207,
+      "learning_rate": 7.749139638433573e-06,
+      "loss": 0.3139,
+      "step": 10184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.586844077895301,
+      "learning_rate": 7.747326517895662e-06,
+      "loss": 0.135,
+      "step": 10185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32530862088865625,
+      "learning_rate": 7.74551347536774e-06,
+      "loss": 0.2699,
+      "step": 10186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1672584420814525,
+      "learning_rate": 7.743700510912588e-06,
+      "loss": 0.6975,
+      "step": 10187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24495780308060752,
+      "learning_rate": 7.741887624592992e-06,
+      "loss": 0.2082,
+      "step": 10188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36084956108999233,
+      "learning_rate": 7.740074816471727e-06,
+      "loss": 0.277,
+      "step": 10189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40954337704045063,
+      "learning_rate": 7.738262086611578e-06,
+      "loss": 0.2866,
+      "step": 10190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36418034114752734,
+      "learning_rate": 7.736449435075314e-06,
+      "loss": 0.2365,
+      "step": 10191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35156018638606407,
+      "learning_rate": 7.734636861925706e-06,
+      "loss": 0.2822,
+      "step": 10192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4751770010443096,
+      "learning_rate": 7.732824367225531e-06,
+      "loss": 0.3951,
+      "step": 10193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6841245197795643,
+      "learning_rate": 7.731011951037547e-06,
+      "loss": 0.2865,
+      "step": 10194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3483252023779255,
+      "learning_rate": 7.729199613424523e-06,
+      "loss": 0.2767,
+      "step": 10195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24313842780450307,
+      "learning_rate": 7.727387354449217e-06,
+      "loss": 0.2301,
+      "step": 10196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0930305495887416,
+      "learning_rate": 7.725575174174395e-06,
+      "loss": 0.6972,
+      "step": 10197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28099806807697314,
+      "learning_rate": 7.723763072662804e-06,
+      "loss": 0.2012,
+      "step": 10198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8655932878922294,
+      "learning_rate": 7.721951049977196e-06,
+      "loss": 0.4398,
+      "step": 10199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36319675934158235,
+      "learning_rate": 7.72013910618033e-06,
+      "loss": 0.3273,
+      "step": 10200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31370298470256497,
+      "learning_rate": 7.718327241334944e-06,
+      "loss": 0.2333,
+      "step": 10201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2954558356084548,
+      "learning_rate": 7.716515455503791e-06,
+      "loss": 0.201,
+      "step": 10202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4505771810527777,
+      "learning_rate": 7.71470374874961e-06,
+      "loss": 0.322,
+      "step": 10203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27661425856954885,
+      "learning_rate": 7.712892121135136e-06,
+      "loss": 0.2182,
+      "step": 10204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2487777153308748,
+      "learning_rate": 7.711080572723113e-06,
+      "loss": 0.7325,
+      "step": 10205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7749108515324832,
+      "learning_rate": 7.709269103576269e-06,
+      "loss": 0.3924,
+      "step": 10206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31900576203015873,
+      "learning_rate": 7.70745771375734e-06,
+      "loss": 0.1904,
+      "step": 10207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3050746601489578,
+      "learning_rate": 7.70564640332905e-06,
+      "loss": 0.2411,
+      "step": 10208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4378828524958338,
+      "learning_rate": 7.703835172354127e-06,
+      "loss": 0.251,
+      "step": 10209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32933250069623154,
+      "learning_rate": 7.702024020895292e-06,
+      "loss": 0.241,
+      "step": 10210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8241027953814687,
+      "learning_rate": 7.70021294901527e-06,
+      "loss": 0.3369,
+      "step": 10211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3543263899051379,
+      "learning_rate": 7.69840195677677e-06,
+      "loss": 0.3201,
+      "step": 10212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4099252481168575,
+      "learning_rate": 7.696591044242513e-06,
+      "loss": 0.3004,
+      "step": 10213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2820915527848439,
+      "learning_rate": 7.694780211475209e-06,
+      "loss": 0.166,
+      "step": 10214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3524674573849591,
+      "learning_rate": 7.692969458537568e-06,
+      "loss": 0.3248,
+      "step": 10215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43080167749690207,
+      "learning_rate": 7.691158785492294e-06,
+      "loss": 0.3089,
+      "step": 10216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6922394291237297,
+      "learning_rate": 7.689348192402095e-06,
+      "loss": 0.2959,
+      "step": 10217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9169189348266817,
+      "learning_rate": 7.687537679329668e-06,
+      "loss": 0.507,
+      "step": 10218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3090444070036527,
+      "learning_rate": 7.685727246337709e-06,
+      "loss": 0.2482,
+      "step": 10219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26794700844517866,
+      "learning_rate": 7.683916893488918e-06,
+      "loss": 0.2302,
+      "step": 10220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29467839115121064,
+      "learning_rate": 7.682106620845984e-06,
+      "loss": 0.1529,
+      "step": 10221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35073917502149066,
+      "learning_rate": 7.6802964284716e-06,
+      "loss": 0.2598,
+      "step": 10222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.202371112662339,
+      "learning_rate": 7.678486316428449e-06,
+      "loss": 0.4024,
+      "step": 10223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3608808093489747,
+      "learning_rate": 7.676676284779217e-06,
+      "loss": 0.276,
+      "step": 10224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32842709343076865,
+      "learning_rate": 7.674866333586586e-06,
+      "loss": 0.2837,
+      "step": 10225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9377059450455212,
+      "learning_rate": 7.673056462913235e-06,
+      "loss": 0.5578,
+      "step": 10226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22635475620796228,
+      "learning_rate": 7.671246672821837e-06,
+      "loss": 0.1744,
+      "step": 10227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.34285947837407843,
+      "learning_rate": 7.669436963375067e-06,
+      "loss": 0.2717,
+      "step": 10228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1313131442190563,
+      "learning_rate": 7.667627334635595e-06,
+      "loss": 0.46,
+      "step": 10229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8438766470568757,
+      "learning_rate": 7.665817786666088e-06,
+      "loss": 0.3498,
+      "step": 10230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4222669207775579,
+      "learning_rate": 7.664008319529215e-06,
+      "loss": 0.3328,
+      "step": 10231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3097461985519247,
+      "learning_rate": 7.66219893328763e-06,
+      "loss": 0.2693,
+      "step": 10232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4686433548216791,
+      "learning_rate": 7.660389628003993e-06,
+      "loss": 0.3129,
+      "step": 10233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42214270318871294,
+      "learning_rate": 7.658580403740965e-06,
+      "loss": 0.2034,
+      "step": 10234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5099855157917578,
+      "learning_rate": 7.656771260561195e-06,
+      "loss": 0.3126,
+      "step": 10235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2866231632582778,
+      "learning_rate": 7.654962198527338e-06,
+      "loss": 0.2458,
+      "step": 10236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2955888837904505,
+      "learning_rate": 7.653153217702036e-06,
+      "loss": 0.1982,
+      "step": 10237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38594128734646455,
+      "learning_rate": 7.651344318147941e-06,
+      "loss": 0.24,
+      "step": 10238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.493756437636117,
+      "learning_rate": 7.649535499927688e-06,
+      "loss": 0.3752,
+      "step": 10239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25207716691183124,
+      "learning_rate": 7.647726763103923e-06,
+      "loss": 0.2299,
+      "step": 10240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7780560932689861,
+      "learning_rate": 7.645918107739274e-06,
+      "loss": 0.4704,
+      "step": 10241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3710784879455275,
+      "learning_rate": 7.644109533896384e-06,
+      "loss": 0.2544,
+      "step": 10242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28552083656844823,
+      "learning_rate": 7.642301041637879e-06,
+      "loss": 0.2177,
+      "step": 10243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47267057079111247,
+      "learning_rate": 7.640492631026387e-06,
+      "loss": 0.3323,
+      "step": 10244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7320353696301287,
+      "learning_rate": 7.638684302124533e-06,
+      "loss": 0.4111,
+      "step": 10245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4219758171502595,
+      "learning_rate": 7.63687605499494e-06,
+      "loss": 0.2955,
+      "step": 10246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36371007108240666,
+      "learning_rate": 7.635067889700228e-06,
+      "loss": 0.2585,
+      "step": 10247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24237484085577543,
+      "learning_rate": 7.633259806303012e-06,
+      "loss": 0.2022,
+      "step": 10248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5130982347891825,
+      "learning_rate": 7.63145180486591e-06,
+      "loss": 0.2591,
+      "step": 10249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3943639466503955,
+      "learning_rate": 7.629643885451527e-06,
+      "loss": 0.212,
+      "step": 10250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42780601563784476,
+      "learning_rate": 7.627836048122477e-06,
+      "loss": 0.3231,
+      "step": 10251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5485089590488623,
+      "learning_rate": 7.626028292941361e-06,
+      "loss": 0.3429,
+      "step": 10252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38528395368808604,
+      "learning_rate": 7.624220619970784e-06,
+      "loss": 0.2701,
+      "step": 10253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27098911193826375,
+      "learning_rate": 7.622413029273343e-06,
+      "loss": 0.1638,
+      "step": 10254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.301421858486742,
+      "learning_rate": 7.62060552091164e-06,
+      "loss": 0.2811,
+      "step": 10255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3772924937493043,
+      "learning_rate": 7.618798094948262e-06,
+      "loss": 0.2382,
+      "step": 10256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6013924440847807,
+      "learning_rate": 7.616990751445806e-06,
+      "loss": 0.4171,
+      "step": 10257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39066029970772426,
+      "learning_rate": 7.615183490466858e-06,
+      "loss": 0.2633,
+      "step": 10258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36730340180357374,
+      "learning_rate": 7.613376312074001e-06,
+      "loss": 0.3297,
+      "step": 10259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2195929862317336,
+      "learning_rate": 7.611569216329821e-06,
+      "loss": 0.1535,
+      "step": 10260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40569141532500297,
+      "learning_rate": 7.609762203296896e-06,
+      "loss": 0.2855,
+      "step": 10261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.566404154555934,
+      "learning_rate": 7.607955273037804e-06,
+      "loss": 0.4164,
+      "step": 10262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3857204802136643,
+      "learning_rate": 7.606148425615117e-06,
+      "loss": 0.2772,
+      "step": 10263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.608997704118891,
+      "learning_rate": 7.604341661091409e-06,
+      "loss": 0.3627,
+      "step": 10264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4045959883896538,
+      "learning_rate": 7.602534979529246e-06,
+      "loss": 0.3597,
+      "step": 10265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.20847234205880777,
+      "learning_rate": 7.600728380991191e-06,
+      "loss": 0.1583,
+      "step": 10266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39509917853189525,
+      "learning_rate": 7.598921865539811e-06,
+      "loss": 0.2954,
+      "step": 10267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33854641736778623,
+      "learning_rate": 7.597115433237664e-06,
+      "loss": 0.2913,
+      "step": 10268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5218405160890431,
+      "learning_rate": 7.5953090841473035e-06,
+      "loss": 0.3258,
+      "step": 10269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5757741501759367,
+      "learning_rate": 7.593502818331289e-06,
+      "loss": 0.3423,
+      "step": 10270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2601837400754058,
+      "learning_rate": 7.5916966358521645e-06,
+      "loss": 0.2893,
+      "step": 10271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43141156338626285,
+      "learning_rate": 7.589890536772486e-06,
+      "loss": 0.2517,
+      "step": 10272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2912064861117238,
+      "learning_rate": 7.588084521154791e-06,
+      "loss": 0.1262,
+      "step": 10273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39090096704594307,
+      "learning_rate": 7.586278589061628e-06,
+      "loss": 0.3112,
+      "step": 10274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.50511253032244,
+      "learning_rate": 7.584472740555533e-06,
+      "loss": 0.3057,
+      "step": 10275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3452619407210834,
+      "learning_rate": 7.582666975699043e-06,
+      "loss": 0.1574,
+      "step": 10276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37024431661053503,
+      "learning_rate": 7.5808612945546915e-06,
+      "loss": 0.3077,
+      "step": 10277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44955413310834985,
+      "learning_rate": 7.5790556971850095e-06,
+      "loss": 0.2907,
+      "step": 10278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22616688185308217,
+      "learning_rate": 7.577250183652523e-06,
+      "loss": 0.1954,
+      "step": 10279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4488134009283226,
+      "learning_rate": 7.575444754019762e-06,
+      "loss": 0.3517,
+      "step": 10280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6806739153782752,
+      "learning_rate": 7.5736394083492414e-06,
+      "loss": 0.4589,
+      "step": 10281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38788290607606046,
+      "learning_rate": 7.571834146703486e-06,
+      "loss": 0.1904,
+      "step": 10282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3014316610003963,
+      "learning_rate": 7.57002896914501e-06,
+      "loss": 0.2887,
+      "step": 10283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4034575913681455,
+      "learning_rate": 7.568223875736325e-06,
+      "loss": 0.2604,
+      "step": 10284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6416524913164146,
+      "learning_rate": 7.566418866539944e-06,
+      "loss": 0.3879,
+      "step": 10285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23814333645697722,
+      "learning_rate": 7.5646139416183705e-06,
+      "loss": 0.1818,
+      "step": 10286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35517257625022275,
+      "learning_rate": 7.562809101034114e-06,
+      "loss": 0.3117,
+      "step": 10287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2153446890355475,
+      "learning_rate": 7.56100434484967e-06,
+      "loss": 0.7712,
+      "step": 10288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30753262429141587,
+      "learning_rate": 7.559199673127545e-06,
+      "loss": 0.1986,
+      "step": 10289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7300295591145592,
+      "learning_rate": 7.557395085930227e-06,
+      "loss": 0.4217,
+      "step": 10290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32301624520605615,
+      "learning_rate": 7.555590583320214e-06,
+      "loss": 0.2986,
+      "step": 10291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.255707287279116,
+      "learning_rate": 7.553786165359993e-06,
+      "loss": 0.1879,
+      "step": 10292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41385565707671423,
+      "learning_rate": 7.551981832112054e-06,
+      "loss": 0.2826,
+      "step": 10293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38640746910827395,
+      "learning_rate": 7.550177583638876e-06,
+      "loss": 0.3198,
+      "step": 10294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2565067070850015,
+      "learning_rate": 7.548373420002945e-06,
+      "loss": 0.2217,
+      "step": 10295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7864258593626743,
+      "learning_rate": 7.546569341266737e-06,
+      "loss": 0.4573,
+      "step": 10296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5094781014286576,
+      "learning_rate": 7.544765347492727e-06,
+      "loss": 0.3626,
+      "step": 10297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3531764204079477,
+      "learning_rate": 7.542961438743389e-06,
+      "loss": 0.2905,
+      "step": 10298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22608108952594327,
+      "learning_rate": 7.54115761508119e-06,
+      "loss": 0.1933,
+      "step": 10299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5899693029472266,
+      "learning_rate": 7.539353876568594e-06,
+      "loss": 0.4156,
+      "step": 10300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3815342442871858,
+      "learning_rate": 7.537550223268071e-06,
+      "loss": 0.3133,
+      "step": 10301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7859397443567046,
+      "learning_rate": 7.5357466552420745e-06,
+      "loss": 0.2873,
+      "step": 10302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.46196064535471065,
+      "learning_rate": 7.533943172553068e-06,
+      "loss": 0.336,
+      "step": 10303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3143589100536943,
+      "learning_rate": 7.5321397752635e-06,
+      "loss": 0.2493,
+      "step": 10304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1956301211002063,
+      "learning_rate": 7.53033646343583e-06,
+      "loss": 0.116,
+      "step": 10305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33209653400560535,
+      "learning_rate": 7.528533237132498e-06,
+      "loss": 0.3106,
+      "step": 10306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.324432066107435,
+      "learning_rate": 7.526730096415957e-06,
+      "loss": 0.2727,
+      "step": 10307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6904170847854164,
+      "learning_rate": 7.524927041348646e-06,
+      "loss": 0.4153,
+      "step": 10308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5469584121601104,
+      "learning_rate": 7.523124071993004e-06,
+      "loss": 0.2054,
+      "step": 10309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.34453117937491873,
+      "learning_rate": 7.521321188411469e-06,
+      "loss": 0.2789,
+      "step": 10310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2578278431274199,
+      "learning_rate": 7.519518390666474e-06,
+      "loss": 0.237,
+      "step": 10311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22170453038599644,
+      "learning_rate": 7.517715678820452e-06,
+      "loss": 0.1511,
+      "step": 10312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3583110798163475,
+      "learning_rate": 7.515913052935827e-06,
+      "loss": 0.2837,
+      "step": 10313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7549106305530019,
+      "learning_rate": 7.514110513075028e-06,
+      "loss": 0.4292,
+      "step": 10314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.291936518934735,
+      "learning_rate": 7.512308059300474e-06,
+      "loss": 0.2353,
+      "step": 10315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3369054517703914,
+      "learning_rate": 7.510505691674586e-06,
+      "loss": 0.2834,
+      "step": 10316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2868584847338976,
+      "learning_rate": 7.5087034102597775e-06,
+      "loss": 0.1644,
+      "step": 10317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2998619310436137,
+      "learning_rate": 7.506901215118465e-06,
+      "loss": 0.2527,
+      "step": 10318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36454420205043037,
+      "learning_rate": 7.505099106313053e-06,
+      "loss": 0.2659,
+      "step": 10319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0795147065212092,
+      "learning_rate": 7.503297083905955e-06,
+      "loss": 0.3916,
+      "step": 10320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1310884841179982,
+      "learning_rate": 7.5014951479595684e-06,
+      "loss": 0.5808,
+      "step": 10321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2958718654952075,
+      "learning_rate": 7.499693298536301e-06,
+      "loss": 0.1898,
+      "step": 10322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28992196328142583,
+      "learning_rate": 7.497891535698546e-06,
+      "loss": 0.2596,
+      "step": 10323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1602792649842537,
+      "learning_rate": 7.496089859508697e-06,
+      "loss": 0.5329,
+      "step": 10324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3745377074556588,
+      "learning_rate": 7.494288270029152e-06,
+      "loss": 0.2368,
+      "step": 10325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31340930333139033,
+      "learning_rate": 7.492486767322293e-06,
+      "loss": 0.2712,
+      "step": 10326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.46084638494659036,
+      "learning_rate": 7.490685351450513e-06,
+      "loss": 0.3427,
+      "step": 10327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3007749876753642,
+      "learning_rate": 7.488884022476189e-06,
+      "loss": 0.1969,
+      "step": 10328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0453565752421365,
+      "learning_rate": 7.487082780461704e-06,
+      "loss": 0.5567,
+      "step": 10329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40266828324910653,
+      "learning_rate": 7.485281625469432e-06,
+      "loss": 0.3253,
+      "step": 10330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33743676652775795,
+      "learning_rate": 7.483480557561753e-06,
+      "loss": 0.2332,
+      "step": 10331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4111867303618263,
+      "learning_rate": 7.481679576801035e-06,
+      "loss": 0.2775,
+      "step": 10332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4693763211900533,
+      "learning_rate": 7.479878683249642e-06,
+      "loss": 0.2668,
+      "step": 10333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.503106972794314,
+      "learning_rate": 7.478077876969943e-06,
+      "loss": 0.2492,
+      "step": 10334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32620964008429154,
+      "learning_rate": 7.476277158024299e-06,
+      "loss": 0.2576,
+      "step": 10335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6618248472899589,
+      "learning_rate": 7.474476526475066e-06,
+      "loss": 0.4087,
+      "step": 10336
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42152985701301704,
+      "learning_rate": 7.4726759823846054e-06,
+      "loss": 0.2839,
+      "step": 10337
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2510070165287194,
+      "learning_rate": 7.470875525815263e-06,
+      "loss": 0.2232,
+      "step": 10338
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42724280486977506,
+      "learning_rate": 7.4690751568293955e-06,
+      "loss": 0.2797,
+      "step": 10339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37480903963553386,
+      "learning_rate": 7.467274875489345e-06,
+      "loss": 0.2427,
+      "step": 10340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41897775646501684,
+      "learning_rate": 7.465474681857459e-06,
+      "loss": 0.2865,
+      "step": 10341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3516394732341963,
+      "learning_rate": 7.463674575996072e-06,
+      "loss": 0.3245,
+      "step": 10342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3376477205658888,
+      "learning_rate": 7.461874557967528e-06,
+      "loss": 0.2722,
+      "step": 10343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2938787637187035,
+      "learning_rate": 7.4600746278341575e-06,
+      "loss": 0.1601,
+      "step": 10344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4272505094482967,
+      "learning_rate": 7.458274785658295e-06,
+      "loss": 0.2518,
+      "step": 10345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32516909890328166,
+      "learning_rate": 7.4564750315022645e-06,
+      "loss": 0.2598,
+      "step": 10346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36114353203460814,
+      "learning_rate": 7.454675365428397e-06,
+      "loss": 0.305,
+      "step": 10347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5091804731487372,
+      "learning_rate": 7.452875787499012e-06,
+      "loss": 0.3558,
+      "step": 10348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3474769886170156,
+      "learning_rate": 7.451076297776427e-06,
+      "loss": 0.2804,
+      "step": 10349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.45404733187295016,
+      "learning_rate": 7.4492768963229635e-06,
+      "loss": 0.3461,
+      "step": 10350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22564809690050838,
+      "learning_rate": 7.447477583200928e-06,
+      "loss": 0.1469,
+      "step": 10351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41250335520063286,
+      "learning_rate": 7.445678358472637e-06,
+      "loss": 0.2752,
+      "step": 10352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43262651395981655,
+      "learning_rate": 7.443879222200392e-06,
+      "loss": 0.3522,
+      "step": 10353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3365479924421112,
+      "learning_rate": 7.442080174446502e-06,
+      "loss": 0.2761,
+      "step": 10354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39423987473222083,
+      "learning_rate": 7.440281215273262e-06,
+      "loss": 0.2763,
+      "step": 10355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4763702671153097,
+      "learning_rate": 7.438482344742977e-06,
+      "loss": 0.3736,
+      "step": 10356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24311212151397302,
+      "learning_rate": 7.436683562917937e-06,
+      "loss": 0.1081,
+      "step": 10357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31815626651918666,
+      "learning_rate": 7.4348848698604345e-06,
+      "loss": 0.2684,
+      "step": 10358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3554260506520077,
+      "learning_rate": 7.433086265632759e-06,
+      "loss": 0.3245,
+      "step": 10359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7899975431739951,
+      "learning_rate": 7.431287750297196e-06,
+      "loss": 0.5236,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.324058350565275,
+      "learning_rate": 7.429489323916028e-06,
+      "loss": 0.2124,
+      "step": 10361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30049014342614966,
+      "learning_rate": 7.427690986551534e-06,
+      "loss": 0.2998,
+      "step": 10362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2735467911648441,
+      "learning_rate": 7.42589273826599e-06,
+      "loss": 0.1753,
+      "step": 10363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3070856253154252,
+      "learning_rate": 7.42409457912167e-06,
+      "loss": 0.214,
+      "step": 10364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7289098597123872,
+      "learning_rate": 7.422296509180844e-06,
+      "loss": 0.3994,
+      "step": 10365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3466921745387132,
+      "learning_rate": 7.420498528505783e-06,
+      "loss": 0.3249,
+      "step": 10366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30005732279309993,
+      "learning_rate": 7.418700637158742e-06,
+      "loss": 0.2126,
+      "step": 10367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4382862328192985,
+      "learning_rate": 7.416902835201989e-06,
+      "loss": 0.7743,
+      "step": 10368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3143363107928643,
+      "learning_rate": 7.415105122697777e-06,
+      "loss": 0.2053,
+      "step": 10369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27691108069319204,
+      "learning_rate": 7.413307499708367e-06,
+      "loss": 0.2016,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47985595670757103,
+      "learning_rate": 7.411509966296004e-06,
+      "loss": 0.297,
+      "step": 10371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9858514760671856,
+      "learning_rate": 7.409712522522942e-06,
+      "loss": 0.7043,
+      "step": 10372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6016526088054295,
+      "learning_rate": 7.407915168451423e-06,
+      "loss": 0.3614,
+      "step": 10373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29457207851404604,
+      "learning_rate": 7.40611790414369e-06,
+      "loss": 0.24,
+      "step": 10374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5271657134892388,
+      "learning_rate": 7.404320729661982e-06,
+      "loss": 0.2722,
+      "step": 10375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4621286015570544,
+      "learning_rate": 7.402523645068536e-06,
+      "loss": 0.3243,
+      "step": 10376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26337030099005204,
+      "learning_rate": 7.400726650425585e-06,
+      "loss": 0.2056,
+      "step": 10377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.337834575079801,
+      "learning_rate": 7.3989297457953565e-06,
+      "loss": 0.3024,
+      "step": 10378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5600185258598154,
+      "learning_rate": 7.3971329312400805e-06,
+      "loss": 0.4083,
+      "step": 10379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4112617760836703,
+      "learning_rate": 7.395336206821979e-06,
+      "loss": 0.2426,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.615001228721209,
+      "learning_rate": 7.393539572603274e-06,
+      "loss": 0.3951,
+      "step": 10381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2863126488000933,
+      "learning_rate": 7.391743028646179e-06,
+      "loss": 0.2512,
+      "step": 10382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31806544952047155,
+      "learning_rate": 7.3899465750129116e-06,
+      "loss": 0.2348,
+      "step": 10383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27767063661119495,
+      "learning_rate": 7.388150211765682e-06,
+      "loss": 0.1522,
+      "step": 10384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.40574913552979075,
+      "learning_rate": 7.3863539389667e-06,
+      "loss": 0.2952,
+      "step": 10385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3449818810488741,
+      "learning_rate": 7.384557756678166e-06,
+      "loss": 0.2971,
+      "step": 10386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4392748651396246,
+      "learning_rate": 7.382761664962287e-06,
+      "loss": 0.2963,
+      "step": 10387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5556219375952675,
+      "learning_rate": 7.380965663881259e-06,
+      "loss": 0.3425,
+      "step": 10388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2330582030181791,
+      "learning_rate": 7.379169753497275e-06,
+      "loss": 0.2051,
+      "step": 10389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3359612146899458,
+      "learning_rate": 7.377373933872531e-06,
+      "loss": 0.2831,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7637438385972759,
+      "learning_rate": 7.375578205069213e-06,
+      "loss": 0.3904,
+      "step": 10391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3545649951392472,
+      "learning_rate": 7.373782567149514e-06,
+      "loss": 0.2763,
+      "step": 10392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6164382213937144,
+      "learning_rate": 7.371987020175606e-06,
+      "loss": 0.3046,
+      "step": 10393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3547602337809164,
+      "learning_rate": 7.370191564209679e-06,
+      "loss": 0.2542,
+      "step": 10394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2775262885159086,
+      "learning_rate": 7.368396199313901e-06,
+      "loss": 0.2352,
+      "step": 10395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4603446729991688,
+      "learning_rate": 7.3666009255504534e-06,
+      "loss": 0.2823,
+      "step": 10396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41311951116192697,
+      "learning_rate": 7.364805742981499e-06,
+      "loss": 0.2084,
+      "step": 10397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26981663565088704,
+      "learning_rate": 7.363010651669211e-06,
+      "loss": 0.2726,
+      "step": 10398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7117455054970516,
+      "learning_rate": 7.361215651675753e-06,
+      "loss": 0.398,
+      "step": 10399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4345400774009875,
+      "learning_rate": 7.359420743063282e-06,
+      "loss": 0.1569,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3219544373698642,
+      "learning_rate": 7.357625925893954e-06,
+      "loss": 0.2259,
+      "step": 10401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2688178658293408,
+      "learning_rate": 7.355831200229928e-06,
+      "loss": 0.2649,
+      "step": 10402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3136728201989251,
+      "learning_rate": 7.354036566133354e-06,
+      "loss": 0.2091,
+      "step": 10403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4715407097290975,
+      "learning_rate": 7.3522420236663805e-06,
+      "loss": 0.3276,
+      "step": 10404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4956804964246323,
+      "learning_rate": 7.350447572891148e-06,
+      "loss": 0.3307,
+      "step": 10405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32953209301390884,
+      "learning_rate": 7.348653213869807e-06,
+      "loss": 0.1928,
+      "step": 10406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.40875599533641027,
+      "learning_rate": 7.346858946664488e-06,
+      "loss": 0.3065,
+      "step": 10407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2591224485745581,
+      "learning_rate": 7.345064771337332e-06,
+      "loss": 0.1801,
+      "step": 10408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4714728778705214,
+      "learning_rate": 7.343270687950468e-06,
+      "loss": 0.36,
+      "step": 10409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28898646326817645,
+      "learning_rate": 7.341476696566026e-06,
+      "loss": 0.2228,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7426477752866582,
+      "learning_rate": 7.33968279724613e-06,
+      "loss": 0.4396,
+      "step": 10411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1024879194552417,
+      "learning_rate": 7.337888990052906e-06,
+      "loss": 0.5185,
+      "step": 10412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23906818831035612,
+      "learning_rate": 7.336095275048474e-06,
+      "loss": 0.1846,
+      "step": 10413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3227646272577517,
+      "learning_rate": 7.334301652294944e-06,
+      "loss": 0.2678,
+      "step": 10414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5806326984574692,
+      "learning_rate": 7.332508121854435e-06,
+      "loss": 0.4073,
+      "step": 10415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34300519661195633,
+      "learning_rate": 7.330714683789053e-06,
+      "loss": 0.2575,
+      "step": 10416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5015264244296673,
+      "learning_rate": 7.32892133816091e-06,
+      "loss": 0.3357,
+      "step": 10417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42152132629199246,
+      "learning_rate": 7.327128085032103e-06,
+      "loss": 0.3183,
+      "step": 10418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3411341644741504,
+      "learning_rate": 7.325334924464737e-06,
+      "loss": 0.2233,
+      "step": 10419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4057527596698589,
+      "learning_rate": 7.323541856520908e-06,
+      "loss": 0.3004,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3473167164819613,
+      "learning_rate": 7.32174888126271e-06,
+      "loss": 0.2923,
+      "step": 10421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36276706144912185,
+      "learning_rate": 7.3199559987522305e-06,
+      "loss": 0.288,
+      "step": 10422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41822234452984713,
+      "learning_rate": 7.3181632090515635e-06,
+      "loss": 0.2161,
+      "step": 10423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5414425489075652,
+      "learning_rate": 7.316370512222785e-06,
+      "loss": 0.3239,
+      "step": 10424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.378916321206655,
+      "learning_rate": 7.314577908327982e-06,
+      "loss": 0.2869,
+      "step": 10425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33481596176071216,
+      "learning_rate": 7.312785397429231e-06,
+      "loss": 0.2687,
+      "step": 10426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5608026074266522,
+      "learning_rate": 7.310992979588607e-06,
+      "loss": 0.3707,
+      "step": 10427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.319216032928723,
+      "learning_rate": 7.30920065486818e-06,
+      "loss": 0.2487,
+      "step": 10428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23822935077111004,
+      "learning_rate": 7.307408423330016e-06,
+      "loss": 0.1923,
+      "step": 10429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1734266520628216,
+      "learning_rate": 7.305616285036186e-06,
+      "loss": 0.5906,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.306891867740223,
+      "learning_rate": 7.303824240048744e-06,
+      "loss": 0.2423,
+      "step": 10431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6941208094628137,
+      "learning_rate": 7.3020322884297565e-06,
+      "loss": 0.3179,
+      "step": 10432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31808162658254446,
+      "learning_rate": 7.300240430241278e-06,
+      "loss": 0.2968,
+      "step": 10433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4021974465762369,
+      "learning_rate": 7.298448665545352e-06,
+      "loss": 0.2697,
+      "step": 10434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2515530206184095,
+      "learning_rate": 7.296656994404034e-06,
+      "loss": 0.1669,
+      "step": 10435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5861578574807791,
+      "learning_rate": 7.294865416879366e-06,
+      "loss": 0.1531,
+      "step": 10436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34651374064157503,
+      "learning_rate": 7.293073933033394e-06,
+      "loss": 0.2677,
+      "step": 10437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37290784368695534,
+      "learning_rate": 7.291282542928158e-06,
+      "loss": 0.3219,
+      "step": 10438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8520457116127256,
+      "learning_rate": 7.289491246625686e-06,
+      "loss": 0.3167,
+      "step": 10439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3610971456454583,
+      "learning_rate": 7.287700044188019e-06,
+      "loss": 0.2736,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27075648165360194,
+      "learning_rate": 7.28590893567718e-06,
+      "loss": 0.2578,
+      "step": 10441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.467437287940649,
+      "learning_rate": 7.2841179211552005e-06,
+      "loss": 0.1746,
+      "step": 10442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3203387589486802,
+      "learning_rate": 7.282327000684099e-06,
+      "loss": 0.2521,
+      "step": 10443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0366281064937823,
+      "learning_rate": 7.280536174325897e-06,
+      "loss": 0.4184,
+      "step": 10444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32022139792761206,
+      "learning_rate": 7.27874544214261e-06,
+      "loss": 0.2512,
+      "step": 10445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31124170915094695,
+      "learning_rate": 7.276954804196252e-06,
+      "loss": 0.2607,
+      "step": 10446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27313298584475565,
+      "learning_rate": 7.2751642605488305e-06,
+      "loss": 0.1748,
+      "step": 10447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4671493965457703,
+      "learning_rate": 7.273373811262356e-06,
+      "loss": 0.7225,
+      "step": 10448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30145172575689483,
+      "learning_rate": 7.271583456398827e-06,
+      "loss": 0.2257,
+      "step": 10449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7093270985040872,
+      "learning_rate": 7.269793196020247e-06,
+      "loss": 0.3663,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0567717494520565,
+      "learning_rate": 7.26800303018861e-06,
+      "loss": 0.7185,
+      "step": 10451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3041414305207806,
+      "learning_rate": 7.266212958965912e-06,
+      "loss": 0.2066,
+      "step": 10452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37389481781956213,
+      "learning_rate": 7.264422982414143e-06,
+      "loss": 0.3267,
+      "step": 10453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25075241124657155,
+      "learning_rate": 7.2626331005952845e-06,
+      "loss": 0.1912,
+      "step": 10454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3700868988947285,
+      "learning_rate": 7.260843313571328e-06,
+      "loss": 0.2145,
+      "step": 10455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9142862153982831,
+      "learning_rate": 7.259053621404246e-06,
+      "loss": 0.4353,
+      "step": 10456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3873001010445203,
+      "learning_rate": 7.2572640241560225e-06,
+      "loss": 0.3062,
+      "step": 10457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3067948858367847,
+      "learning_rate": 7.255474521888624e-06,
+      "loss": 0.1958,
+      "step": 10458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3374418148179951,
+      "learning_rate": 7.253685114664029e-06,
+      "loss": 0.232,
+      "step": 10459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45806778271637055,
+      "learning_rate": 7.251895802544197e-06,
+      "loss": 0.3377,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4224168796644813,
+      "learning_rate": 7.250106585591098e-06,
+      "loss": 0.2808,
+      "step": 10461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4315638186581728,
+      "learning_rate": 7.2483174638666876e-06,
+      "loss": 0.2695,
+      "step": 10462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1670725205025345,
+      "learning_rate": 7.246528437432927e-06,
+      "loss": 0.7757,
+      "step": 10463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3893678908800406,
+      "learning_rate": 7.244739506351765e-06,
+      "loss": 0.2822,
+      "step": 10464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27797550100587226,
+      "learning_rate": 7.242950670685159e-06,
+      "loss": 0.222,
+      "step": 10465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4390055988054794,
+      "learning_rate": 7.2411619304950535e-06,
+      "loss": 0.26,
+      "step": 10466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32031797829067743,
+      "learning_rate": 7.239373285843392e-06,
+      "loss": 0.2501,
+      "step": 10467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6696812537335045,
+      "learning_rate": 7.237584736792112e-06,
+      "loss": 0.3122,
+      "step": 10468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34998214322910576,
+      "learning_rate": 7.235796283403153e-06,
+      "loss": 0.3351,
+      "step": 10469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3469237876049126,
+      "learning_rate": 7.234007925738451e-06,
+      "loss": 0.2644,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4179928240007145,
+      "learning_rate": 7.2322196638599365e-06,
+      "loss": 0.5999,
+      "step": 10471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4999890912597717,
+      "learning_rate": 7.230431497829533e-06,
+      "loss": 0.2942,
+      "step": 10472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2984927447908153,
+      "learning_rate": 7.228643427709172e-06,
+      "loss": 0.2701,
+      "step": 10473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28578890500066373,
+      "learning_rate": 7.226855453560766e-06,
+      "loss": 0.1861,
+      "step": 10474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8567557080685262,
+      "learning_rate": 7.2250675754462384e-06,
+      "loss": 0.4243,
+      "step": 10475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38207266662063444,
+      "learning_rate": 7.2232797934275e-06,
+      "loss": 0.2831,
+      "step": 10476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32210840344326147,
+      "learning_rate": 7.221492107566466e-06,
+      "loss": 0.285,
+      "step": 10477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.527301928692191,
+      "learning_rate": 7.2197045179250395e-06,
+      "loss": 0.2243,
+      "step": 10478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3561268181099593,
+      "learning_rate": 7.217917024565124e-06,
+      "loss": 0.2425,
+      "step": 10479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38431117081131044,
+      "learning_rate": 7.216129627548625e-06,
+      "loss": 0.271,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.35747867342515993,
+      "learning_rate": 7.214342326937434e-06,
+      "loss": 0.2597,
+      "step": 10481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3926495922665949,
+      "learning_rate": 7.212555122793452e-06,
+      "loss": 0.2489,
+      "step": 10482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4978920915555641,
+      "learning_rate": 7.210768015178563e-06,
+      "loss": 0.3426,
+      "step": 10483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7036657391861827,
+      "learning_rate": 7.208981004154661e-06,
+      "loss": 0.4456,
+      "step": 10484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26724477915832684,
+      "learning_rate": 7.2071940897836235e-06,
+      "loss": 0.2103,
+      "step": 10485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24639680661311067,
+      "learning_rate": 7.205407272127336e-06,
+      "loss": 0.181,
+      "step": 10486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1345625130187684,
+      "learning_rate": 7.203620551247675e-06,
+      "loss": 0.7537,
+      "step": 10487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32662252851637746,
+      "learning_rate": 7.201833927206514e-06,
+      "loss": 0.206,
+      "step": 10488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3779858157814377,
+      "learning_rate": 7.200047400065722e-06,
+      "loss": 0.2848,
+      "step": 10489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7855434602278417,
+      "learning_rate": 7.198260969887171e-06,
+      "loss": 0.4231,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22071700621294713,
+      "learning_rate": 7.196474636732722e-06,
+      "loss": 0.1673,
+      "step": 10491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41724723056633095,
+      "learning_rate": 7.194688400664232e-06,
+      "loss": 0.2648,
+      "step": 10492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33254747854591676,
+      "learning_rate": 7.192902261743566e-06,
+      "loss": 0.3185,
+      "step": 10493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23978557568988323,
+      "learning_rate": 7.191116220032572e-06,
+      "loss": 0.0635,
+      "step": 10494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45877974911373315,
+      "learning_rate": 7.189330275593104e-06,
+      "loss": 0.2984,
+      "step": 10495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36561107076275134,
+      "learning_rate": 7.187544428487006e-06,
+      "loss": 0.2965,
+      "step": 10496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6271544762259579,
+      "learning_rate": 7.1857586787761246e-06,
+      "loss": 0.3096,
+      "step": 10497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22321373342280323,
+      "learning_rate": 7.183973026522297e-06,
+      "loss": 0.1597,
+      "step": 10498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1306791168483041,
+      "learning_rate": 7.182187471787365e-06,
+      "loss": 0.7098,
+      "step": 10499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5794004017247366,
+      "learning_rate": 7.180402014633159e-06,
+      "loss": 0.2908,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27354208770268323,
+      "learning_rate": 7.178616655121513e-06,
+      "loss": 0.2454,
+      "step": 10501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.158877404681521,
+      "learning_rate": 7.176831393314248e-06,
+      "loss": 0.6185,
+      "step": 10502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5729698953409218,
+      "learning_rate": 7.175046229273191e-06,
+      "loss": 0.3404,
+      "step": 10503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2731478718166219,
+      "learning_rate": 7.17326116306016e-06,
+      "loss": 0.196,
+      "step": 10504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34421103264516534,
+      "learning_rate": 7.171476194736975e-06,
+      "loss": 0.3172,
+      "step": 10505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4021245776566278,
+      "learning_rate": 7.169691324365447e-06,
+      "loss": 0.2735,
+      "step": 10506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4057698150407605,
+      "learning_rate": 7.167906552007387e-06,
+      "loss": 0.2176,
+      "step": 10507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39037278889321536,
+      "learning_rate": 7.166121877724599e-06,
+      "loss": 0.3269,
+      "step": 10508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42136169762460707,
+      "learning_rate": 7.164337301578892e-06,
+      "loss": 0.2538,
+      "step": 10509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3061269281320481,
+      "learning_rate": 7.162552823632059e-06,
+      "loss": 0.2145,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6243819662075308,
+      "learning_rate": 7.160768443945902e-06,
+      "loss": 0.3622,
+      "step": 10511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39212060454543934,
+      "learning_rate": 7.15898416258221e-06,
+      "loss": 0.3269,
+      "step": 10512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29126800036976597,
+      "learning_rate": 7.157199979602777e-06,
+      "loss": 0.2787,
+      "step": 10513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4578357233133374,
+      "learning_rate": 7.155415895069385e-06,
+      "loss": 0.1526,
+      "step": 10514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7504728210353762,
+      "learning_rate": 7.153631909043818e-06,
+      "loss": 0.3463,
+      "step": 10515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39710388219731657,
+      "learning_rate": 7.151848021587855e-06,
+      "loss": 0.2964,
+      "step": 10516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3785129422319976,
+      "learning_rate": 7.150064232763274e-06,
+      "loss": 0.2846,
+      "step": 10517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5141174513877448,
+      "learning_rate": 7.1482805426318465e-06,
+      "loss": 0.314,
+      "step": 10518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2443119966552638,
+      "learning_rate": 7.146496951255339e-06,
+      "loss": 0.2207,
+      "step": 10519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30675794437787113,
+      "learning_rate": 7.144713458695521e-06,
+      "loss": 0.2244,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5894451000930164,
+      "learning_rate": 7.1429300650141505e-06,
+      "loss": 0.3308,
+      "step": 10521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3686690194975854,
+      "learning_rate": 7.141146770272993e-06,
+      "loss": 0.3187,
+      "step": 10522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6833298869205651,
+      "learning_rate": 7.139363574533797e-06,
+      "loss": 0.3821,
+      "step": 10523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28139648241122606,
+      "learning_rate": 7.137580477858319e-06,
+      "loss": 0.231,
+      "step": 10524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41488090474727446,
+      "learning_rate": 7.1357974803083044e-06,
+      "loss": 0.3398,
+      "step": 10525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30323806929317104,
+      "learning_rate": 7.134014581945501e-06,
+      "loss": 0.1871,
+      "step": 10526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3600629089927518,
+      "learning_rate": 7.132231782831649e-06,
+      "loss": 0.1966,
+      "step": 10527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.50493454952615,
+      "learning_rate": 7.130449083028488e-06,
+      "loss": 0.3759,
+      "step": 10528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4093641362730516,
+      "learning_rate": 7.1286664825977505e-06,
+      "loss": 0.3145,
+      "step": 10529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6043612819812036,
+      "learning_rate": 7.1268839816011695e-06,
+      "loss": 0.2636,
+      "step": 10530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24807553607158486,
+      "learning_rate": 7.125101580100474e-06,
+      "loss": 0.2093,
+      "step": 10531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2676598452867699,
+      "learning_rate": 7.123319278157385e-06,
+      "loss": 0.2648,
+      "step": 10532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5550732423895852,
+      "learning_rate": 7.121537075833629e-06,
+      "loss": 0.1338,
+      "step": 10533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36036269402932336,
+      "learning_rate": 7.119754973190915e-06,
+      "loss": 0.2997,
+      "step": 10534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9122378994044027,
+      "learning_rate": 7.11797297029097e-06,
+      "loss": 0.3795,
+      "step": 10535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3485600601021822,
+      "learning_rate": 7.116191067195494e-06,
+      "loss": 0.3105,
+      "step": 10536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30727339279065635,
+      "learning_rate": 7.114409263966195e-06,
+      "loss": 0.1943,
+      "step": 10537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2655701703598823,
+      "learning_rate": 7.11262756066478e-06,
+      "loss": 0.1737,
+      "step": 10538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6735221155033528,
+      "learning_rate": 7.110845957352948e-06,
+      "loss": 0.3591,
+      "step": 10539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27521311307189994,
+      "learning_rate": 7.109064454092398e-06,
+      "loss": 0.2416,
+      "step": 10540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.854134216843187,
+      "learning_rate": 7.1072830509448185e-06,
+      "loss": 0.4508,
+      "step": 10541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6999782759176724,
+      "learning_rate": 7.105501747971906e-06,
+      "loss": 0.4854,
+      "step": 10542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3321506116907094,
+      "learning_rate": 7.103720545235342e-06,
+      "loss": 0.2268,
+      "step": 10543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.233533515493403,
+      "learning_rate": 7.10193944279681e-06,
+      "loss": 0.2295,
+      "step": 10544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7011227141792372,
+      "learning_rate": 7.100158440717993e-06,
+      "loss": 0.3943,
+      "step": 10545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3676451737884998,
+      "learning_rate": 7.098377539060562e-06,
+      "loss": 0.273,
+      "step": 10546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7447326581606627,
+      "learning_rate": 7.096596737886194e-06,
+      "loss": 0.3583,
+      "step": 10547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32197795833723225,
+      "learning_rate": 7.0948160372565534e-06,
+      "loss": 0.2779,
+      "step": 10548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3593098276921128,
+      "learning_rate": 7.093035437233311e-06,
+      "loss": 0.2799,
+      "step": 10549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1792310565543837,
+      "learning_rate": 7.091254937878125e-06,
+      "loss": 0.084,
+      "step": 10550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0643531416224592,
+      "learning_rate": 7.089474539252656e-06,
+      "loss": 0.3914,
+      "step": 10551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.250677757331347,
+      "learning_rate": 7.087694241418558e-06,
+      "loss": 0.2396,
+      "step": 10552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3975599441810498,
+      "learning_rate": 7.085914044437485e-06,
+      "loss": 0.2871,
+      "step": 10553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9224701306896524,
+      "learning_rate": 7.084133948371081e-06,
+      "loss": 0.6096,
+      "step": 10554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3374955882849087,
+      "learning_rate": 7.082353953280995e-06,
+      "loss": 0.2439,
+      "step": 10555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.21337914040021594,
+      "learning_rate": 7.080574059228866e-06,
+      "loss": 0.1718,
+      "step": 10556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2038444194000044,
+      "learning_rate": 7.07879426627633e-06,
+      "loss": 0.3988,
+      "step": 10557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30269484417471876,
+      "learning_rate": 7.077014574485025e-06,
+      "loss": 0.24,
+      "step": 10558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2401651526461381,
+      "learning_rate": 7.075234983916577e-06,
+      "loss": 0.7724,
+      "step": 10559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30485211566640963,
+      "learning_rate": 7.073455494632618e-06,
+      "loss": 0.2483,
+      "step": 10560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.46309273195641676,
+      "learning_rate": 7.071676106694767e-06,
+      "loss": 0.2658,
+      "step": 10561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32683809943240216,
+      "learning_rate": 7.06989682016465e-06,
+      "loss": 0.2232,
+      "step": 10562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7879418679957149,
+      "learning_rate": 7.068117635103877e-06,
+      "loss": 0.2834,
+      "step": 10563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3973588991375235,
+      "learning_rate": 7.066338551574066e-06,
+      "loss": 0.29,
+      "step": 10564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4464487646966575,
+      "learning_rate": 7.064559569636824e-06,
+      "loss": 0.3641,
+      "step": 10565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.885248157283508,
+      "learning_rate": 7.062780689353758e-06,
+      "loss": 0.3694,
+      "step": 10566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3276928966908981,
+      "learning_rate": 7.06100191078647e-06,
+      "loss": 0.2578,
+      "step": 10567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35618267701930156,
+      "learning_rate": 7.0592232339965664e-06,
+      "loss": 0.3178,
+      "step": 10568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3768634547209685,
+      "learning_rate": 7.057444659045627e-06,
+      "loss": 0.2148,
+      "step": 10569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25590767906812234,
+      "learning_rate": 7.055666185995256e-06,
+      "loss": 0.2129,
+      "step": 10570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5079144727780354,
+      "learning_rate": 7.053887814907036e-06,
+      "loss": 0.3775,
+      "step": 10571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47995334935170597,
+      "learning_rate": 7.0521095458425555e-06,
+      "loss": 0.3491,
+      "step": 10572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3018031483350368,
+      "learning_rate": 7.050331378863395e-06,
+      "loss": 0.1862,
+      "step": 10573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7799879504359073,
+      "learning_rate": 7.048553314031132e-06,
+      "loss": 0.4413,
+      "step": 10574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34979493541325096,
+      "learning_rate": 7.04677535140734e-06,
+      "loss": 0.2612,
+      "step": 10575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2285568510508859,
+      "learning_rate": 7.0449974910535916e-06,
+      "loss": 0.1851,
+      "step": 10576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0304051087015562,
+      "learning_rate": 7.043219733031452e-06,
+      "loss": 0.627,
+      "step": 10577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0139546657916503,
+      "learning_rate": 7.041442077402487e-06,
+      "loss": 0.6937,
+      "step": 10578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30342367295446254,
+      "learning_rate": 7.0396645242282535e-06,
+      "loss": 0.2022,
+      "step": 10579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3986109508000254,
+      "learning_rate": 7.037887073570313e-06,
+      "loss": 0.3156,
+      "step": 10580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4877704840722932,
+      "learning_rate": 7.036109725490214e-06,
+      "loss": 0.2859,
+      "step": 10581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30519393626632985,
+      "learning_rate": 7.03433248004951e-06,
+      "loss": 0.1873,
+      "step": 10582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41057454427454104,
+      "learning_rate": 7.032555337309743e-06,
+      "loss": 0.3229,
+      "step": 10583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3539347623622069,
+      "learning_rate": 7.030778297332457e-06,
+      "loss": 0.3001,
+      "step": 10584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3887406352092294,
+      "learning_rate": 7.0290013601791905e-06,
+      "loss": 0.2601,
+      "step": 10585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47380536198285,
+      "learning_rate": 7.027224525911479e-06,
+      "loss": 0.2907,
+      "step": 10586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4561919825253446,
+      "learning_rate": 7.025447794590856e-06,
+      "loss": 0.3529,
+      "step": 10587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24317420310642943,
+      "learning_rate": 7.023671166278845e-06,
+      "loss": 0.2204,
+      "step": 10588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30300578946259576,
+      "learning_rate": 7.021894641036977e-06,
+      "loss": 0.1961,
+      "step": 10589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7928562012811823,
+      "learning_rate": 7.020118218926767e-06,
+      "loss": 0.5509,
+      "step": 10590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31281127151054094,
+      "learning_rate": 7.018341900009738e-06,
+      "loss": 0.2595,
+      "step": 10591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.378034705279621,
+      "learning_rate": 7.0165656843473965e-06,
+      "loss": 0.2803,
+      "step": 10592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9236198996788825,
+      "learning_rate": 7.0147895720012596e-06,
+      "loss": 0.5131,
+      "step": 10593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24966395258526514,
+      "learning_rate": 7.01301356303283e-06,
+      "loss": 0.1917,
+      "step": 10594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9200006479669353,
+      "learning_rate": 7.011237657503615e-06,
+      "loss": 0.5364,
+      "step": 10595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2708229012641202,
+      "learning_rate": 7.009461855475111e-06,
+      "loss": 0.2601,
+      "step": 10596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42431345222551275,
+      "learning_rate": 7.00768615700881e-06,
+      "loss": 0.3102,
+      "step": 10597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49933535085134173,
+      "learning_rate": 7.005910562166213e-06,
+      "loss": 0.3538,
+      "step": 10598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32711074269410184,
+      "learning_rate": 7.004135071008803e-06,
+      "loss": 0.244,
+      "step": 10599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3911638518785674,
+      "learning_rate": 7.0023596835980676e-06,
+      "loss": 0.2621,
+      "step": 10600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.549319709524697,
+      "learning_rate": 7.000584399995486e-06,
+      "loss": 0.4431,
+      "step": 10601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4383392566950631,
+      "learning_rate": 6.998809220262541e-06,
+      "loss": 0.2549,
+      "step": 10602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.39079099029956776,
+      "learning_rate": 6.997034144460702e-06,
+      "loss": 0.2778,
+      "step": 10603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24881046372954682,
+      "learning_rate": 6.995259172651441e-06,
+      "loss": 0.2452,
+      "step": 10604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0911990822931978,
+      "learning_rate": 6.993484304896225e-06,
+      "loss": 0.2585,
+      "step": 10605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3859135519020631,
+      "learning_rate": 6.991709541256517e-06,
+      "loss": 0.271,
+      "step": 10606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38547240651856923,
+      "learning_rate": 6.98993488179378e-06,
+      "loss": 0.328,
+      "step": 10607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.496427375438408,
+      "learning_rate": 6.988160326569471e-06,
+      "loss": 0.2881,
+      "step": 10608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2807034932115483,
+      "learning_rate": 6.986385875645036e-06,
+      "loss": 0.2477,
+      "step": 10609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36666726706200375,
+      "learning_rate": 6.984611529081931e-06,
+      "loss": 0.2671,
+      "step": 10610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3784004456555454,
+      "learning_rate": 6.982837286941598e-06,
+      "loss": 0.2992,
+      "step": 10611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2947344402549206,
+      "learning_rate": 6.981063149285481e-06,
+      "loss": 0.1862,
+      "step": 10612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2011332586742978,
+      "learning_rate": 6.979289116175014e-06,
+      "loss": 0.7797,
+      "step": 10613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6269447977839719,
+      "learning_rate": 6.977515187671639e-06,
+      "loss": 0.3863,
+      "step": 10614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28525954215061056,
+      "learning_rate": 6.975741363836781e-06,
+      "loss": 0.2034,
+      "step": 10615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23251089488001686,
+      "learning_rate": 6.973967644731872e-06,
+      "loss": 0.2167,
+      "step": 10616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.136268898483301,
+      "learning_rate": 6.972194030418329e-06,
+      "loss": 0.4865,
+      "step": 10617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35214424867594674,
+      "learning_rate": 6.97042052095758e-06,
+      "loss": 0.1561,
+      "step": 10618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3186430841898647,
+      "learning_rate": 6.968647116411036e-06,
+      "loss": 0.2863,
+      "step": 10619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5067307018719389,
+      "learning_rate": 6.966873816840114e-06,
+      "loss": 0.3595,
+      "step": 10620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33263369625653055,
+      "learning_rate": 6.96510062230622e-06,
+      "loss": 0.1733,
+      "step": 10621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23286758847398575,
+      "learning_rate": 6.963327532870763e-06,
+      "loss": 0.1742,
+      "step": 10622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4730282585689458,
+      "learning_rate": 6.961554548595142e-06,
+      "loss": 0.326,
+      "step": 10623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3619899467460046,
+      "learning_rate": 6.959781669540754e-06,
+      "loss": 0.2461,
+      "step": 10624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3841805210107079,
+      "learning_rate": 6.958008895769e-06,
+      "loss": 0.2792,
+      "step": 10625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7631851504994283,
+      "learning_rate": 6.956236227341262e-06,
+      "loss": 0.5202,
+      "step": 10626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2999124611334779,
+      "learning_rate": 6.954463664318937e-06,
+      "loss": 0.2546,
+      "step": 10627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24820162596488865,
+      "learning_rate": 6.952691206763402e-06,
+      "loss": 0.1693,
+      "step": 10628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2347189617039587,
+      "learning_rate": 6.950918854736041e-06,
+      "loss": 0.5083,
+      "step": 10629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5522730367103457,
+      "learning_rate": 6.949146608298227e-06,
+      "loss": 0.3568,
+      "step": 10630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3104954162518697,
+      "learning_rate": 6.947374467511336e-06,
+      "loss": 0.2486,
+      "step": 10631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4520113873626869,
+      "learning_rate": 6.945602432436736e-06,
+      "loss": 0.3982,
+      "step": 10632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4093916598564017,
+      "learning_rate": 6.9438305031357935e-06,
+      "loss": 0.2927,
+      "step": 10633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3338504922517524,
+      "learning_rate": 6.9420586796698655e-06,
+      "loss": 0.219,
+      "step": 10634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25665018793378996,
+      "learning_rate": 6.940286962100318e-06,
+      "loss": 0.1841,
+      "step": 10635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6598078007493531,
+      "learning_rate": 6.938515350488503e-06,
+      "loss": 0.4409,
+      "step": 10636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3300845858986062,
+      "learning_rate": 6.936743844895768e-06,
+      "loss": 0.2797,
+      "step": 10637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4680762857712185,
+      "learning_rate": 6.934972445383459e-06,
+      "loss": 0.2779,
+      "step": 10638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4200145663298372,
+      "learning_rate": 6.933201152012925e-06,
+      "loss": 0.3078,
+      "step": 10639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27962340097798727,
+      "learning_rate": 6.931429964845501e-06,
+      "loss": 0.2241,
+      "step": 10640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53746389745485,
+      "learning_rate": 6.929658883942527e-06,
+      "loss": 0.1577,
+      "step": 10641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4959989835213774,
+      "learning_rate": 6.927887909365333e-06,
+      "loss": 0.3364,
+      "step": 10642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29723094294413305,
+      "learning_rate": 6.92611704117525e-06,
+      "loss": 0.2759,
+      "step": 10643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8817731319299101,
+      "learning_rate": 6.924346279433599e-06,
+      "loss": 0.4836,
+      "step": 10644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.603693846570787,
+      "learning_rate": 6.922575624201706e-06,
+      "loss": 0.379,
+      "step": 10645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31016181140642807,
+      "learning_rate": 6.920805075540886e-06,
+      "loss": 0.2332,
+      "step": 10646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27221079855815977,
+      "learning_rate": 6.919034633512456e-06,
+      "loss": 0.2433,
+      "step": 10647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6010295063712859,
+      "learning_rate": 6.917264298177724e-06,
+      "loss": 0.267,
+      "step": 10648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4000427804407078,
+      "learning_rate": 6.915494069597993e-06,
+      "loss": 0.3225,
+      "step": 10649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0407707069428398,
+      "learning_rate": 6.913723947834574e-06,
+      "loss": 0.6438,
+      "step": 10650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2765627603584837,
+      "learning_rate": 6.9119539329487585e-06,
+      "loss": 0.2274,
+      "step": 10651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5224307700746306,
+      "learning_rate": 6.9101840250018485e-06,
+      "loss": 0.3428,
+      "step": 10652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35898777209765914,
+      "learning_rate": 6.908414224055129e-06,
+      "loss": 0.2134,
+      "step": 10653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7316390060143079,
+      "learning_rate": 6.906644530169896e-06,
+      "loss": 0.2693,
+      "step": 10654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2562741289721085,
+      "learning_rate": 6.904874943407427e-06,
+      "loss": 0.2569,
+      "step": 10655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.057082723143132,
+      "learning_rate": 6.903105463829007e-06,
+      "loss": 0.7312,
+      "step": 10656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6832390986635581,
+      "learning_rate": 6.901336091495912e-06,
+      "loss": 0.159,
+      "step": 10657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2674851553546089,
+      "learning_rate": 6.899566826469415e-06,
+      "loss": 0.2218,
+      "step": 10658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3579968138322811,
+      "learning_rate": 6.897797668810784e-06,
+      "loss": 0.2911,
+      "step": 10659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5464163869475679,
+      "learning_rate": 6.896028618581287e-06,
+      "loss": 0.3132,
+      "step": 10660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3000321311735532,
+      "learning_rate": 6.894259675842188e-06,
+      "loss": 0.2106,
+      "step": 10661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2299271449853884,
+      "learning_rate": 6.892490840654739e-06,
+      "loss": 0.8311,
+      "step": 10662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30179860641260736,
+      "learning_rate": 6.890722113080201e-06,
+      "loss": 0.2821,
+      "step": 10663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34845039622272467,
+      "learning_rate": 6.888953493179819e-06,
+      "loss": 0.2284,
+      "step": 10664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5995945108100336,
+      "learning_rate": 6.88718498101485e-06,
+      "loss": 0.3831,
+      "step": 10665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25283872062952506,
+      "learning_rate": 6.885416576646525e-06,
+      "loss": 0.1791,
+      "step": 10666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2785774684194344,
+      "learning_rate": 6.883648280136094e-06,
+      "loss": 0.223,
+      "step": 10667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1132804366543598,
+      "learning_rate": 6.881880091544786e-06,
+      "loss": 0.7453,
+      "step": 10668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8623000767888052,
+      "learning_rate": 6.880112010933839e-06,
+      "loss": 0.5088,
+      "step": 10669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3251227489998159,
+      "learning_rate": 6.878344038364481e-06,
+      "loss": 0.2274,
+      "step": 10670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36645085483371054,
+      "learning_rate": 6.8765761738979305e-06,
+      "loss": 0.3253,
+      "step": 10671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26849276242470155,
+      "learning_rate": 6.874808417595415e-06,
+      "loss": 0.1771,
+      "step": 10672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32862638825590396,
+      "learning_rate": 6.87304076951815e-06,
+      "loss": 0.2591,
+      "step": 10673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4298652986509392,
+      "learning_rate": 6.871273229727346e-06,
+      "loss": 0.2987,
+      "step": 10674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5045099892284276,
+      "learning_rate": 6.869505798284217e-06,
+      "loss": 0.3762,
+      "step": 10675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32738198225492715,
+      "learning_rate": 6.867738475249967e-06,
+      "loss": 0.28,
+      "step": 10676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6053803182315074,
+      "learning_rate": 6.8659712606858e-06,
+      "loss": 0.292,
+      "step": 10677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2784146187816311,
+      "learning_rate": 6.8642041546529115e-06,
+      "loss": 0.2397,
+      "step": 10678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25915374026684634,
+      "learning_rate": 6.8624371572125e-06,
+      "loss": 0.1908,
+      "step": 10679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1034569399626564,
+      "learning_rate": 6.860670268425754e-06,
+      "loss": 0.5397,
+      "step": 10680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6580196847714953,
+      "learning_rate": 6.858903488353863e-06,
+      "loss": 0.3914,
+      "step": 10681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3087985735861544,
+      "learning_rate": 6.857136817058007e-06,
+      "loss": 0.2731,
+      "step": 10682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30721390171367746,
+      "learning_rate": 6.855370254599369e-06,
+      "loss": 0.2289,
+      "step": 10683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22683232782377963,
+      "learning_rate": 6.853603801039124e-06,
+      "loss": 0.156,
+      "step": 10684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3619861963806639,
+      "learning_rate": 6.8518374564384434e-06,
+      "loss": 0.2869,
+      "step": 10685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2769919962248444,
+      "learning_rate": 6.850071220858496e-06,
+      "loss": 0.6362,
+      "step": 10686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36211924335233864,
+      "learning_rate": 6.84830509436045e-06,
+      "loss": 0.2483,
+      "step": 10687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.372473918353896,
+      "learning_rate": 6.846539077005461e-06,
+      "loss": 0.2733,
+      "step": 10688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7227237641851568,
+      "learning_rate": 6.844773168854686e-06,
+      "loss": 0.4258,
+      "step": 10689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25478526632383197,
+      "learning_rate": 6.843007369969283e-06,
+      "loss": 0.1721,
+      "step": 10690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2592583923802607,
+      "learning_rate": 6.841241680410398e-06,
+      "loss": 0.2386,
+      "step": 10691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.417897430422118,
+      "learning_rate": 6.83947610023918e-06,
+      "loss": 0.3193,
+      "step": 10692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8573806813691548,
+      "learning_rate": 6.837710629516765e-06,
+      "loss": 0.3226,
+      "step": 10693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32550275471629125,
+      "learning_rate": 6.835945268304298e-06,
+      "loss": 0.2556,
+      "step": 10694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3535437138265614,
+      "learning_rate": 6.834180016662908e-06,
+      "loss": 0.3155,
+      "step": 10695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9949706011291491,
+      "learning_rate": 6.8324148746537286e-06,
+      "loss": 0.3106,
+      "step": 10696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31547765324720395,
+      "learning_rate": 6.830649842337885e-06,
+      "loss": 0.2539,
+      "step": 10697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0524343916133838,
+      "learning_rate": 6.828884919776504e-06,
+      "loss": 0.5325,
+      "step": 10698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44576274356069756,
+      "learning_rate": 6.827120107030698e-06,
+      "loss": 0.3259,
+      "step": 10699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22639385024816566,
+      "learning_rate": 6.82535540416159e-06,
+      "loss": 0.1372,
+      "step": 10700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1446799361802587,
+      "learning_rate": 6.823590811230287e-06,
+      "loss": 0.6339,
+      "step": 10701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3551463442681128,
+      "learning_rate": 6.821826328297896e-06,
+      "loss": 0.3097,
+      "step": 10702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34226727600361334,
+      "learning_rate": 6.820061955425527e-06,
+      "loss": 0.1847,
+      "step": 10703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5462335453113937,
+      "learning_rate": 6.818297692674273e-06,
+      "loss": 0.3826,
+      "step": 10704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41512218730788686,
+      "learning_rate": 6.81653354010523e-06,
+      "loss": 0.2327,
+      "step": 10705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2849103841623909,
+      "learning_rate": 6.8147694977794975e-06,
+      "loss": 0.1901,
+      "step": 10706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35804096430942806,
+      "learning_rate": 6.813005565758158e-06,
+      "loss": 0.3045,
+      "step": 10707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.959727916273778,
+      "learning_rate": 6.8112417441022995e-06,
+      "loss": 0.5161,
+      "step": 10708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41141512882499015,
+      "learning_rate": 6.809478032873002e-06,
+      "loss": 0.3103,
+      "step": 10709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28673660446727955,
+      "learning_rate": 6.807714432131343e-06,
+      "loss": 0.2599,
+      "step": 10710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6706818173645254,
+      "learning_rate": 6.805950941938395e-06,
+      "loss": 0.4362,
+      "step": 10711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2853954047662548,
+      "learning_rate": 6.804187562355231e-06,
+      "loss": 0.2041,
+      "step": 10712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3040463174783864,
+      "learning_rate": 6.802424293442914e-06,
+      "loss": 0.1679,
+      "step": 10713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3824192547178403,
+      "learning_rate": 6.800661135262505e-06,
+      "loss": 0.3014,
+      "step": 10714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3310073163699963,
+      "learning_rate": 6.7988980878750636e-06,
+      "loss": 0.2803,
+      "step": 10715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.900982088435711,
+      "learning_rate": 6.797135151341643e-06,
+      "loss": 0.3219,
+      "step": 10716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42156896906947466,
+      "learning_rate": 6.7953723257232955e-06,
+      "loss": 0.3128,
+      "step": 10717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2688101831184305,
+      "learning_rate": 6.793609611081064e-06,
+      "loss": 0.2541,
+      "step": 10718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3140732036560362,
+      "learning_rate": 6.791847007475998e-06,
+      "loss": 0.1823,
+      "step": 10719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0008202280028524,
+      "learning_rate": 6.7900845149691285e-06,
+      "loss": 0.631,
+      "step": 10720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42576299687385705,
+      "learning_rate": 6.7883221336214965e-06,
+      "loss": 0.2948,
+      "step": 10721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42623039582871874,
+      "learning_rate": 6.7865598634941295e-06,
+      "loss": 0.3172,
+      "step": 10722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45379008885941363,
+      "learning_rate": 6.784797704648058e-06,
+      "loss": 0.2807,
+      "step": 10723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3814091792954231,
+      "learning_rate": 6.7830356571443016e-06,
+      "loss": 0.2453,
+      "step": 10724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2705471683167067,
+      "learning_rate": 6.7812737210438836e-06,
+      "loss": 0.1876,
+      "step": 10725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3377882881037869,
+      "learning_rate": 6.77951189640782e-06,
+      "loss": 0.2592,
+      "step": 10726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3912153744933908,
+      "learning_rate": 6.777750183297117e-06,
+      "loss": 0.26,
+      "step": 10727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43342392314247236,
+      "learning_rate": 6.77598858177279e-06,
+      "loss": 0.3417,
+      "step": 10728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9386708011142606,
+      "learning_rate": 6.774227091895835e-06,
+      "loss": 0.3769,
+      "step": 10729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3215869104097005,
+      "learning_rate": 6.772465713727262e-06,
+      "loss": 0.2572,
+      "step": 10730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21444487048175678,
+      "learning_rate": 6.77070444732806e-06,
+      "loss": 0.1924,
+      "step": 10731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7105654768008676,
+      "learning_rate": 6.768943292759226e-06,
+      "loss": 0.3145,
+      "step": 10732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8871355072259255,
+      "learning_rate": 6.767182250081744e-06,
+      "loss": 0.2807,
+      "step": 10733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36136097231280245,
+      "learning_rate": 6.765421319356605e-06,
+      "loss": 0.3122,
+      "step": 10734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45744692480747945,
+      "learning_rate": 6.763660500644783e-06,
+      "loss": 0.3692,
+      "step": 10735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31629793854850224,
+      "learning_rate": 6.761899794007262e-06,
+      "loss": 0.2086,
+      "step": 10736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25178578119165373,
+      "learning_rate": 6.760139199505014e-06,
+      "loss": 0.1469,
+      "step": 10737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3273390465451645,
+      "learning_rate": 6.758378717199004e-06,
+      "loss": 0.2983,
+      "step": 10738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.58071175528015,
+      "learning_rate": 6.756618347150196e-06,
+      "loss": 0.2238,
+      "step": 10739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3900166059812892,
+      "learning_rate": 6.7548580894195585e-06,
+      "loss": 0.3206,
+      "step": 10740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5014095471100632,
+      "learning_rate": 6.753097944068043e-06,
+      "loss": 0.6091,
+      "step": 10741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27374873890951196,
+      "learning_rate": 6.7513379111566105e-06,
+      "loss": 0.2153,
+      "step": 10742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2331727972943417,
+      "learning_rate": 6.749577990746202e-06,
+      "loss": 0.2058,
+      "step": 10743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6106539407307946,
+      "learning_rate": 6.74781818289777e-06,
+      "loss": 0.3983,
+      "step": 10744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3191253890892391,
+      "learning_rate": 6.746058487672253e-06,
+      "loss": 0.1622,
+      "step": 10745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2832733787647948,
+      "learning_rate": 6.744298905130593e-06,
+      "loss": 0.2955,
+      "step": 10746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2000591687611528,
+      "learning_rate": 6.74253943533372e-06,
+      "loss": 0.7825,
+      "step": 10747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5827279940418081,
+      "learning_rate": 6.740780078342568e-06,
+      "loss": 0.304,
+      "step": 10748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2421944177584202,
+      "learning_rate": 6.7390208342180595e-06,
+      "loss": 0.1781,
+      "step": 10749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3432118231078927,
+      "learning_rate": 6.737261703021123e-06,
+      "loss": 0.3119,
+      "step": 10750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6051458957393929,
+      "learning_rate": 6.735502684812669e-06,
+      "loss": 0.371,
+      "step": 10751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4037052100199354,
+      "learning_rate": 6.73374377965362e-06,
+      "loss": 0.2492,
+      "step": 10752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9394090165245456,
+      "learning_rate": 6.731984987604882e-06,
+      "loss": 0.5348,
+      "step": 10753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29799092639937913,
+      "learning_rate": 6.730226308727363e-06,
+      "loss": 0.2664,
+      "step": 10754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44911316464676126,
+      "learning_rate": 6.728467743081968e-06,
+      "loss": 0.2381,
+      "step": 10755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3233394168263766,
+      "learning_rate": 6.726709290729592e-06,
+      "loss": 0.2382,
+      "step": 10756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5916566364028347,
+      "learning_rate": 6.724950951731135e-06,
+      "loss": 0.3323,
+      "step": 10757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2714572194543101,
+      "learning_rate": 6.723192726147482e-06,
+      "loss": 0.2394,
+      "step": 10758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0438216384131889,
+      "learning_rate": 6.721434614039528e-06,
+      "loss": 0.6214,
+      "step": 10759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8915962859834544,
+      "learning_rate": 6.719676615468149e-06,
+      "loss": 0.4534,
+      "step": 10760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3335692130553307,
+      "learning_rate": 6.717918730494231e-06,
+      "loss": 0.2893,
+      "step": 10761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23904715452470016,
+      "learning_rate": 6.716160959178644e-06,
+      "loss": 0.2044,
+      "step": 10762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.77472616741585,
+      "learning_rate": 6.714403301582263e-06,
+      "loss": 0.4212,
+      "step": 10763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3341935232637106,
+      "learning_rate": 6.712645757765952e-06,
+      "loss": 0.272,
+      "step": 10764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8951143831388909,
+      "learning_rate": 6.710888327790581e-06,
+      "loss": 0.3213,
+      "step": 10765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29522917110829133,
+      "learning_rate": 6.709131011717005e-06,
+      "loss": 0.2728,
+      "step": 10766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3995050553221213,
+      "learning_rate": 6.707373809606077e-06,
+      "loss": 0.3016,
+      "step": 10767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33594474411352,
+      "learning_rate": 6.705616721518655e-06,
+      "loss": 0.1888,
+      "step": 10768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2893697612282927,
+      "learning_rate": 6.703859747515584e-06,
+      "loss": 0.1952,
+      "step": 10769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27035597075734846,
+      "learning_rate": 6.702102887657709e-06,
+      "loss": 0.2757,
+      "step": 10770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8535729454967663,
+      "learning_rate": 6.7003461420058715e-06,
+      "loss": 0.3506,
+      "step": 10771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5549308852351893,
+      "learning_rate": 6.6985895106209005e-06,
+      "loss": 0.3496,
+      "step": 10772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36265593609148106,
+      "learning_rate": 6.696832993563636e-06,
+      "loss": 0.2992,
+      "step": 10773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3523666206537711,
+      "learning_rate": 6.695076590894899e-06,
+      "loss": 0.2899,
+      "step": 10774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2912704248423908,
+      "learning_rate": 6.693320302675521e-06,
+      "loss": 0.1405,
+      "step": 10775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3451346608345089,
+      "learning_rate": 6.6915641289663154e-06,
+      "loss": 0.2796,
+      "step": 10776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3192817030035515,
+      "learning_rate": 6.689808069828105e-06,
+      "loss": 0.2486,
+      "step": 10777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38933740283548857,
+      "learning_rate": 6.688052125321698e-06,
+      "loss": 0.2439,
+      "step": 10778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3694365237432283,
+      "learning_rate": 6.686296295507903e-06,
+      "loss": 0.2851,
+      "step": 10779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4782734090964818,
+      "learning_rate": 6.684540580447525e-06,
+      "loss": 0.2538,
+      "step": 10780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4149927816314323,
+      "learning_rate": 6.682784980201363e-06,
+      "loss": 0.2624,
+      "step": 10781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2933724186560608,
+      "learning_rate": 6.6810294948302165e-06,
+      "loss": 0.2648,
+      "step": 10782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.46882515485188336,
+      "learning_rate": 6.679274124394874e-06,
+      "loss": 0.2259,
+      "step": 10783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5844321532824087,
+      "learning_rate": 6.677518868956128e-06,
+      "loss": 0.2892,
+      "step": 10784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29395991528266874,
+      "learning_rate": 6.675763728574758e-06,
+      "loss": 0.2712,
+      "step": 10785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.47841665979336606,
+      "learning_rate": 6.67400870331155e-06,
+      "loss": 0.3708,
+      "step": 10786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7244194775477389,
+      "learning_rate": 6.672253793227273e-06,
+      "loss": 0.4866,
+      "step": 10787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3039119401991573,
+      "learning_rate": 6.670498998382708e-06,
+      "loss": 0.2247,
+      "step": 10788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35386920706771574,
+      "learning_rate": 6.668744318838618e-06,
+      "loss": 0.2584,
+      "step": 10789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31571310808178393,
+      "learning_rate": 6.66698975465577e-06,
+      "loss": 0.2504,
+      "step": 10790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3214123531686496,
+      "learning_rate": 6.665235305894925e-06,
+      "loss": 0.2037,
+      "step": 10791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.902925238719976,
+      "learning_rate": 6.663480972616835e-06,
+      "loss": 0.4962,
+      "step": 10792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3389188225905232,
+      "learning_rate": 6.661726754882256e-06,
+      "loss": 0.2932,
+      "step": 10793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3137494677922765,
+      "learning_rate": 6.659972652751936e-06,
+      "loss": 0.2132,
+      "step": 10794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39839423898137716,
+      "learning_rate": 6.658218666286621e-06,
+      "loss": 0.3217,
+      "step": 10795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.40810260556052064,
+      "learning_rate": 6.656464795547048e-06,
+      "loss": 0.2755,
+      "step": 10796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34635085809256516,
+      "learning_rate": 6.654711040593957e-06,
+      "loss": 0.2657,
+      "step": 10797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3303025437976195,
+      "learning_rate": 6.652957401488076e-06,
+      "loss": 0.2489,
+      "step": 10798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1771107826528375,
+      "learning_rate": 6.651203878290139e-06,
+      "loss": 0.7729,
+      "step": 10799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3086966114240212,
+      "learning_rate": 6.649450471060865e-06,
+      "loss": 0.245,
+      "step": 10800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42800613630194834,
+      "learning_rate": 6.64769717986098e-06,
+      "loss": 0.2697,
+      "step": 10801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26776956632235843,
+      "learning_rate": 6.6459440047511955e-06,
+      "loss": 0.2168,
+      "step": 10802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3370355453206775,
+      "learning_rate": 6.6441909457922286e-06,
+      "loss": 0.2665,
+      "step": 10803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0228746361906365,
+      "learning_rate": 6.642438003044781e-06,
+      "loss": 0.3131,
+      "step": 10804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35770512175069535,
+      "learning_rate": 6.640685176569568e-06,
+      "loss": 0.324,
+      "step": 10805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31252442730763696,
+      "learning_rate": 6.638932466427277e-06,
+      "loss": 0.248,
+      "step": 10806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7175209362515863,
+      "learning_rate": 6.637179872678612e-06,
+      "loss": 0.3296,
+      "step": 10807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23272726866411692,
+      "learning_rate": 6.635427395384262e-06,
+      "loss": 0.1369,
+      "step": 10808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4007091085654962,
+      "learning_rate": 6.633675034604918e-06,
+      "loss": 0.2966,
+      "step": 10809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.348395608259044,
+      "learning_rate": 6.6319227904012605e-06,
+      "loss": 0.2952,
+      "step": 10810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7075777072812005,
+      "learning_rate": 6.630170662833974e-06,
+      "loss": 0.2966,
+      "step": 10811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.359803224019601,
+      "learning_rate": 6.62841865196373e-06,
+      "loss": 0.2858,
+      "step": 10812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3924621489778972,
+      "learning_rate": 6.626666757851208e-06,
+      "loss": 0.3285,
+      "step": 10813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3379307119515423,
+      "learning_rate": 6.624914980557067e-06,
+      "loss": 0.1459,
+      "step": 10814
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25824750797348944,
+      "learning_rate": 6.623163320141977e-06,
+      "loss": 0.2055,
+      "step": 10815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9254204327897717,
+      "learning_rate": 6.621411776666593e-06,
+      "loss": 0.5132,
+      "step": 10816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3290210750926486,
+      "learning_rate": 6.619660350191577e-06,
+      "loss": 0.2528,
+      "step": 10817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3258960929973952,
+      "learning_rate": 6.617909040777578e-06,
+      "loss": 0.2723,
+      "step": 10818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6898961033010748,
+      "learning_rate": 6.6161578484852405e-06,
+      "loss": 0.4741,
+      "step": 10819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1955724790230162,
+      "learning_rate": 6.614406773375215e-06,
+      "loss": 0.0904,
+      "step": 10820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31062886283654306,
+      "learning_rate": 6.612655815508135e-06,
+      "loss": 0.2714,
+      "step": 10821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5003777175917457,
+      "learning_rate": 6.610904974944638e-06,
+      "loss": 0.3493,
+      "step": 10822
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6498721179269221,
+      "learning_rate": 6.609154251745356e-06,
+      "loss": 0.4102,
+      "step": 10823
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31094275540668453,
+      "learning_rate": 6.607403645970919e-06,
+      "loss": 0.2138,
+      "step": 10824
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35222845924002294,
+      "learning_rate": 6.605653157681945e-06,
+      "loss": 0.308,
+      "step": 10825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5014401149672492,
+      "learning_rate": 6.603902786939058e-06,
+      "loss": 0.3462,
+      "step": 10826
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23552220497581033,
+      "learning_rate": 6.60215253380287e-06,
+      "loss": 0.1319,
+      "step": 10827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7787992405401837,
+      "learning_rate": 6.600402398333995e-06,
+      "loss": 0.3407,
+      "step": 10828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36508668218733975,
+      "learning_rate": 6.598652380593037e-06,
+      "loss": 0.3031,
+      "step": 10829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3085726407226889,
+      "learning_rate": 6.596902480640603e-06,
+      "loss": 0.1881,
+      "step": 10830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7405743438567102,
+      "learning_rate": 6.595152698537289e-06,
+      "loss": 0.5434,
+      "step": 10831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4452690524635368,
+      "learning_rate": 6.59340303434369e-06,
+      "loss": 0.7422,
+      "step": 10832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21649306649578795,
+      "learning_rate": 6.591653488120398e-06,
+      "loss": 0.1843,
+      "step": 10833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34346378810527417,
+      "learning_rate": 6.589904059927998e-06,
+      "loss": 0.2565,
+      "step": 10834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.692637088445479,
+      "learning_rate": 6.588154749827076e-06,
+      "loss": 0.4296,
+      "step": 10835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41281261424206633,
+      "learning_rate": 6.586405557878206e-06,
+      "loss": 0.287,
+      "step": 10836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32134511517364006,
+      "learning_rate": 6.584656484141967e-06,
+      "loss": 0.2603,
+      "step": 10837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.80282010542682,
+      "learning_rate": 6.582907528678928e-06,
+      "loss": 0.4662,
+      "step": 10838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33148783656373965,
+      "learning_rate": 6.5811586915496515e-06,
+      "loss": 0.2478,
+      "step": 10839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37101099026547674,
+      "learning_rate": 6.579409972814703e-06,
+      "loss": 0.1527,
+      "step": 10840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3358065314278554,
+      "learning_rate": 6.577661372534639e-06,
+      "loss": 0.3153,
+      "step": 10841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3190398698999832,
+      "learning_rate": 6.575912890770017e-06,
+      "loss": 0.2436,
+      "step": 10842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8837230043317458,
+      "learning_rate": 6.574164527581383e-06,
+      "loss": 0.4759,
+      "step": 10843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.47649999430816803,
+      "learning_rate": 6.5724162830292835e-06,
+      "loss": 0.3134,
+      "step": 10844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3922580154206632,
+      "learning_rate": 6.570668157174263e-06,
+      "loss": 0.2656,
+      "step": 10845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23499303324710186,
+      "learning_rate": 6.568920150076854e-06,
+      "loss": 0.1829,
+      "step": 10846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6347531042335608,
+      "learning_rate": 6.567172261797594e-06,
+      "loss": 0.417,
+      "step": 10847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39361758639994104,
+      "learning_rate": 6.5654244923970105e-06,
+      "loss": 0.2663,
+      "step": 10848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3087842118794902,
+      "learning_rate": 6.56367684193563e-06,
+      "loss": 0.2918,
+      "step": 10849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0242043905970053,
+      "learning_rate": 6.561929310473971e-06,
+      "loss": 0.2512,
+      "step": 10850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4242460543027238,
+      "learning_rate": 6.560181898072554e-06,
+      "loss": 0.3042,
+      "step": 10851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5240664744635831,
+      "learning_rate": 6.558434604791888e-06,
+      "loss": 0.3468,
+      "step": 10852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22341814224458637,
+      "learning_rate": 6.556687430692486e-06,
+      "loss": 0.1698,
+      "step": 10853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39096154916868614,
+      "learning_rate": 6.5549403758348485e-06,
+      "loss": 0.2807,
+      "step": 10854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5309980404520158,
+      "learning_rate": 6.553193440279479e-06,
+      "loss": 0.3881,
+      "step": 10855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4725538900448848,
+      "learning_rate": 6.551446624086873e-06,
+      "loss": 0.2522,
+      "step": 10856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3080050015821684,
+      "learning_rate": 6.549699927317519e-06,
+      "loss": 0.2818,
+      "step": 10857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5715657627569951,
+      "learning_rate": 6.5479533500319105e-06,
+      "loss": 0.35,
+      "step": 10858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2846118138159484,
+      "learning_rate": 6.546206892290527e-06,
+      "loss": 0.1529,
+      "step": 10859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3320757102260107,
+      "learning_rate": 6.544460554153853e-06,
+      "loss": 0.2813,
+      "step": 10860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36196437928371844,
+      "learning_rate": 6.542714335682359e-06,
+      "loss": 0.2973,
+      "step": 10861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.141389722693649,
+      "learning_rate": 6.54096823693652e-06,
+      "loss": 0.4734,
+      "step": 10862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6271265112294695,
+      "learning_rate": 6.5392222579768015e-06,
+      "loss": 0.2178,
+      "step": 10863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3753408561849779,
+      "learning_rate": 6.537476398863669e-06,
+      "loss": 0.313,
+      "step": 10864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33153725793470623,
+      "learning_rate": 6.535730659657577e-06,
+      "loss": 0.2618,
+      "step": 10865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32391570737586556,
+      "learning_rate": 6.533985040418988e-06,
+      "loss": 0.1546,
+      "step": 10866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3128620256105588,
+      "learning_rate": 6.532239541208343e-06,
+      "loss": 0.2564,
+      "step": 10867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4533892575766774,
+      "learning_rate": 6.530494162086098e-06,
+      "loss": 0.3441,
+      "step": 10868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28938108521272216,
+      "learning_rate": 6.5287489031126875e-06,
+      "loss": 0.2152,
+      "step": 10869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.46521166367836747,
+      "learning_rate": 6.527003764348555e-06,
+      "loss": 0.3301,
+      "step": 10870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1228207484521606,
+      "learning_rate": 6.5252587458541325e-06,
+      "loss": 0.5712,
+      "step": 10871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5512502762042788,
+      "learning_rate": 6.523513847689854e-06,
+      "loss": 0.3708,
+      "step": 10872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19166979867318232,
+      "learning_rate": 6.521769069916136e-06,
+      "loss": 0.1914,
+      "step": 10873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9560087983933109,
+      "learning_rate": 6.520024412593409e-06,
+      "loss": 0.4243,
+      "step": 10874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5336182223969617,
+      "learning_rate": 6.518279875782083e-06,
+      "loss": 0.3538,
+      "step": 10875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3886374618315691,
+      "learning_rate": 6.516535459542579e-06,
+      "loss": 0.2293,
+      "step": 10876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.40788284876238157,
+      "learning_rate": 6.514791163935299e-06,
+      "loss": 0.3177,
+      "step": 10877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5755561972390177,
+      "learning_rate": 6.513046989020653e-06,
+      "loss": 0.3556,
+      "step": 10878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23691729198980177,
+      "learning_rate": 6.5113029348590384e-06,
+      "loss": 0.152,
+      "step": 10879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4475169294764376,
+      "learning_rate": 6.509559001510854e-06,
+      "loss": 0.2962,
+      "step": 10880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7213839327976953,
+      "learning_rate": 6.5078151890364916e-06,
+      "loss": 0.3129,
+      "step": 10881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3567362019743413,
+      "learning_rate": 6.50607149749634e-06,
+      "loss": 0.2346,
+      "step": 10882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8685527444154926,
+      "learning_rate": 6.504327926950782e-06,
+      "loss": 0.5479,
+      "step": 10883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42927802250755426,
+      "learning_rate": 6.502584477460195e-06,
+      "loss": 0.3064,
+      "step": 10884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2919765365381944,
+      "learning_rate": 6.50084114908496e-06,
+      "loss": 0.2674,
+      "step": 10885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5838456550127608,
+      "learning_rate": 6.4990979418854436e-06,
+      "loss": 0.2418,
+      "step": 10886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33539401157915516,
+      "learning_rate": 6.497354855922016e-06,
+      "loss": 0.2074,
+      "step": 10887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.36912521934440096,
+      "learning_rate": 6.495611891255038e-06,
+      "loss": 0.2864,
+      "step": 10888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.332611184205583,
+      "learning_rate": 6.493869047944872e-06,
+      "loss": 0.2492,
+      "step": 10889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5711716515617774,
+      "learning_rate": 6.4921263260518664e-06,
+      "loss": 0.413,
+      "step": 10890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34918005553161635,
+      "learning_rate": 6.490383725636377e-06,
+      "loss": 0.2892,
+      "step": 10891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.321025479817607,
+      "learning_rate": 6.488641246758749e-06,
+      "loss": 0.2723,
+      "step": 10892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2324885317833231,
+      "learning_rate": 6.486898889479323e-06,
+      "loss": 0.1738,
+      "step": 10893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.38333854647224463,
+      "learning_rate": 6.485156653858438e-06,
+      "loss": 0.3174,
+      "step": 10894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.008010824991491,
+      "learning_rate": 6.483414539956426e-06,
+      "loss": 0.3672,
+      "step": 10895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27193926530077744,
+      "learning_rate": 6.48167254783362e-06,
+      "loss": 0.2538,
+      "step": 10896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41061131285869107,
+      "learning_rate": 6.479930677550338e-06,
+      "loss": 0.3058,
+      "step": 10897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6100862999061782,
+      "learning_rate": 6.478188929166909e-06,
+      "loss": 0.3987,
+      "step": 10898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.15737671574549414,
+      "learning_rate": 6.476447302743643e-06,
+      "loss": 0.0739,
+      "step": 10899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3672614455951174,
+      "learning_rate": 6.474705798340857e-06,
+      "loss": 0.3108,
+      "step": 10900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37121567073539297,
+      "learning_rate": 6.472964416018857e-06,
+      "loss": 0.3178,
+      "step": 10901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5386380238554062,
+      "learning_rate": 6.471223155837949e-06,
+      "loss": 0.2583,
+      "step": 10902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3702006272772363,
+      "learning_rate": 6.469482017858428e-06,
+      "loss": 0.2944,
+      "step": 10903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3343193983572874,
+      "learning_rate": 6.4677410021405975e-06,
+      "loss": 0.3281,
+      "step": 10904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.15994451590807493,
+      "learning_rate": 6.46600010874474e-06,
+      "loss": 0.0712,
+      "step": 10905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35321148685659404,
+      "learning_rate": 6.4642593377311515e-06,
+      "loss": 0.2867,
+      "step": 10906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1587678178893002,
+      "learning_rate": 6.462518689160109e-06,
+      "loss": 0.568,
+      "step": 10907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30479496232056075,
+      "learning_rate": 6.460778163091891e-06,
+      "loss": 0.2539,
+      "step": 10908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35727687245385015,
+      "learning_rate": 6.45903775958677e-06,
+      "loss": 0.2846,
+      "step": 10909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4649919245936944,
+      "learning_rate": 6.457297478705023e-06,
+      "loss": 0.7402,
+      "step": 10910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2619218553658152,
+      "learning_rate": 6.45555732050691e-06,
+      "loss": 0.1705,
+      "step": 10911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33591317094101153,
+      "learning_rate": 6.4538172850526955e-06,
+      "loss": 0.2024,
+      "step": 10912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.341833146387076,
+      "learning_rate": 6.452077372402634e-06,
+      "loss": 0.3078,
+      "step": 10913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5891021497920335,
+      "learning_rate": 6.450337582616983e-06,
+      "loss": 0.3681,
+      "step": 10914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34289810526304676,
+      "learning_rate": 6.448597915755988e-06,
+      "loss": 0.2137,
+      "step": 10915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.359263118619319,
+      "learning_rate": 6.446858371879896e-06,
+      "loss": 0.312,
+      "step": 10916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28334773046172257,
+      "learning_rate": 6.445118951048942e-06,
+      "loss": 0.173,
+      "step": 10917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30036622389977835,
+      "learning_rate": 6.44337965332337e-06,
+      "loss": 0.19,
+      "step": 10918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.749494957815434,
+      "learning_rate": 6.4416404787634045e-06,
+      "loss": 0.4418,
+      "step": 10919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3391009390947701,
+      "learning_rate": 6.439901427429278e-06,
+      "loss": 0.3278,
+      "step": 10920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4741224416315272,
+      "learning_rate": 6.438162499381212e-06,
+      "loss": 0.2293,
+      "step": 10921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9750360327136708,
+      "learning_rate": 6.4364236946794234e-06,
+      "loss": 0.6323,
+      "step": 10922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24827351464326222,
+      "learning_rate": 6.434685013384132e-06,
+      "loss": 0.1511,
+      "step": 10923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.288203382381012,
+      "learning_rate": 6.432946455555542e-06,
+      "loss": 0.2516,
+      "step": 10924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.463203852610262,
+      "learning_rate": 6.4312080212538665e-06,
+      "loss": 0.2672,
+      "step": 10925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7318010513536188,
+      "learning_rate": 6.4294697105393e-06,
+      "loss": 0.4274,
+      "step": 10926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3059976691995668,
+      "learning_rate": 6.427731523472047e-06,
+      "loss": 0.2783,
+      "step": 10927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3061553802177348,
+      "learning_rate": 6.425993460112297e-06,
+      "loss": 0.2512,
+      "step": 10928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2644528421377624,
+      "learning_rate": 6.424255520520239e-06,
+      "loss": 0.1606,
+      "step": 10929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3216069885908651,
+      "learning_rate": 6.422517704756057e-06,
+      "loss": 0.2645,
+      "step": 10930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0873383787909738,
+      "learning_rate": 6.420780012879937e-06,
+      "loss": 0.3225,
+      "step": 10931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32131116306748664,
+      "learning_rate": 6.419042444952048e-06,
+      "loss": 0.2816,
+      "step": 10932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3490888556275218,
+      "learning_rate": 6.417305001032567e-06,
+      "loss": 0.2824,
+      "step": 10933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0053545569436766,
+      "learning_rate": 6.415567681181658e-06,
+      "loss": 0.4126,
+      "step": 10934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3229247233746252,
+      "learning_rate": 6.413830485459488e-06,
+      "loss": 0.2473,
+      "step": 10935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3380905219858116,
+      "learning_rate": 6.412093413926213e-06,
+      "loss": 0.2688,
+      "step": 10936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.554045753814118,
+      "learning_rate": 6.410356466641989e-06,
+      "loss": 0.298,
+      "step": 10937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9802965063335598,
+      "learning_rate": 6.408619643666967e-06,
+      "loss": 0.3922,
+      "step": 10938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31719321797306044,
+      "learning_rate": 6.40688294506129e-06,
+      "loss": 0.2581,
+      "step": 10939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3305320623675954,
+      "learning_rate": 6.405146370885107e-06,
+      "loss": 0.2959,
+      "step": 10940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2577489342671599,
+      "learning_rate": 6.403409921198548e-06,
+      "loss": 0.1013,
+      "step": 10941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3603206692947732,
+      "learning_rate": 6.401673596061747e-06,
+      "loss": 0.2713,
+      "step": 10942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.004731951323438,
+      "learning_rate": 6.399937395534837e-06,
+      "loss": 0.4131,
+      "step": 10943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26455269733315795,
+      "learning_rate": 6.398201319677937e-06,
+      "loss": 0.2145,
+      "step": 10944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31962065915692844,
+      "learning_rate": 6.396465368551172e-06,
+      "loss": 0.2746,
+      "step": 10945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1313165310387425,
+      "learning_rate": 6.394729542214657e-06,
+      "loss": 0.606,
+      "step": 10946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33812898844419714,
+      "learning_rate": 6.392993840728503e-06,
+      "loss": 0.2592,
+      "step": 10947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3277772485530417,
+      "learning_rate": 6.391258264152818e-06,
+      "loss": 0.2676,
+      "step": 10948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7622917250054192,
+      "learning_rate": 6.389522812547701e-06,
+      "loss": 0.4147,
+      "step": 10949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2837893256968949,
+      "learning_rate": 6.3877874859732556e-06,
+      "loss": 0.1997,
+      "step": 10950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3097014346057528,
+      "learning_rate": 6.386052284489575e-06,
+      "loss": 0.1805,
+      "step": 10951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3016138484836072,
+      "learning_rate": 6.3843172081567474e-06,
+      "loss": 0.2933,
+      "step": 10952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7273665127231179,
+      "learning_rate": 6.382582257034858e-06,
+      "loss": 0.4313,
+      "step": 10953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3243297248965466,
+      "learning_rate": 6.380847431183992e-06,
+      "loss": 0.188,
+      "step": 10954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3351138381356627,
+      "learning_rate": 6.379112730664222e-06,
+      "loss": 0.2968,
+      "step": 10955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3233328725092727,
+      "learning_rate": 6.377378155535625e-06,
+      "loss": 0.2423,
+      "step": 10956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2918832243355452,
+      "learning_rate": 6.375643705858263e-06,
+      "loss": 0.2093,
+      "step": 10957
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0029877128585176,
+      "learning_rate": 6.373909381692207e-06,
+      "loss": 0.5905,
+      "step": 10958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5027694083886098,
+      "learning_rate": 6.372175183097511e-06,
+      "loss": 0.3369,
+      "step": 10959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3078451694449207,
+      "learning_rate": 6.370441110134233e-06,
+      "loss": 0.2701,
+      "step": 10960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41308557058341894,
+      "learning_rate": 6.3687071628624244e-06,
+      "loss": 0.2483,
+      "step": 10961
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45906823035018285,
+      "learning_rate": 6.366973341342128e-06,
+      "loss": 0.3574,
+      "step": 10962
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34703098652586084,
+      "learning_rate": 6.365239645633392e-06,
+      "loss": 0.2705,
+      "step": 10963
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26359333923079414,
+      "learning_rate": 6.3635060757962485e-06,
+      "loss": 0.1914,
+      "step": 10964
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5789237996497145,
+      "learning_rate": 6.361772631890735e-06,
+      "loss": 0.3915,
+      "step": 10965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3167200194383066,
+      "learning_rate": 6.360039313976875e-06,
+      "loss": 0.2645,
+      "step": 10966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3052671121934587,
+      "learning_rate": 6.3583061221147015e-06,
+      "loss": 0.2044,
+      "step": 10967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3444639163727557,
+      "learning_rate": 6.356573056364227e-06,
+      "loss": 0.3192,
+      "step": 10968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5307240095449588,
+      "learning_rate": 6.354840116785473e-06,
+      "loss": 0.3267,
+      "step": 10969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25389779069515805,
+      "learning_rate": 6.353107303438447e-06,
+      "loss": 0.1643,
+      "step": 10970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4045083838798069,
+      "learning_rate": 6.351374616383161e-06,
+      "loss": 0.3156,
+      "step": 10971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6028023578671253,
+      "learning_rate": 6.349642055679613e-06,
+      "loss": 0.3478,
+      "step": 10972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4182115635796257,
+      "learning_rate": 6.347909621387809e-06,
+      "loss": 0.2903,
+      "step": 10973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0543676184663318,
+      "learning_rate": 6.346177313567732e-06,
+      "loss": 0.4733,
+      "step": 10974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2926592072070201,
+      "learning_rate": 6.34444513227938e-06,
+      "loss": 0.2497,
+      "step": 10975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24753098131598456,
+      "learning_rate": 6.342713077582733e-06,
+      "loss": 0.2406,
+      "step": 10976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5753767227553651,
+      "learning_rate": 6.340981149537777e-06,
+      "loss": 0.3157,
+      "step": 10977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3953220853323611,
+      "learning_rate": 6.339249348204485e-06,
+      "loss": 0.2584,
+      "step": 10978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5473609875755407,
+      "learning_rate": 6.337517673642833e-06,
+      "loss": 0.3299,
+      "step": 10979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3745830109479958,
+      "learning_rate": 6.335786125912784e-06,
+      "loss": 0.2747,
+      "step": 10980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40675845180689163,
+      "learning_rate": 6.3340547050743055e-06,
+      "loss": 0.3007,
+      "step": 10981
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.690730580115114,
+      "learning_rate": 6.332323411187353e-06,
+      "loss": 0.3359,
+      "step": 10982
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2285847486188215,
+      "learning_rate": 6.330592244311885e-06,
+      "loss": 0.1904,
+      "step": 10983
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3287690816760705,
+      "learning_rate": 6.328861204507848e-06,
+      "loss": 0.2549,
+      "step": 10984
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4816162350887843,
+      "learning_rate": 6.327130291835192e-06,
+      "loss": 0.8568,
+      "step": 10985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1107702877033343,
+      "learning_rate": 6.325399506353855e-06,
+      "loss": 0.8058,
+      "step": 10986
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32706815193793987,
+      "learning_rate": 6.323668848123774e-06,
+      "loss": 0.19,
+      "step": 10987
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3039169646645536,
+      "learning_rate": 6.321938317204886e-06,
+      "loss": 0.2826,
+      "step": 10988
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34350298686577224,
+      "learning_rate": 6.320207913657111e-06,
+      "loss": 0.223,
+      "step": 10989
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43173203039681673,
+      "learning_rate": 6.3184776375403814e-06,
+      "loss": 0.1458,
+      "step": 10990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32354358590641585,
+      "learning_rate": 6.3167474889146096e-06,
+      "loss": 0.292,
+      "step": 10991
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4753435688955507,
+      "learning_rate": 6.315017467839717e-06,
+      "loss": 0.3936,
+      "step": 10992
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37915678416069587,
+      "learning_rate": 6.313287574375609e-06,
+      "loss": 0.2156,
+      "step": 10993
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4090258131741896,
+      "learning_rate": 6.311557808582196e-06,
+      "loss": 0.3467,
+      "step": 10994
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3414960965925519,
+      "learning_rate": 6.309828170519376e-06,
+      "loss": 0.2537,
+      "step": 10995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2555111542785431,
+      "learning_rate": 6.308098660247049e-06,
+      "loss": 0.1574,
+      "step": 10996
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53907874280059,
+      "learning_rate": 6.306369277825104e-06,
+      "loss": 0.3552,
+      "step": 10997
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8653136044186649,
+      "learning_rate": 6.304640023313435e-06,
+      "loss": 0.5482,
+      "step": 10998
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2936106670828241,
+      "learning_rate": 6.302910896771921e-06,
+      "loss": 0.2673,
+      "step": 10999
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3666631543606978,
+      "learning_rate": 6.301181898260444e-06,
+      "loss": 0.2542,
+      "step": 11000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42116700454476547,
+      "learning_rate": 6.299453027838881e-06,
+      "loss": 0.2626,
+      "step": 11001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33832310295296864,
+      "learning_rate": 6.297724285567098e-06,
+      "loss": 0.2159,
+      "step": 11002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3384603298577127,
+      "learning_rate": 6.295995671504965e-06,
+      "loss": 0.2478,
+      "step": 11003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4787703629140393,
+      "learning_rate": 6.294267185712342e-06,
+      "loss": 0.361,
+      "step": 11004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.588450481399667,
+      "learning_rate": 6.292538828249087e-06,
+      "loss": 0.355,
+      "step": 11005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32064689647937217,
+      "learning_rate": 6.290810599175052e-06,
+      "loss": 0.234,
+      "step": 11006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3023253847261538,
+      "learning_rate": 6.289082498550091e-06,
+      "loss": 0.2633,
+      "step": 11007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33301628269246675,
+      "learning_rate": 6.287354526434042e-06,
+      "loss": 0.1976,
+      "step": 11008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3279184303807924,
+      "learning_rate": 6.285626682886743e-06,
+      "loss": 0.2236,
+      "step": 11009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7138001020794481,
+      "learning_rate": 6.283898967968034e-06,
+      "loss": 0.4225,
+      "step": 11010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35085811541230716,
+      "learning_rate": 6.282171381737742e-06,
+      "loss": 0.2892,
+      "step": 11011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35646638300198263,
+      "learning_rate": 6.280443924255697e-06,
+      "loss": 0.3308,
+      "step": 11012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9594251689271127,
+      "learning_rate": 6.27871659558172e-06,
+      "loss": 0.2799,
+      "step": 11013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2198364833708466,
+      "learning_rate": 6.276989395775625e-06,
+      "loss": 0.1496,
+      "step": 11014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28285740443505136,
+      "learning_rate": 6.275262324897229e-06,
+      "loss": 0.2867,
+      "step": 11015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7428097831158089,
+      "learning_rate": 6.273535383006336e-06,
+      "loss": 0.319,
+      "step": 11016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5076884926327849,
+      "learning_rate": 6.271808570162754e-06,
+      "loss": 0.3514,
+      "step": 11017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3783077303287608,
+      "learning_rate": 6.27008188642628e-06,
+      "loss": 0.3001,
+      "step": 11018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3307956499614347,
+      "learning_rate": 6.268355331856713e-06,
+      "loss": 0.2285,
+      "step": 11019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2362548513026544,
+      "learning_rate": 6.266628906513836e-06,
+      "loss": 0.1491,
+      "step": 11020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42279259430228205,
+      "learning_rate": 6.264902610457442e-06,
+      "loss": 0.316,
+      "step": 11021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.575956482245312,
+      "learning_rate": 6.263176443747309e-06,
+      "loss": 0.3045,
+      "step": 11022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.48818995665095016,
+      "learning_rate": 6.261450406443217e-06,
+      "loss": 0.3887,
+      "step": 11023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3348148241071795,
+      "learning_rate": 6.259724498604933e-06,
+      "loss": 0.2703,
+      "step": 11024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1273760167300912,
+      "learning_rate": 6.257998720292233e-06,
+      "loss": 0.5308,
+      "step": 11025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1848169085706313,
+      "learning_rate": 6.256273071564874e-06,
+      "loss": 0.088,
+      "step": 11026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25545609687182125,
+      "learning_rate": 6.254547552482617e-06,
+      "loss": 0.2522,
+      "step": 11027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7623754548706887,
+      "learning_rate": 6.25282216310522e-06,
+      "loss": 0.368,
+      "step": 11028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4726985840995557,
+      "learning_rate": 6.2510969034924265e-06,
+      "loss": 0.3078,
+      "step": 11029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3572847810692785,
+      "learning_rate": 6.249371773703989e-06,
+      "loss": 0.276,
+      "step": 11030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3708677390096252,
+      "learning_rate": 6.247646773799645e-06,
+      "loss": 0.3205,
+      "step": 11031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25304006716718513,
+      "learning_rate": 6.245921903839132e-06,
+      "loss": 0.123,
+      "step": 11032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3925028301065731,
+      "learning_rate": 6.24419716388218e-06,
+      "loss": 0.2976,
+      "step": 11033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8941776485349109,
+      "learning_rate": 6.242472553988521e-06,
+      "loss": 0.4912,
+      "step": 11034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3501794684964001,
+      "learning_rate": 6.240748074217875e-06,
+      "loss": 0.3189,
+      "step": 11035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3330008477936822,
+      "learning_rate": 6.239023724629962e-06,
+      "loss": 0.217,
+      "step": 11036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1036007794647276,
+      "learning_rate": 6.237299505284495e-06,
+      "loss": 0.616,
+      "step": 11037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3554645171715663,
+      "learning_rate": 6.235575416241185e-06,
+      "loss": 0.3099,
+      "step": 11038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26021400716218845,
+      "learning_rate": 6.233851457559736e-06,
+      "loss": 0.2198,
+      "step": 11039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35160884390360947,
+      "learning_rate": 6.232127629299849e-06,
+      "loss": 0.2377,
+      "step": 11040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0184502874359644,
+      "learning_rate": 6.230403931521224e-06,
+      "loss": 0.7252,
+      "step": 11041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27161208909389467,
+      "learning_rate": 6.228680364283546e-06,
+      "loss": 0.1799,
+      "step": 11042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3497995921510243,
+      "learning_rate": 6.226956927646504e-06,
+      "loss": 0.3126,
+      "step": 11043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6248331696318262,
+      "learning_rate": 6.225233621669782e-06,
+      "loss": 0.4057,
+      "step": 11044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3119749624889606,
+      "learning_rate": 6.2235104464130545e-06,
+      "loss": 0.2302,
+      "step": 11045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3756060506238427,
+      "learning_rate": 6.221787401936002e-06,
+      "loss": 0.2444,
+      "step": 11046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34670055217013945,
+      "learning_rate": 6.220064488298285e-06,
+      "loss": 0.3051,
+      "step": 11047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.249174577275586,
+      "learning_rate": 6.2183417055595765e-06,
+      "loss": 0.2063,
+      "step": 11048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0123588001422943,
+      "learning_rate": 6.216619053779529e-06,
+      "loss": 0.2817,
+      "step": 11049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.36445402991388354,
+      "learning_rate": 6.214896533017803e-06,
+      "loss": 0.3022,
+      "step": 11050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3092688596231202,
+      "learning_rate": 6.213174143334046e-06,
+      "loss": 0.2614,
+      "step": 11051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25530111498148117,
+      "learning_rate": 6.211451884787907e-06,
+      "loss": 0.1012,
+      "step": 11052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4294403584004139,
+      "learning_rate": 6.209729757439026e-06,
+      "loss": 0.3003,
+      "step": 11053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4008704554780664,
+      "learning_rate": 6.208007761347039e-06,
+      "loss": 0.2886,
+      "step": 11054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27196718819293497,
+      "learning_rate": 6.206285896571582e-06,
+      "loss": 0.24,
+      "step": 11055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5460900718518568,
+      "learning_rate": 6.20456416317228e-06,
+      "loss": 0.332,
+      "step": 11056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.369685863562834,
+      "learning_rate": 6.202842561208759e-06,
+      "loss": 0.2815,
+      "step": 11057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5100637075537648,
+      "learning_rate": 6.201121090740634e-06,
+      "loss": 0.2929,
+      "step": 11058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.316403402827174,
+      "learning_rate": 6.199399751827525e-06,
+      "loss": 0.2596,
+      "step": 11059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3212154547207368,
+      "learning_rate": 6.197678544529037e-06,
+      "loss": 0.2387,
+      "step": 11060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5344441277836296,
+      "learning_rate": 6.195957468904781e-06,
+      "loss": 0.3082,
+      "step": 11061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3320258149067593,
+      "learning_rate": 6.19423652501435e-06,
+      "loss": 0.2633,
+      "step": 11062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3367919993055683,
+      "learning_rate": 6.192515712917348e-06,
+      "loss": 0.2523,
+      "step": 11063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0809686276937551,
+      "learning_rate": 6.19079503267336e-06,
+      "loss": 0.6319,
+      "step": 11064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19407826999737932,
+      "learning_rate": 6.189074484341979e-06,
+      "loss": 0.098,
+      "step": 11065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31577201724467685,
+      "learning_rate": 6.187354067982785e-06,
+      "loss": 0.2449,
+      "step": 11066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3403924479356509,
+      "learning_rate": 6.185633783655354e-06,
+      "loss": 0.302,
+      "step": 11067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6402462185739675,
+      "learning_rate": 6.183913631419263e-06,
+      "loss": 0.3312,
+      "step": 11068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3382422319781781,
+      "learning_rate": 6.182193611334075e-06,
+      "loss": 0.258,
+      "step": 11069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2736662713881572,
+      "learning_rate": 6.180473723459361e-06,
+      "loss": 0.7406,
+      "step": 11070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25041480842921693,
+      "learning_rate": 6.178753967854677e-06,
+      "loss": 0.2126,
+      "step": 11071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4063472874123882,
+      "learning_rate": 6.17703434457958e-06,
+      "loss": 0.2927,
+      "step": 11072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.586628542560072,
+      "learning_rate": 6.175314853693617e-06,
+      "loss": 0.3441,
+      "step": 11073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27702603918539737,
+      "learning_rate": 6.173595495256338e-06,
+      "loss": 0.2467,
+      "step": 11074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35689733543792085,
+      "learning_rate": 6.1718762693272846e-06,
+      "loss": 0.1613,
+      "step": 11075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46985107766144624,
+      "learning_rate": 6.170157175965988e-06,
+      "loss": 0.3795,
+      "step": 11076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7798022987856814,
+      "learning_rate": 6.168438215231984e-06,
+      "loss": 0.499,
+      "step": 11077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2888764256268748,
+      "learning_rate": 6.166719387184802e-06,
+      "loss": 0.1906,
+      "step": 11078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32604639861053947,
+      "learning_rate": 6.16500069188396e-06,
+      "loss": 0.3296,
+      "step": 11079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2507700628659636,
+      "learning_rate": 6.163282129388981e-06,
+      "loss": 0.1362,
+      "step": 11080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31590120501365204,
+      "learning_rate": 6.1615636997593745e-06,
+      "loss": 0.2005,
+      "step": 11081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3610564542876327,
+      "learning_rate": 6.159845403054654e-06,
+      "loss": 0.332,
+      "step": 11082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2608533543659348,
+      "learning_rate": 6.15812723933432e-06,
+      "loss": 0.7569,
+      "step": 11083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32773053299135346,
+      "learning_rate": 6.1564092086578765e-06,
+      "loss": 0.2223,
+      "step": 11084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8451695808115459,
+      "learning_rate": 6.154691311084816e-06,
+      "loss": 0.4394,
+      "step": 11085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21285824327119043,
+      "learning_rate": 6.152973546674631e-06,
+      "loss": 0.2111,
+      "step": 11086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3162598405639108,
+      "learning_rate": 6.151255915486804e-06,
+      "loss": 0.2425,
+      "step": 11087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9304264514798442,
+      "learning_rate": 6.1495384175808224e-06,
+      "loss": 0.3778,
+      "step": 11088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7140149435298279,
+      "learning_rate": 6.147821053016159e-06,
+      "loss": 0.4178,
+      "step": 11089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3320040093447702,
+      "learning_rate": 6.146103821852286e-06,
+      "loss": 0.245,
+      "step": 11090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35658129823456214,
+      "learning_rate": 6.144386724148674e-06,
+      "loss": 0.2434,
+      "step": 11091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27249855103507575,
+      "learning_rate": 6.142669759964781e-06,
+      "loss": 0.1735,
+      "step": 11092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5892126845666474,
+      "learning_rate": 6.140952929360071e-06,
+      "loss": 0.3012,
+      "step": 11093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24465733364449624,
+      "learning_rate": 6.139236232393993e-06,
+      "loss": 0.2379,
+      "step": 11094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9972780676501973,
+      "learning_rate": 6.137519669126e-06,
+      "loss": 0.4295,
+      "step": 11095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4816808218850385,
+      "learning_rate": 6.135803239615532e-06,
+      "loss": 0.337,
+      "step": 11096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2635477301210972,
+      "learning_rate": 6.134086943922034e-06,
+      "loss": 0.183,
+      "step": 11097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2615090099141512,
+      "learning_rate": 6.132370782104937e-06,
+      "loss": 0.2292,
+      "step": 11098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38708784016928843,
+      "learning_rate": 6.130654754223676e-06,
+      "loss": 0.2688,
+      "step": 11099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5112058574827448,
+      "learning_rate": 6.128938860337672e-06,
+      "loss": 0.3646,
+      "step": 11100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7461043466785471,
+      "learning_rate": 6.127223100506351e-06,
+      "loss": 0.3124,
+      "step": 11101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27170610461361683,
+      "learning_rate": 6.125507474789125e-06,
+      "loss": 0.2624,
+      "step": 11102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5161349562940573,
+      "learning_rate": 6.123791983245411e-06,
+      "loss": 0.3863,
+      "step": 11103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21589302078484102,
+      "learning_rate": 6.122076625934612e-06,
+      "loss": 0.1027,
+      "step": 11104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3764534243966187,
+      "learning_rate": 6.120361402916135e-06,
+      "loss": 0.2836,
+      "step": 11105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37975502883,
+      "learning_rate": 6.118646314249376e-06,
+      "loss": 0.3184,
+      "step": 11106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6739943039680442,
+      "learning_rate": 6.116931359993725e-06,
+      "loss": 0.3041,
+      "step": 11107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37209910595351725,
+      "learning_rate": 6.115216540208577e-06,
+      "loss": 0.2577,
+      "step": 11108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5528497756841371,
+      "learning_rate": 6.1135018549533146e-06,
+      "loss": 0.3213,
+      "step": 11109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21630491935510923,
+      "learning_rate": 6.111787304287312e-06,
+      "loss": 0.1668,
+      "step": 11110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6091154648621666,
+      "learning_rate": 6.11007288826995e-06,
+      "loss": 0.3458,
+      "step": 11111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35893455503756927,
+      "learning_rate": 6.108358606960595e-06,
+      "loss": 0.31,
+      "step": 11112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7157112727794527,
+      "learning_rate": 6.1066444604186156e-06,
+      "loss": 0.4223,
+      "step": 11113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26562757365475465,
+      "learning_rate": 6.104930448703369e-06,
+      "loss": 0.2132,
+      "step": 11114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3765400989158581,
+      "learning_rate": 6.1032165718742154e-06,
+      "loss": 0.296,
+      "step": 11115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4711689009780174,
+      "learning_rate": 6.1015028299905025e-06,
+      "loss": 0.2469,
+      "step": 11116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27771652468962005,
+      "learning_rate": 6.0997892231115805e-06,
+      "loss": 0.1337,
+      "step": 11117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2506367775544241,
+      "learning_rate": 6.098075751296792e-06,
+      "loss": 0.272,
+      "step": 11118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7113697594382398,
+      "learning_rate": 6.096362414605468e-06,
+      "loss": 0.4545,
+      "step": 11119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31511420097201853,
+      "learning_rate": 6.0946492130969494e-06,
+      "loss": 0.1979,
+      "step": 11120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48430785170819585,
+      "learning_rate": 6.092936146830557e-06,
+      "loss": 0.3365,
+      "step": 11121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34387547840728383,
+      "learning_rate": 6.091223215865621e-06,
+      "loss": 0.2893,
+      "step": 11122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23731111105713615,
+      "learning_rate": 6.089510420261455e-06,
+      "loss": 0.1881,
+      "step": 11123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3714432234815818,
+      "learning_rate": 6.087797760077376e-06,
+      "loss": 0.265,
+      "step": 11124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7696874757906164,
+      "learning_rate": 6.086085235372692e-06,
+      "loss": 0.488,
+      "step": 11125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29488211422426086,
+      "learning_rate": 6.084372846206709e-06,
+      "loss": 0.2643,
+      "step": 11126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3843671048783373,
+      "learning_rate": 6.0826605926387226e-06,
+      "loss": 0.2295,
+      "step": 11127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28084906133884785,
+      "learning_rate": 6.080948474728036e-06,
+      "loss": 0.1574,
+      "step": 11128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8237786686226553,
+      "learning_rate": 6.079236492533931e-06,
+      "loss": 0.3501,
+      "step": 11129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2653302823814511,
+      "learning_rate": 6.077524646115701e-06,
+      "loss": 0.2377,
+      "step": 11130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.790921659400872,
+      "learning_rate": 6.075812935532623e-06,
+      "loss": 0.5031,
+      "step": 11131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.732432702456682,
+      "learning_rate": 6.074101360843973e-06,
+      "loss": 0.4053,
+      "step": 11132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3450032202174695,
+      "learning_rate": 6.072389922109027e-06,
+      "loss": 0.2106,
+      "step": 11133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3755738306929151,
+      "learning_rate": 6.070678619387045e-06,
+      "loss": 0.2833,
+      "step": 11134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.923733611804049,
+      "learning_rate": 6.068967452737296e-06,
+      "loss": 0.3767,
+      "step": 11135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2933452155571611,
+      "learning_rate": 6.067256422219034e-06,
+      "loss": 0.2224,
+      "step": 11136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29408326213010355,
+      "learning_rate": 6.065545527891514e-06,
+      "loss": 0.1626,
+      "step": 11137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3024053064991005,
+      "learning_rate": 6.063834769813982e-06,
+      "loss": 0.2829,
+      "step": 11138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.441073941280975,
+      "learning_rate": 6.062124148045685e-06,
+      "loss": 0.3102,
+      "step": 11139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.033460670183201,
+      "learning_rate": 6.060413662645856e-06,
+      "loss": 0.3068,
+      "step": 11140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3800496285695335,
+      "learning_rate": 6.058703313673735e-06,
+      "loss": 0.3042,
+      "step": 11141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3069885259774946,
+      "learning_rate": 6.0569931011885504e-06,
+      "loss": 0.2701,
+      "step": 11142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.261025683645621,
+      "learning_rate": 6.055283025249526e-06,
+      "loss": 0.1452,
+      "step": 11143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30579045165578644,
+      "learning_rate": 6.053573085915875e-06,
+      "loss": 0.1788,
+      "step": 11144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41347157047586297,
+      "learning_rate": 6.0518632832468215e-06,
+      "loss": 0.3075,
+      "step": 11145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.509072555332188,
+      "learning_rate": 6.050153617301571e-06,
+      "loss": 0.2722,
+      "step": 11146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0137824837962006,
+      "learning_rate": 6.048444088139334e-06,
+      "loss": 0.4353,
+      "step": 11147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32813531727045664,
+      "learning_rate": 6.0467346958193056e-06,
+      "loss": 0.253,
+      "step": 11148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3297149666866975,
+      "learning_rate": 6.045025440400684e-06,
+      "loss": 0.2803,
+      "step": 11149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2517094708835535,
+      "learning_rate": 6.043316321942663e-06,
+      "loss": 0.1736,
+      "step": 11150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.36579356484526454,
+      "learning_rate": 6.0416073405044274e-06,
+      "loss": 0.2721,
+      "step": 11151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.009514354771467,
+      "learning_rate": 6.039898496145159e-06,
+      "loss": 0.4483,
+      "step": 11152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46789848161172504,
+      "learning_rate": 6.038189788924036e-06,
+      "loss": 0.2603,
+      "step": 11153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28716040798789005,
+      "learning_rate": 6.03648121890023e-06,
+      "loss": 0.2665,
+      "step": 11154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2217392161585139,
+      "learning_rate": 6.03477278613291e-06,
+      "loss": 0.7532,
+      "step": 11155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20730714099122807,
+      "learning_rate": 6.033064490681238e-06,
+      "loss": 0.1146,
+      "step": 11156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4212369317329507,
+      "learning_rate": 6.031356332604369e-06,
+      "loss": 0.2847,
+      "step": 11157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3950415550882442,
+      "learning_rate": 6.029648311961462e-06,
+      "loss": 0.3112,
+      "step": 11158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5599324152074959,
+      "learning_rate": 6.027940428811662e-06,
+      "loss": 0.1964,
+      "step": 11159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3724627817967568,
+      "learning_rate": 6.026232683214115e-06,
+      "loss": 0.2767,
+      "step": 11160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4752167622283187,
+      "learning_rate": 6.024525075227959e-06,
+      "loss": 0.3687,
+      "step": 11161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21954940331324588,
+      "learning_rate": 6.02281760491233e-06,
+      "loss": 0.1958,
+      "step": 11162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33151403738859675,
+      "learning_rate": 6.021110272326354e-06,
+      "loss": 0.2124,
+      "step": 11163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515384369202005,
+      "learning_rate": 6.0194030775291605e-06,
+      "loss": 0.3731,
+      "step": 11164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4626758506358267,
+      "learning_rate": 6.017696020579864e-06,
+      "loss": 0.3026,
+      "step": 11165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26893646286578476,
+      "learning_rate": 6.015989101537586e-06,
+      "loss": 0.2204,
+      "step": 11166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1058572603376675,
+      "learning_rate": 6.0142823204614335e-06,
+      "loss": 0.703,
+      "step": 11167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3964445682878911,
+      "learning_rate": 6.012575677410512e-06,
+      "loss": 0.2333,
+      "step": 11168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27704603202525324,
+      "learning_rate": 6.010869172443923e-06,
+      "loss": 0.2202,
+      "step": 11169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3433573699670649,
+      "learning_rate": 6.0091628056207655e-06,
+      "loss": 0.2727,
+      "step": 11170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9905163871846923,
+      "learning_rate": 6.007456577000128e-06,
+      "loss": 0.4821,
+      "step": 11171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35401947805745243,
+      "learning_rate": 6.005750486641095e-06,
+      "loss": 0.2056,
+      "step": 11172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5022193592434819,
+      "learning_rate": 6.004044534602753e-06,
+      "loss": 0.3671,
+      "step": 11173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3889968137655978,
+      "learning_rate": 6.002338720944174e-06,
+      "loss": 0.3389,
+      "step": 11174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32816842990029704,
+      "learning_rate": 6.000633045724438e-06,
+      "loss": 0.2712,
+      "step": 11175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2643109087774337,
+      "learning_rate": 5.998927509002608e-06,
+      "loss": 0.136,
+      "step": 11176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34977099868176076,
+      "learning_rate": 5.997222110837742e-06,
+      "loss": 0.2911,
+      "step": 11177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41325953968084655,
+      "learning_rate": 5.995516851288904e-06,
+      "loss": 0.2771,
+      "step": 11178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5048678658640868,
+      "learning_rate": 5.9938117304151445e-06,
+      "loss": 0.3181,
+      "step": 11179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5855639790646341,
+      "learning_rate": 5.992106748275513e-06,
+      "loss": 0.3861,
+      "step": 11180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3731060722176844,
+      "learning_rate": 5.990401904929051e-06,
+      "loss": 0.2846,
+      "step": 11181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2450716350813472,
+      "learning_rate": 5.988697200434801e-06,
+      "loss": 0.1784,
+      "step": 11182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.588488452674224,
+      "learning_rate": 5.986992634851794e-06,
+      "loss": 0.7367,
+      "step": 11183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33778700717578847,
+      "learning_rate": 5.985288208239057e-06,
+      "loss": 0.2592,
+      "step": 11184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38914509602391384,
+      "learning_rate": 5.98358392065562e-06,
+      "loss": 0.2963,
+      "step": 11185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.780635173262054,
+      "learning_rate": 5.981879772160497e-06,
+      "loss": 0.3924,
+      "step": 11186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31361656933869875,
+      "learning_rate": 5.980175762812705e-06,
+      "loss": 0.2679,
+      "step": 11187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4583142651536605,
+      "learning_rate": 5.978471892671254e-06,
+      "loss": 0.2751,
+      "step": 11188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2585065185246987,
+      "learning_rate": 5.976768161795149e-06,
+      "loss": 0.1919,
+      "step": 11189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3744518103024148,
+      "learning_rate": 5.975064570243387e-06,
+      "loss": 0.2544,
+      "step": 11190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2344220965569719,
+      "learning_rate": 5.973361118074969e-06,
+      "loss": 0.8546,
+      "step": 11191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9938392858124002,
+      "learning_rate": 5.97165780534888e-06,
+      "loss": 0.3239,
+      "step": 11192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2816535477742484,
+      "learning_rate": 5.969954632124111e-06,
+      "loss": 0.248,
+      "step": 11193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5002961541221576,
+      "learning_rate": 5.968251598459636e-06,
+      "loss": 0.3412,
+      "step": 11194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2849071332633207,
+      "learning_rate": 5.966548704414436e-06,
+      "loss": 0.1259,
+      "step": 11195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38572866452620297,
+      "learning_rate": 5.964845950047484e-06,
+      "loss": 0.2577,
+      "step": 11196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3144011371244218,
+      "learning_rate": 5.96314333541774e-06,
+      "loss": 0.2866,
+      "step": 11197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0859733305344872,
+      "learning_rate": 5.961440860584169e-06,
+      "loss": 0.4253,
+      "step": 11198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31667468432575874,
+      "learning_rate": 5.959738525605727e-06,
+      "loss": 0.1599,
+      "step": 11199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3265305092885602,
+      "learning_rate": 5.958036330541368e-06,
+      "loss": 0.2316,
+      "step": 11200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2730721162752112,
+      "learning_rate": 5.956334275450035e-06,
+      "loss": 0.2342,
+      "step": 11201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.315786099286474,
+      "learning_rate": 5.954632360390673e-06,
+      "loss": 0.1946,
+      "step": 11202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8586307279331272,
+      "learning_rate": 5.9529305854222185e-06,
+      "loss": 0.4156,
+      "step": 11203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8222346678946091,
+      "learning_rate": 5.951228950603605e-06,
+      "loss": 0.5293,
+      "step": 11204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26079077888105623,
+      "learning_rate": 5.949527455993756e-06,
+      "loss": 0.2104,
+      "step": 11205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4512259533268354,
+      "learning_rate": 5.947826101651599e-06,
+      "loss": 0.3257,
+      "step": 11206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32723592880119196,
+      "learning_rate": 5.946124887636049e-06,
+      "loss": 0.2213,
+      "step": 11207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29865161019025654,
+      "learning_rate": 5.944423814006022e-06,
+      "loss": 0.2226,
+      "step": 11208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35149010732275404,
+      "learning_rate": 5.9427228808204216e-06,
+      "loss": 0.3059,
+      "step": 11209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7682225597339026,
+      "learning_rate": 5.941022088138158e-06,
+      "loss": 0.4489,
+      "step": 11210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30994659665733854,
+      "learning_rate": 5.939321436018119e-06,
+      "loss": 0.2518,
+      "step": 11211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5923893515591114,
+      "learning_rate": 5.937620924519207e-06,
+      "loss": 0.0215,
+      "step": 11212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22704968985868573,
+      "learning_rate": 5.935920553700305e-06,
+      "loss": 0.216,
+      "step": 11213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5706346113528733,
+      "learning_rate": 5.934220323620303e-06,
+      "loss": 0.3533,
+      "step": 11214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37401496655785443,
+      "learning_rate": 5.932520234338073e-06,
+      "loss": 0.2737,
+      "step": 11215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4734541768133188,
+      "learning_rate": 5.930820285912495e-06,
+      "loss": 0.4071,
+      "step": 11216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3539907526938965,
+      "learning_rate": 5.9291204784024335e-06,
+      "loss": 0.2553,
+      "step": 11217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3659412287344784,
+      "learning_rate": 5.9274208118667565e-06,
+      "loss": 0.2484,
+      "step": 11218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3113800892023193,
+      "learning_rate": 5.92572128636432e-06,
+      "loss": 0.1939,
+      "step": 11219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5020546282550093,
+      "learning_rate": 5.924021901953983e-06,
+      "loss": 0.2901,
+      "step": 11220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.286313888428949,
+      "learning_rate": 5.922322658694591e-06,
+      "loss": 0.2489,
+      "step": 11221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2063295574915158,
+      "learning_rate": 5.920623556644987e-06,
+      "loss": 0.7685,
+      "step": 11222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.422377259307182,
+      "learning_rate": 5.918924595864017e-06,
+      "loss": 0.3199,
+      "step": 11223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5829496855100268,
+      "learning_rate": 5.917225776410511e-06,
+      "loss": 0.3525,
+      "step": 11224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2512331675050159,
+      "learning_rate": 5.915527098343302e-06,
+      "loss": 0.206,
+      "step": 11225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3745193433663806,
+      "learning_rate": 5.913828561721214e-06,
+      "loss": 0.2792,
+      "step": 11226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3169721307067816,
+      "learning_rate": 5.912130166603066e-06,
+      "loss": 0.2322,
+      "step": 11227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4240458248846178,
+      "learning_rate": 5.910431913047674e-06,
+      "loss": 0.2787,
+      "step": 11228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30928589966688147,
+      "learning_rate": 5.908733801113851e-06,
+      "loss": 0.2619,
+      "step": 11229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48813601012960506,
+      "learning_rate": 5.907035830860399e-06,
+      "loss": 0.3456,
+      "step": 11230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5823485197085368,
+      "learning_rate": 5.905338002346122e-06,
+      "loss": 0.314,
+      "step": 11231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5900398729201102,
+      "learning_rate": 5.9036403156298125e-06,
+      "loss": 0.2937,
+      "step": 11232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23746694604233817,
+      "learning_rate": 5.901942770770264e-06,
+      "loss": 0.2338,
+      "step": 11233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30105969621280293,
+      "learning_rate": 5.900245367826258e-06,
+      "loss": 0.1899,
+      "step": 11234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6304257335183989,
+      "learning_rate": 5.898548106856583e-06,
+      "loss": 0.3072,
+      "step": 11235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4284540690270265,
+      "learning_rate": 5.896850987920009e-06,
+      "loss": 0.2833,
+      "step": 11236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38906981985827127,
+      "learning_rate": 5.895154011075308e-06,
+      "loss": 0.3207,
+      "step": 11237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5208798117829698,
+      "learning_rate": 5.893457176381248e-06,
+      "loss": 0.1669,
+      "step": 11238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39622311730384707,
+      "learning_rate": 5.891760483896587e-06,
+      "loss": 0.3249,
+      "step": 11239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38496094157001054,
+      "learning_rate": 5.890063933680087e-06,
+      "loss": 0.2616,
+      "step": 11240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22651439846232713,
+      "learning_rate": 5.8883675257904936e-06,
+      "loss": 0.1797,
+      "step": 11241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38679464880782893,
+      "learning_rate": 5.886671260286558e-06,
+      "loss": 0.3031,
+      "step": 11242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7959092880908866,
+      "learning_rate": 5.884975137227018e-06,
+      "loss": 0.3909,
+      "step": 11243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3346253622241024,
+      "learning_rate": 5.883279156670616e-06,
+      "loss": 0.1724,
+      "step": 11244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30811296607418726,
+      "learning_rate": 5.881583318676078e-06,
+      "loss": 0.2709,
+      "step": 11245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3766558287583714,
+      "learning_rate": 5.879887623302131e-06,
+      "loss": 0.3055,
+      "step": 11246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23333225379841346,
+      "learning_rate": 5.8781920706075e-06,
+      "loss": 0.1275,
+      "step": 11247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5988756712228822,
+      "learning_rate": 5.876496660650899e-06,
+      "loss": 0.3427,
+      "step": 11248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37202182636931236,
+      "learning_rate": 5.874801393491041e-06,
+      "loss": 0.301,
+      "step": 11249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.705704673592431,
+      "learning_rate": 5.873106269186635e-06,
+      "loss": 0.3855,
+      "step": 11250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3430225343190707,
+      "learning_rate": 5.871411287796379e-06,
+      "loss": 0.229,
+      "step": 11251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3532476705395196,
+      "learning_rate": 5.869716449378975e-06,
+      "loss": 0.3249,
+      "step": 11252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21846891869803547,
+      "learning_rate": 5.8680217539931106e-06,
+      "loss": 0.147,
+      "step": 11253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3593853724638047,
+      "learning_rate": 5.866327201697477e-06,
+      "loss": 0.2326,
+      "step": 11254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7948326466630459,
+      "learning_rate": 5.864632792550753e-06,
+      "loss": 0.4388,
+      "step": 11255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4301961477693214,
+      "learning_rate": 5.862938526611619e-06,
+      "loss": 0.2876,
+      "step": 11256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27854050842859307,
+      "learning_rate": 5.861244403938744e-06,
+      "loss": 0.232,
+      "step": 11257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2624624061673055,
+      "learning_rate": 5.859550424590801e-06,
+      "loss": 0.7603,
+      "step": 11258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3089760684352078,
+      "learning_rate": 5.857856588626445e-06,
+      "loss": 0.182,
+      "step": 11259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23799079783274543,
+      "learning_rate": 5.856162896104339e-06,
+      "loss": 0.2428,
+      "step": 11260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.874747782088808,
+      "learning_rate": 5.854469347083134e-06,
+      "loss": 0.5215,
+      "step": 11261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5987070815263701,
+      "learning_rate": 5.852775941621476e-06,
+      "loss": 0.2821,
+      "step": 11262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3813507046247867,
+      "learning_rate": 5.851082679778011e-06,
+      "loss": 0.2997,
+      "step": 11263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3432165598746957,
+      "learning_rate": 5.8493895616113714e-06,
+      "loss": 0.2648,
+      "step": 11264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49326946299327873,
+      "learning_rate": 5.847696587180195e-06,
+      "loss": 0.3269,
+      "step": 11265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3997606558563779,
+      "learning_rate": 5.846003756543106e-06,
+      "loss": 0.309,
+      "step": 11266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2854817732641904,
+      "learning_rate": 5.844311069758729e-06,
+      "loss": 0.138,
+      "step": 11267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.359787125128394,
+      "learning_rate": 5.842618526885679e-06,
+      "loss": 0.2986,
+      "step": 11268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34338489317305765,
+      "learning_rate": 5.840926127982573e-06,
+      "loss": 0.2825,
+      "step": 11269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7416336040716597,
+      "learning_rate": 5.839233873108016e-06,
+      "loss": 0.3722,
+      "step": 11270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6592265725324533,
+      "learning_rate": 5.837541762320609e-06,
+      "loss": 0.4155,
+      "step": 11271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3190564215746574,
+      "learning_rate": 5.835849795678954e-06,
+      "loss": 0.2622,
+      "step": 11272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19298277235341782,
+      "learning_rate": 5.834157973241643e-06,
+      "loss": 0.1614,
+      "step": 11273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7677838234055707,
+      "learning_rate": 5.83246629506726e-06,
+      "loss": 0.4286,
+      "step": 11274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3608993586517985,
+      "learning_rate": 5.830774761214392e-06,
+      "loss": 0.303,
+      "step": 11275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5166381381403379,
+      "learning_rate": 5.829083371741609e-06,
+      "loss": 0.3991,
+      "step": 11276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5044253316790134,
+      "learning_rate": 5.827392126707499e-06,
+      "loss": 0.2718,
+      "step": 11277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3718745113889626,
+      "learning_rate": 5.825701026170616e-06,
+      "loss": 0.3091,
+      "step": 11278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28428241777519586,
+      "learning_rate": 5.824010070189523e-06,
+      "loss": 0.2016,
+      "step": 11279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30819648744024025,
+      "learning_rate": 5.8223192588227836e-06,
+      "loss": 0.2175,
+      "step": 11280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3475980694959061,
+      "learning_rate": 5.820628592128952e-06,
+      "loss": 0.275,
+      "step": 11281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.711060174788503,
+      "learning_rate": 5.81893807016657e-06,
+      "loss": 0.4232,
+      "step": 11282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5564311587922617,
+      "learning_rate": 5.817247692994179e-06,
+      "loss": 0.2324,
+      "step": 11283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39698374618945803,
+      "learning_rate": 5.815557460670326e-06,
+      "loss": 0.2794,
+      "step": 11284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24435255765312222,
+      "learning_rate": 5.813867373253537e-06,
+      "loss": 0.2001,
+      "step": 11285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7686243817440529,
+      "learning_rate": 5.8121774308023415e-06,
+      "loss": 0.4895,
+      "step": 11286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31839227350161425,
+      "learning_rate": 5.810487633375261e-06,
+      "loss": 0.2133,
+      "step": 11287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4042563230540896,
+      "learning_rate": 5.80879798103081e-06,
+      "loss": 0.3192,
+      "step": 11288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.76019850506445,
+      "learning_rate": 5.807108473827508e-06,
+      "loss": 0.4039,
+      "step": 11289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30882169987209035,
+      "learning_rate": 5.80541911182386e-06,
+      "loss": 0.2067,
+      "step": 11290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2653860564848561,
+      "learning_rate": 5.803729895078368e-06,
+      "loss": 0.1633,
+      "step": 11291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37243490114112704,
+      "learning_rate": 5.802040823649524e-06,
+      "loss": 0.3116,
+      "step": 11292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2908543475298292,
+      "learning_rate": 5.800351897595832e-06,
+      "loss": 0.2062,
+      "step": 11293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7543086085745012,
+      "learning_rate": 5.7986631169757715e-06,
+      "loss": 0.4082,
+      "step": 11294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7486544921576171,
+      "learning_rate": 5.796974481847827e-06,
+      "loss": 0.4599,
+      "step": 11295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2232167232788598,
+      "learning_rate": 5.795285992270472e-06,
+      "loss": 0.216,
+      "step": 11296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47861387869738203,
+      "learning_rate": 5.793597648302185e-06,
+      "loss": 0.258,
+      "step": 11297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4226253650595357,
+      "learning_rate": 5.791909450001432e-06,
+      "loss": 0.3018,
+      "step": 11298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32740022436539834,
+      "learning_rate": 5.790221397426672e-06,
+      "loss": 0.2581,
+      "step": 11299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37377043137715343,
+      "learning_rate": 5.7885334906363656e-06,
+      "loss": 0.2573,
+      "step": 11300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0508988726333814,
+      "learning_rate": 5.786845729688958e-06,
+      "loss": 0.5739,
+      "step": 11301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39097791401524284,
+      "learning_rate": 5.785158114642906e-06,
+      "loss": 0.2412,
+      "step": 11302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2146535619035754,
+      "learning_rate": 5.783470645556648e-06,
+      "loss": 0.1288,
+      "step": 11303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30364120884793583,
+      "learning_rate": 5.781783322488619e-06,
+      "loss": 0.2825,
+      "step": 11304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35964780048498673,
+      "learning_rate": 5.78009614549725e-06,
+      "loss": 0.2912,
+      "step": 11305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7982606825690333,
+      "learning_rate": 5.778409114640973e-06,
+      "loss": 0.3289,
+      "step": 11306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2002096946727734,
+      "learning_rate": 5.776722229978206e-06,
+      "loss": 0.5363,
+      "step": 11307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26494917021712633,
+      "learning_rate": 5.775035491567367e-06,
+      "loss": 0.2465,
+      "step": 11308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2207873813549719,
+      "learning_rate": 5.773348899466864e-06,
+      "loss": 0.1466,
+      "step": 11309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8187241696416505,
+      "learning_rate": 5.7716624537351105e-06,
+      "loss": 0.4156,
+      "step": 11310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3221642458006116,
+      "learning_rate": 5.769976154430507e-06,
+      "loss": 0.2634,
+      "step": 11311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3845684244489153,
+      "learning_rate": 5.768290001611446e-06,
+      "loss": 0.2996,
+      "step": 11312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9866515649350825,
+      "learning_rate": 5.7666039953363155e-06,
+      "loss": 0.3581,
+      "step": 11313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3284899038074513,
+      "learning_rate": 5.76491813566351e-06,
+      "loss": 0.2465,
+      "step": 11314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3074311625669295,
+      "learning_rate": 5.763232422651407e-06,
+      "loss": 0.5833,
+      "step": 11315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2768885144280665,
+      "learning_rate": 5.761546856358384e-06,
+      "loss": 0.2403,
+      "step": 11316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34299468161640334,
+      "learning_rate": 5.759861436842806e-06,
+      "loss": 0.2699,
+      "step": 11317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.846850459365096,
+      "learning_rate": 5.7581761641630485e-06,
+      "loss": 0.5027,
+      "step": 11318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25050625975393576,
+      "learning_rate": 5.756491038377469e-06,
+      "loss": 0.211,
+      "step": 11319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39637983568228946,
+      "learning_rate": 5.754806059544421e-06,
+      "loss": 0.2666,
+      "step": 11320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5585156954349844,
+      "learning_rate": 5.753121227722254e-06,
+      "loss": 0.3317,
+      "step": 11321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4894209756295622,
+      "learning_rate": 5.7514365429693186e-06,
+      "loss": 0.2484,
+      "step": 11322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39527617870681714,
+      "learning_rate": 5.749752005343954e-06,
+      "loss": 0.2635,
+      "step": 11323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33485572046557144,
+      "learning_rate": 5.7480676149044945e-06,
+      "loss": 0.2856,
+      "step": 11324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4181430086739952,
+      "learning_rate": 5.746383371709267e-06,
+      "loss": 0.2432,
+      "step": 11325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28865695519714385,
+      "learning_rate": 5.7446992758166035e-06,
+      "loss": 0.1886,
+      "step": 11326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.43098022642011,
+      "learning_rate": 5.743015327284822e-06,
+      "loss": 0.332,
+      "step": 11327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46720475592837557,
+      "learning_rate": 5.7413315261722355e-06,
+      "loss": 0.3363,
+      "step": 11328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3041795818931702,
+      "learning_rate": 5.739647872537157e-06,
+      "loss": 0.2173,
+      "step": 11329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.087779922116115,
+      "learning_rate": 5.737964366437885e-06,
+      "loss": 0.8105,
+      "step": 11330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23944258282874695,
+      "learning_rate": 5.736281007932727e-06,
+      "loss": 0.181,
+      "step": 11331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27234929845391487,
+      "learning_rate": 5.734597797079974e-06,
+      "loss": 0.2099,
+      "step": 11332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7745522496107576,
+      "learning_rate": 5.732914733937917e-06,
+      "loss": 0.4524,
+      "step": 11333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7024537449114042,
+      "learning_rate": 5.731231818564834e-06,
+      "loss": 0.4207,
+      "step": 11334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3230484251573659,
+      "learning_rate": 5.729549051019014e-06,
+      "loss": 0.2026,
+      "step": 11335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3556886103604453,
+      "learning_rate": 5.7278664313587275e-06,
+      "loss": 0.3461,
+      "step": 11336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30108299641494074,
+      "learning_rate": 5.726183959642242e-06,
+      "loss": 0.1966,
+      "step": 11337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6097747958982636,
+      "learning_rate": 5.724501635927818e-06,
+      "loss": 0.3249,
+      "step": 11338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37914027796049266,
+      "learning_rate": 5.722819460273723e-06,
+      "loss": 0.2742,
+      "step": 11339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35501758807427714,
+      "learning_rate": 5.7211374327382066e-06,
+      "loss": 0.2916,
+      "step": 11340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5940137811131664,
+      "learning_rate": 5.719455553379516e-06,
+      "loss": 0.3368,
+      "step": 11341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36694725493194763,
+      "learning_rate": 5.717773822255896e-06,
+      "loss": 0.2532,
+      "step": 11342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23800875625112516,
+      "learning_rate": 5.71609223942558e-06,
+      "loss": 0.2069,
+      "step": 11343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37304610142960154,
+      "learning_rate": 5.7144108049468106e-06,
+      "loss": 0.2867,
+      "step": 11344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36018737364414755,
+      "learning_rate": 5.712729518877813e-06,
+      "loss": 0.2725,
+      "step": 11345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.454786122363037,
+      "learning_rate": 5.711048381276801e-06,
+      "loss": 0.838,
+      "step": 11346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4161241033055131,
+      "learning_rate": 5.709367392202003e-06,
+      "loss": 0.2913,
+      "step": 11347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28370028246480244,
+      "learning_rate": 5.707686551711628e-06,
+      "loss": 0.2552,
+      "step": 11348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2784506630346675,
+      "learning_rate": 5.706005859863883e-06,
+      "loss": 0.1635,
+      "step": 11349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36016396685166263,
+      "learning_rate": 5.704325316716966e-06,
+      "loss": 0.3082,
+      "step": 11350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4467132067514661,
+      "learning_rate": 5.702644922329083e-06,
+      "loss": 0.3476,
+      "step": 11351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.327846838982423,
+      "learning_rate": 5.70096467675842e-06,
+      "loss": 0.2569,
+      "step": 11352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.367416999026494,
+      "learning_rate": 5.699284580063167e-06,
+      "loss": 0.2946,
+      "step": 11353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5459229122762558,
+      "learning_rate": 5.697604632301504e-06,
+      "loss": 0.4332,
+      "step": 11354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25668900132390404,
+      "learning_rate": 5.695924833531603e-06,
+      "loss": 0.2039,
+      "step": 11355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4629903377649034,
+      "learning_rate": 5.6942451838116445e-06,
+      "loss": 0.2769,
+      "step": 11356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2880039976122195,
+      "learning_rate": 5.69256568319979e-06,
+      "loss": 0.2513,
+      "step": 11357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0645379082384252,
+      "learning_rate": 5.6908863317542e-06,
+      "loss": 0.3649,
+      "step": 11358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3809085828908202,
+      "learning_rate": 5.689207129533027e-06,
+      "loss": 0.3278,
+      "step": 11359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33329868938159685,
+      "learning_rate": 5.687528076594432e-06,
+      "loss": 0.3125,
+      "step": 11360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7300849087139614,
+      "learning_rate": 5.685849172996551e-06,
+      "loss": 0.411,
+      "step": 11361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4960419001943459,
+      "learning_rate": 5.6841704187975296e-06,
+      "loss": 0.2599,
+      "step": 11362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21882316289564832,
+      "learning_rate": 5.682491814055497e-06,
+      "loss": 0.2088,
+      "step": 11363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4726847855526382,
+      "learning_rate": 5.680813358828592e-06,
+      "loss": 0.3169,
+      "step": 11364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30338445447650797,
+      "learning_rate": 5.679135053174932e-06,
+      "loss": 0.1485,
+      "step": 11365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4310514017451592,
+      "learning_rate": 5.677456897152641e-06,
+      "loss": 0.3493,
+      "step": 11366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33338075483482793,
+      "learning_rate": 5.6757788908198316e-06,
+      "loss": 0.2944,
+      "step": 11367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4127582369100299,
+      "learning_rate": 5.674101034234609e-06,
+      "loss": 0.239,
+      "step": 11368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3351444220704729,
+      "learning_rate": 5.672423327455085e-06,
+      "loss": 0.2188,
+      "step": 11369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4345390366631468,
+      "learning_rate": 5.670745770539356e-06,
+      "loss": 0.2443,
+      "step": 11370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24478960709092923,
+      "learning_rate": 5.669068363545516e-06,
+      "loss": 0.2213,
+      "step": 11371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4821758480038891,
+      "learning_rate": 5.667391106531647e-06,
+      "loss": 0.3277,
+      "step": 11372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7428114884812085,
+      "learning_rate": 5.665713999555842e-06,
+      "loss": 0.3687,
+      "step": 11373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7485401252645069,
+      "learning_rate": 5.6640370426761735e-06,
+      "loss": 0.4149,
+      "step": 11374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23688874268890042,
+      "learning_rate": 5.662360235950717e-06,
+      "loss": 0.1964,
+      "step": 11375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31576975671951585,
+      "learning_rate": 5.6606835794375346e-06,
+      "loss": 0.254,
+      "step": 11376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7108945036865616,
+      "learning_rate": 5.659007073194697e-06,
+      "loss": 0.4271,
+      "step": 11377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33881004573315643,
+      "learning_rate": 5.657330717280258e-06,
+      "loss": 0.2427,
+      "step": 11378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35024537485757967,
+      "learning_rate": 5.655654511752274e-06,
+      "loss": 0.2942,
+      "step": 11379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6233162286325228,
+      "learning_rate": 5.653978456668779e-06,
+      "loss": 0.3727,
+      "step": 11380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23743888501996108,
+      "learning_rate": 5.652302552087827e-06,
+      "loss": 0.1682,
+      "step": 11381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42937799121144055,
+      "learning_rate": 5.6506267980674515e-06,
+      "loss": 0.2655,
+      "step": 11382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3103419789202479,
+      "learning_rate": 5.648951194665683e-06,
+      "loss": 0.294,
+      "step": 11383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32653676277691335,
+      "learning_rate": 5.647275741940543e-06,
+      "loss": 0.2316,
+      "step": 11384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6254068439076284,
+      "learning_rate": 5.645600439950061e-06,
+      "loss": 0.4076,
+      "step": 11385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8407360463120334,
+      "learning_rate": 5.643925288752248e-06,
+      "loss": 0.4822,
+      "step": 11386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35286540111542963,
+      "learning_rate": 5.642250288405116e-06,
+      "loss": 0.2868,
+      "step": 11387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2163730802370492,
+      "learning_rate": 5.6405754389666635e-06,
+      "loss": 0.1894,
+      "step": 11388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5102026088195443,
+      "learning_rate": 5.638900740494901e-06,
+      "loss": 0.3567,
+      "step": 11389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39499275164180747,
+      "learning_rate": 5.637226193047818e-06,
+      "loss": 0.293,
+      "step": 11390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36184693966404863,
+      "learning_rate": 5.635551796683405e-06,
+      "loss": 0.2894,
+      "step": 11391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1693194311109523,
+      "learning_rate": 5.633877551459646e-06,
+      "loss": 0.5964,
+      "step": 11392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3155464664914508,
+      "learning_rate": 5.6322034574345145e-06,
+      "loss": 0.2347,
+      "step": 11393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20528795990945872,
+      "learning_rate": 5.630529514665993e-06,
+      "loss": 0.1466,
+      "step": 11394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4960690183431173,
+      "learning_rate": 5.628855723212048e-06,
+      "loss": 0.3488,
+      "step": 11395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32121279561481897,
+      "learning_rate": 5.62718208313064e-06,
+      "loss": 0.2645,
+      "step": 11396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7035742236195769,
+      "learning_rate": 5.625508594479725e-06,
+      "loss": 0.3335,
+      "step": 11397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7735193256482944,
+      "learning_rate": 5.6238352573172635e-06,
+      "loss": 0.4442,
+      "step": 11398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26150894556463467,
+      "learning_rate": 5.622162071701198e-06,
+      "loss": 0.2663,
+      "step": 11399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27248367312206967,
+      "learning_rate": 5.6204890376894735e-06,
+      "loss": 0.1802,
+      "step": 11400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6700607475194417,
+      "learning_rate": 5.61881615534002e-06,
+      "loss": 0.2683,
+      "step": 11401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3534029225730532,
+      "learning_rate": 5.617143424710778e-06,
+      "loss": 0.3028,
+      "step": 11402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3681851782516326,
+      "learning_rate": 5.615470845859672e-06,
+      "loss": 0.2829,
+      "step": 11403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5710242786946949,
+      "learning_rate": 5.613798418844623e-06,
+      "loss": 0.2027,
+      "step": 11404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37252093370364825,
+      "learning_rate": 5.6121261437235445e-06,
+      "loss": 0.2886,
+      "step": 11405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30991829143518573,
+      "learning_rate": 5.6104540205543445e-06,
+      "loss": 0.1872,
+      "step": 11406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3134833617394962,
+      "learning_rate": 5.608782049394938e-06,
+      "loss": 0.2361,
+      "step": 11407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3926136787003498,
+      "learning_rate": 5.60711023030322e-06,
+      "loss": 0.3061,
+      "step": 11408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9496778327004386,
+      "learning_rate": 5.605438563337087e-06,
+      "loss": 0.7436,
+      "step": 11409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4551020413608209,
+      "learning_rate": 5.6037670485544215e-06,
+      "loss": 0.2394,
+      "step": 11410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2614793357489382,
+      "learning_rate": 5.60209568601312e-06,
+      "loss": 0.2574,
+      "step": 11411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4548863557191528,
+      "learning_rate": 5.600424475771058e-06,
+      "loss": 0.2345,
+      "step": 11412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6171749330064369,
+      "learning_rate": 5.5987534178861e-06,
+      "loss": 0.3804,
+      "step": 11413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.310695729368847,
+      "learning_rate": 5.5970825124161255e-06,
+      "loss": 0.2084,
+      "step": 11414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2857621372988274,
+      "learning_rate": 5.595411759418995e-06,
+      "loss": 0.2615,
+      "step": 11415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9880804735860319,
+      "learning_rate": 5.5937411589525655e-06,
+      "loss": 0.5304,
+      "step": 11416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29446071991324213,
+      "learning_rate": 5.592070711074691e-06,
+      "loss": 0.1841,
+      "step": 11417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7950217222504905,
+      "learning_rate": 5.590400415843214e-06,
+      "loss": 0.4106,
+      "step": 11418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2692860474214037,
+      "learning_rate": 5.5887302733159835e-06,
+      "loss": 0.2516,
+      "step": 11419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28812730174273965,
+      "learning_rate": 5.587060283550835e-06,
+      "loss": 0.2093,
+      "step": 11420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4135254933127519,
+      "learning_rate": 5.585390446605598e-06,
+      "loss": 0.2756,
+      "step": 11421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3391811932650513,
+      "learning_rate": 5.583720762538097e-06,
+      "loss": 0.302,
+      "step": 11422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5581481167123312,
+      "learning_rate": 5.58205123140616e-06,
+      "loss": 0.1832,
+      "step": 11423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8472758677182067,
+      "learning_rate": 5.5803818532676e-06,
+      "loss": 0.402,
+      "step": 11424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8200728581102539,
+      "learning_rate": 5.578712628180225e-06,
+      "loss": 0.4662,
+      "step": 11425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3875008712683409,
+      "learning_rate": 5.577043556201838e-06,
+      "loss": 0.2686,
+      "step": 11426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2416269514785201,
+      "learning_rate": 5.575374637390246e-06,
+      "loss": 0.2104,
+      "step": 11427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4898967331695952,
+      "learning_rate": 5.573705871803241e-06,
+      "loss": 0.2438,
+      "step": 11428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3172534747512469,
+      "learning_rate": 5.57203725949861e-06,
+      "loss": 0.2663,
+      "step": 11429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9321900723666333,
+      "learning_rate": 5.570368800534139e-06,
+      "loss": 0.3019,
+      "step": 11430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4761349234786449,
+      "learning_rate": 5.568700494967603e-06,
+      "loss": 0.3314,
+      "step": 11431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32465852015725005,
+      "learning_rate": 5.567032342856781e-06,
+      "loss": 0.2739,
+      "step": 11432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1793227954302825,
+      "learning_rate": 5.565364344259438e-06,
+      "loss": 0.1204,
+      "step": 11433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3672817087183113,
+      "learning_rate": 5.563696499233337e-06,
+      "loss": 0.3124,
+      "step": 11434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33119602560147793,
+      "learning_rate": 5.562028807836233e-06,
+      "loss": 0.2328,
+      "step": 11435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7227754858322918,
+      "learning_rate": 5.560361270125884e-06,
+      "loss": 0.3104,
+      "step": 11436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0485837440591883,
+      "learning_rate": 5.558693886160032e-06,
+      "loss": 0.5809,
+      "step": 11437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32761246481860695,
+      "learning_rate": 5.557026655996422e-06,
+      "loss": 0.2612,
+      "step": 11438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37388839229218224,
+      "learning_rate": 5.555359579692782e-06,
+      "loss": 0.3131,
+      "step": 11439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.339682737117197,
+      "learning_rate": 5.553692657306853e-06,
+      "loss": 0.1579,
+      "step": 11440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33306031343411013,
+      "learning_rate": 5.552025888896356e-06,
+      "loss": 0.2583,
+      "step": 11441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9423663180233574,
+      "learning_rate": 5.550359274519012e-06,
+      "loss": 0.4231,
+      "step": 11442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3801574172959991,
+      "learning_rate": 5.54869281423253e-06,
+      "loss": 0.2682,
+      "step": 11443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37924059353551987,
+      "learning_rate": 5.547026508094629e-06,
+      "loss": 0.2664,
+      "step": 11444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5090379498547105,
+      "learning_rate": 5.545360356163009e-06,
+      "loss": 0.3892,
+      "step": 11445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.22891711398429887,
+      "learning_rate": 5.54369435849537e-06,
+      "loss": 0.182,
+      "step": 11446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47266754852632326,
+      "learning_rate": 5.5420285151493995e-06,
+      "loss": 0.3223,
+      "step": 11447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6746597708337676,
+      "learning_rate": 5.540362826182791e-06,
+      "loss": 0.3292,
+      "step": 11448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2132277880660471,
+      "learning_rate": 5.538697291653228e-06,
+      "loss": 0.7311,
+      "step": 11449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3082460870951119,
+      "learning_rate": 5.537031911618385e-06,
+      "loss": 0.2031,
+      "step": 11450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3597103355681424,
+      "learning_rate": 5.53536668613593e-06,
+      "loss": 0.3081,
+      "step": 11451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41272000121983643,
+      "learning_rate": 5.5337016152635396e-06,
+      "loss": 0.2683,
+      "step": 11452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3017020610938978,
+      "learning_rate": 5.53203669905887e-06,
+      "loss": 0.2304,
+      "step": 11453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43437243190707603,
+      "learning_rate": 5.530371937579577e-06,
+      "loss": 0.2326,
+      "step": 11454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3317204639922343,
+      "learning_rate": 5.528707330883308e-06,
+      "loss": 0.2883,
+      "step": 11455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3327247393119727,
+      "learning_rate": 5.527042879027715e-06,
+      "loss": 0.2176,
+      "step": 11456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3365246378891962,
+      "learning_rate": 5.525378582070438e-06,
+      "loss": 0.7339,
+      "step": 11457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3413124951106821,
+      "learning_rate": 5.523714440069104e-06,
+      "loss": 0.3221,
+      "step": 11458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3312639650942997,
+      "learning_rate": 5.522050453081349e-06,
+      "loss": 0.2081,
+      "step": 11459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25778157585553163,
+      "learning_rate": 5.5203866211647904e-06,
+      "loss": 0.1958,
+      "step": 11460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9463391804603914,
+      "learning_rate": 5.518722944377053e-06,
+      "loss": 0.507,
+      "step": 11461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37823758384435996,
+      "learning_rate": 5.517059422775748e-06,
+      "loss": 0.3032,
+      "step": 11462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2921793595846805,
+      "learning_rate": 5.515396056418482e-06,
+      "loss": 0.2408,
+      "step": 11463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5883154221839674,
+      "learning_rate": 5.513732845362856e-06,
+      "loss": 0.4082,
+      "step": 11464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2867425278106484,
+      "learning_rate": 5.51206978966647e-06,
+      "loss": 0.1848,
+      "step": 11465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24205597069931772,
+      "learning_rate": 5.510406889386914e-06,
+      "loss": 0.2085,
+      "step": 11466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2740952812449393,
+      "learning_rate": 5.5087441445817765e-06,
+      "loss": 0.7298,
+      "step": 11467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5389920094979618,
+      "learning_rate": 5.507081555308631e-06,
+      "loss": 0.3645,
+      "step": 11468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3339186256662309,
+      "learning_rate": 5.505419121625062e-06,
+      "loss": 0.2649,
+      "step": 11469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3448488174392606,
+      "learning_rate": 5.503756843588635e-06,
+      "loss": 0.3087,
+      "step": 11470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3839012951403879,
+      "learning_rate": 5.502094721256916e-06,
+      "loss": 0.2503,
+      "step": 11471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24422334705090812,
+      "learning_rate": 5.500432754687464e-06,
+      "loss": 0.1323,
+      "step": 11472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0341671776790258,
+      "learning_rate": 5.498770943937828e-06,
+      "loss": 0.6798,
+      "step": 11473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2849180917522527,
+      "learning_rate": 5.497109289065563e-06,
+      "loss": 0.2485,
+      "step": 11474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5072468134724325,
+      "learning_rate": 5.495447790128211e-06,
+      "loss": 0.3332,
+      "step": 11475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6364321717273036,
+      "learning_rate": 5.493786447183308e-06,
+      "loss": 0.3118,
+      "step": 11476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.603457958394203,
+      "learning_rate": 5.4921252602883834e-06,
+      "loss": 0.2669,
+      "step": 11477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24790100570896947,
+      "learning_rate": 5.490464229500969e-06,
+      "loss": 0.2388,
+      "step": 11478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26144460223892885,
+      "learning_rate": 5.488803354878587e-06,
+      "loss": 0.2021,
+      "step": 11479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5581526805294116,
+      "learning_rate": 5.487142636478749e-06,
+      "loss": 0.3137,
+      "step": 11480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.46154960985058663,
+      "learning_rate": 5.485482074358968e-06,
+      "loss": 0.3153,
+      "step": 11481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34216719846377985,
+      "learning_rate": 5.48382166857675e-06,
+      "loss": 0.2507,
+      "step": 11482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5615865278512538,
+      "learning_rate": 5.482161419189591e-06,
+      "loss": 0.2432,
+      "step": 11483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23731768796120098,
+      "learning_rate": 5.4805013262549885e-06,
+      "loss": 0.1988,
+      "step": 11484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.286312030117045,
+      "learning_rate": 5.478841389830427e-06,
+      "loss": 0.4339,
+      "step": 11485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30921205394689105,
+      "learning_rate": 5.477181609973399e-06,
+      "loss": 0.2737,
+      "step": 11486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.38646319079933794,
+      "learning_rate": 5.475521986741377e-06,
+      "loss": 0.2982,
+      "step": 11487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0393640176508394,
+      "learning_rate": 5.4738625201918324e-06,
+      "loss": 0.5382,
+      "step": 11488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2587507359924883,
+      "learning_rate": 5.472203210382231e-06,
+      "loss": 0.0732,
+      "step": 11489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25179914104749707,
+      "learning_rate": 5.470544057370042e-06,
+      "loss": 0.2673,
+      "step": 11490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26847000724967895,
+      "learning_rate": 5.468885061212716e-06,
+      "loss": 0.1861,
+      "step": 11491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33831228269888364,
+      "learning_rate": 5.467226221967707e-06,
+      "loss": 0.2408,
+      "step": 11492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49982395528851653,
+      "learning_rate": 5.465567539692455e-06,
+      "loss": 0.3434,
+      "step": 11493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3667710672250758,
+      "learning_rate": 5.463909014444409e-06,
+      "loss": 0.316,
+      "step": 11494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5919667616125639,
+      "learning_rate": 5.462250646280997e-06,
+      "loss": 0.174,
+      "step": 11495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2919001787559771,
+      "learning_rate": 5.460592435259651e-06,
+      "loss": 0.2295,
+      "step": 11496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45198607078143405,
+      "learning_rate": 5.458934381437793e-06,
+      "loss": 0.3138,
+      "step": 11497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29738566743482925,
+      "learning_rate": 5.457276484872839e-06,
+      "loss": 0.2406,
+      "step": 11498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34804859280565137,
+      "learning_rate": 5.455618745622209e-06,
+      "loss": 0.2933,
+      "step": 11499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1076657921418147,
+      "learning_rate": 5.453961163743304e-06,
+      "loss": 0.6148,
+      "step": 11500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7959871319468619,
+      "learning_rate": 5.452303739293532e-06,
+      "loss": 0.3982,
+      "step": 11501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2500213796943707,
+      "learning_rate": 5.4506464723302784e-06,
+      "loss": 0.2233,
+      "step": 11502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.39779781556021504,
+      "learning_rate": 5.448989362910949e-06,
+      "loss": 0.2594,
+      "step": 11503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5295136834773769,
+      "learning_rate": 5.447332411092921e-06,
+      "loss": 0.3786,
+      "step": 11504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23936075071175678,
+      "learning_rate": 5.445675616933576e-06,
+      "loss": 0.1819,
+      "step": 11505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34567033680043985,
+      "learning_rate": 5.444018980490284e-06,
+      "loss": 0.2925,
+      "step": 11506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7141315652138195,
+      "learning_rate": 5.4423625018204226e-06,
+      "loss": 0.3823,
+      "step": 11507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32578121816347494,
+      "learning_rate": 5.440706180981352e-06,
+      "loss": 0.2104,
+      "step": 11508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.633359243897076,
+      "learning_rate": 5.439050018030432e-06,
+      "loss": 0.368,
+      "step": 11509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3036594850395294,
+      "learning_rate": 5.437394013025012e-06,
+      "loss": 0.2951,
+      "step": 11510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3218116260567253,
+      "learning_rate": 5.435738166022437e-06,
+      "loss": 0.2107,
+      "step": 11511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27971976076190586,
+      "learning_rate": 5.434082477080058e-06,
+      "loss": 0.1691,
+      "step": 11512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37628575012172416,
+      "learning_rate": 5.432426946255206e-06,
+      "loss": 0.2813,
+      "step": 11513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29629151440877577,
+      "learning_rate": 5.4307715736052125e-06,
+      "loss": 0.2711,
+      "step": 11514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7761394695030567,
+      "learning_rate": 5.429116359187403e-06,
+      "loss": 0.3176,
+      "step": 11515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1914877581468113,
+      "learning_rate": 5.427461303059096e-06,
+      "loss": 0.8457,
+      "step": 11516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3173092424468873,
+      "learning_rate": 5.425806405277609e-06,
+      "loss": 0.2569,
+      "step": 11517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24185940664168293,
+      "learning_rate": 5.424151665900246e-06,
+      "loss": 0.1908,
+      "step": 11518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2671168707668683,
+      "learning_rate": 5.422497084984317e-06,
+      "loss": 0.5888,
+      "step": 11519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3204691857610319,
+      "learning_rate": 5.420842662587118e-06,
+      "loss": 0.2625,
+      "step": 11520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.837128435683929,
+      "learning_rate": 5.41918839876594e-06,
+      "loss": 0.3109,
+      "step": 11521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28996702322416396,
+      "learning_rate": 5.41753429357807e-06,
+      "loss": 0.2688,
+      "step": 11522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3468731094237857,
+      "learning_rate": 5.4158803470807875e-06,
+      "loss": 0.2796,
+      "step": 11523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25427661796028395,
+      "learning_rate": 5.414226559331375e-06,
+      "loss": 0.1582,
+      "step": 11524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28924380253126614,
+      "learning_rate": 5.4125729303871e-06,
+      "loss": 0.229,
+      "step": 11525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.371977089760644,
+      "learning_rate": 5.410919460305226e-06,
+      "loss": 0.2852,
+      "step": 11526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7166682064653169,
+      "learning_rate": 5.409266149143011e-06,
+      "loss": 0.4062,
+      "step": 11527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5786444654582379,
+      "learning_rate": 5.407612996957716e-06,
+      "loss": 0.3038,
+      "step": 11528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34804236614724643,
+      "learning_rate": 5.405960003806585e-06,
+      "loss": 0.2652,
+      "step": 11529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.22885662878947813,
+      "learning_rate": 5.4043071697468604e-06,
+      "loss": 0.2186,
+      "step": 11530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6317423866420934,
+      "learning_rate": 5.4026544948357795e-06,
+      "loss": 0.2647,
+      "step": 11531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3832186602134028,
+      "learning_rate": 5.401001979130578e-06,
+      "loss": 0.273,
+      "step": 11532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.46543784414296346,
+      "learning_rate": 5.399349622688479e-06,
+      "loss": 0.3423,
+      "step": 11533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.444348249465614,
+      "learning_rate": 5.397697425566707e-06,
+      "loss": 0.2682,
+      "step": 11534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32841907606679893,
+      "learning_rate": 5.396045387822474e-06,
+      "loss": 0.2644,
+      "step": 11535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29513111111681184,
+      "learning_rate": 5.394393509512987e-06,
+      "loss": 0.1942,
+      "step": 11536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4622638865061429,
+      "learning_rate": 5.392741790695459e-06,
+      "loss": 0.346,
+      "step": 11537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2652974325826972,
+      "learning_rate": 5.391090231427086e-06,
+      "loss": 0.2254,
+      "step": 11538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8429674842171873,
+      "learning_rate": 5.389438831765059e-06,
+      "loss": 0.4517,
+      "step": 11539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2820194430463543,
+      "learning_rate": 5.387787591766562e-06,
+      "loss": 0.8453,
+      "step": 11540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25818239711603863,
+      "learning_rate": 5.386136511488789e-06,
+      "loss": 0.211,
+      "step": 11541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4809465676464306,
+      "learning_rate": 5.384485590988908e-06,
+      "loss": 0.3567,
+      "step": 11542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5760262213588997,
+      "learning_rate": 5.382834830324093e-06,
+      "loss": 0.2892,
+      "step": 11543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2266235624622428,
+      "learning_rate": 5.381184229551506e-06,
+      "loss": 0.1549,
+      "step": 11544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5227017467399918,
+      "learning_rate": 5.379533788728313e-06,
+      "loss": 0.3697,
+      "step": 11545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48189898826149535,
+      "learning_rate": 5.377883507911668e-06,
+      "loss": 0.4144,
+      "step": 11546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2966848658482328,
+      "learning_rate": 5.376233387158722e-06,
+      "loss": 0.1897,
+      "step": 11547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6944979318698845,
+      "learning_rate": 5.3745834265266054e-06,
+      "loss": 0.4308,
+      "step": 11548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2628428496235638,
+      "learning_rate": 5.372933626072472e-06,
+      "loss": 0.2565,
+      "step": 11549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3982433001251939,
+      "learning_rate": 5.371283985853446e-06,
+      "loss": 0.2565,
+      "step": 11550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31675267485780206,
+      "learning_rate": 5.369634505926658e-06,
+      "loss": 0.1933,
+      "step": 11551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1224053933698808,
+      "learning_rate": 5.367985186349223e-06,
+      "loss": 0.7319,
+      "step": 11552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3152515760305445,
+      "learning_rate": 5.3663360271782675e-06,
+      "loss": 0.2678,
+      "step": 11553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3356411560582424,
+      "learning_rate": 5.364687028470894e-06,
+      "loss": 0.2476,
+      "step": 11554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6719810096510042,
+      "learning_rate": 5.363038190284211e-06,
+      "loss": 0.4357,
+      "step": 11555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2516818101597573,
+      "learning_rate": 5.36138951267531e-06,
+      "loss": 0.2007,
+      "step": 11556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2570646063899158,
+      "learning_rate": 5.359740995701297e-06,
+      "loss": 0.1896,
+      "step": 11557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0684404877013456,
+      "learning_rate": 5.358092639419252e-06,
+      "loss": 0.7644,
+      "step": 11558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.313824020741263,
+      "learning_rate": 5.356444443886262e-06,
+      "loss": 0.2709,
+      "step": 11559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7877441786641255,
+      "learning_rate": 5.3547964091593955e-06,
+      "loss": 0.3024,
+      "step": 11560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3567833071601899,
+      "learning_rate": 5.353148535295733e-06,
+      "loss": 0.3206,
+      "step": 11561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24311132372342875,
+      "learning_rate": 5.351500822352338e-06,
+      "loss": 0.1998,
+      "step": 11562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4527933293358373,
+      "learning_rate": 5.3498532703862685e-06,
+      "loss": 0.2622,
+      "step": 11563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4517449777092792,
+      "learning_rate": 5.34820587945458e-06,
+      "loss": 0.2963,
+      "step": 11564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4205026001874601,
+      "learning_rate": 5.34655864961432e-06,
+      "loss": 0.2904,
+      "step": 11565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5018741672324546,
+      "learning_rate": 5.344911580922536e-06,
+      "loss": 0.3114,
+      "step": 11566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3871254278340785,
+      "learning_rate": 5.343264673436264e-06,
+      "loss": 0.1549,
+      "step": 11567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.392704311459362,
+      "learning_rate": 5.341617927212537e-06,
+      "loss": 0.2629,
+      "step": 11568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2654887678301144,
+      "learning_rate": 5.339971342308377e-06,
+      "loss": 0.2367,
+      "step": 11569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9922960293520513,
+      "learning_rate": 5.33832491878081e-06,
+      "loss": 0.5467,
+      "step": 11570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41226520432516683,
+      "learning_rate": 5.3366786566868545e-06,
+      "loss": 0.2878,
+      "step": 11571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4855856005335587,
+      "learning_rate": 5.335032556083515e-06,
+      "loss": 0.3625,
+      "step": 11572
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37447574962277336,
+      "learning_rate": 5.333386617027793e-06,
+      "loss": 0.2506,
+      "step": 11573
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32366398711854405,
+      "learning_rate": 5.331740839576697e-06,
+      "loss": 0.2546,
+      "step": 11574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2625217824484109,
+      "learning_rate": 5.330095223787214e-06,
+      "loss": 0.1951,
+      "step": 11575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.779074955761385,
+      "learning_rate": 5.3284497697163325e-06,
+      "loss": 0.5442,
+      "step": 11576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27009579975714293,
+      "learning_rate": 5.326804477421035e-06,
+      "loss": 0.2266,
+      "step": 11577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.42870519581357974,
+      "learning_rate": 5.325159346958293e-06,
+      "loss": 0.3498,
+      "step": 11578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0770530022845466,
+      "learning_rate": 5.323514378385086e-06,
+      "loss": 0.5653,
+      "step": 11579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25141483196515735,
+      "learning_rate": 5.321869571758375e-06,
+      "loss": 0.1638,
+      "step": 11580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31187570399237746,
+      "learning_rate": 5.32022492713512e-06,
+      "loss": 0.2445,
+      "step": 11581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7460901109990338,
+      "learning_rate": 5.318580444572276e-06,
+      "loss": 0.4426,
+      "step": 11582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43048688929882034,
+      "learning_rate": 5.316936124126788e-06,
+      "loss": 0.2262,
+      "step": 11583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40015207220498034,
+      "learning_rate": 5.3152919658556e-06,
+      "loss": 0.3021,
+      "step": 11584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3664291203534989,
+      "learning_rate": 5.313647969815647e-06,
+      "loss": 0.2887,
+      "step": 11585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.351670785679463,
+      "learning_rate": 5.312004136063866e-06,
+      "loss": 0.184,
+      "step": 11586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31282841073476086,
+      "learning_rate": 5.310360464657183e-06,
+      "loss": 0.2668,
+      "step": 11587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9626155482266147,
+      "learning_rate": 5.308716955652513e-06,
+      "loss": 0.4174,
+      "step": 11588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2985410671209383,
+      "learning_rate": 5.3070736091067734e-06,
+      "loss": 0.267,
+      "step": 11589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30113331654362685,
+      "learning_rate": 5.30543042507687e-06,
+      "loss": 0.2067,
+      "step": 11590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0546116846708513,
+      "learning_rate": 5.303787403619711e-06,
+      "loss": 0.4606,
+      "step": 11591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3379662424665475,
+      "learning_rate": 5.302144544792194e-06,
+      "loss": 0.2209,
+      "step": 11592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2875890580534507,
+      "learning_rate": 5.300501848651209e-06,
+      "loss": 0.244,
+      "step": 11593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0134444255063182,
+      "learning_rate": 5.298859315253639e-06,
+      "loss": 0.3835,
+      "step": 11594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41250760331534936,
+      "learning_rate": 5.297216944656371e-06,
+      "loss": 0.3064,
+      "step": 11595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30618938735708207,
+      "learning_rate": 5.29557473691628e-06,
+      "loss": 0.1658,
+      "step": 11596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3628548944099711,
+      "learning_rate": 5.293932692090233e-06,
+      "loss": 0.3127,
+      "step": 11597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3740872155732483,
+      "learning_rate": 5.292290810235092e-06,
+      "loss": 0.2613,
+      "step": 11598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5642476827593211,
+      "learning_rate": 5.29064909140772e-06,
+      "loss": 0.2667,
+      "step": 11599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5899163594087162,
+      "learning_rate": 5.289007535664967e-06,
+      "loss": 0.3414,
+      "step": 11600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40825253228739095,
+      "learning_rate": 5.287366143063682e-06,
+      "loss": 0.3243,
+      "step": 11601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2748846662378343,
+      "learning_rate": 5.285724913660704e-06,
+      "loss": 0.2326,
+      "step": 11602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43119599601784436,
+      "learning_rate": 5.284083847512866e-06,
+      "loss": 0.1601,
+      "step": 11603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8477040777808765,
+      "learning_rate": 5.282442944677005e-06,
+      "loss": 0.4426,
+      "step": 11604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27256531636671766,
+      "learning_rate": 5.280802205209943e-06,
+      "loss": 0.2666,
+      "step": 11605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5546628272888353,
+      "learning_rate": 5.279161629168497e-06,
+      "loss": 0.3214,
+      "step": 11606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5307811121875399,
+      "learning_rate": 5.2775212166094755e-06,
+      "loss": 0.3207,
+      "step": 11607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2547089293698558,
+      "learning_rate": 5.275880967589697e-06,
+      "loss": 0.2159,
+      "step": 11608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3482597708001019,
+      "learning_rate": 5.274240882165958e-06,
+      "loss": 0.2253,
+      "step": 11609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5292646057639918,
+      "learning_rate": 5.272600960395051e-06,
+      "loss": 0.283,
+      "step": 11610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3723792760127392,
+      "learning_rate": 5.270961202333769e-06,
+      "loss": 0.292,
+      "step": 11611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7733747358805955,
+      "learning_rate": 5.2693216080388984e-06,
+      "loss": 0.4955,
+      "step": 11612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29065692377208907,
+      "learning_rate": 5.267682177567219e-06,
+      "loss": 0.2262,
+      "step": 11613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.37946091941589666,
+      "learning_rate": 5.266042910975501e-06,
+      "loss": 0.3012,
+      "step": 11614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28954371908158133,
+      "learning_rate": 5.264403808320514e-06,
+      "loss": 0.1924,
+      "step": 11615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36618728513841026,
+      "learning_rate": 5.26276486965902e-06,
+      "loss": 0.2213,
+      "step": 11616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3013556099336767,
+      "learning_rate": 5.261126095047774e-06,
+      "loss": 0.2854,
+      "step": 11617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6303827060308433,
+      "learning_rate": 5.259487484543528e-06,
+      "loss": 0.4274,
+      "step": 11618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3363842194928553,
+      "learning_rate": 5.257849038203022e-06,
+      "loss": 0.1905,
+      "step": 11619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2881946093992692,
+      "learning_rate": 5.256210756083004e-06,
+      "loss": 0.2326,
+      "step": 11620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2732752542283715,
+      "learning_rate": 5.254572638240204e-06,
+      "loss": 0.2327,
+      "step": 11621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6126870326894789,
+      "learning_rate": 5.252934684731349e-06,
+      "loss": 0.2707,
+      "step": 11622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38004426464239194,
+      "learning_rate": 5.251296895613158e-06,
+      "loss": 0.2675,
+      "step": 11623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3487591686090327,
+      "learning_rate": 5.249659270942355e-06,
+      "loss": 0.3091,
+      "step": 11624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.809507380773592,
+      "learning_rate": 5.248021810775647e-06,
+      "loss": 0.4803,
+      "step": 11625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24443885048236402,
+      "learning_rate": 5.2463845151697404e-06,
+      "loss": 0.1772,
+      "step": 11626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4368490630538348,
+      "learning_rate": 5.2447473841813335e-06,
+      "loss": 0.2457,
+      "step": 11627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4995512757478243,
+      "learning_rate": 5.243110417867117e-06,
+      "loss": 0.3012,
+      "step": 11628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29444216072671153,
+      "learning_rate": 5.241473616283783e-06,
+      "loss": 0.2354,
+      "step": 11629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4348647745786949,
+      "learning_rate": 5.239836979488015e-06,
+      "loss": 0.7923,
+      "step": 11630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7504570213240344,
+      "learning_rate": 5.238200507536488e-06,
+      "loss": 0.4943,
+      "step": 11631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31611857707224705,
+      "learning_rate": 5.23656420048587e-06,
+      "loss": 0.2007,
+      "step": 11632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2280572315656847,
+      "learning_rate": 5.23492805839283e-06,
+      "loss": 0.2107,
+      "step": 11633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5290475222067625,
+      "learning_rate": 5.233292081314027e-06,
+      "loss": 0.3361,
+      "step": 11634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33595680323497873,
+      "learning_rate": 5.231656269306116e-06,
+      "loss": 0.2121,
+      "step": 11635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34149735897754147,
+      "learning_rate": 5.230020622425738e-06,
+      "loss": 0.2989,
+      "step": 11636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5854348227990583,
+      "learning_rate": 5.228385140729545e-06,
+      "loss": 0.4182,
+      "step": 11637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.37979442690169474,
+      "learning_rate": 5.226749824274169e-06,
+      "loss": 0.2815,
+      "step": 11638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4566569153966813,
+      "learning_rate": 5.225114673116243e-06,
+      "loss": 0.239,
+      "step": 11639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2744528609459148,
+      "learning_rate": 5.223479687312388e-06,
+      "loss": 0.2327,
+      "step": 11640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36138822589171504,
+      "learning_rate": 5.2218448669192235e-06,
+      "loss": 0.2638,
+      "step": 11641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8235685157243273,
+      "learning_rate": 5.220210211993371e-06,
+      "loss": 0.457,
+      "step": 11642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.145506050031028,
+      "learning_rate": 5.21857572259143e-06,
+      "loss": 0.6852,
+      "step": 11643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32843776235279604,
+      "learning_rate": 5.216941398770009e-06,
+      "loss": 0.2588,
+      "step": 11644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33328103317540847,
+      "learning_rate": 5.215307240585696e-06,
+      "loss": 0.2655,
+      "step": 11645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34161222318740736,
+      "learning_rate": 5.213673248095092e-06,
+      "loss": 0.2008,
+      "step": 11646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34405362854914645,
+      "learning_rate": 5.212039421354779e-06,
+      "loss": 0.2812,
+      "step": 11647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3367695864863821,
+      "learning_rate": 5.2104057604213335e-06,
+      "loss": 0.2562,
+      "step": 11648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7079692534150634,
+      "learning_rate": 5.208772265351332e-06,
+      "loss": 0.426,
+      "step": 11649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35211947295926666,
+      "learning_rate": 5.207138936201339e-06,
+      "loss": 0.2587,
+      "step": 11650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6453440156091795,
+      "learning_rate": 5.205505773027919e-06,
+      "loss": 0.3904,
+      "step": 11651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2491559057765381,
+      "learning_rate": 5.203872775887628e-06,
+      "loss": 0.1939,
+      "step": 11652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.362328629160573,
+      "learning_rate": 5.202239944837013e-06,
+      "loss": 0.283,
+      "step": 11653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4889478459733575,
+      "learning_rate": 5.200607279932626e-06,
+      "loss": 0.3449,
+      "step": 11654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6228592787338032,
+      "learning_rate": 5.198974781231003e-06,
+      "loss": 0.308,
+      "step": 11655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30679920540472044,
+      "learning_rate": 5.197342448788676e-06,
+      "loss": 0.2397,
+      "step": 11656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3359198585706779,
+      "learning_rate": 5.19571028266217e-06,
+      "loss": 0.3035,
+      "step": 11657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22988331423007832,
+      "learning_rate": 5.194078282908015e-06,
+      "loss": 0.0879,
+      "step": 11658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32083574318893576,
+      "learning_rate": 5.192446449582722e-06,
+      "loss": 0.2683,
+      "step": 11659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.37311945503442323,
+      "learning_rate": 5.190814782742801e-06,
+      "loss": 0.3295,
+      "step": 11660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9174082847992806,
+      "learning_rate": 5.1891832824447545e-06,
+      "loss": 0.3407,
+      "step": 11661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.311522227954505,
+      "learning_rate": 5.1875519487450865e-06,
+      "loss": 0.2712,
+      "step": 11662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7068267725620065,
+      "learning_rate": 5.185920781700288e-06,
+      "loss": 0.4194,
+      "step": 11663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3690138293865834,
+      "learning_rate": 5.184289781366847e-06,
+      "loss": 0.2892,
+      "step": 11664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2054627742441291,
+      "learning_rate": 5.182658947801242e-06,
+      "loss": 0.1386,
+      "step": 11665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2891895203008483,
+      "learning_rate": 5.1810282810599475e-06,
+      "loss": 0.6783,
+      "step": 11666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8395194551653319,
+      "learning_rate": 5.17939778119944e-06,
+      "loss": 0.412,
+      "step": 11667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2798409401034943,
+      "learning_rate": 5.1777674482761805e-06,
+      "loss": 0.2069,
+      "step": 11668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44920481661776707,
+      "learning_rate": 5.176137282346627e-06,
+      "loss": 0.3641,
+      "step": 11669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29918314203834345,
+      "learning_rate": 5.174507283467228e-06,
+      "loss": 0.1879,
+      "step": 11670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32543040157224057,
+      "learning_rate": 5.172877451694438e-06,
+      "loss": 0.197,
+      "step": 11671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36804171214608106,
+      "learning_rate": 5.171247787084694e-06,
+      "loss": 0.3286,
+      "step": 11672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9862885552626823,
+      "learning_rate": 5.169618289694432e-06,
+      "loss": 0.3905,
+      "step": 11673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2984786136741482,
+      "learning_rate": 5.167988959580077e-06,
+      "loss": 0.1931,
+      "step": 11674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0670030322829498,
+      "learning_rate": 5.16635979679806e-06,
+      "loss": 0.5376,
+      "step": 11675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34866352029633485,
+      "learning_rate": 5.1647308014047955e-06,
+      "loss": 0.2992,
+      "step": 11676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33441485196590653,
+      "learning_rate": 5.163101973456696e-06,
+      "loss": 0.2794,
+      "step": 11677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2906375868270081,
+      "learning_rate": 5.161473313010162e-06,
+      "loss": 0.15,
+      "step": 11678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9677833792997231,
+      "learning_rate": 5.159844820121605e-06,
+      "loss": 0.4,
+      "step": 11679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2778530544704572,
+      "learning_rate": 5.158216494847412e-06,
+      "loss": 0.2421,
+      "step": 11680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44108767628976453,
+      "learning_rate": 5.156588337243974e-06,
+      "loss": 0.2371,
+      "step": 11681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9779485327404348,
+      "learning_rate": 5.154960347367675e-06,
+      "loss": 0.4578,
+      "step": 11682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27024184345027874,
+      "learning_rate": 5.153332525274888e-06,
+      "loss": 0.225,
+      "step": 11683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3015393644257497,
+      "learning_rate": 5.1517048710219895e-06,
+      "loss": 0.2489,
+      "step": 11684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8142713116682633,
+      "learning_rate": 5.150077384665342e-06,
+      "loss": 0.397,
+      "step": 11685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.333100189552544,
+      "learning_rate": 5.148450066261303e-06,
+      "loss": 0.2676,
+      "step": 11686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4376687450485571,
+      "learning_rate": 5.146822915866232e-06,
+      "loss": 0.24,
+      "step": 11687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3483112994923384,
+      "learning_rate": 5.145195933536476e-06,
+      "loss": 0.2511,
+      "step": 11688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4132548196316583,
+      "learning_rate": 5.143569119328376e-06,
+      "loss": 0.2765,
+      "step": 11689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7860273994083556,
+      "learning_rate": 5.141942473298264e-06,
+      "loss": 0.2614,
+      "step": 11690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3396154895768527,
+      "learning_rate": 5.140315995502478e-06,
+      "loss": 0.2806,
+      "step": 11691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4005484919825302,
+      "learning_rate": 5.1386896859973425e-06,
+      "loss": 0.2488,
+      "step": 11692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33537309338762605,
+      "learning_rate": 5.1370635448391736e-06,
+      "loss": 0.2494,
+      "step": 11693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2551963101728314,
+      "learning_rate": 5.135437572084284e-06,
+      "loss": 0.2223,
+      "step": 11694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39248745343059227,
+      "learning_rate": 5.133811767788979e-06,
+      "loss": 0.2952,
+      "step": 11695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29648525882978316,
+      "learning_rate": 5.132186132009567e-06,
+      "loss": 0.2726,
+      "step": 11696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9013794775813733,
+      "learning_rate": 5.13056066480234e-06,
+      "loss": 0.357,
+      "step": 11697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34343934566226364,
+      "learning_rate": 5.128935366223588e-06,
+      "loss": 0.2625,
+      "step": 11698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2714716915290179,
+      "learning_rate": 5.12731023632959e-06,
+      "loss": 0.1737,
+      "step": 11699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3910270132526932,
+      "learning_rate": 5.125685275176633e-06,
+      "loss": 0.3165,
+      "step": 11700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35071460165870216,
+      "learning_rate": 5.124060482820986e-06,
+      "loss": 0.1882,
+      "step": 11701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9215729327737139,
+      "learning_rate": 5.122435859318915e-06,
+      "loss": 0.4102,
+      "step": 11702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45771396649945395,
+      "learning_rate": 5.120811404726675e-06,
+      "loss": 0.3709,
+      "step": 11703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27670028601173474,
+      "learning_rate": 5.119187119100533e-06,
+      "loss": 0.2096,
+      "step": 11704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26230396682039764,
+      "learning_rate": 5.117563002496728e-06,
+      "loss": 0.168,
+      "step": 11705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0274803578898541,
+      "learning_rate": 5.115939054971508e-06,
+      "loss": 0.4535,
+      "step": 11706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6142974067442903,
+      "learning_rate": 5.114315276581108e-06,
+      "loss": 0.2165,
+      "step": 11707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2637060378536606,
+      "learning_rate": 5.1126916673817575e-06,
+      "loss": 0.2753,
+      "step": 11708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.119474613305846,
+      "learning_rate": 5.111068227429686e-06,
+      "loss": 0.6991,
+      "step": 11709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3746375472716624,
+      "learning_rate": 5.109444956781113e-06,
+      "loss": 0.1687,
+      "step": 11710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26177824937585215,
+      "learning_rate": 5.10782185549225e-06,
+      "loss": 0.2256,
+      "step": 11711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48526546688989247,
+      "learning_rate": 5.106198923619302e-06,
+      "loss": 0.3127,
+      "step": 11712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6618222141302967,
+      "learning_rate": 5.10457616121848e-06,
+      "loss": 0.2913,
+      "step": 11713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45600963586374926,
+      "learning_rate": 5.102953568345973e-06,
+      "loss": 0.2749,
+      "step": 11714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4655809092972486,
+      "learning_rate": 5.101331145057975e-06,
+      "loss": 0.3785,
+      "step": 11715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36798216280033647,
+      "learning_rate": 5.0997088914106685e-06,
+      "loss": 0.2713,
+      "step": 11716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21140015509665294,
+      "learning_rate": 5.098086807460232e-06,
+      "loss": 0.1602,
+      "step": 11717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7371500660216942,
+      "learning_rate": 5.096464893262838e-06,
+      "loss": 0.3948,
+      "step": 11718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3635716936363465,
+      "learning_rate": 5.094843148874654e-06,
+      "loss": 0.2994,
+      "step": 11719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3033816677570989,
+      "learning_rate": 5.0932215743518375e-06,
+      "loss": 0.2358,
+      "step": 11720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0560147653471157,
+      "learning_rate": 5.0916001697505506e-06,
+      "loss": 0.6911,
+      "step": 11721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3864890302897584,
+      "learning_rate": 5.089978935126939e-06,
+      "loss": 0.2657,
+      "step": 11722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22419003265245574,
+      "learning_rate": 5.088357870537146e-06,
+      "loss": 0.1348,
+      "step": 11723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3287953682638831,
+      "learning_rate": 5.086736976037304e-06,
+      "loss": 0.3103,
+      "step": 11724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7126230773256188,
+      "learning_rate": 5.085116251683554e-06,
+      "loss": 0.3805,
+      "step": 11725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34080945340313146,
+      "learning_rate": 5.083495697532016e-06,
+      "loss": 0.2898,
+      "step": 11726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32289359281894736,
+      "learning_rate": 5.081875313638811e-06,
+      "loss": 0.2766,
+      "step": 11727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7495934803198708,
+      "learning_rate": 5.080255100060048e-06,
+      "loss": 0.3571,
+      "step": 11728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2572948506856289,
+      "learning_rate": 5.078635056851844e-06,
+      "loss": 0.2007,
+      "step": 11729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5697537663129776,
+      "learning_rate": 5.077015184070296e-06,
+      "loss": 0.2481,
+      "step": 11730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33024815771221205,
+      "learning_rate": 5.075395481771501e-06,
+      "loss": 0.2487,
+      "step": 11731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29132435493008624,
+      "learning_rate": 5.073775950011548e-06,
+      "loss": 0.2737,
+      "step": 11732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1874298610861767,
+      "learning_rate": 5.072156588846519e-06,
+      "loss": 0.5421,
+      "step": 11733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5883251543734425,
+      "learning_rate": 5.070537398332498e-06,
+      "loss": 0.3345,
+      "step": 11734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27343202258408916,
+      "learning_rate": 5.068918378525555e-06,
+      "loss": 0.2571,
+      "step": 11735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6437468881803801,
+      "learning_rate": 5.067299529481758e-06,
+      "loss": 0.2749,
+      "step": 11736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34159851863097196,
+      "learning_rate": 5.065680851257162e-06,
+      "loss": 0.1786,
+      "step": 11737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.37613981223297227,
+      "learning_rate": 5.0640623439078285e-06,
+      "loss": 0.2851,
+      "step": 11738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.365571260834238,
+      "learning_rate": 5.062444007489804e-06,
+      "loss": 0.2933,
+      "step": 11739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.346912717482984,
+      "learning_rate": 5.060825842059132e-06,
+      "loss": 0.1916,
+      "step": 11740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.47141640995945183,
+      "learning_rate": 5.059207847671845e-06,
+      "loss": 0.3235,
+      "step": 11741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45849898304823045,
+      "learning_rate": 5.05759002438398e-06,
+      "loss": 0.2979,
+      "step": 11742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25376962264594627,
+      "learning_rate": 5.055972372251562e-06,
+      "loss": 0.1864,
+      "step": 11743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3635867761463181,
+      "learning_rate": 5.054354891330607e-06,
+      "loss": 0.2763,
+      "step": 11744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.209100047476115,
+      "learning_rate": 5.05273758167713e-06,
+      "loss": 0.7096,
+      "step": 11745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5569931214330035,
+      "learning_rate": 5.051120443347134e-06,
+      "loss": 0.2018,
+      "step": 11746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2892336497399014,
+      "learning_rate": 5.049503476396627e-06,
+      "loss": 0.2789,
+      "step": 11747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43334647063447024,
+      "learning_rate": 5.047886680881603e-06,
+      "loss": 0.3457,
+      "step": 11748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.165429228208656,
+      "learning_rate": 5.0462700568580495e-06,
+      "loss": 0.0847,
+      "step": 11749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4192615834692852,
+      "learning_rate": 5.044653604381952e-06,
+      "loss": 0.3126,
+      "step": 11750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3438659632511523,
+      "learning_rate": 5.043037323509285e-06,
+      "loss": 0.3159,
+      "step": 11751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6009103614653322,
+      "learning_rate": 5.041421214296025e-06,
+      "loss": 0.375,
+      "step": 11752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32373127606272445,
+      "learning_rate": 5.039805276798128e-06,
+      "loss": 0.2451,
+      "step": 11753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0671526753829008,
+      "learning_rate": 5.0381895110715676e-06,
+      "loss": 0.5263,
+      "step": 11754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23259453773066574,
+      "learning_rate": 5.03657391717229e-06,
+      "loss": 0.1934,
+      "step": 11755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3365964910715124,
+      "learning_rate": 5.0349584951562445e-06,
+      "loss": 0.2178,
+      "step": 11756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0692943385938543,
+      "learning_rate": 5.033343245079373e-06,
+      "loss": 0.7361,
+      "step": 11757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5576953322811902,
+      "learning_rate": 5.031728166997607e-06,
+      "loss": 0.3471,
+      "step": 11758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2808777177692286,
+      "learning_rate": 5.0301132609668845e-06,
+      "loss": 0.2165,
+      "step": 11759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5072273706254729,
+      "learning_rate": 5.028498527043126e-06,
+      "loss": 0.3641,
+      "step": 11760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2782055835137195,
+      "learning_rate": 5.026883965282252e-06,
+      "loss": 0.1837,
+      "step": 11761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.322179200138884,
+      "learning_rate": 5.025269575740166e-06,
+      "loss": 0.222,
+      "step": 11762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3592214488272502,
+      "learning_rate": 5.023655358472786e-06,
+      "loss": 0.3299,
+      "step": 11763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6839481155341293,
+      "learning_rate": 5.022041313536006e-06,
+      "loss": 0.4145,
+      "step": 11764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.302911265362752,
+      "learning_rate": 5.020427440985721e-06,
+      "loss": 0.2343,
+      "step": 11765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2681543614082758,
+      "learning_rate": 5.018813740877817e-06,
+      "loss": 0.234,
+      "step": 11766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23444856503788336,
+      "learning_rate": 5.0172002132681815e-06,
+      "loss": 0.2076,
+      "step": 11767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32344719208450956,
+      "learning_rate": 5.0155868582126886e-06,
+      "loss": 0.262,
+      "step": 11768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8788977231247462,
+      "learning_rate": 5.0139736757672095e-06,
+      "loss": 0.3568,
+      "step": 11769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5589684828125013,
+      "learning_rate": 5.012360665987607e-06,
+      "loss": 0.3362,
+      "step": 11770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2517561774348396,
+      "learning_rate": 5.010747828929736e-06,
+      "loss": 0.2556,
+      "step": 11771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1552259575630777,
+      "learning_rate": 5.009135164649457e-06,
+      "loss": 0.2947,
+      "step": 11772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.35356777969622866,
+      "learning_rate": 5.007522673202613e-06,
+      "loss": 0.2245,
+      "step": 11773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33910329035869946,
+      "learning_rate": 5.005910354645043e-06,
+      "loss": 0.2669,
+      "step": 11774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3495175522409411,
+      "learning_rate": 5.0042982090325805e-06,
+      "loss": 0.3104,
+      "step": 11775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.622377722237779,
+      "learning_rate": 5.002686236421059e-06,
+      "loss": 0.2137,
+      "step": 11776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34094064564462573,
+      "learning_rate": 5.0010744368663e-06,
+      "loss": 0.2648,
+      "step": 11777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1026202721985212,
+      "learning_rate": 4.999462810424116e-06,
+      "loss": 0.4879,
+      "step": 11778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3465454930026479,
+      "learning_rate": 4.9978513571503175e-06,
+      "loss": 0.2476,
+      "step": 11779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.295672774198805,
+      "learning_rate": 4.996240077100713e-06,
+      "loss": 0.242,
+      "step": 11780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6049472071811616,
+      "learning_rate": 4.994628970331102e-06,
+      "loss": 0.3999,
+      "step": 11781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29679226403336495,
+      "learning_rate": 4.993018036897274e-06,
+      "loss": 0.1961,
+      "step": 11782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26024884929211345,
+      "learning_rate": 4.991407276855016e-06,
+      "loss": 0.2608,
+      "step": 11783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41637954941656796,
+      "learning_rate": 4.989796690260108e-06,
+      "loss": 0.2302,
+      "step": 11784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0573213732138746,
+      "learning_rate": 4.988186277168325e-06,
+      "loss": 0.3642,
+      "step": 11785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32321682893392756,
+      "learning_rate": 4.9865760376354365e-06,
+      "loss": 0.2565,
+      "step": 11786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3490489850595347,
+      "learning_rate": 4.9849659717172e-06,
+      "loss": 0.2924,
+      "step": 11787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4404451474161239,
+      "learning_rate": 4.98335607946938e-06,
+      "loss": 0.3545,
+      "step": 11788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23778656116027025,
+      "learning_rate": 4.981746360947724e-06,
+      "loss": 0.1595,
+      "step": 11789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3534561202280206,
+      "learning_rate": 4.980136816207974e-06,
+      "loss": 0.6599,
+      "step": 11790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3459827769144183,
+      "learning_rate": 4.978527445305869e-06,
+      "loss": 0.3084,
+      "step": 11791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2924247907379295,
+      "learning_rate": 4.976918248297145e-06,
+      "loss": 0.1902,
+      "step": 11792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.524691825002447,
+      "learning_rate": 4.9753092252375245e-06,
+      "loss": 0.396,
+      "step": 11793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36268677838139596,
+      "learning_rate": 4.973700376182732e-06,
+      "loss": 0.3425,
+      "step": 11794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2732204910654022,
+      "learning_rate": 4.972091701188478e-06,
+      "loss": 0.1814,
+      "step": 11795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2662254490724855,
+      "learning_rate": 4.970483200310468e-06,
+      "loss": 0.1567,
+      "step": 11796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6628178672031443,
+      "learning_rate": 4.968874873604414e-06,
+      "loss": 0.4117,
+      "step": 11797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31714829171919484,
+      "learning_rate": 4.967266721126005e-06,
+      "loss": 0.1911,
+      "step": 11798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3405285791476989,
+      "learning_rate": 4.965658742930934e-06,
+      "loss": 0.3089,
+      "step": 11799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0963642128915825,
+      "learning_rate": 4.964050939074881e-06,
+      "loss": 0.6414,
+      "step": 11800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2131071516902079,
+      "learning_rate": 4.962443309613529e-06,
+      "loss": 0.1658,
+      "step": 11801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3574377659432091,
+      "learning_rate": 4.96083585460255e-06,
+      "loss": 0.2421,
+      "step": 11802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4570436088730508,
+      "learning_rate": 4.95922857409761e-06,
+      "loss": 0.3248,
+      "step": 11803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3235139120968666,
+      "learning_rate": 4.9576214681543626e-06,
+      "loss": 0.2579,
+      "step": 11804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8548199042981579,
+      "learning_rate": 4.956014536828471e-06,
+      "loss": 0.3425,
+      "step": 11805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3368131841542381,
+      "learning_rate": 4.954407780175578e-06,
+      "loss": 0.3074,
+      "step": 11806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3103485250750534,
+      "learning_rate": 4.952801198251328e-06,
+      "loss": 0.2388,
+      "step": 11807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2644081194182972,
+      "learning_rate": 4.95119479111135e-06,
+      "loss": 0.091,
+      "step": 11808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.811557495156051,
+      "learning_rate": 4.949588558811285e-06,
+      "loss": 0.3806,
+      "step": 11809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4076560589755643,
+      "learning_rate": 4.947982501406749e-06,
+      "loss": 0.2817,
+      "step": 11810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29270911302935926,
+      "learning_rate": 4.946376618953364e-06,
+      "loss": 0.2591,
+      "step": 11811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9784414407077597,
+      "learning_rate": 4.944770911506739e-06,
+      "loss": 0.536,
+      "step": 11812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3713203487743198,
+      "learning_rate": 4.9431653791224744e-06,
+      "loss": 0.2449,
+      "step": 11813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26768673340053445,
+      "learning_rate": 4.941560021856181e-06,
+      "loss": 0.2029,
+      "step": 11814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4008652783655156,
+      "learning_rate": 4.9399548397634455e-06,
+      "loss": 0.2715,
+      "step": 11815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3935508821109515,
+      "learning_rate": 4.938349832899856e-06,
+      "loss": 0.2394,
+      "step": 11816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49750616813772636,
+      "learning_rate": 4.9367450013209905e-06,
+      "loss": 0.3878,
+      "step": 11817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3181744075806371,
+      "learning_rate": 4.935140345082436e-06,
+      "loss": 0.2385,
+      "step": 11818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42534101653733136,
+      "learning_rate": 4.93353586423975e-06,
+      "loss": 0.3122,
+      "step": 11819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27271478591220527,
+      "learning_rate": 4.9319315588484954e-06,
+      "loss": 0.2234,
+      "step": 11820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6360729424545637,
+      "learning_rate": 4.930327428964235e-06,
+      "loss": 0.3097,
+      "step": 11821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32253078555839093,
+      "learning_rate": 4.9287234746425195e-06,
+      "loss": 0.2473,
+      "step": 11822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34936474152769675,
+      "learning_rate": 4.927119695938891e-06,
+      "loss": 0.3168,
+      "step": 11823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9379810581274167,
+      "learning_rate": 4.925516092908891e-06,
+      "loss": 0.3485,
+      "step": 11824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31432849488872344,
+      "learning_rate": 4.923912665608045e-06,
+      "loss": 0.2603,
+      "step": 11825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4927456068221821,
+      "learning_rate": 4.9223094140918894e-06,
+      "loss": 0.2566,
+      "step": 11826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31740940794607303,
+      "learning_rate": 4.920706338415941e-06,
+      "loss": 0.2725,
+      "step": 11827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.35754575878697037,
+      "learning_rate": 4.919103438635713e-06,
+      "loss": 0.1741,
+      "step": 11828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4985427045345408,
+      "learning_rate": 4.91750071480671e-06,
+      "loss": 0.3825,
+      "step": 11829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33708352874029796,
+      "learning_rate": 4.915898166984443e-06,
+      "loss": 0.2834,
+      "step": 11830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5475661290159132,
+      "learning_rate": 4.914295795224404e-06,
+      "loss": 0.1531,
+      "step": 11831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38665851593133643,
+      "learning_rate": 4.912693599582083e-06,
+      "loss": 0.3186,
+      "step": 11832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2821804515532111,
+      "learning_rate": 4.91109158011296e-06,
+      "loss": 0.191,
+      "step": 11833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2936069143811596,
+      "learning_rate": 4.909489736872521e-06,
+      "loss": 0.1984,
+      "step": 11834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5737718239366725,
+      "learning_rate": 4.907888069916234e-06,
+      "loss": 0.317,
+      "step": 11835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2137512039405558,
+      "learning_rate": 4.906286579299563e-06,
+      "loss": 0.6646,
+      "step": 11836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36728729290760725,
+      "learning_rate": 4.904685265077969e-06,
+      "loss": 0.19,
+      "step": 11837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30145925210114205,
+      "learning_rate": 4.903084127306901e-06,
+      "loss": 0.2932,
+      "step": 11838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44813104620717403,
+      "learning_rate": 4.901483166041815e-06,
+      "loss": 0.272,
+      "step": 11839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3188875124250454,
+      "learning_rate": 4.899882381338147e-06,
+      "loss": 0.2182,
+      "step": 11840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3501168527999765,
+      "learning_rate": 4.898281773251333e-06,
+      "loss": 0.254,
+      "step": 11841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3725052385902562,
+      "learning_rate": 4.896681341836798e-06,
+      "loss": 0.3149,
+      "step": 11842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3919431187346392,
+      "learning_rate": 4.895081087149974e-06,
+      "loss": 0.2764,
+      "step": 11843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.447679970754077,
+      "learning_rate": 4.8934810092462705e-06,
+      "loss": 0.283,
+      "step": 11844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27528776173796865,
+      "learning_rate": 4.891881108181101e-06,
+      "loss": 0.1694,
+      "step": 11845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2515033888633789,
+      "learning_rate": 4.890281384009865e-06,
+      "loss": 0.2506,
+      "step": 11846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4647181444523978,
+      "learning_rate": 4.8886818367879686e-06,
+      "loss": 0.2796,
+      "step": 11847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6205120870967442,
+      "learning_rate": 4.8870824665708e-06,
+      "loss": 0.4248,
+      "step": 11848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6572729053267969,
+      "learning_rate": 4.885483273413747e-06,
+      "loss": 0.3524,
+      "step": 11849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26634984300357595,
+      "learning_rate": 4.883884257372188e-06,
+      "loss": 0.2635,
+      "step": 11850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21217690977582984,
+      "learning_rate": 4.882285418501497e-06,
+      "loss": 0.1465,
+      "step": 11851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.823238938356525,
+      "learning_rate": 4.88068675685704e-06,
+      "loss": 0.5252,
+      "step": 11852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.346625383613797,
+      "learning_rate": 4.879088272494184e-06,
+      "loss": 0.2784,
+      "step": 11853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3474056837835713,
+      "learning_rate": 4.877489965468274e-06,
+      "loss": 0.2554,
+      "step": 11854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6079999000177458,
+      "learning_rate": 4.875891835834672e-06,
+      "loss": 0.3113,
+      "step": 11855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37799226612572256,
+      "learning_rate": 4.874293883648714e-06,
+      "loss": 0.3038,
+      "step": 11856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.43480049777111607,
+      "learning_rate": 4.8726961089657385e-06,
+      "loss": 0.1556,
+      "step": 11857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.255060595881029,
+      "learning_rate": 4.871098511841073e-06,
+      "loss": 0.2094,
+      "step": 11858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37794696995047256,
+      "learning_rate": 4.8695010923300505e-06,
+      "loss": 0.2933,
+      "step": 11859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.703634639952293,
+      "learning_rate": 4.867903850487983e-06,
+      "loss": 0.3383,
+      "step": 11860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5711648018980142,
+      "learning_rate": 4.866306786370184e-06,
+      "loss": 0.3568,
+      "step": 11861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3153893572774383,
+      "learning_rate": 4.864709900031961e-06,
+      "loss": 0.3094,
+      "step": 11862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.46738359154640297,
+      "learning_rate": 4.86311319152861e-06,
+      "loss": 0.3345,
+      "step": 11863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.174118788912665,
+      "learning_rate": 4.8615166609154315e-06,
+      "loss": 0.1225,
+      "step": 11864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5155715979717324,
+      "learning_rate": 4.85992030824771e-06,
+      "loss": 0.3517,
+      "step": 11865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39699707131238365,
+      "learning_rate": 4.858324133580727e-06,
+      "loss": 0.3108,
+      "step": 11866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.723771118865917,
+      "learning_rate": 4.856728136969755e-06,
+      "loss": 0.265,
+      "step": 11867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34812703384012267,
+      "learning_rate": 4.85513231847007e-06,
+      "loss": 0.258,
+      "step": 11868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.46778392750169046,
+      "learning_rate": 4.853536678136932e-06,
+      "loss": 0.3343,
+      "step": 11869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26524386390299387,
+      "learning_rate": 4.851941216025597e-06,
+      "loss": 0.1666,
+      "step": 11870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3529335727394295,
+      "learning_rate": 4.850345932191313e-06,
+      "loss": 0.2873,
+      "step": 11871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6736113996854561,
+      "learning_rate": 4.848750826689332e-06,
+      "loss": 0.373,
+      "step": 11872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2454207027025233,
+      "learning_rate": 4.8471558995748865e-06,
+      "loss": 0.1576,
+      "step": 11873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29709806190197335,
+      "learning_rate": 4.845561150903212e-06,
+      "loss": 0.2612,
+      "step": 11874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1809594524008666,
+      "learning_rate": 4.843966580729533e-06,
+      "loss": 0.6127,
+      "step": 11875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6862481533740115,
+      "learning_rate": 4.842372189109066e-06,
+      "loss": 0.4077,
+      "step": 11876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6072253682769295,
+      "learning_rate": 4.840777976097032e-06,
+      "loss": 0.2227,
+      "step": 11877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.35301852333241995,
+      "learning_rate": 4.839183941748635e-06,
+      "loss": 0.3078,
+      "step": 11878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2620722298074846,
+      "learning_rate": 4.837590086119076e-06,
+      "loss": 0.1697,
+      "step": 11879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32669040181183934,
+      "learning_rate": 4.835996409263546e-06,
+      "loss": 0.2134,
+      "step": 11880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.089646351240924,
+      "learning_rate": 4.834402911237243e-06,
+      "loss": 0.4845,
+      "step": 11881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29818673852673244,
+      "learning_rate": 4.832809592095344e-06,
+      "loss": 0.2822,
+      "step": 11882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.329051440364525,
+      "learning_rate": 4.831216451893027e-06,
+      "loss": 0.2178,
+      "step": 11883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6968295293375228,
+      "learning_rate": 4.829623490685459e-06,
+      "loss": 0.4373,
+      "step": 11884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23540721441096973,
+      "learning_rate": 4.828030708527814e-06,
+      "loss": 0.1897,
+      "step": 11885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34736558596523004,
+      "learning_rate": 4.826438105475239e-06,
+      "loss": 0.2225,
+      "step": 11886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0944632104089678,
+      "learning_rate": 4.824845681582892e-06,
+      "loss": 0.5581,
+      "step": 11887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8324240460651077,
+      "learning_rate": 4.82325343690591e-06,
+      "loss": 0.414,
+      "step": 11888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32586291593223726,
+      "learning_rate": 4.821661371499444e-06,
+      "loss": 0.2691,
+      "step": 11889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33201111450208354,
+      "learning_rate": 4.820069485418622e-06,
+      "loss": 0.259,
+      "step": 11890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2894218031621524,
+      "learning_rate": 4.818477778718571e-06,
+      "loss": 0.1965,
+      "step": 11891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34646058218222653,
+      "learning_rate": 4.8168862514544075e-06,
+      "loss": 0.2612,
+      "step": 11892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9011059324528964,
+      "learning_rate": 4.815294903681254e-06,
+      "loss": 0.0465,
+      "step": 11893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38189133211873966,
+      "learning_rate": 4.813703735454216e-06,
+      "loss": 0.2818,
+      "step": 11894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37774095648872624,
+      "learning_rate": 4.812112746828394e-06,
+      "loss": 0.2809,
+      "step": 11895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9196950014411767,
+      "learning_rate": 4.810521937858881e-06,
+      "loss": 0.5367,
+      "step": 11896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25821957577363874,
+      "learning_rate": 4.808931308600774e-06,
+      "loss": 0.2529,
+      "step": 11897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33995182083802594,
+      "learning_rate": 4.807340859109152e-06,
+      "loss": 0.2719,
+      "step": 11898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.408344759125669,
+      "learning_rate": 4.805750589439092e-06,
+      "loss": 0.1797,
+      "step": 11899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.577276240726866,
+      "learning_rate": 4.804160499645667e-06,
+      "loss": 0.3866,
+      "step": 11900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33678995159423364,
+      "learning_rate": 4.802570589783937e-06,
+      "loss": 0.2425,
+      "step": 11901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3749231267150535,
+      "learning_rate": 4.800980859908967e-06,
+      "loss": 0.3353,
+      "step": 11902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39713015624552944,
+      "learning_rate": 4.799391310075806e-06,
+      "loss": 0.1311,
+      "step": 11903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2695977078762357,
+      "learning_rate": 4.7978019403395e-06,
+      "loss": 0.2272,
+      "step": 11904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5527853027413822,
+      "learning_rate": 4.796212750755087e-06,
+      "loss": 0.3386,
+      "step": 11905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5264787695999982,
+      "learning_rate": 4.794623741377605e-06,
+      "loss": 0.2782,
+      "step": 11906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3201557404278092,
+      "learning_rate": 4.79303491226208e-06,
+      "loss": 0.2416,
+      "step": 11907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1192785579323365,
+      "learning_rate": 4.791446263463531e-06,
+      "loss": 0.7994,
+      "step": 11908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26545596616417333,
+      "learning_rate": 4.7898577950369704e-06,
+      "loss": 0.2056,
+      "step": 11909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24110527748871047,
+      "learning_rate": 4.788269507037415e-06,
+      "loss": 0.1904,
+      "step": 11910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8106310137327224,
+      "learning_rate": 4.786681399519862e-06,
+      "loss": 0.416,
+      "step": 11911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7495867803826378,
+      "learning_rate": 4.785093472539307e-06,
+      "loss": 0.3306,
+      "step": 11912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2882924469905718,
+      "learning_rate": 4.783505726150738e-06,
+      "loss": 0.2652,
+      "step": 11913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4408666504381776,
+      "learning_rate": 4.781918160409145e-06,
+      "loss": 0.3201,
+      "step": 11914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4572909264973247,
+      "learning_rate": 4.780330775369501e-06,
+      "loss": 0.2594,
+      "step": 11915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2702213205694998,
+      "learning_rate": 4.778743571086779e-06,
+      "loss": 0.1873,
+      "step": 11916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390198271391285,
+      "learning_rate": 4.777156547615942e-06,
+      "loss": 0.2586,
+      "step": 11917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5337637998050275,
+      "learning_rate": 4.775569705011945e-06,
+      "loss": 0.361,
+      "step": 11918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30304924911307957,
+      "learning_rate": 4.773983043329753e-06,
+      "loss": 0.1909,
+      "step": 11919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1874428884612134,
+      "learning_rate": 4.7723965626243e-06,
+      "loss": 0.7611,
+      "step": 11920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33172829141311116,
+      "learning_rate": 4.770810262950524e-06,
+      "loss": 0.2931,
+      "step": 11921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31447195769036773,
+      "learning_rate": 4.769224144363368e-06,
+      "loss": 0.2002,
+      "step": 11922
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49881291519374765,
+      "learning_rate": 4.767638206917755e-06,
+      "loss": 0.2958,
+      "step": 11923
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3918596183860913,
+      "learning_rate": 4.766052450668606e-06,
+      "loss": 0.2039,
+      "step": 11924
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27808781938656596,
+      "learning_rate": 4.764466875670836e-06,
+      "loss": 0.2249,
+      "step": 11925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4956137686581436,
+      "learning_rate": 4.762881481979349e-06,
+      "loss": 0.3887,
+      "step": 11926
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.111755849842336,
+      "learning_rate": 4.761296269649054e-06,
+      "loss": 0.4972,
+      "step": 11927
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30684184917618834,
+      "learning_rate": 4.759711238734844e-06,
+      "loss": 0.2521,
+      "step": 11928
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4735692273389909,
+      "learning_rate": 4.75812638929161e-06,
+      "loss": 0.2749,
+      "step": 11929
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2613868405056514,
+      "learning_rate": 4.756541721374228e-06,
+      "loss": 0.2068,
+      "step": 11930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33203601499476443,
+      "learning_rate": 4.7549572350375864e-06,
+      "loss": 0.248,
+      "step": 11931
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9132904905924238,
+      "learning_rate": 4.753372930336548e-06,
+      "loss": 0.4699,
+      "step": 11932
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.32872137130223583,
+      "learning_rate": 4.751788807325981e-06,
+      "loss": 0.2959,
+      "step": 11933
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.396983128779614,
+      "learning_rate": 4.750204866060738e-06,
+      "loss": 0.2726,
+      "step": 11934
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5185433454178232,
+      "learning_rate": 4.748621106595679e-06,
+      "loss": 0.2699,
+      "step": 11935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2455213486640146,
+      "learning_rate": 4.747037528985644e-06,
+      "loss": 0.1981,
+      "step": 11936
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4123647936906814,
+      "learning_rate": 4.745454133285474e-06,
+      "loss": 0.2777,
+      "step": 11937
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3633484612973121,
+      "learning_rate": 4.743870919549998e-06,
+      "loss": 0.3307,
+      "step": 11938
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6143787094869488,
+      "learning_rate": 4.74228788783405e-06,
+      "loss": 0.3097,
+      "step": 11939
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3848915895137807,
+      "learning_rate": 4.740705038192444e-06,
+      "loss": 0.2994,
+      "step": 11940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.303569188063324,
+      "learning_rate": 4.7391223706799994e-06,
+      "loss": 0.2996,
+      "step": 11941
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22025983968170426,
+      "learning_rate": 4.73753988535152e-06,
+      "loss": 0.0816,
+      "step": 11942
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3947879607288585,
+      "learning_rate": 4.735957582261803e-06,
+      "loss": 0.2905,
+      "step": 11943
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5295208092043296,
+      "learning_rate": 4.7343754614656536e-06,
+      "loss": 0.3748,
+      "step": 11944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3434557916596737,
+      "learning_rate": 4.732793523017856e-06,
+      "loss": 0.2484,
+      "step": 11945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4718786351336028,
+      "learning_rate": 4.73121176697319e-06,
+      "loss": 0.3152,
+      "step": 11946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5218283957356089,
+      "learning_rate": 4.729630193386433e-06,
+      "loss": 0.3238,
+      "step": 11947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.158404783519397,
+      "learning_rate": 4.728048802312358e-06,
+      "loss": 0.0912,
+      "step": 11948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27999339984469956,
+      "learning_rate": 4.726467593805726e-06,
+      "loss": 0.2589,
+      "step": 11949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5390317862693036,
+      "learning_rate": 4.724886567921295e-06,
+      "loss": 0.4137,
+      "step": 11950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5653285205882527,
+      "learning_rate": 4.723305724713812e-06,
+      "loss": 0.3685,
+      "step": 11951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4366394060085917,
+      "learning_rate": 4.721725064238028e-06,
+      "loss": 0.1037,
+      "step": 11952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28421369940574165,
+      "learning_rate": 4.720144586548681e-06,
+      "loss": 0.2764,
+      "step": 11953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3944676448912269,
+      "learning_rate": 4.718564291700497e-06,
+      "loss": 0.2295,
+      "step": 11954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34010771117451527,
+      "learning_rate": 4.7169841797482005e-06,
+      "loss": 0.1541,
+      "step": 11955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3757324807553204,
+      "learning_rate": 4.7154042507465195e-06,
+      "loss": 0.2925,
+      "step": 11956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2737937072179737,
+      "learning_rate": 4.713824504750161e-06,
+      "loss": 0.2567,
+      "step": 11957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33085688299480903,
+      "learning_rate": 4.7122449418138325e-06,
+      "loss": 0.1568,
+      "step": 11958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5549992341677822,
+      "learning_rate": 4.710665561992232e-06,
+      "loss": 0.3311,
+      "step": 11959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.43458610126481007,
+      "learning_rate": 4.709086365340057e-06,
+      "loss": 0.2455,
+      "step": 11960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24458519823897504,
+      "learning_rate": 4.707507351911995e-06,
+      "loss": 0.2202,
+      "step": 11961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4809860354275576,
+      "learning_rate": 4.705928521762726e-06,
+      "loss": 0.3509,
+      "step": 11962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6523562611627876,
+      "learning_rate": 4.7043498749469204e-06,
+      "loss": 0.2662,
+      "step": 11963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6412526764903204,
+      "learning_rate": 4.702771411519256e-06,
+      "loss": 0.3287,
+      "step": 11964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2694814585384461,
+      "learning_rate": 4.701193131534389e-06,
+      "loss": 0.2403,
+      "step": 11965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9540802506457705,
+      "learning_rate": 4.699615035046975e-06,
+      "loss": 0.5154,
+      "step": 11966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3832799734593423,
+      "learning_rate": 4.698037122111665e-06,
+      "loss": 0.3095,
+      "step": 11967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5151002559842178,
+      "learning_rate": 4.696459392783098e-06,
+      "loss": 0.2608,
+      "step": 11968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2970906609548814,
+      "learning_rate": 4.694881847115918e-06,
+      "loss": 0.2769,
+      "step": 11969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3673952699013813,
+      "learning_rate": 4.69330448516475e-06,
+      "loss": 0.2208,
+      "step": 11970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3823198217572113,
+      "learning_rate": 4.691727306984222e-06,
+      "loss": 0.2333,
+      "step": 11971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4797977748585798,
+      "learning_rate": 4.690150312628944e-06,
+      "loss": 0.3882,
+      "step": 11972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42047099420085404,
+      "learning_rate": 4.688573502153536e-06,
+      "loss": 0.3205,
+      "step": 11973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35808423918623256,
+      "learning_rate": 4.6869968756126e-06,
+      "loss": 0.2417,
+      "step": 11974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2642813721183281,
+      "learning_rate": 4.685420433060732e-06,
+      "loss": 0.1715,
+      "step": 11975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42438116021427397,
+      "learning_rate": 4.683844174552523e-06,
+      "loss": 0.2682,
+      "step": 11976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2952126454257907,
+      "learning_rate": 4.682268100142567e-06,
+      "loss": 0.2697,
+      "step": 11977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3042987725864317,
+      "learning_rate": 4.680692209885436e-06,
+      "loss": 0.3308,
+      "step": 11978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50954991915773,
+      "learning_rate": 4.679116503835706e-06,
+      "loss": 0.3273,
+      "step": 11979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35979928673775013,
+      "learning_rate": 4.6775409820479415e-06,
+      "loss": 0.311,
+      "step": 11980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34865340009068274,
+      "learning_rate": 4.675965644576701e-06,
+      "loss": 0.2457,
+      "step": 11981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2343614065070297,
+      "learning_rate": 4.674390491476545e-06,
+      "loss": 0.161,
+      "step": 11982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.39514869929001734,
+      "learning_rate": 4.672815522802018e-06,
+      "loss": 0.3022,
+      "step": 11983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42193017601516436,
+      "learning_rate": 4.671240738607659e-06,
+      "loss": 0.2649,
+      "step": 11984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3575772895676162,
+      "learning_rate": 4.669666138948001e-06,
+      "loss": 0.3064,
+      "step": 11985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37733439670751967,
+      "learning_rate": 4.668091723877584e-06,
+      "loss": 0.2944,
+      "step": 11986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.363482865984895,
+      "learning_rate": 4.666517493450916e-06,
+      "loss": 0.1889,
+      "step": 11987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2920184473283872,
+      "learning_rate": 4.664943447722514e-06,
+      "loss": 0.1943,
+      "step": 11988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.268324034496146,
+      "learning_rate": 4.6633695867468955e-06,
+      "loss": 0.2651,
+      "step": 11989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6601723005250368,
+      "learning_rate": 4.661795910578558e-06,
+      "loss": 0.382,
+      "step": 11990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6064327864030062,
+      "learning_rate": 4.660222419271999e-06,
+      "loss": 0.2746,
+      "step": 11991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3675959255251272,
+      "learning_rate": 4.658649112881709e-06,
+      "loss": 0.3073,
+      "step": 11992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.36307823733203226,
+      "learning_rate": 4.657075991462165e-06,
+      "loss": 0.3006,
+      "step": 11993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2183623643397165,
+      "learning_rate": 4.6555030550678544e-06,
+      "loss": 0.0842,
+      "step": 11994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35620400135270036,
+      "learning_rate": 4.6539303037532435e-06,
+      "loss": 0.2911,
+      "step": 11995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9075387144091024,
+      "learning_rate": 4.652357737572796e-06,
+      "loss": 0.392,
+      "step": 11996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30833975615137077,
+      "learning_rate": 4.650785356580967e-06,
+      "loss": 0.2674,
+      "step": 11997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.36252082852686,
+      "learning_rate": 4.649213160832213e-06,
+      "loss": 0.3089,
+      "step": 11998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0088006244221261,
+      "learning_rate": 4.647641150380978e-06,
+      "loss": 0.562,
+      "step": 11999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17551431395096048,
+      "learning_rate": 4.6460693252817e-06,
+      "loss": 0.1311,
+      "step": 12000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.32796221667677156,
+      "learning_rate": 4.644497685588808e-06,
+      "loss": 0.2919,
+      "step": 12001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.211749796355846,
+      "learning_rate": 4.642926231356734e-06,
+      "loss": 0.3921,
+      "step": 12002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6840428587949849,
+      "learning_rate": 4.641354962639894e-06,
+      "loss": 0.4082,
+      "step": 12003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3276160490754549,
+      "learning_rate": 4.639783879492701e-06,
+      "loss": 0.2112,
+      "step": 12004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3450703208904191,
+      "learning_rate": 4.638212981969562e-06,
+      "loss": 0.2976,
+      "step": 12005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22299481634481944,
+      "learning_rate": 4.636642270124874e-06,
+      "loss": 0.1401,
+      "step": 12006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4736605076802758,
+      "learning_rate": 4.6350717440130366e-06,
+      "loss": 0.2279,
+      "step": 12007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6337811756742268,
+      "learning_rate": 4.633501403688434e-06,
+      "loss": 0.3114,
+      "step": 12008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0475717388652215,
+      "learning_rate": 4.631931249205447e-06,
+      "loss": 0.5844,
+      "step": 12009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29755432704987433,
+      "learning_rate": 4.630361280618446e-06,
+      "loss": 0.2081,
+      "step": 12010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1889278166919943,
+      "learning_rate": 4.628791497981807e-06,
+      "loss": 0.6442,
+      "step": 12011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2764738843094473,
+      "learning_rate": 4.627221901349887e-06,
+      "loss": 0.2578,
+      "step": 12012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24776928880783564,
+      "learning_rate": 4.625652490777042e-06,
+      "loss": 0.1999,
+      "step": 12013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7327164556440673,
+      "learning_rate": 4.624083266317616e-06,
+      "loss": 0.3085,
+      "step": 12014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1927932070668756,
+      "learning_rate": 4.62251422802596e-06,
+      "loss": 0.7523,
+      "step": 12015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3228198567523802,
+      "learning_rate": 4.620945375956404e-06,
+      "loss": 0.2437,
+      "step": 12016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5191991671756184,
+      "learning_rate": 4.619376710163279e-06,
+      "loss": 0.2589,
+      "step": 12017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7583758410617182,
+      "learning_rate": 4.617808230700907e-06,
+      "loss": 0.42,
+      "step": 12018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33941374151587894,
+      "learning_rate": 4.6162399376236e-06,
+      "loss": 0.2747,
+      "step": 12019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2758703659931883,
+      "learning_rate": 4.614671830985681e-06,
+      "loss": 0.2094,
+      "step": 12020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.354847088036182,
+      "learning_rate": 4.613103910841441e-06,
+      "loss": 0.21,
+      "step": 12021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3544508126717817,
+      "learning_rate": 4.611536177245176e-06,
+      "loss": 0.2592,
+      "step": 12022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0195307111162017,
+      "learning_rate": 4.609968630251187e-06,
+      "loss": 0.3661,
+      "step": 12023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3860554430872647,
+      "learning_rate": 4.608401269913751e-06,
+      "loss": 0.2795,
+      "step": 12024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.39062012651266825,
+      "learning_rate": 4.606834096287148e-06,
+      "loss": 0.2738,
+      "step": 12025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3576834430909374,
+      "learning_rate": 4.605267109425645e-06,
+      "loss": 0.2074,
+      "step": 12026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.050741505415178,
+      "learning_rate": 4.6037003093835135e-06,
+      "loss": 0.4651,
+      "step": 12027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3340140311716859,
+      "learning_rate": 4.602133696215007e-06,
+      "loss": 0.2685,
+      "step": 12028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3868998595979161,
+      "learning_rate": 4.6005672699743795e-06,
+      "loss": 0.3129,
+      "step": 12029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6958039333913612,
+      "learning_rate": 4.599001030715876e-06,
+      "loss": 0.2917,
+      "step": 12030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3372161814167908,
+      "learning_rate": 4.59743497849373e-06,
+      "loss": 0.274,
+      "step": 12031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3879938736639436,
+      "learning_rate": 4.5958691133621815e-06,
+      "loss": 0.3127,
+      "step": 12032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.14895293830728926,
+      "learning_rate": 4.594303435375454e-06,
+      "loss": 0.1142,
+      "step": 12033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.322727915954783,
+      "learning_rate": 4.592737944587766e-06,
+      "loss": 0.2661,
+      "step": 12034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1372554738979654,
+      "learning_rate": 4.591172641053326e-06,
+      "loss": 0.5845,
+      "step": 12035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42201688444677404,
+      "learning_rate": 4.589607524826351e-06,
+      "loss": 0.2639,
+      "step": 12036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3165929336205737,
+      "learning_rate": 4.588042595961032e-06,
+      "loss": 0.2747,
+      "step": 12037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40563172992912694,
+      "learning_rate": 4.586477854511566e-06,
+      "loss": 0.2555,
+      "step": 12038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3947859210351605,
+      "learning_rate": 4.584913300532135e-06,
+      "loss": 0.2764,
+      "step": 12039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28751873344914913,
+      "learning_rate": 4.583348934076929e-06,
+      "loss": 0.1917,
+      "step": 12040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3649164575475577,
+      "learning_rate": 4.581784755200115e-06,
+      "loss": 0.2989,
+      "step": 12041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7260664375548328,
+      "learning_rate": 4.580220763955863e-06,
+      "loss": 0.3636,
+      "step": 12042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29090161691547395,
+      "learning_rate": 4.578656960398328e-06,
+      "loss": 0.1897,
+      "step": 12043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3837560896918546,
+      "learning_rate": 4.577093344581674e-06,
+      "loss": 0.316,
+      "step": 12044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3944502241833138,
+      "learning_rate": 4.575529916560043e-06,
+      "loss": 0.2643,
+      "step": 12045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28381455429202385,
+      "learning_rate": 4.573966676387579e-06,
+      "loss": 0.2047,
+      "step": 12046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5857048817229604,
+      "learning_rate": 4.5724036241184144e-06,
+      "loss": 0.2819,
+      "step": 12047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3373748374080497,
+      "learning_rate": 4.5708407598066766e-06,
+      "loss": 0.2939,
+      "step": 12048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.285983173917611,
+      "learning_rate": 4.569278083506492e-06,
+      "loss": 0.1824,
+      "step": 12049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1284028289918933,
+      "learning_rate": 4.567715595271976e-06,
+      "loss": 0.5352,
+      "step": 12050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1099867271284871,
+      "learning_rate": 4.566153295157233e-06,
+      "loss": 0.7988,
+      "step": 12051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2915399836846607,
+      "learning_rate": 4.5645911832163654e-06,
+      "loss": 0.2437,
+      "step": 12052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35039432313772084,
+      "learning_rate": 4.563029259503474e-06,
+      "loss": 0.2095,
+      "step": 12053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.41625909045888376,
+      "learning_rate": 4.561467524072651e-06,
+      "loss": 0.2736,
+      "step": 12054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.308228128875875,
+      "learning_rate": 4.5599059769779654e-06,
+      "loss": 0.2516,
+      "step": 12055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34633389816855337,
+      "learning_rate": 4.558344618273506e-06,
+      "loss": 0.2532,
+      "step": 12056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1119503132303514,
+      "learning_rate": 4.556783448013338e-06,
+      "loss": 0.7745,
+      "step": 12057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.505912833656344,
+      "learning_rate": 4.555222466251525e-06,
+      "loss": 0.3091,
+      "step": 12058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30427640882311047,
+      "learning_rate": 4.553661673042123e-06,
+      "loss": 0.2245,
+      "step": 12059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2822146065378598,
+      "learning_rate": 4.552101068439181e-06,
+      "loss": 0.2482,
+      "step": 12060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3780466110606745,
+      "learning_rate": 4.550540652496748e-06,
+      "loss": 0.2474,
+      "step": 12061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5094716111081256,
+      "learning_rate": 4.548980425268857e-06,
+      "loss": 0.2451,
+      "step": 12062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2888871968821543,
+      "learning_rate": 4.5474203868095415e-06,
+      "loss": 0.4701,
+      "step": 12063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2587492497087311,
+      "learning_rate": 4.545860537172818e-06,
+      "loss": 0.2568,
+      "step": 12064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.46174224389574064,
+      "learning_rate": 4.5443008764127135e-06,
+      "loss": 0.3415,
+      "step": 12065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24079285479608747,
+      "learning_rate": 4.542741404583235e-06,
+      "loss": 0.1051,
+      "step": 12066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42779305694074776,
+      "learning_rate": 4.541182121738388e-06,
+      "loss": 0.2387,
+      "step": 12067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28243846220049185,
+      "learning_rate": 4.539623027932165e-06,
+      "loss": 0.2727,
+      "step": 12068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7645806896247914,
+      "learning_rate": 4.538064123218565e-06,
+      "loss": 0.34,
+      "step": 12069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34980810066617957,
+      "learning_rate": 4.53650540765157e-06,
+      "loss": 0.2661,
+      "step": 12070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5862188882028873,
+      "learning_rate": 4.534946881285158e-06,
+      "loss": 0.3972,
+      "step": 12071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21256587958188933,
+      "learning_rate": 4.533388544173301e-06,
+      "loss": 0.169,
+      "step": 12072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3133687014658715,
+      "learning_rate": 4.531830396369959e-06,
+      "loss": 0.2401,
+      "step": 12073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1903642275444457,
+      "learning_rate": 4.530272437929099e-06,
+      "loss": 0.6457,
+      "step": 12074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3660557403373204,
+      "learning_rate": 4.528714668904669e-06,
+      "loss": 0.2687,
+      "step": 12075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5739283505931541,
+      "learning_rate": 4.527157089350616e-06,
+      "loss": 0.3292,
+      "step": 12076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4039211294031608,
+      "learning_rate": 4.525599699320873e-06,
+      "loss": 0.3117,
+      "step": 12077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30685207485541593,
+      "learning_rate": 4.52404249886938e-06,
+      "loss": 0.1924,
+      "step": 12078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.287329467106963,
+      "learning_rate": 4.5224854880500615e-06,
+      "loss": 0.0706,
+      "step": 12079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27929117817412435,
+      "learning_rate": 4.520928666916834e-06,
+      "loss": 0.2665,
+      "step": 12080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7506304080374441,
+      "learning_rate": 4.519372035523607e-06,
+      "loss": 0.3991,
+      "step": 12081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4259104980903786,
+      "learning_rate": 4.517815593924295e-06,
+      "loss": 0.2043,
+      "step": 12082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3686775706552464,
+      "learning_rate": 4.5162593421727926e-06,
+      "loss": 0.3217,
+      "step": 12083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4937075304767455,
+      "learning_rate": 4.514703280322995e-06,
+      "loss": 0.3287,
+      "step": 12084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2030857822006487,
+      "learning_rate": 4.513147408428786e-06,
+      "loss": 0.1249,
+      "step": 12085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5124655651149372,
+      "learning_rate": 4.5115917265440425e-06,
+      "loss": 0.3465,
+      "step": 12086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4821076032344685,
+      "learning_rate": 4.510036234722645e-06,
+      "loss": 0.3284,
+      "step": 12087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2927910612808501,
+      "learning_rate": 4.5084809330184605e-06,
+      "loss": 0.2206,
+      "step": 12088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5136650655164572,
+      "learning_rate": 4.506925821485338e-06,
+      "loss": 0.3624,
+      "step": 12089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2936563962334066,
+      "learning_rate": 4.505370900177142e-06,
+      "loss": 0.1928,
+      "step": 12090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3934100319552002,
+      "learning_rate": 4.503816169147715e-06,
+      "loss": 0.2727,
+      "step": 12091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28160661063648673,
+      "learning_rate": 4.502261628450898e-06,
+      "loss": 0.2303,
+      "step": 12092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5788519867248023,
+      "learning_rate": 4.5007072781405205e-06,
+      "loss": 0.3512,
+      "step": 12093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6096906100436412,
+      "learning_rate": 4.4991531182704166e-06,
+      "loss": 0.3835,
+      "step": 12094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28201262052309894,
+      "learning_rate": 4.497599148894404e-06,
+      "loss": 0.2467,
+      "step": 12095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33592745821692854,
+      "learning_rate": 4.496045370066296e-06,
+      "loss": 0.2624,
+      "step": 12096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3753798828666515,
+      "learning_rate": 4.494491781839901e-06,
+      "loss": 0.2396,
+      "step": 12097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3387213970720124,
+      "learning_rate": 4.492938384269015e-06,
+      "loss": 0.249,
+      "step": 12098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33277551615459394,
+      "learning_rate": 4.491385177407439e-06,
+      "loss": 0.293,
+      "step": 12099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7391401050418194,
+      "learning_rate": 4.489832161308958e-06,
+      "loss": 0.4181,
+      "step": 12100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3603244564137188,
+      "learning_rate": 4.488279336027353e-06,
+      "loss": 0.3054,
+      "step": 12101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3353997447238762,
+      "learning_rate": 4.486726701616393e-06,
+      "loss": 0.0981,
+      "step": 12102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2523472651585536,
+      "learning_rate": 4.485174258129854e-06,
+      "loss": 0.2128,
+      "step": 12103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29579376641255173,
+      "learning_rate": 4.483622005621493e-06,
+      "loss": 0.2761,
+      "step": 12104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.67653447878386,
+      "learning_rate": 4.4820699441450655e-06,
+      "loss": 0.3091,
+      "step": 12105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6401876821418024,
+      "learning_rate": 4.4805180737543145e-06,
+      "loss": 0.4376,
+      "step": 12106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3921815615134285,
+      "learning_rate": 4.478966394502988e-06,
+      "loss": 0.3295,
+      "step": 12107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3475348263674097,
+      "learning_rate": 4.4774149064448195e-06,
+      "loss": 0.2545,
+      "step": 12108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36534583192677034,
+      "learning_rate": 4.475863609633534e-06,
+      "loss": 0.2564,
+      "step": 12109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3668623445667178,
+      "learning_rate": 4.474312504122854e-06,
+      "loss": 0.2889,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25610872600300666,
+      "learning_rate": 4.472761589966493e-06,
+      "loss": 0.2137,
+      "step": 12111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.561497219037444,
+      "learning_rate": 4.471210867218161e-06,
+      "loss": 0.3434,
+      "step": 12112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37026942810244867,
+      "learning_rate": 4.4696603359315604e-06,
+      "loss": 0.3116,
+      "step": 12113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7497924302268585,
+      "learning_rate": 4.468109996160385e-06,
+      "loss": 0.4706,
+      "step": 12114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3076550596172886,
+      "learning_rate": 4.466559847958318e-06,
+      "loss": 0.2392,
+      "step": 12115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3388503851948695,
+      "learning_rate": 4.46500989137905e-06,
+      "loss": 0.2912,
+      "step": 12116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3243942487719419,
+      "learning_rate": 4.463460126476251e-06,
+      "loss": 0.2058,
+      "step": 12117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46445889232908727,
+      "learning_rate": 4.46191055330359e-06,
+      "loss": 0.2573,
+      "step": 12118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26405584070347937,
+      "learning_rate": 4.460361171914724e-06,
+      "loss": 0.2552,
+      "step": 12119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48086867532454386,
+      "learning_rate": 4.458811982363317e-06,
+      "loss": 0.2995,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6352621011962262,
+      "learning_rate": 4.457262984703015e-06,
+      "loss": 0.2686,
+      "step": 12121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2669809765909583,
+      "learning_rate": 4.455714178987456e-06,
+      "loss": 0.2236,
+      "step": 12122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26373342863482124,
+      "learning_rate": 4.454165565270272e-06,
+      "loss": 0.2267,
+      "step": 12123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.47415110258209775,
+      "learning_rate": 4.452617143605099e-06,
+      "loss": 0.2042,
+      "step": 12124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37038910052018265,
+      "learning_rate": 4.451068914045556e-06,
+      "loss": 0.3001,
+      "step": 12125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7943156231991123,
+      "learning_rate": 4.449520876645258e-06,
+      "loss": 0.372,
+      "step": 12126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30751059252070245,
+      "learning_rate": 4.447973031457809e-06,
+      "loss": 0.2835,
+      "step": 12127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31206151442044566,
+      "learning_rate": 4.4464253785368205e-06,
+      "loss": 0.1877,
+      "step": 12128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2591005552248067,
+      "learning_rate": 4.4448779179358815e-06,
+      "loss": 0.1728,
+      "step": 12129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.248204418833787,
+      "learning_rate": 4.443330649708581e-06,
+      "loss": 0.7986,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2736973391515441,
+      "learning_rate": 4.441783573908498e-06,
+      "loss": 0.2117,
+      "step": 12131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5426384659039937,
+      "learning_rate": 4.440236690589215e-06,
+      "loss": 0.3368,
+      "step": 12132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9915066429385122,
+      "learning_rate": 4.438689999804295e-06,
+      "loss": 0.4082,
+      "step": 12133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2815498231479696,
+      "learning_rate": 4.437143501607302e-06,
+      "loss": 0.1841,
+      "step": 12134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21873970118674596,
+      "learning_rate": 4.435597196051789e-06,
+      "loss": 0.2132,
+      "step": 12135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2154872539731227,
+      "learning_rate": 4.434051083191304e-06,
+      "loss": 0.742,
+      "step": 12136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3401125230433799,
+      "learning_rate": 4.432505163079394e-06,
+      "loss": 0.1953,
+      "step": 12137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0219315897519718,
+      "learning_rate": 4.4309594357695895e-06,
+      "loss": 0.4175,
+      "step": 12138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37693401492937234,
+      "learning_rate": 4.429413901315421e-06,
+      "loss": 0.3112,
+      "step": 12139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3591687661022238,
+      "learning_rate": 4.4278685597704065e-06,
+      "loss": 0.2951,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.434384821341122,
+      "learning_rate": 4.426323411188067e-06,
+      "loss": 0.1519,
+      "step": 12141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2933368604620994,
+      "learning_rate": 4.424778455621908e-06,
+      "loss": 0.25,
+      "step": 12142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32729484470318493,
+      "learning_rate": 4.4232336931254324e-06,
+      "loss": 0.2727,
+      "step": 12143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7742472208938666,
+      "learning_rate": 4.42168912375213e-06,
+      "loss": 0.2935,
+      "step": 12144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.790094629215079,
+      "learning_rate": 4.420144747555497e-06,
+      "loss": 0.3164,
+      "step": 12145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.355497378756577,
+      "learning_rate": 4.418600564589012e-06,
+      "loss": 0.273,
+      "step": 12146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33562205802520717,
+      "learning_rate": 4.417056574906148e-06,
+      "loss": 0.2513,
+      "step": 12147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9439148205438117,
+      "learning_rate": 4.415512778560376e-06,
+      "loss": 0.5202,
+      "step": 12148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31573271535824526,
+      "learning_rate": 4.413969175605152e-06,
+      "loss": 0.2504,
+      "step": 12149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30110495517556174,
+      "learning_rate": 4.412425766093939e-06,
+      "loss": 0.1944,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5043638552372725,
+      "learning_rate": 4.410882550080182e-06,
+      "loss": 0.3179,
+      "step": 12151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33032605242595947,
+      "learning_rate": 4.409339527617321e-06,
+      "loss": 0.2619,
+      "step": 12152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2303501312225467,
+      "learning_rate": 4.407796698758788e-06,
+      "loss": 0.6729,
+      "step": 12153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46335850324152866,
+      "learning_rate": 4.40625406355802e-06,
+      "loss": 0.2766,
+      "step": 12154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2846449360546112,
+      "learning_rate": 4.404711622068436e-06,
+      "loss": 0.2599,
+      "step": 12155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7503421917007584,
+      "learning_rate": 4.40316937434344e-06,
+      "loss": 0.4376,
+      "step": 12156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2820811862011241,
+      "learning_rate": 4.401627320436453e-06,
+      "loss": 0.1031,
+      "step": 12157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33872988149067856,
+      "learning_rate": 4.40008546040087e-06,
+      "loss": 0.2569,
+      "step": 12158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3510877902653597,
+      "learning_rate": 4.3985437942900865e-06,
+      "loss": 0.3007,
+      "step": 12159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8283159971853021,
+      "learning_rate": 4.397002322157492e-06,
+      "loss": 0.3384,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3278628379074249,
+      "learning_rate": 4.395461044056462e-06,
+      "loss": 0.2431,
+      "step": 12161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4197960471554202,
+      "learning_rate": 4.393919960040377e-06,
+      "loss": 0.2814,
+      "step": 12162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2417208244068282,
+      "learning_rate": 4.392379070162604e-06,
+      "loss": 0.1892,
+      "step": 12163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3290819278571764,
+      "learning_rate": 4.390838374476503e-06,
+      "loss": 0.2373,
+      "step": 12164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9639347889069623,
+      "learning_rate": 4.3892978730354245e-06,
+      "loss": 0.5774,
+      "step": 12165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32514156426728147,
+      "learning_rate": 4.387757565892722e-06,
+      "loss": 0.3137,
+      "step": 12166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2997269568806797,
+      "learning_rate": 4.386217453101735e-06,
+      "loss": 0.1819,
+      "step": 12167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.44106553840301743,
+      "learning_rate": 4.384677534715794e-06,
+      "loss": 0.2737,
+      "step": 12168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40065246533447957,
+      "learning_rate": 4.383137810788226e-06,
+      "loss": 0.2289,
+      "step": 12169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3319847681859942,
+      "learning_rate": 4.381598281372358e-06,
+      "loss": 0.1977,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3532188786608427,
+      "learning_rate": 4.3800589465215e-06,
+      "loss": 0.3221,
+      "step": 12171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8397267713928125,
+      "learning_rate": 4.378519806288959e-06,
+      "loss": 0.3961,
+      "step": 12172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32192933397243806,
+      "learning_rate": 4.376980860728031e-06,
+      "loss": 0.2236,
+      "step": 12173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0490275553766555,
+      "learning_rate": 4.375442109892019e-06,
+      "loss": 0.527,
+      "step": 12174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22301783924245172,
+      "learning_rate": 4.373903553834203e-06,
+      "loss": 0.2092,
+      "step": 12175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3305834602905949,
+      "learning_rate": 4.372365192607866e-06,
+      "loss": 0.2539,
+      "step": 12176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0340970094704132,
+      "learning_rate": 4.370827026266281e-06,
+      "loss": 0.2975,
+      "step": 12177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37116083450859394,
+      "learning_rate": 4.36928905486271e-06,
+      "loss": 0.299,
+      "step": 12178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29651541573485524,
+      "learning_rate": 4.3677512784504195e-06,
+      "loss": 0.2427,
+      "step": 12179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0371072230107874,
+      "learning_rate": 4.366213697082661e-06,
+      "loss": 0.3959,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3476298224017206,
+      "learning_rate": 4.3646763108126796e-06,
+      "loss": 0.1951,
+      "step": 12181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.651854837110252,
+      "learning_rate": 4.363139119693712e-06,
+      "loss": 0.3743,
+      "step": 12182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3424674523302693,
+      "learning_rate": 4.361602123778998e-06,
+      "loss": 0.2393,
+      "step": 12183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.009983692526136,
+      "learning_rate": 4.360065323121759e-06,
+      "loss": 0.4591,
+      "step": 12184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4010361679084553,
+      "learning_rate": 4.358528717775217e-06,
+      "loss": 0.2785,
+      "step": 12185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35234124130496747,
+      "learning_rate": 4.356992307792578e-06,
+      "loss": 0.2476,
+      "step": 12186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23746666293127477,
+      "learning_rate": 4.355456093227056e-06,
+      "loss": 0.1907,
+      "step": 12187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39902282031332437,
+      "learning_rate": 4.353920074131848e-06,
+      "loss": 0.2616,
+      "step": 12188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.437743857169606,
+      "learning_rate": 4.352384250560147e-06,
+      "loss": 0.3208,
+      "step": 12189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3195775949375206,
+      "learning_rate": 4.350848622565131e-06,
+      "loss": 0.239,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3694916234759852,
+      "learning_rate": 4.349313190199988e-06,
+      "loss": 0.28,
+      "step": 12191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5313708496767161,
+      "learning_rate": 4.347777953517885e-06,
+      "loss": 0.3363,
+      "step": 12192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19723128611524043,
+      "learning_rate": 4.3462429125719884e-06,
+      "loss": 0.0851,
+      "step": 12193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3500548154240149,
+      "learning_rate": 4.344708067415454e-06,
+      "loss": 0.263,
+      "step": 12194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.38467142893601863,
+      "learning_rate": 4.34317341810144e-06,
+      "loss": 0.2988,
+      "step": 12195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6932149267264256,
+      "learning_rate": 4.341638964683086e-06,
+      "loss": 0.3442,
+      "step": 12196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3163438844619936,
+      "learning_rate": 4.3401047072135315e-06,
+      "loss": 0.2652,
+      "step": 12197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2479930111845223,
+      "learning_rate": 4.338570645745904e-06,
+      "loss": 0.5147,
+      "step": 12198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21571739388095768,
+      "learning_rate": 4.337036780333336e-06,
+      "loss": 0.1895,
+      "step": 12199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5681859271955713,
+      "learning_rate": 4.33550311102894e-06,
+      "loss": 0.3091,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4052305180015126,
+      "learning_rate": 4.333969637885827e-06,
+      "loss": 0.3169,
+      "step": 12201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33747729303466445,
+      "learning_rate": 4.332436360957104e-06,
+      "loss": 0.3031,
+      "step": 12202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34873869056387957,
+      "learning_rate": 4.3309032802958605e-06,
+      "loss": 0.1527,
+      "step": 12203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4009960812786949,
+      "learning_rate": 4.329370395955198e-06,
+      "loss": 0.301,
+      "step": 12204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6642800794115589,
+      "learning_rate": 4.3278377079881935e-06,
+      "loss": 0.3961,
+      "step": 12205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30282343673125794,
+      "learning_rate": 4.326305216447926e-06,
+      "loss": 0.2192,
+      "step": 12206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23522905895128113,
+      "learning_rate": 4.32477292138746e-06,
+      "loss": 0.2117,
+      "step": 12207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2941628507913063,
+      "learning_rate": 4.3232408228598685e-06,
+      "loss": 0.6561,
+      "step": 12208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35384968435253333,
+      "learning_rate": 4.321708920918203e-06,
+      "loss": 0.2012,
+      "step": 12209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3530687951144725,
+      "learning_rate": 4.320177215615513e-06,
+      "loss": 0.2766,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5241546712916147,
+      "learning_rate": 4.318645707004839e-06,
+      "loss": 0.3251,
+      "step": 12211
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36823411666661837,
+      "learning_rate": 4.317114395139222e-06,
+      "loss": 0.2413,
+      "step": 12212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2259296402410964,
+      "learning_rate": 4.3155832800716905e-06,
+      "loss": 0.1746,
+      "step": 12213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35860128935561664,
+      "learning_rate": 4.314052361855265e-06,
+      "loss": 0.3109,
+      "step": 12214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5859399504943746,
+      "learning_rate": 4.312521640542961e-06,
+      "loss": 0.4153,
+      "step": 12215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35181741535289196,
+      "learning_rate": 4.310991116187786e-06,
+      "loss": 0.2138,
+      "step": 12216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5705186817412079,
+      "learning_rate": 4.309460788842747e-06,
+      "loss": 0.3709,
+      "step": 12217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3280362062848473,
+      "learning_rate": 4.307930658560836e-06,
+      "loss": 0.2838,
+      "step": 12218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25938678514674396,
+      "learning_rate": 4.306400725395041e-06,
+      "loss": 0.1739,
+      "step": 12219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3705889257477343,
+      "learning_rate": 4.304870989398341e-06,
+      "loss": 0.2753,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7826792297259539,
+      "learning_rate": 4.303341450623717e-06,
+      "loss": 0.3297,
+      "step": 12221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25656625473960787,
+      "learning_rate": 4.301812109124134e-06,
+      "loss": 0.2215,
+      "step": 12222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7552748960134755,
+      "learning_rate": 4.300282964952553e-06,
+      "loss": 0.3884,
+      "step": 12223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6690995984223479,
+      "learning_rate": 4.2987540181619265e-06,
+      "loss": 0.3853,
+      "step": 12224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2465990758728293,
+      "learning_rate": 4.2972252688052055e-06,
+      "loss": 0.1742,
+      "step": 12225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3394607609597814,
+      "learning_rate": 4.295696716935326e-06,
+      "loss": 0.2924,
+      "step": 12226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3112479870105338,
+      "learning_rate": 4.294168362605224e-06,
+      "loss": 0.2147,
+      "step": 12227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3997767864323401,
+      "learning_rate": 4.292640205867824e-06,
+      "loss": 0.2892,
+      "step": 12228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.072222696374089,
+      "learning_rate": 4.291112246776052e-06,
+      "loss": 0.2978,
+      "step": 12229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2996227549076089,
+      "learning_rate": 4.2895844853828165e-06,
+      "loss": 0.2741,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40168932524681566,
+      "learning_rate": 4.288056921741024e-06,
+      "loss": 0.3001,
+      "step": 12231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2768315986265087,
+      "learning_rate": 4.286529555903572e-06,
+      "loss": 0.1643,
+      "step": 12232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41305074420474364,
+      "learning_rate": 4.285002387923359e-06,
+      "loss": 0.2157,
+      "step": 12233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3232120840888913,
+      "learning_rate": 4.283475417853268e-06,
+      "loss": 0.2766,
+      "step": 12234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7148956900870683,
+      "learning_rate": 4.2819486457461765e-06,
+      "loss": 0.2733,
+      "step": 12235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8019061600739739,
+      "learning_rate": 4.280422071654955e-06,
+      "loss": 0.4053,
+      "step": 12236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3394041675400556,
+      "learning_rate": 4.278895695632474e-06,
+      "loss": 0.2784,
+      "step": 12237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3341011586741274,
+      "learning_rate": 4.27736951773159e-06,
+      "loss": 0.2744,
+      "step": 12238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35152470282886145,
+      "learning_rate": 4.275843538005153e-06,
+      "loss": 0.2208,
+      "step": 12239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.370400992897311,
+      "learning_rate": 4.274317756506008e-06,
+      "loss": 0.2785,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4693231268558691,
+      "learning_rate": 4.2727921732869894e-06,
+      "loss": 0.2495,
+      "step": 12241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3241862213090096,
+      "learning_rate": 4.271266788400935e-06,
+      "loss": 0.2379,
+      "step": 12242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3449514730274378,
+      "learning_rate": 4.269741601900667e-06,
+      "loss": 0.2681,
+      "step": 12243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1807650540650891,
+      "learning_rate": 4.268216613838998e-06,
+      "loss": 0.6882,
+      "step": 12244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36468184086103056,
+      "learning_rate": 4.266691824268739e-06,
+      "loss": 0.2161,
+      "step": 12245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29744735921142973,
+      "learning_rate": 4.2651672332427e-06,
+      "loss": 0.264,
+      "step": 12246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.476773513987374,
+      "learning_rate": 4.263642840813672e-06,
+      "loss": 0.2558,
+      "step": 12247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5281422302262536,
+      "learning_rate": 4.262118647034447e-06,
+      "loss": 0.132,
+      "step": 12248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3918244540090564,
+      "learning_rate": 4.260594651957801e-06,
+      "loss": 0.3159,
+      "step": 12249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3831510721886654,
+      "learning_rate": 4.25907085563652e-06,
+      "loss": 0.3368,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.518415226887272,
+      "learning_rate": 4.257547258123369e-06,
+      "loss": 0.2439,
+      "step": 12251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3912430449227991,
+      "learning_rate": 4.256023859471109e-06,
+      "loss": 0.2949,
+      "step": 12252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31137797694610614,
+      "learning_rate": 4.254500659732496e-06,
+      "loss": 0.1849,
+      "step": 12253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36367056958104427,
+      "learning_rate": 4.2529776589602735e-06,
+      "loss": 0.3097,
+      "step": 12254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28821543574715613,
+      "learning_rate": 4.251454857207193e-06,
+      "loss": 0.1862,
+      "step": 12255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7304920210259426,
+      "learning_rate": 4.249932254525985e-06,
+      "loss": 0.4968,
+      "step": 12256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4805174118924019,
+      "learning_rate": 4.24840985096937e-06,
+      "loss": 0.3333,
+      "step": 12257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.24509749223680857,
+      "learning_rate": 4.246887646590077e-06,
+      "loss": 0.2109,
+      "step": 12258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30668412572681175,
+      "learning_rate": 4.245365641440818e-06,
+      "loss": 0.2188,
+      "step": 12259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0420482711700545,
+      "learning_rate": 4.243843835574299e-06,
+      "loss": 0.4492,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29947922607947813,
+      "learning_rate": 4.242322229043218e-06,
+      "loss": 0.2126,
+      "step": 12261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34042241000669604,
+      "learning_rate": 4.240800821900274e-06,
+      "loss": 0.2878,
+      "step": 12262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8679278842448671,
+      "learning_rate": 4.23927961419815e-06,
+      "loss": 0.4032,
+      "step": 12263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3370249737144881,
+      "learning_rate": 4.237758605989523e-06,
+      "loss": 0.247,
+      "step": 12264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.15104224638072863,
+      "learning_rate": 4.236237797327071e-06,
+      "loss": 0.071,
+      "step": 12265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37245443877274914,
+      "learning_rate": 4.2347171882634505e-06,
+      "loss": 0.3221,
+      "step": 12266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3957859482382533,
+      "learning_rate": 4.2331967788513295e-06,
+      "loss": 0.2676,
+      "step": 12267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4817458267951734,
+      "learning_rate": 4.231676569143357e-06,
+      "loss": 0.2875,
+      "step": 12268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33385576628286906,
+      "learning_rate": 4.230156559192177e-06,
+      "loss": 0.3,
+      "step": 12269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3606609119285642,
+      "learning_rate": 4.228636749050422e-06,
+      "loss": 0.2815,
+      "step": 12270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28662475275203025,
+      "learning_rate": 4.227117138770733e-06,
+      "loss": 0.1133,
+      "step": 12271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1924756559482446,
+      "learning_rate": 4.225597728405729e-06,
+      "loss": 0.5055,
+      "step": 12272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33127274692223163,
+      "learning_rate": 4.224078518008028e-06,
+      "loss": 0.2566,
+      "step": 12273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39603641748968726,
+      "learning_rate": 4.222559507630235e-06,
+      "loss": 0.2764,
+      "step": 12274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1371302801557202,
+      "learning_rate": 4.221040697324962e-06,
+      "loss": 0.8109,
+      "step": 12275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.35190049784212624,
+      "learning_rate": 4.2195220871448005e-06,
+      "loss": 0.2424,
+      "step": 12276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27023187100195273,
+      "learning_rate": 4.218003677142342e-06,
+      "loss": 0.2132,
+      "step": 12277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.48459025177133447,
+      "learning_rate": 4.216485467370163e-06,
+      "loss": 0.2421,
+      "step": 12278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30829697404074013,
+      "learning_rate": 4.214967457880846e-06,
+      "loss": 0.2431,
+      "step": 12279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6722535491136014,
+      "learning_rate": 4.213449648726958e-06,
+      "loss": 0.4095,
+      "step": 12280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30788458676053476,
+      "learning_rate": 4.211932039961061e-06,
+      "loss": 0.2747,
+      "step": 12281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3495341382200821,
+      "learning_rate": 4.210414631635707e-06,
+      "loss": 0.2409,
+      "step": 12282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1231503038430486,
+      "learning_rate": 4.208897423803443e-06,
+      "loss": 0.5617,
+      "step": 12283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3074058591745467,
+      "learning_rate": 4.207380416516815e-06,
+      "loss": 0.162,
+      "step": 12284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4244766221986541,
+      "learning_rate": 4.2058636098283545e-06,
+      "loss": 0.294,
+      "step": 12285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31602488202105283,
+      "learning_rate": 4.204347003790588e-06,
+      "loss": 0.3086,
+      "step": 12286
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9753506360078198,
+      "learning_rate": 4.202830598456032e-06,
+      "loss": 0.5152,
+      "step": 12287
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4134773260026491,
+      "learning_rate": 4.201314393877206e-06,
+      "loss": 0.2844,
+      "step": 12288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3680992182220188,
+      "learning_rate": 4.199798390106613e-06,
+      "loss": 0.2835,
+      "step": 12289
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29708444907868753,
+      "learning_rate": 4.198282587196757e-06,
+      "loss": 0.2302,
+      "step": 12290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2937281945675394,
+      "learning_rate": 4.196766985200118e-06,
+      "loss": 0.1866,
+      "step": 12291
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6844793590759544,
+      "learning_rate": 4.195251584169192e-06,
+      "loss": 0.4549,
+      "step": 12292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5098854869023844,
+      "learning_rate": 4.193736384156455e-06,
+      "loss": 0.3973,
+      "step": 12293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2564448270712793,
+      "learning_rate": 4.192221385214377e-06,
+      "loss": 0.2072,
+      "step": 12294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5101193979062211,
+      "learning_rate": 4.190706587395418e-06,
+      "loss": 0.2752,
+      "step": 12295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4351332109330603,
+      "learning_rate": 4.189191990752044e-06,
+      "loss": 0.271,
+      "step": 12296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26640761033236676,
+      "learning_rate": 4.187677595336702e-06,
+      "loss": 0.2164,
+      "step": 12297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7739642089022813,
+      "learning_rate": 4.186163401201835e-06,
+      "loss": 0.3333,
+      "step": 12298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9586181627625986,
+      "learning_rate": 4.184649408399876e-06,
+      "loss": 0.6132,
+      "step": 12299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3301622040533963,
+      "learning_rate": 4.183135616983261e-06,
+      "loss": 0.1892,
+      "step": 12300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4567133495984465,
+      "learning_rate": 4.181622027004409e-06,
+      "loss": 0.3046,
+      "step": 12301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.46370032678492656,
+      "learning_rate": 4.1801086385157366e-06,
+      "loss": 0.3539,
+      "step": 12302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4066621177522798,
+      "learning_rate": 4.178595451569648e-06,
+      "loss": 0.2692,
+      "step": 12303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20560129040918335,
+      "learning_rate": 4.177082466218553e-06,
+      "loss": 0.1409,
+      "step": 12304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.35498442687133175,
+      "learning_rate": 4.17556968251484e-06,
+      "loss": 0.3217,
+      "step": 12305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6429202460112974,
+      "learning_rate": 4.1740571005109e-06,
+      "loss": 0.3263,
+      "step": 12306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.35829501157032695,
+      "learning_rate": 4.1725447202591115e-06,
+      "loss": 0.2361,
+      "step": 12307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6702766382148628,
+      "learning_rate": 4.171032541811846e-06,
+      "loss": 0.3579,
+      "step": 12308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.302468040201284,
+      "learning_rate": 4.169520565221476e-06,
+      "loss": 0.2649,
+      "step": 12309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19513857944378782,
+      "learning_rate": 4.1680087905403575e-06,
+      "loss": 0.1667,
+      "step": 12310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3169755916645531,
+      "learning_rate": 4.166497217820844e-06,
+      "loss": 0.7212,
+      "step": 12311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3903916545542141,
+      "learning_rate": 4.164985847115279e-06,
+      "loss": 0.2538,
+      "step": 12312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3307900193032403,
+      "learning_rate": 4.163474678476004e-06,
+      "loss": 0.2559,
+      "step": 12313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7210475091920264,
+      "learning_rate": 4.161963711955351e-06,
+      "loss": 0.3969,
+      "step": 12314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4130701912582865,
+      "learning_rate": 4.1604529476056446e-06,
+      "loss": 0.2791,
+      "step": 12315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33158894769484887,
+      "learning_rate": 4.158942385479198e-06,
+      "loss": 0.2499,
+      "step": 12316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26527730099168056,
+      "learning_rate": 4.157432025628327e-06,
+      "loss": 0.2226,
+      "step": 12317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4024841563367093,
+      "learning_rate": 4.155921868105336e-06,
+      "loss": 0.2536,
+      "step": 12318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4715939185767074,
+      "learning_rate": 4.154411912962518e-06,
+      "loss": 0.3245,
+      "step": 12319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49784484653712924,
+      "learning_rate": 4.152902160252165e-06,
+      "loss": 0.2754,
+      "step": 12320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3167715593401071,
+      "learning_rate": 4.151392610026554e-06,
+      "loss": 0.2556,
+      "step": 12321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5362943498624293,
+      "learning_rate": 4.149883262337969e-06,
+      "loss": 0.361,
+      "step": 12322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2270767119225587,
+      "learning_rate": 4.148374117238676e-06,
+      "loss": 0.1251,
+      "step": 12323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5482866300378484,
+      "learning_rate": 4.1468651747809366e-06,
+      "loss": 0.2773,
+      "step": 12324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26991337180134006,
+      "learning_rate": 4.145356435017003e-06,
+      "loss": 0.262,
+      "step": 12325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8129780685542745,
+      "learning_rate": 4.143847897999124e-06,
+      "loss": 0.3138,
+      "step": 12326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6230381356562327,
+      "learning_rate": 4.142339563779542e-06,
+      "loss": 0.3043,
+      "step": 12327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4264045585072286,
+      "learning_rate": 4.140831432410484e-06,
+      "loss": 0.3266,
+      "step": 12328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3738143640469887,
+      "learning_rate": 4.139323503944186e-06,
+      "loss": 0.3328,
+      "step": 12329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2431647354371285,
+      "learning_rate": 4.1378157784328625e-06,
+      "loss": 0.1254,
+      "step": 12330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5894238839413188,
+      "learning_rate": 4.136308255928726e-06,
+      "loss": 0.3434,
+      "step": 12331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2000110180565327,
+      "learning_rate": 4.134800936483983e-06,
+      "loss": 0.6542,
+      "step": 12332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.267646159596712,
+      "learning_rate": 4.1332938201508285e-06,
+      "loss": 0.2183,
+      "step": 12333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5490669982217542,
+      "learning_rate": 4.13178690698146e-06,
+      "loss": 0.3078,
+      "step": 12334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.40224921477044767,
+      "learning_rate": 4.130280197028058e-06,
+      "loss": 0.2917,
+      "step": 12335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.398436666474072,
+      "learning_rate": 4.128773690342801e-06,
+      "loss": 0.2357,
+      "step": 12336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3085871901913578,
+      "learning_rate": 4.127267386977854e-06,
+      "loss": 0.2875,
+      "step": 12337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41757727862480615,
+      "learning_rate": 4.125761286985389e-06,
+      "loss": 0.2504,
+      "step": 12338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5953337494643126,
+      "learning_rate": 4.124255390417558e-06,
+      "loss": 0.2831,
+      "step": 12339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3973615565133707,
+      "learning_rate": 4.122749697326511e-06,
+      "loss": 0.2354,
+      "step": 12340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3576969019418588,
+      "learning_rate": 4.121244207764384e-06,
+      "loss": 0.3025,
+      "step": 12341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6161027575033992,
+      "learning_rate": 4.119738921783323e-06,
+      "loss": 0.3604,
+      "step": 12342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22322602570107805,
+      "learning_rate": 4.118233839435449e-06,
+      "loss": 0.1689,
+      "step": 12343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5065158568778143,
+      "learning_rate": 4.1167289607728845e-06,
+      "loss": 0.3368,
+      "step": 12344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4057898445010308,
+      "learning_rate": 4.1152242858477435e-06,
+      "loss": 0.2837,
+      "step": 12345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31512309786509607,
+      "learning_rate": 4.113719814712127e-06,
+      "loss": 0.2346,
+      "step": 12346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6702265380981722,
+      "learning_rate": 4.112215547418145e-06,
+      "loss": 0.4688,
+      "step": 12347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.38723962767239795,
+      "learning_rate": 4.110711484017886e-06,
+      "loss": 0.3277,
+      "step": 12348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.238084010943153,
+      "learning_rate": 4.1092076245634346e-06,
+      "loss": 0.1993,
+      "step": 12349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4857992938940668,
+      "learning_rate": 4.107703969106867e-06,
+      "loss": 0.2588,
+      "step": 12350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.617699439344278,
+      "learning_rate": 4.10620051770026e-06,
+      "loss": 0.3235,
+      "step": 12351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4087980308933262,
+      "learning_rate": 4.104697270395676e-06,
+      "loss": 0.3254,
+      "step": 12352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34914826637736407,
+      "learning_rate": 4.103194227245172e-06,
+      "loss": 0.2729,
+      "step": 12353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5393994201005855,
+      "learning_rate": 4.101691388300795e-06,
+      "loss": 0.3091,
+      "step": 12354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42117791947230276,
+      "learning_rate": 4.100188753614595e-06,
+      "loss": 0.2889,
+      "step": 12355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2679960829980229,
+      "learning_rate": 4.098686323238604e-06,
+      "loss": 0.1591,
+      "step": 12356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3767455945961782,
+      "learning_rate": 4.097184097224853e-06,
+      "loss": 0.263,
+      "step": 12357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3979800959449096,
+      "learning_rate": 4.095682075625363e-06,
+      "loss": 0.2972,
+      "step": 12358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7153844837257711,
+      "learning_rate": 4.094180258492147e-06,
+      "loss": 0.3261,
+      "step": 12359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33698757282514463,
+      "learning_rate": 4.092678645877217e-06,
+      "loss": 0.3312,
+      "step": 12360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32736161830074756,
+      "learning_rate": 4.09117723783257e-06,
+      "loss": 0.2544,
+      "step": 12361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29749907521333085,
+      "learning_rate": 4.089676034410198e-06,
+      "loss": 0.1178,
+      "step": 12362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7603494708438945,
+      "learning_rate": 4.088175035662095e-06,
+      "loss": 0.4403,
+      "step": 12363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3803334125043377,
+      "learning_rate": 4.086674241640235e-06,
+      "loss": 0.2635,
+      "step": 12364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3682851850725568,
+      "learning_rate": 4.085173652396593e-06,
+      "loss": 0.3203,
+      "step": 12365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5811058200485942,
+      "learning_rate": 4.083673267983128e-06,
+      "loss": 0.2581,
+      "step": 12366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34836499493050266,
+      "learning_rate": 4.0821730884518085e-06,
+      "loss": 0.2758,
+      "step": 12367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2767104843102272,
+      "learning_rate": 4.08067311385458e-06,
+      "loss": 0.1553,
+      "step": 12368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30352089800280146,
+      "learning_rate": 4.079173344243387e-06,
+      "loss": 0.2286,
+      "step": 12369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32423430858686303,
+      "learning_rate": 4.077673779670166e-06,
+      "loss": 0.2643,
+      "step": 12370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6190946400648354,
+      "learning_rate": 4.076174420186844e-06,
+      "loss": 0.4023,
+      "step": 12371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3159074215735569,
+      "learning_rate": 4.07467526584535e-06,
+      "loss": 0.2145,
+      "step": 12372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30664118077870395,
+      "learning_rate": 4.073176316697598e-06,
+      "loss": 0.2746,
+      "step": 12373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3017502296489269,
+      "learning_rate": 4.071677572795495e-06,
+      "loss": 0.1723,
+      "step": 12374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4615780722507267,
+      "learning_rate": 4.0701790341909386e-06,
+      "loss": 0.2075,
+      "step": 12375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3855982088594073,
+      "learning_rate": 4.068680700935831e-06,
+      "loss": 0.3036,
+      "step": 12376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36671811376397273,
+      "learning_rate": 4.0671825730820555e-06,
+      "loss": 0.3023,
+      "step": 12377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2575854114457314,
+      "learning_rate": 4.065684650681493e-06,
+      "loss": 0.75,
+      "step": 12378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34502650981498256,
+      "learning_rate": 4.064186933786012e-06,
+      "loss": 0.2024,
+      "step": 12379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22705977337178385,
+      "learning_rate": 4.062689422447487e-06,
+      "loss": 0.2112,
+      "step": 12380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.640804257150291,
+      "learning_rate": 4.061192116717771e-06,
+      "loss": 0.4038,
+      "step": 12381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33188078306411595,
+      "learning_rate": 4.0596950166487146e-06,
+      "loss": 0.2323,
+      "step": 12382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2416745783711352,
+      "learning_rate": 4.058198122292167e-06,
+      "loss": 0.6851,
+      "step": 12383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33509544879795256,
+      "learning_rate": 4.0567014336999584e-06,
+      "loss": 0.2977,
+      "step": 12384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.35947747921768164,
+      "learning_rate": 4.055204950923927e-06,
+      "loss": 0.205,
+      "step": 12385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32272880351486805,
+      "learning_rate": 4.053708674015893e-06,
+      "loss": 0.1713,
+      "step": 12386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5147264850139122,
+      "learning_rate": 4.052212603027672e-06,
+      "loss": 0.3099,
+      "step": 12387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2537057854401847,
+      "learning_rate": 4.050716738011068e-06,
+      "loss": 0.1977,
+      "step": 12388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5015595237780548,
+      "learning_rate": 4.049221079017892e-06,
+      "loss": 0.3651,
+      "step": 12389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2806204433781172,
+      "learning_rate": 4.0477256260999344e-06,
+      "loss": 0.6888,
+      "step": 12390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3961070936444837,
+      "learning_rate": 4.046230379308982e-06,
+      "loss": 0.2544,
+      "step": 12391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24351026173987872,
+      "learning_rate": 4.0447353386968155e-06,
+      "loss": 0.2007,
+      "step": 12392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5317384655177135,
+      "learning_rate": 4.043240504315209e-06,
+      "loss": 0.3628,
+      "step": 12393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36782911635286475,
+      "learning_rate": 4.041745876215927e-06,
+      "loss": 0.3094,
+      "step": 12394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26601632760932586,
+      "learning_rate": 4.040251454450729e-06,
+      "loss": 0.1666,
+      "step": 12395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36052225817436,
+      "learning_rate": 4.038757239071364e-06,
+      "loss": 0.3037,
+      "step": 12396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3159879747015827,
+      "learning_rate": 4.037263230129583e-06,
+      "loss": 0.2501,
+      "step": 12397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6144153528165102,
+      "learning_rate": 4.035769427677118e-06,
+      "loss": 0.2974,
+      "step": 12398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4335431228327503,
+      "learning_rate": 4.034275831765702e-06,
+      "loss": 0.2698,
+      "step": 12399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2705356897806853,
+      "learning_rate": 4.032782442447055e-06,
+      "loss": 0.249,
+      "step": 12400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27198063920594956,
+      "learning_rate": 4.031289259772898e-06,
+      "loss": 0.1899,
+      "step": 12401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9726204979369423,
+      "learning_rate": 4.029796283794938e-06,
+      "loss": 0.5632,
+      "step": 12402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3511151019796598,
+      "learning_rate": 4.028303514564876e-06,
+      "loss": 0.2668,
+      "step": 12403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.46440585552108493,
+      "learning_rate": 4.026810952134402e-06,
+      "loss": 0.2985,
+      "step": 12404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3997705620102345,
+      "learning_rate": 4.025318596555212e-06,
+      "loss": 0.2284,
+      "step": 12405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24602939683462624,
+      "learning_rate": 4.023826447878982e-06,
+      "loss": 0.2192,
+      "step": 12406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1948345929606112,
+      "learning_rate": 4.022334506157386e-06,
+      "loss": 0.8185,
+      "step": 12407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33598971886925083,
+      "learning_rate": 4.020842771442085e-06,
+      "loss": 0.2542,
+      "step": 12408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42240133665361956,
+      "learning_rate": 4.019351243784745e-06,
+      "loss": 0.2873,
+      "step": 12409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5062951214860231,
+      "learning_rate": 4.017859923237014e-06,
+      "loss": 0.3348,
+      "step": 12410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34173636455748163,
+      "learning_rate": 4.016368809850537e-06,
+      "loss": 0.2076,
+      "step": 12411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3865772364660087,
+      "learning_rate": 4.01487790367695e-06,
+      "loss": 0.2696,
+      "step": 12412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3557344917335959,
+      "learning_rate": 4.013387204767881e-06,
+      "loss": 0.3212,
+      "step": 12413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23541794800684201,
+      "learning_rate": 4.01189671317496e-06,
+      "loss": 0.0956,
+      "step": 12414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2951339678252151,
+      "learning_rate": 4.0104064289497965e-06,
+      "loss": 0.2571,
+      "step": 12415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33535513867717087,
+      "learning_rate": 4.008916352144002e-06,
+      "loss": 0.302,
+      "step": 12416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.12245964652745,
+      "learning_rate": 4.007426482809172e-06,
+      "loss": 0.5287,
+      "step": 12417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29781734362720846,
+      "learning_rate": 4.0059368209969106e-06,
+      "loss": 0.186,
+      "step": 12418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0779840391036732,
+      "learning_rate": 4.004447366758798e-06,
+      "loss": 0.7548,
+      "step": 12419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22319618891727672,
+      "learning_rate": 4.002958120146415e-06,
+      "loss": 0.2061,
+      "step": 12420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3003774906221277,
+      "learning_rate": 4.001469081211332e-06,
+      "loss": 0.2077,
+      "step": 12421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6334985632628138,
+      "learning_rate": 3.99998025000512e-06,
+      "loss": 0.3586,
+      "step": 12422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0323028337477798,
+      "learning_rate": 3.998491626579334e-06,
+      "loss": 0.4385,
+      "step": 12423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22590105407607644,
+      "learning_rate": 3.997003210985524e-06,
+      "loss": 0.2142,
+      "step": 12424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2186487143817915,
+      "learning_rate": 3.995515003275235e-06,
+      "loss": 0.7572,
+      "step": 12425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.320271494595398,
+      "learning_rate": 3.9940270035000036e-06,
+      "loss": 0.2219,
+      "step": 12426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5789444426466166,
+      "learning_rate": 3.992539211711359e-06,
+      "loss": 0.3055,
+      "step": 12427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2750210705197789,
+      "learning_rate": 3.991051627960822e-06,
+      "loss": 0.2382,
+      "step": 12428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1626005918199633,
+      "learning_rate": 3.989564252299907e-06,
+      "loss": 0.6162,
+      "step": 12429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6137501022840672,
+      "learning_rate": 3.988077084780126e-06,
+      "loss": 0.3585,
+      "step": 12430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3564797669004418,
+      "learning_rate": 3.986590125452977e-06,
+      "loss": 0.2574,
+      "step": 12431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3595173618942061,
+      "learning_rate": 3.985103374369954e-06,
+      "loss": 0.3004,
+      "step": 12432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5610988564310685,
+      "learning_rate": 3.983616831582538e-06,
+      "loss": 0.3092,
+      "step": 12433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22865917764490962,
+      "learning_rate": 3.9821304971422155e-06,
+      "loss": 0.153,
+      "step": 12434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0396124580607748,
+      "learning_rate": 3.980644371100457e-06,
+      "loss": 0.4856,
+      "step": 12435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2712024160349323,
+      "learning_rate": 3.979158453508724e-06,
+      "loss": 0.2664,
+      "step": 12436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.45818436095554166,
+      "learning_rate": 3.977672744418475e-06,
+      "loss": 0.2934,
+      "step": 12437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7060226833636093,
+      "learning_rate": 3.976187243881156e-06,
+      "loss": 0.3856,
+      "step": 12438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28119753881737464,
+      "learning_rate": 3.974701951948218e-06,
+      "loss": 0.207,
+      "step": 12439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3156999954213079,
+      "learning_rate": 3.973216868671092e-06,
+      "loss": 0.2725,
+      "step": 12440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3275821325716865,
+      "learning_rate": 3.9717319941012054e-06,
+      "loss": 0.0955,
+      "step": 12441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4230111041588665,
+      "learning_rate": 3.970247328289979e-06,
+      "loss": 0.3222,
+      "step": 12442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5330638370371196,
+      "learning_rate": 3.96876287128883e-06,
+      "loss": 0.3889,
+      "step": 12443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.48238254940743125,
+      "learning_rate": 3.967278623149165e-06,
+      "loss": 0.2652,
+      "step": 12444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6377281456624062,
+      "learning_rate": 3.965794583922382e-06,
+      "loss": 0.27,
+      "step": 12445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24998486209393883,
+      "learning_rate": 3.964310753659869e-06,
+      "loss": 0.1951,
+      "step": 12446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3267863704261433,
+      "learning_rate": 3.9628271324130185e-06,
+      "loss": 0.2329,
+      "step": 12447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5376672392151259,
+      "learning_rate": 3.961343720233204e-06,
+      "loss": 0.2847,
+      "step": 12448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41943787136242416,
+      "learning_rate": 3.9598605171717976e-06,
+      "loss": 0.3178,
+      "step": 12449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6504493113239271,
+      "learning_rate": 3.958377523280162e-06,
+      "loss": 0.3003,
+      "step": 12450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6739410311061353,
+      "learning_rate": 3.956894738609649e-06,
+      "loss": 0.3241,
+      "step": 12451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25881186952173607,
+      "learning_rate": 3.955412163211615e-06,
+      "loss": 0.2511,
+      "step": 12452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4730776211188233,
+      "learning_rate": 3.953929797137398e-06,
+      "loss": 0.2741,
+      "step": 12453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32294979003618474,
+      "learning_rate": 3.9524476404383324e-06,
+      "loss": 0.1982,
+      "step": 12454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4427233885442675,
+      "learning_rate": 3.9509656931657405e-06,
+      "loss": 0.3236,
+      "step": 12455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5624680637454126,
+      "learning_rate": 3.949483955370951e-06,
+      "loss": 0.3349,
+      "step": 12456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3926441797137969,
+      "learning_rate": 3.9480024271052715e-06,
+      "loss": 0.1955,
+      "step": 12457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33987897951718266,
+      "learning_rate": 3.946521108420008e-06,
+      "loss": 0.2365,
+      "step": 12458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3372391079565309,
+      "learning_rate": 3.945039999366458e-06,
+      "loss": 0.2494,
+      "step": 12459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3119762747576032,
+      "learning_rate": 3.9435590999959115e-06,
+      "loss": 0.2094,
+      "step": 12460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5354947724202691,
+      "learning_rate": 3.942078410359655e-06,
+      "loss": 0.3259,
+      "step": 12461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7620520172406171,
+      "learning_rate": 3.940597930508962e-06,
+      "loss": 0.4483,
+      "step": 12462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3951157539623328,
+      "learning_rate": 3.939117660495098e-06,
+      "loss": 0.2072,
+      "step": 12463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30130478030699875,
+      "learning_rate": 3.937637600369332e-06,
+      "loss": 0.2685,
+      "step": 12464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.317880848548765,
+      "learning_rate": 3.936157750182915e-06,
+      "loss": 0.1775,
+      "step": 12465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7654264548014176,
+      "learning_rate": 3.934678109987096e-06,
+      "loss": 0.3159,
+      "step": 12466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35211292954046547,
+      "learning_rate": 3.933198679833108e-06,
+      "loss": 0.2436,
+      "step": 12467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48058464826750036,
+      "learning_rate": 3.931719459772193e-06,
+      "loss": 0.4169,
+      "step": 12468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7896326662596463,
+      "learning_rate": 3.9302404498555725e-06,
+      "loss": 0.3491,
+      "step": 12469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1967202253818277,
+      "learning_rate": 3.928761650134464e-06,
+      "loss": 0.149,
+      "step": 12470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5161953619328937,
+      "learning_rate": 3.927283060660075e-06,
+      "loss": 0.326,
+      "step": 12471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4444720336390856,
+      "learning_rate": 3.925804681483614e-06,
+      "loss": 0.3244,
+      "step": 12472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36218587264417024,
+      "learning_rate": 3.924326512656279e-06,
+      "loss": 0.243,
+      "step": 12473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.308785521351059,
+      "learning_rate": 3.922848554229254e-06,
+      "loss": 0.7986,
+      "step": 12474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37676328938957093,
+      "learning_rate": 3.921370806253722e-06,
+      "loss": 0.2643,
+      "step": 12475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29437300765355345,
+      "learning_rate": 3.919893268780854e-06,
+      "loss": 0.2193,
+      "step": 12476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3144016115644878,
+      "learning_rate": 3.918415941861825e-06,
+      "loss": 0.2123,
+      "step": 12477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49162002682484635,
+      "learning_rate": 3.91693882554779e-06,
+      "loss": 0.2872,
+      "step": 12478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36777889700838373,
+      "learning_rate": 3.915461919889903e-06,
+      "loss": 0.3064,
+      "step": 12479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3299043967705451,
+      "learning_rate": 3.913985224939303e-06,
+      "loss": 0.2594,
+      "step": 12480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.790198056679378,
+      "learning_rate": 3.912508740747137e-06,
+      "loss": 0.3766,
+      "step": 12481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34173548287831607,
+      "learning_rate": 3.911032467364531e-06,
+      "loss": 0.2535,
+      "step": 12482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2538296150592013,
+      "learning_rate": 3.909556404842609e-06,
+      "loss": 0.1983,
+      "step": 12483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8236400542230081,
+      "learning_rate": 3.908080553232484e-06,
+      "loss": 0.4423,
+      "step": 12484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3028413178151663,
+      "learning_rate": 3.906604912585271e-06,
+      "loss": 0.2541,
+      "step": 12485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.951138237415222,
+      "learning_rate": 3.905129482952067e-06,
+      "loss": 0.4526,
+      "step": 12486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3129589276357312,
+      "learning_rate": 3.903654264383967e-06,
+      "loss": 0.2508,
+      "step": 12487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3663928277836569,
+      "learning_rate": 3.902179256932058e-06,
+      "loss": 0.2743,
+      "step": 12488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.776248988057081,
+      "learning_rate": 3.900704460647416e-06,
+      "loss": 0.29,
+      "step": 12489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32241816415937624,
+      "learning_rate": 3.89922987558112e-06,
+      "loss": 0.1965,
+      "step": 12490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2956955229384876,
+      "learning_rate": 3.897755501784231e-06,
+      "loss": 0.2823,
+      "step": 12491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49019798055890323,
+      "learning_rate": 3.896281339307805e-06,
+      "loss": 0.3925,
+      "step": 12492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8539817933659437,
+      "learning_rate": 3.8948073882028945e-06,
+      "loss": 0.1781,
+      "step": 12493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3545876480577167,
+      "learning_rate": 3.893333648520542e-06,
+      "loss": 0.2658,
+      "step": 12494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3817268627937381,
+      "learning_rate": 3.891860120311784e-06,
+      "loss": 0.3108,
+      "step": 12495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4087311870618779,
+      "learning_rate": 3.890386803627642e-06,
+      "loss": 0.1709,
+      "step": 12496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4015110688746216,
+      "learning_rate": 3.888913698519145e-06,
+      "loss": 0.3142,
+      "step": 12497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4151975450528388,
+      "learning_rate": 3.887440805037306e-06,
+      "loss": 0.2699,
+      "step": 12498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3269623807011317,
+      "learning_rate": 3.885968123233128e-06,
+      "loss": 0.233,
+      "step": 12499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35990831564539366,
+      "learning_rate": 3.884495653157611e-06,
+      "loss": 0.292,
+      "step": 12500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.67394425063061,
+      "learning_rate": 3.883023394861742e-06,
+      "loss": 0.3572,
+      "step": 12501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33215648058639724,
+      "learning_rate": 3.881551348396515e-06,
+      "loss": 0.1395,
+      "step": 12502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2825531167285477,
+      "learning_rate": 3.880079513812901e-06,
+      "loss": 0.2577,
+      "step": 12503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3434631773890578,
+      "learning_rate": 3.878607891161871e-06,
+      "loss": 0.2822,
+      "step": 12504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8360101546679122,
+      "learning_rate": 3.8771364804943825e-06,
+      "loss": 0.4662,
+      "step": 12505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31128367410087177,
+      "learning_rate": 3.8756652818613975e-06,
+      "loss": 0.2256,
+      "step": 12506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3889226384732306,
+      "learning_rate": 3.8741942953138616e-06,
+      "loss": 0.3229,
+      "step": 12507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4681783448097931,
+      "learning_rate": 3.872723520902713e-06,
+      "loss": 0.7554,
+      "step": 12508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2748888158937971,
+      "learning_rate": 3.87125295867888e-06,
+      "loss": 0.2006,
+      "step": 12509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27882312944220194,
+      "learning_rate": 3.8697826086933e-06,
+      "loss": 0.1742,
+      "step": 12510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3533427566146823,
+      "learning_rate": 3.868312470996884e-06,
+      "loss": 0.3008,
+      "step": 12511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3111571689105808,
+      "learning_rate": 3.866842545640542e-06,
+      "loss": 0.1881,
+      "step": 12512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6455308980072993,
+      "learning_rate": 3.86537283267518e-06,
+      "loss": 0.3791,
+      "step": 12513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4809467316313205,
+      "learning_rate": 3.863903332151689e-06,
+      "loss": 0.3647,
+      "step": 12514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40863296174877306,
+      "learning_rate": 3.862434044120966e-06,
+      "loss": 0.2974,
+      "step": 12515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20722922763128526,
+      "learning_rate": 3.860964968633888e-06,
+      "loss": 0.1662,
+      "step": 12516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5038681834454446,
+      "learning_rate": 3.859496105741328e-06,
+      "loss": 0.3576,
+      "step": 12517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3730892496342433,
+      "learning_rate": 3.858027455494152e-06,
+      "loss": 0.271,
+      "step": 12518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37789847744452215,
+      "learning_rate": 3.856559017943223e-06,
+      "loss": 0.2646,
+      "step": 12519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2120655893657206,
+      "learning_rate": 3.8550907931393925e-06,
+      "loss": 0.5642,
+      "step": 12520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32447720363858357,
+      "learning_rate": 3.853622781133503e-06,
+      "loss": 0.2848,
+      "step": 12521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3132306682340447,
+      "learning_rate": 3.852154981976388e-06,
+      "loss": 0.2106,
+      "step": 12522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43512513969447275,
+      "learning_rate": 3.8506873957188865e-06,
+      "loss": 0.3212,
+      "step": 12523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32854106565529423,
+      "learning_rate": 3.849220022411815e-06,
+      "loss": 0.261,
+      "step": 12524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4033956109292242,
+      "learning_rate": 3.84775286210599e-06,
+      "loss": 0.1076,
+      "step": 12525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4484468926497038,
+      "learning_rate": 3.846285914852216e-06,
+      "loss": 0.3562,
+      "step": 12526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2682409012322284,
+      "learning_rate": 3.844819180701302e-06,
+      "loss": 0.2414,
+      "step": 12527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2722056250509972,
+      "learning_rate": 3.843352659704032e-06,
+      "loss": 0.6141,
+      "step": 12528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4828470045708207,
+      "learning_rate": 3.841886351911195e-06,
+      "loss": 0.227,
+      "step": 12529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30225508543561413,
+      "learning_rate": 3.840420257373565e-06,
+      "loss": 0.245,
+      "step": 12530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29599291538165057,
+      "learning_rate": 3.83895437614192e-06,
+      "loss": 0.246,
+      "step": 12531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1752706306155982,
+      "learning_rate": 3.837488708267021e-06,
+      "loss": 0.3813,
+      "step": 12532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43554275734913983,
+      "learning_rate": 3.836023253799621e-06,
+      "loss": 0.2608,
+      "step": 12533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4608820204490836,
+      "learning_rate": 3.834558012790469e-06,
+      "loss": 0.347,
+      "step": 12534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2756554502113018,
+      "learning_rate": 3.833092985290311e-06,
+      "loss": 0.2264,
+      "step": 12535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39145909647593713,
+      "learning_rate": 3.831628171349877e-06,
+      "loss": 0.2618,
+      "step": 12536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49976557663593313,
+      "learning_rate": 3.8301635710198946e-06,
+      "loss": 0.2346,
+      "step": 12537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4174397744132169,
+      "learning_rate": 3.828699184351079e-06,
+      "loss": 0.2437,
+      "step": 12538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.276496262939729,
+      "learning_rate": 3.8272350113941494e-06,
+      "loss": 0.2565,
+      "step": 12539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5865133389721945,
+      "learning_rate": 3.825771052199805e-06,
+      "loss": 0.4208,
+      "step": 12540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8880953147875467,
+      "learning_rate": 3.824307306818745e-06,
+      "loss": 0.4872,
+      "step": 12541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2268528697619566,
+      "learning_rate": 3.822843775301656e-06,
+      "loss": 0.1535,
+      "step": 12542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29010451586633673,
+      "learning_rate": 3.821380457699217e-06,
+      "loss": 0.2439,
+      "step": 12543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2260624615809157,
+      "learning_rate": 3.819917354062113e-06,
+      "loss": 0.6806,
+      "step": 12544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2824308471763668,
+      "learning_rate": 3.8184544644410026e-06,
+      "loss": 0.2042,
+      "step": 12545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.776619086277665,
+      "learning_rate": 3.816991788886551e-06,
+      "loss": 0.4011,
+      "step": 12546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3530785405930903,
+      "learning_rate": 3.815529327449402e-06,
+      "loss": 0.312,
+      "step": 12547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32265673340370904,
+      "learning_rate": 3.8140670801802114e-06,
+      "loss": 0.1984,
+      "step": 12548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2522603406858445,
+      "learning_rate": 3.8126050471296116e-06,
+      "loss": 0.1534,
+      "step": 12549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3592744062620737,
+      "learning_rate": 3.811143228348233e-06,
+      "loss": 0.3229,
+      "step": 12550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33175709985865126,
+      "learning_rate": 3.809681623886694e-06,
+      "loss": 0.1751,
+      "step": 12551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5046438803543758,
+      "learning_rate": 3.8082202337956187e-06,
+      "loss": 0.3209,
+      "step": 12552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0417354951774573,
+      "learning_rate": 3.80675905812561e-06,
+      "loss": 0.524,
+      "step": 12553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3264891220161526,
+      "learning_rate": 3.805298096927269e-06,
+      "loss": 0.2089,
+      "step": 12554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2294114627721482,
+      "learning_rate": 3.803837350251188e-06,
+      "loss": 0.2131,
+      "step": 12555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2952230032564946,
+      "learning_rate": 3.8023768181479493e-06,
+      "loss": 0.8189,
+      "step": 12556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48200138614488336,
+      "learning_rate": 3.800916500668139e-06,
+      "loss": 0.3078,
+      "step": 12557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2545052188435359,
+      "learning_rate": 3.7994563978623243e-06,
+      "loss": 0.2414,
+      "step": 12558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1399172653767695,
+      "learning_rate": 3.7979965097810667e-06,
+      "loss": 0.5805,
+      "step": 12559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4024592476572722,
+      "learning_rate": 3.7965368364749244e-06,
+      "loss": 0.2655,
+      "step": 12560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2189980848773631,
+      "learning_rate": 3.7950773779944437e-06,
+      "loss": 0.1199,
+      "step": 12561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31760219818640645,
+      "learning_rate": 3.793618134390168e-06,
+      "loss": 0.3045,
+      "step": 12562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3990174933506131,
+      "learning_rate": 3.792159105712625e-06,
+      "loss": 0.2764,
+      "step": 12563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49668825548514706,
+      "learning_rate": 3.7907002920123482e-06,
+      "loss": 0.2711,
+      "step": 12564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2066436895640726,
+      "learning_rate": 3.7892416933398534e-06,
+      "loss": 0.5866,
+      "step": 12565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.325699463496659,
+      "learning_rate": 3.7877833097456527e-06,
+      "loss": 0.259,
+      "step": 12566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2975927184507548,
+      "learning_rate": 3.786325141280248e-06,
+      "loss": 0.2501,
+      "step": 12567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43994328199667015,
+      "learning_rate": 3.7848671879941334e-06,
+      "loss": 0.2036,
+      "step": 12568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6533205368181108,
+      "learning_rate": 3.783409449937804e-06,
+      "loss": 0.3164,
+      "step": 12569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3752405058155906,
+      "learning_rate": 3.7819519271617377e-06,
+      "loss": 0.3023,
+      "step": 12570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3617519561068833,
+      "learning_rate": 3.7804946197164096e-06,
+      "loss": 0.2507,
+      "step": 12571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.594735705756224,
+      "learning_rate": 3.779037527652282e-06,
+      "loss": 0.3228,
+      "step": 12572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2359376307305643,
+      "learning_rate": 3.77758065101982e-06,
+      "loss": 0.1857,
+      "step": 12573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35542427665950116,
+      "learning_rate": 3.7761239898694724e-06,
+      "loss": 0.2582,
+      "step": 12574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5222486191552719,
+      "learning_rate": 3.774667544251683e-06,
+      "loss": 0.309,
+      "step": 12575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41862420835222075,
+      "learning_rate": 3.773211314216887e-06,
+      "loss": 0.3407,
+      "step": 12576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3031113013277529,
+      "learning_rate": 3.7717552998155184e-06,
+      "loss": 0.3212,
+      "step": 12577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3062819204901965,
+      "learning_rate": 3.770299501097995e-06,
+      "loss": 0.2475,
+      "step": 12578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29069925258267515,
+      "learning_rate": 3.768843918114733e-06,
+      "loss": 0.2404,
+      "step": 12579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4692780214015884,
+      "learning_rate": 3.767388550916138e-06,
+      "loss": 0.2927,
+      "step": 12580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3419993754544002,
+      "learning_rate": 3.7659333995526047e-06,
+      "loss": 0.2067,
+      "step": 12581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5372210817123204,
+      "learning_rate": 3.7644784640745346e-06,
+      "loss": 0.3642,
+      "step": 12582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3720534268613179,
+      "learning_rate": 3.763023744532307e-06,
+      "loss": 0.3042,
+      "step": 12583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3256481791308857,
+      "learning_rate": 3.761569240976298e-06,
+      "loss": 0.1504,
+      "step": 12584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4828728938480592,
+      "learning_rate": 3.7601149534568757e-06,
+      "loss": 0.3779,
+      "step": 12585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.313460913744006,
+      "learning_rate": 3.7586608820244076e-06,
+      "loss": 0.3181,
+      "step": 12586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18315300828709036,
+      "learning_rate": 3.7572070267292438e-06,
+      "loss": 0.0854,
+      "step": 12587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33954649548166543,
+      "learning_rate": 3.7557533876217325e-06,
+      "loss": 0.2789,
+      "step": 12588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1214543359616003,
+      "learning_rate": 3.7542999647522094e-06,
+      "loss": 0.5101,
+      "step": 12589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3266141732037362,
+      "learning_rate": 3.7528467581710137e-06,
+      "loss": 0.2917,
+      "step": 12590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3392692953249234,
+      "learning_rate": 3.7513937679284664e-06,
+      "loss": 0.2326,
+      "step": 12591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.704023761433763,
+      "learning_rate": 3.749940994074884e-06,
+      "loss": 0.4103,
+      "step": 12592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3830258897079524,
+      "learning_rate": 3.7484884366605758e-06,
+      "loss": 0.2555,
+      "step": 12593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24571652256923193,
+      "learning_rate": 3.7470360957358442e-06,
+      "loss": 0.2247,
+      "step": 12594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4350436646267506,
+      "learning_rate": 3.7455839713509844e-06,
+      "loss": 0.2822,
+      "step": 12595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5916263926081792,
+      "learning_rate": 3.7441320635562828e-06,
+      "loss": 0.3318,
+      "step": 12596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3253913954279248,
+      "learning_rate": 3.7426803724020143e-06,
+      "loss": 0.2366,
+      "step": 12597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3462154083645354,
+      "learning_rate": 3.7412288979384604e-06,
+      "loss": 0.2957,
+      "step": 12598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36298865887781684,
+      "learning_rate": 3.739777640215879e-06,
+      "loss": 0.2289,
+      "step": 12599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37829378006780723,
+      "learning_rate": 3.7383265992845297e-06,
+      "loss": 0.2398,
+      "step": 12600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4908548396772708,
+      "learning_rate": 3.736875775194657e-06,
+      "loss": 0.2657,
+      "step": 12601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34330907741617966,
+      "learning_rate": 3.7354251679965103e-06,
+      "loss": 0.2757,
+      "step": 12602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3923379452795223,
+      "learning_rate": 3.7339747777403212e-06,
+      "loss": 0.2949,
+      "step": 12603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8516521788185919,
+      "learning_rate": 3.7325246044763164e-06,
+      "loss": 0.2947,
+      "step": 12604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2060849155640518,
+      "learning_rate": 3.7310746482547143e-06,
+      "loss": 0.7756,
+      "step": 12605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2613915751889735,
+      "learning_rate": 3.729624909125724e-06,
+      "loss": 0.2592,
+      "step": 12606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25492878460525664,
+      "learning_rate": 3.7281753871395575e-06,
+      "loss": 0.1748,
+      "step": 12607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6764847935708693,
+      "learning_rate": 3.726726082346408e-06,
+      "loss": 0.3877,
+      "step": 12608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3204321858680907,
+      "learning_rate": 3.725276994796463e-06,
+      "loss": 0.2423,
+      "step": 12609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3596000242651779,
+      "learning_rate": 3.7238281245399032e-06,
+      "loss": 0.2481,
+      "step": 12610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7745774779469131,
+      "learning_rate": 3.72237947162691e-06,
+      "loss": 0.5167,
+      "step": 12611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35232144352690364,
+      "learning_rate": 3.7209310361076445e-06,
+      "loss": 0.2963,
+      "step": 12612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24440360054586446,
+      "learning_rate": 3.719482818032267e-06,
+      "loss": 0.1043,
+      "step": 12613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3260683902895594,
+      "learning_rate": 3.7180348174509275e-06,
+      "loss": 0.2785,
+      "step": 12614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34981957635207345,
+      "learning_rate": 3.7165870344137746e-06,
+      "loss": 0.2824,
+      "step": 12615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.22822645248553,
+      "learning_rate": 3.715139468970942e-06,
+      "loss": 0.6816,
+      "step": 12616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3974630087585372,
+      "learning_rate": 3.7136921211725595e-06,
+      "loss": 0.2532,
+      "step": 12617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34005021932049256,
+      "learning_rate": 3.7122449910687495e-06,
+      "loss": 0.2688,
+      "step": 12618
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3135211846099563,
+      "learning_rate": 3.710798078709621e-06,
+      "loss": 0.2613,
+      "step": 12619
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3350723208691652,
+      "learning_rate": 3.7093513841452876e-06,
+      "loss": 0.1817,
+      "step": 12620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3935505259433406,
+      "learning_rate": 3.7079049074258465e-06,
+      "loss": 0.3123,
+      "step": 12621
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36173769233262987,
+      "learning_rate": 3.7064586486013865e-06,
+      "loss": 0.3023,
+      "step": 12622
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1886936765534302,
+      "learning_rate": 3.7050126077219908e-06,
+      "loss": 0.4164,
+      "step": 12623
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31195240262411766,
+      "learning_rate": 3.70356678483774e-06,
+      "loss": 0.2534,
+      "step": 12624
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.43851398106112804,
+      "learning_rate": 3.702121179998701e-06,
+      "loss": 0.3313,
+      "step": 12625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3608270575366872,
+      "learning_rate": 3.7006757932549355e-06,
+      "loss": 0.2095,
+      "step": 12626
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2572942190550343,
+      "learning_rate": 3.6992306246564923e-06,
+      "loss": 0.2211,
+      "step": 12627
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2467478374867251,
+      "learning_rate": 3.697785674253428e-06,
+      "loss": 0.6807,
+      "step": 12628
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4778900262033747,
+      "learning_rate": 3.696340942095772e-06,
+      "loss": 0.3873,
+      "step": 12629
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2611211344507252,
+      "learning_rate": 3.6948964282335576e-06,
+      "loss": 0.2184,
+      "step": 12630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6519226464029577,
+      "learning_rate": 3.693452132716806e-06,
+      "loss": 0.3749,
+      "step": 12631
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37288735997878936,
+      "learning_rate": 3.6920080555955396e-06,
+      "loss": 0.2098,
+      "step": 12632
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2948673783502911,
+      "learning_rate": 3.6905641969197626e-06,
+      "loss": 0.1943,
+      "step": 12633
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34946669359400256,
+      "learning_rate": 3.689120556739475e-06,
+      "loss": 0.3015,
+      "step": 12634
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.262967952928276,
+      "learning_rate": 3.687677135104669e-06,
+      "loss": 0.7467,
+      "step": 12635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2994023778760764,
+      "learning_rate": 3.6862339320653353e-06,
+      "loss": 0.1952,
+      "step": 12636
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6864311781365865,
+      "learning_rate": 3.6847909476714495e-06,
+      "loss": 0.3597,
+      "step": 12637
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2598362738797895,
+      "learning_rate": 3.683348181972981e-06,
+      "loss": 0.2505,
+      "step": 12638
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3127138752691672,
+      "learning_rate": 3.68190563501989e-06,
+      "loss": 0.1789,
+      "step": 12639
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47588630014745503,
+      "learning_rate": 3.6804633068621388e-06,
+      "loss": 0.3032,
+      "step": 12640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34987437967800145,
+      "learning_rate": 3.6790211975496714e-06,
+      "loss": 0.3093,
+      "step": 12641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3070022736126839,
+      "learning_rate": 3.6775793071324283e-06,
+      "loss": 0.2453,
+      "step": 12642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7542534218968645,
+      "learning_rate": 3.6761376356603385e-06,
+      "loss": 0.3064,
+      "step": 12643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39024087041958594,
+      "learning_rate": 3.674696183183334e-06,
+      "loss": 0.225,
+      "step": 12644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2456900281354834,
+      "learning_rate": 3.6732549497513292e-06,
+      "loss": 0.2106,
+      "step": 12645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34175475930678456,
+      "learning_rate": 3.6718139354142326e-06,
+      "loss": 0.2592,
+      "step": 12646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8023321846699596,
+      "learning_rate": 3.670373140221947e-06,
+      "loss": 0.452,
+      "step": 12647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3495489226786334,
+      "learning_rate": 3.6689325642243643e-06,
+      "loss": 0.2793,
+      "step": 12648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7430728352602514,
+      "learning_rate": 3.6674922074713783e-06,
+      "loss": 0.3452,
+      "step": 12649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3347746450775112,
+      "learning_rate": 3.6660520700128642e-06,
+      "loss": 0.3015,
+      "step": 12650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24771906724986534,
+      "learning_rate": 3.6646121518986954e-06,
+      "loss": 0.2161,
+      "step": 12651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3430615488225278,
+      "learning_rate": 3.6631724531787314e-06,
+      "loss": 0.113,
+      "step": 12652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38006058780149277,
+      "learning_rate": 3.6617329739028373e-06,
+      "loss": 0.2922,
+      "step": 12653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3295860270105801,
+      "learning_rate": 3.660293714120856e-06,
+      "loss": 0.2493,
+      "step": 12654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8069415336157884,
+      "learning_rate": 3.6588546738826325e-06,
+      "loss": 0.4577,
+      "step": 12655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44498482845603343,
+      "learning_rate": 3.6574158532379944e-06,
+      "loss": 0.1469,
+      "step": 12656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.40957331326092905,
+      "learning_rate": 3.6559772522367765e-06,
+      "loss": 0.3075,
+      "step": 12657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24643548781053554,
+      "learning_rate": 3.6545388709287933e-06,
+      "loss": 0.2624,
+      "step": 12658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7845058080505076,
+      "learning_rate": 3.653100709363856e-06,
+      "loss": 0.3092,
+      "step": 12659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38306157889681686,
+      "learning_rate": 3.651662767591768e-06,
+      "loss": 0.2797,
+      "step": 12660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39024767404061234,
+      "learning_rate": 3.650225045662322e-06,
+      "loss": 0.3401,
+      "step": 12661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46207148468081677,
+      "learning_rate": 3.6487875436253173e-06,
+      "loss": 0.2373,
+      "step": 12662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3593077845975231,
+      "learning_rate": 3.6473502615305233e-06,
+      "loss": 0.2678,
+      "step": 12663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3862979083597893,
+      "learning_rate": 3.645913199427713e-06,
+      "loss": 0.2191,
+      "step": 12664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34427342537308164,
+      "learning_rate": 3.6444763573666586e-06,
+      "loss": 0.2569,
+      "step": 12665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3287239171784687,
+      "learning_rate": 3.643039735397115e-06,
+      "loss": 0.2551,
+      "step": 12666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2073091689438713,
+      "learning_rate": 3.6416033335688306e-06,
+      "loss": 0.8047,
+      "step": 12667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.14833529116643,
+      "learning_rate": 3.640167151931547e-06,
+      "loss": 0.4932,
+      "step": 12668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2683170080436344,
+      "learning_rate": 3.6387311905350053e-06,
+      "loss": 0.2103,
+      "step": 12669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27189369694643967,
+      "learning_rate": 3.637295449428928e-06,
+      "loss": 0.2041,
+      "step": 12670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5783715207547905,
+      "learning_rate": 3.6358599286630367e-06,
+      "loss": 0.3914,
+      "step": 12671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2876284884270302,
+      "learning_rate": 3.634424628287041e-06,
+      "loss": 0.191,
+      "step": 12672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47055640385816155,
+      "learning_rate": 3.632989548350645e-06,
+      "loss": 0.3705,
+      "step": 12673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.526309013570079,
+      "learning_rate": 3.631554688903549e-06,
+      "loss": 0.3535,
+      "step": 12674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3751698398541402,
+      "learning_rate": 3.6301200499954416e-06,
+      "loss": 0.1871,
+      "step": 12675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3542442085533244,
+      "learning_rate": 3.6286856316760023e-06,
+      "loss": 0.2631,
+      "step": 12676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.273470471007027,
+      "learning_rate": 3.6272514339949015e-06,
+      "loss": 0.2747,
+      "step": 12677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5734636947485215,
+      "learning_rate": 3.6258174570018133e-06,
+      "loss": 0.3421,
+      "step": 12678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3687365959759761,
+      "learning_rate": 3.6243837007463933e-06,
+      "loss": 0.262,
+      "step": 12679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2296519963082793,
+      "learning_rate": 3.6229501652782904e-06,
+      "loss": 0.5555,
+      "step": 12680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31969153444544296,
+      "learning_rate": 3.6215168506471466e-06,
+      "loss": 0.2874,
+      "step": 12681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4076483909163558,
+      "learning_rate": 3.6200837569026036e-06,
+      "loss": 0.2688,
+      "step": 12682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41300092423729834,
+      "learning_rate": 3.618650884094285e-06,
+      "loss": 0.2668,
+      "step": 12683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3944044972280825,
+      "learning_rate": 3.617218232271812e-06,
+      "loss": 0.2958,
+      "step": 12684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25016697584075415,
+      "learning_rate": 3.615785801484797e-06,
+      "loss": 0.1869,
+      "step": 12685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4950885923845611,
+      "learning_rate": 3.6143535917828422e-06,
+      "loss": 0.3181,
+      "step": 12686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.369686886187919,
+      "learning_rate": 3.612921603215551e-06,
+      "loss": 0.263,
+      "step": 12687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5134965413780258,
+      "learning_rate": 3.6114898358325103e-06,
+      "loss": 0.2762,
+      "step": 12688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3598094168898819,
+      "learning_rate": 3.6100582896833012e-06,
+      "loss": 0.3195,
+      "step": 12689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5903895067372618,
+      "learning_rate": 3.6086269648174965e-06,
+      "loss": 0.3883,
+      "step": 12690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.285502437384372,
+      "learning_rate": 3.607195861284668e-06,
+      "loss": 0.2375,
+      "step": 12691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29724734450311907,
+      "learning_rate": 3.605764979134372e-06,
+      "loss": 0.1829,
+      "step": 12692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38014274107318985,
+      "learning_rate": 3.6043343184161593e-06,
+      "loss": 0.2531,
+      "step": 12693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3741144539894074,
+      "learning_rate": 3.602903879179571e-06,
+      "loss": 0.3121,
+      "step": 12694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9836418738839882,
+      "learning_rate": 3.601473661474154e-06,
+      "loss": 0.3773,
+      "step": 12695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5388543502057055,
+      "learning_rate": 3.600043665349424e-06,
+      "loss": 0.33,
+      "step": 12696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2299485082398649,
+      "learning_rate": 3.5986138908549073e-06,
+      "loss": 0.2319,
+      "step": 12697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4261332556361552,
+      "learning_rate": 3.597184338040114e-06,
+      "loss": 0.166,
+      "step": 12698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4785686820285072,
+      "learning_rate": 3.595755006954553e-06,
+      "loss": 0.3013,
+      "step": 12699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.40973404733494684,
+      "learning_rate": 3.5943258976477226e-06,
+      "loss": 0.2931,
+      "step": 12700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32627133475863995,
+      "learning_rate": 3.5928970101691096e-06,
+      "loss": 0.2481,
+      "step": 12701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5170098162809887,
+      "learning_rate": 3.5914683445681954e-06,
+      "loss": 0.3466,
+      "step": 12702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.287299691816571,
+      "learning_rate": 3.59003990089446e-06,
+      "loss": 0.2333,
+      "step": 12703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3970243877751687,
+      "learning_rate": 3.588611679197366e-06,
+      "loss": 0.2657,
+      "step": 12704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2731068096527432,
+      "learning_rate": 3.587183679526375e-06,
+      "loss": 0.2241,
+      "step": 12705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5270425973716926,
+      "learning_rate": 3.585755901930934e-06,
+      "loss": 0.3451,
+      "step": 12706
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2638374003320183,
+      "learning_rate": 3.5843283464604927e-06,
+      "loss": 0.7912,
+      "step": 12707
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4955313922411486,
+      "learning_rate": 3.582901013164486e-06,
+      "loss": 0.1773,
+      "step": 12708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25806863051136353,
+      "learning_rate": 3.5814739020923405e-06,
+      "loss": 0.2594,
+      "step": 12709
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2920727776909075,
+      "learning_rate": 3.5800470132934785e-06,
+      "loss": 0.2262,
+      "step": 12710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7188461489974537,
+      "learning_rate": 3.5786203468173087e-06,
+      "loss": 0.2492,
+      "step": 12711
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3632276005566369,
+      "learning_rate": 3.5771939027132428e-06,
+      "loss": 0.2854,
+      "step": 12712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3989870298151256,
+      "learning_rate": 3.5757676810306775e-06,
+      "loss": 0.3184,
+      "step": 12713
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3194280116657528,
+      "learning_rate": 3.5743416818189993e-06,
+      "loss": 0.1514,
+      "step": 12714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.353739585397453,
+      "learning_rate": 3.5729159051275895e-06,
+      "loss": 0.2931,
+      "step": 12715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6406462606108349,
+      "learning_rate": 3.5714903510058296e-06,
+      "loss": 0.3994,
+      "step": 12716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2038712219236113,
+      "learning_rate": 3.570065019503082e-06,
+      "loss": 0.1946,
+      "step": 12717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3350444004903638,
+      "learning_rate": 3.5686399106687064e-06,
+      "loss": 0.2076,
+      "step": 12718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.097246714693747,
+      "learning_rate": 3.567215024552051e-06,
+      "loss": 0.6239,
+      "step": 12719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46965475632734666,
+      "learning_rate": 3.5657903612024658e-06,
+      "loss": 0.3749,
+      "step": 12720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.263875280635042,
+      "learning_rate": 3.5643659206692837e-06,
+      "loss": 0.2184,
+      "step": 12721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5866450994990379,
+      "learning_rate": 3.562941703001832e-06,
+      "loss": 0.3771,
+      "step": 12722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19861196590825053,
+      "learning_rate": 3.5615177082494334e-06,
+      "loss": 0.1384,
+      "step": 12723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3572080133883536,
+      "learning_rate": 3.5600939364613963e-06,
+      "loss": 0.217,
+      "step": 12724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41415889915049825,
+      "learning_rate": 3.5586703876870333e-06,
+      "loss": 0.3052,
+      "step": 12725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5154925408780823,
+      "learning_rate": 3.557247061975636e-06,
+      "loss": 0.3153,
+      "step": 12726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34276841548401615,
+      "learning_rate": 3.5558239593764978e-06,
+      "loss": 0.2438,
+      "step": 12727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42795797835213084,
+      "learning_rate": 3.554401079938894e-06,
+      "loss": 0.329,
+      "step": 12728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.234860424746158,
+      "learning_rate": 3.552978423712111e-06,
+      "loss": 0.1943,
+      "step": 12729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36464014173959514,
+      "learning_rate": 3.5515559907454045e-06,
+      "loss": 0.2748,
+      "step": 12730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9060753945623665,
+      "learning_rate": 3.550133781088033e-06,
+      "loss": 0.5051,
+      "step": 12731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3529980523946801,
+      "learning_rate": 3.5487117947892558e-06,
+      "loss": 0.2867,
+      "step": 12732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3674842048475259,
+      "learning_rate": 3.5472900318983105e-06,
+      "loss": 0.2548,
+      "step": 12733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8401740489172301,
+      "learning_rate": 3.545868492464435e-06,
+      "loss": 0.3047,
+      "step": 12734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27141746267388867,
+      "learning_rate": 3.544447176536855e-06,
+      "loss": 0.1553,
+      "step": 12735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32057223006156993,
+      "learning_rate": 3.543026084164789e-06,
+      "loss": 0.2634,
+      "step": 12736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3573243078189275,
+      "learning_rate": 3.5416052153974546e-06,
+      "loss": 0.2645,
+      "step": 12737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6295162887590943,
+      "learning_rate": 3.5401845702840543e-06,
+      "loss": 0.4241,
+      "step": 12738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38271116549580053,
+      "learning_rate": 3.5387641488737855e-06,
+      "loss": 0.2994,
+      "step": 12739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44089101642573486,
+      "learning_rate": 3.5373439512158315e-06,
+      "loss": 0.2251,
+      "step": 12740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2504442391142626,
+      "learning_rate": 3.5359239773593833e-06,
+      "loss": 0.2017,
+      "step": 12741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3315826635141327,
+      "learning_rate": 3.534504227353609e-06,
+      "loss": 0.2718,
+      "step": 12742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0141255217066432,
+      "learning_rate": 3.5330847012476754e-06,
+      "loss": 0.6677,
+      "step": 12743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2908904563347626,
+      "learning_rate": 3.5316653990907367e-06,
+      "loss": 0.2313,
+      "step": 12744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35460691259874044,
+      "learning_rate": 3.5302463209319514e-06,
+      "loss": 0.317,
+      "step": 12745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6951267542723861,
+      "learning_rate": 3.5288274668204568e-06,
+      "loss": 0.4621,
+      "step": 12746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.15127146330290486,
+      "learning_rate": 3.527408836805389e-06,
+      "loss": 0.0716,
+      "step": 12747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.297568465211192,
+      "learning_rate": 3.525990430935876e-06,
+      "loss": 0.2372,
+      "step": 12748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.321758839001798,
+      "learning_rate": 3.524572249261031e-06,
+      "loss": 0.3338,
+      "step": 12749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5864810260206559,
+      "learning_rate": 3.5231542918299753e-06,
+      "loss": 0.3112,
+      "step": 12750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2883308570588384,
+      "learning_rate": 3.5217365586918073e-06,
+      "loss": 0.2406,
+      "step": 12751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47306256806256525,
+      "learning_rate": 3.5203190498956242e-06,
+      "loss": 0.3095,
+      "step": 12752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49735107729836175,
+      "learning_rate": 3.518901765490509e-06,
+      "loss": 0.347,
+      "step": 12753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17686563294193186,
+      "learning_rate": 3.517484705525551e-06,
+      "loss": 0.1394,
+      "step": 12754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8933579186696481,
+      "learning_rate": 3.5160678700498197e-06,
+      "loss": 0.5273,
+      "step": 12755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3495559430890266,
+      "learning_rate": 3.5146512591123783e-06,
+      "loss": 0.3032,
+      "step": 12756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28892578502445293,
+      "learning_rate": 3.513234872762282e-06,
+      "loss": 0.1972,
+      "step": 12757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0657258271815264,
+      "learning_rate": 3.511818711048587e-06,
+      "loss": 0.5283,
+      "step": 12758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2885840837702499,
+      "learning_rate": 3.5104027740203305e-06,
+      "loss": 0.1883,
+      "step": 12759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2869873138895025,
+      "learning_rate": 3.5089870617265465e-06,
+      "loss": 0.1858,
+      "step": 12760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.352194227970117,
+      "learning_rate": 3.5075715742162586e-06,
+      "loss": 0.3088,
+      "step": 12761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7226569774131096,
+      "learning_rate": 3.506156311538491e-06,
+      "loss": 0.4724,
+      "step": 12762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29774972113115467,
+      "learning_rate": 3.504741273742254e-06,
+      "loss": 0.194,
+      "step": 12763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.33947635583751484,
+      "learning_rate": 3.503326460876545e-06,
+      "loss": 0.2951,
+      "step": 12764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1678877911120926,
+      "learning_rate": 3.5019118729903566e-06,
+      "loss": 0.4789,
+      "step": 12765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3045109971875889,
+      "learning_rate": 3.5004975101326854e-06,
+      "loss": 0.2444,
+      "step": 12766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5454583862261513,
+      "learning_rate": 3.4990833723525054e-06,
+      "loss": 0.2266,
+      "step": 12767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3329777623579333,
+      "learning_rate": 3.497669459698788e-06,
+      "loss": 0.3025,
+      "step": 12768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3145139009440519,
+      "learning_rate": 3.496255772220495e-06,
+      "loss": 0.2587,
+      "step": 12769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23984063415861187,
+      "learning_rate": 3.4948423099665883e-06,
+      "loss": 0.1022,
+      "step": 12770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8220955217574739,
+      "learning_rate": 3.493429072986013e-06,
+      "loss": 0.402,
+      "step": 12771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2833979664828243,
+      "learning_rate": 3.492016061327709e-06,
+      "loss": 0.249,
+      "step": 12772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4655948518759838,
+      "learning_rate": 3.490603275040605e-06,
+      "loss": 0.2649,
+      "step": 12773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39936457950351845,
+      "learning_rate": 3.4891907141736324e-06,
+      "loss": 0.289,
+      "step": 12774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2320835341969602,
+      "learning_rate": 3.487778378775707e-06,
+      "loss": 0.1958,
+      "step": 12775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3366992622728923,
+      "learning_rate": 3.4863662688957355e-06,
+      "loss": 0.2385,
+      "step": 12776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8735030497179362,
+      "learning_rate": 3.4849543845826195e-06,
+      "loss": 0.373,
+      "step": 12777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3820901509586576,
+      "learning_rate": 3.4835427258852507e-06,
+      "loss": 0.2542,
+      "step": 12778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4889596251920278,
+      "learning_rate": 3.4821312928525197e-06,
+      "loss": 0.3594,
+      "step": 12779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3395884327853301,
+      "learning_rate": 3.4807200855333024e-06,
+      "loss": 0.2753,
+      "step": 12780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41027269658708754,
+      "learning_rate": 3.479309103976467e-06,
+      "loss": 0.2878,
+      "step": 12781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2492148324473185,
+      "learning_rate": 3.4778983482308746e-06,
+      "loss": 0.1902,
+      "step": 12782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7729463052691775,
+      "learning_rate": 3.4764878183453855e-06,
+      "loss": 0.2883,
+      "step": 12783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2636022559064381,
+      "learning_rate": 3.475077514368842e-06,
+      "loss": 0.2694,
+      "step": 12784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4722767331960266,
+      "learning_rate": 3.4736674363500846e-06,
+      "loss": 0.3413,
+      "step": 12785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9849108722772745,
+      "learning_rate": 3.472257584337939e-06,
+      "loss": 0.527,
+      "step": 12786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2514472326522464,
+      "learning_rate": 3.470847958381236e-06,
+      "loss": 0.2084,
+      "step": 12787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2962379312373279,
+      "learning_rate": 3.469438558528787e-06,
+      "loss": 0.2507,
+      "step": 12788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7197590638637732,
+      "learning_rate": 3.468029384829401e-06,
+      "loss": 0.2857,
+      "step": 12789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34947417036129547,
+      "learning_rate": 3.466620437331876e-06,
+      "loss": 0.2543,
+      "step": 12790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3024804970661898,
+      "learning_rate": 3.4652117160850006e-06,
+      "loss": 0.5634,
+      "step": 12791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35195261144446016,
+      "learning_rate": 3.463803221137566e-06,
+      "loss": 0.2948,
+      "step": 12792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3084578488343683,
+      "learning_rate": 3.462394952538345e-06,
+      "loss": 0.1994,
+      "step": 12793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22815859921668444,
+      "learning_rate": 3.460986910336106e-06,
+      "loss": 0.1535,
+      "step": 12794
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7811335529012565,
+      "learning_rate": 3.459579094579605e-06,
+      "loss": 0.3796,
+      "step": 12795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27942275007513806,
+      "learning_rate": 3.4581715053176023e-06,
+      "loss": 0.1953,
+      "step": 12796
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3920486415648585,
+      "learning_rate": 3.456764142598843e-06,
+      "loss": 0.3051,
+      "step": 12797
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.016852076671105,
+      "learning_rate": 3.455357006472052e-06,
+      "loss": 0.657,
+      "step": 12798
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33966130409641193,
+      "learning_rate": 3.4539500969859706e-06,
+      "loss": 0.1705,
+      "step": 12799
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22172259172292844,
+      "learning_rate": 3.4525434141893166e-06,
+      "loss": 0.2172,
+      "step": 12800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7058709583737947,
+      "learning_rate": 3.4511369581308017e-06,
+      "loss": 0.3947,
+      "step": 12801
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3354166197148681,
+      "learning_rate": 3.449730728859132e-06,
+      "loss": 0.1666,
+      "step": 12802
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5337470482155633,
+      "learning_rate": 3.4483247264230034e-06,
+      "loss": 0.3789,
+      "step": 12803
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33592213852392117,
+      "learning_rate": 3.4469189508711098e-06,
+      "loss": 0.3046,
+      "step": 12804
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6246678196271466,
+      "learning_rate": 3.445513402252132e-06,
+      "loss": 0.3147,
+      "step": 12805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31427524782070476,
+      "learning_rate": 3.444108080614743e-06,
+      "loss": 0.204,
+      "step": 12806
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48310103519188197,
+      "learning_rate": 3.4427029860076056e-06,
+      "loss": 0.2649,
+      "step": 12807
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26865324112297956,
+      "learning_rate": 3.441298118479386e-06,
+      "loss": 0.2608,
+      "step": 12808
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5907160387384897,
+      "learning_rate": 3.4398934780787297e-06,
+      "loss": 0.2304,
+      "step": 12809
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1524510092382603,
+      "learning_rate": 3.43848906485428e-06,
+      "loss": 0.7557,
+      "step": 12810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38459895702479086,
+      "learning_rate": 3.4370848788546695e-06,
+      "loss": 0.2584,
+      "step": 12811
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2760382853690984,
+      "learning_rate": 3.4356809201285303e-06,
+      "loss": 0.241,
+      "step": 12812
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32014199369104235,
+      "learning_rate": 3.4342771887244784e-06,
+      "loss": 0.2341,
+      "step": 12813
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5749750430673263,
+      "learning_rate": 3.4328736846911247e-06,
+      "loss": 0.3045,
+      "step": 12814
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38232768236464837,
+      "learning_rate": 3.4314704080770744e-06,
+      "loss": 0.2291,
+      "step": 12815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3446190128141687,
+      "learning_rate": 3.4300673589309163e-06,
+      "loss": 0.3249,
+      "step": 12816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5823714362679814,
+      "learning_rate": 3.428664537301247e-06,
+      "loss": 0.2962,
+      "step": 12817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3685845728880524,
+      "learning_rate": 3.4272619432366427e-06,
+      "loss": 0.3151,
+      "step": 12818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2341879964186778,
+      "learning_rate": 3.425859576785674e-06,
+      "loss": 0.1659,
+      "step": 12819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4030323576618036,
+      "learning_rate": 3.4244574379969032e-06,
+      "loss": 0.2683,
+      "step": 12820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4474160431672498,
+      "learning_rate": 3.4230555269188903e-06,
+      "loss": 0.2913,
+      "step": 12821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0297083176155641,
+      "learning_rate": 3.4216538436001836e-06,
+      "loss": 0.5263,
+      "step": 12822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38588698705698504,
+      "learning_rate": 3.4202523880893202e-06,
+      "loss": 0.2781,
+      "step": 12823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30214966289449274,
+      "learning_rate": 3.4188511604348297e-06,
+      "loss": 0.2844,
+      "step": 12824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21761652521172148,
+      "learning_rate": 3.417450160685245e-06,
+      "loss": 0.0645,
+      "step": 12825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32300800479962954,
+      "learning_rate": 3.416049388889078e-06,
+      "loss": 0.1809,
+      "step": 12826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3752187308252932,
+      "learning_rate": 3.4146488450948367e-06,
+      "loss": 0.2808,
+      "step": 12827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3238753458144825,
+      "learning_rate": 3.413248529351023e-06,
+      "loss": 0.2652,
+      "step": 12828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4298149619699671,
+      "learning_rate": 3.411848441706127e-06,
+      "loss": 0.2901,
+      "step": 12829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4810941246561356,
+      "learning_rate": 3.410448582208642e-06,
+      "loss": 0.3489,
+      "step": 12830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5252502309396361,
+      "learning_rate": 3.409048950907037e-06,
+      "loss": 0.3361,
+      "step": 12831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1994109208173046,
+      "learning_rate": 3.4076495478497795e-06,
+      "loss": 0.1346,
+      "step": 12832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4101863021105691,
+      "learning_rate": 3.406250373085337e-06,
+      "loss": 0.3074,
+      "step": 12833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6969992266110001,
+      "learning_rate": 3.4048514266621612e-06,
+      "loss": 0.4448,
+      "step": 12834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4233808236326501,
+      "learning_rate": 3.403452708628697e-06,
+      "loss": 0.238,
+      "step": 12835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29572459303480436,
+      "learning_rate": 3.4020542190333795e-06,
+      "loss": 0.2596,
+      "step": 12836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49815984479914355,
+      "learning_rate": 3.4006559579246425e-06,
+      "loss": 0.2174,
+      "step": 12837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2761033725607673,
+      "learning_rate": 3.3992579253509062e-06,
+      "loss": 0.0935,
+      "step": 12838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32754610459515865,
+      "learning_rate": 3.3978601213605842e-06,
+      "loss": 0.2857,
+      "step": 12839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4294008648810787,
+      "learning_rate": 3.3964625460020827e-06,
+      "loss": 0.311,
+      "step": 12840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6594774691330908,
+      "learning_rate": 3.395065199323796e-06,
+      "loss": 0.3494,
+      "step": 12841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32425053227691586,
+      "learning_rate": 3.393668081374121e-06,
+      "loss": 0.2053,
+      "step": 12842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5089151987838674,
+      "learning_rate": 3.3922711922014352e-06,
+      "loss": 0.3626,
+      "step": 12843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21503853351438224,
+      "learning_rate": 3.3908745318541146e-06,
+      "loss": 0.1865,
+      "step": 12844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31100684636950715,
+      "learning_rate": 3.389478100380521e-06,
+      "loss": 0.2,
+      "step": 12845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7679876511139743,
+      "learning_rate": 3.3880818978290196e-06,
+      "loss": 0.3937,
+      "step": 12846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32560271064088603,
+      "learning_rate": 3.386685924247959e-06,
+      "loss": 0.2838,
+      "step": 12847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32516907110621057,
+      "learning_rate": 3.3852901796856796e-06,
+      "loss": 0.21,
+      "step": 12848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2717231476692796,
+      "learning_rate": 3.3838946641905134e-06,
+      "loss": 0.4662,
+      "step": 12849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2335607078138597,
+      "learning_rate": 3.382499377810794e-06,
+      "loss": 0.1513,
+      "step": 12850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2814240890847466,
+      "learning_rate": 3.3811043205948366e-06,
+      "loss": 0.2342,
+      "step": 12851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5990969852514557,
+      "learning_rate": 3.3797094925909526e-06,
+      "loss": 0.3364,
+      "step": 12852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1668272839736702,
+      "learning_rate": 3.378314893847443e-06,
+      "loss": 0.7382,
+      "step": 12853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3361253693355029,
+      "learning_rate": 3.3769205244126013e-06,
+      "loss": 0.2619,
+      "step": 12854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.357664507868556,
+      "learning_rate": 3.3755263843347196e-06,
+      "loss": 0.2275,
+      "step": 12855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26691235584553286,
+      "learning_rate": 3.3741324736620752e-06,
+      "loss": 0.1768,
+      "step": 12856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3316452448222894,
+      "learning_rate": 3.3727387924429377e-06,
+      "loss": 0.2648,
+      "step": 12857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.095159106391374,
+      "learning_rate": 3.371345340725568e-06,
+      "loss": 0.2964,
+      "step": 12858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29582495886406085,
+      "learning_rate": 3.3699521185582274e-06,
+      "loss": 0.2747,
+      "step": 12859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34966377054219355,
+      "learning_rate": 3.3685591259891592e-06,
+      "loss": 0.2708,
+      "step": 12860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3699008142557696,
+      "learning_rate": 3.367166363066604e-06,
+      "loss": 0.2283,
+      "step": 12861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37940865289293624,
+      "learning_rate": 3.3657738298387886e-06,
+      "loss": 0.2389,
+      "step": 12862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29870327609057534,
+      "learning_rate": 3.3643815263539438e-06,
+      "loss": 0.2815,
+      "step": 12863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6121185536613202,
+      "learning_rate": 3.3629894526602847e-06,
+      "loss": 0.2784,
+      "step": 12864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39370032195369115,
+      "learning_rate": 3.361597608806012e-06,
+      "loss": 0.2998,
+      "step": 12865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3425273598020543,
+      "learning_rate": 3.360205994839326e-06,
+      "loss": 0.2651,
+      "step": 12866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38555273129152684,
+      "learning_rate": 3.358814610808424e-06,
+      "loss": 0.3122,
+      "step": 12867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6651039322509703,
+      "learning_rate": 3.3574234567614862e-06,
+      "loss": 0.2088,
+      "step": 12868
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3512050670033464,
+      "learning_rate": 3.356032532746688e-06,
+      "loss": 0.2761,
+      "step": 12869
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5160563391559385,
+      "learning_rate": 3.354641838812195e-06,
+      "loss": 0.3093,
+      "step": 12870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2944111295460084,
+      "learning_rate": 3.353251375006171e-06,
+      "loss": 0.2444,
+      "step": 12871
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23733746897925087,
+      "learning_rate": 3.3518611413767675e-06,
+      "loss": 0.2007,
+      "step": 12872
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2606823606756283,
+      "learning_rate": 3.3504711379721267e-06,
+      "loss": 0.4898,
+      "step": 12873
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8636552913098463,
+      "learning_rate": 3.3490813648403808e-06,
+      "loss": 0.3021,
+      "step": 12874
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2671159928686584,
+      "learning_rate": 3.347691822029665e-06,
+      "loss": 0.2538,
+      "step": 12875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5173905617386279,
+      "learning_rate": 3.346302509588095e-06,
+      "loss": 0.3694,
+      "step": 12876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17909842675389065,
+      "learning_rate": 3.344913427563784e-06,
+      "loss": 0.1285,
+      "step": 12877
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30187736747915783,
+      "learning_rate": 3.343524576004833e-06,
+      "loss": 0.2529,
+      "step": 12878
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34252991641349273,
+      "learning_rate": 3.342135954959338e-06,
+      "loss": 0.2783,
+      "step": 12879
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9488231871291415,
+      "learning_rate": 3.3407475644753907e-06,
+      "loss": 0.3745,
+      "step": 12880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29818740197209903,
+      "learning_rate": 3.3393594046010693e-06,
+      "loss": 0.208,
+      "step": 12881
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.210855898826414,
+      "learning_rate": 3.3379714753844463e-06,
+      "loss": 0.633,
+      "step": 12882
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25512255050159116,
+      "learning_rate": 3.3365837768735798e-06,
+      "loss": 0.2485,
+      "step": 12883
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29905505580538666,
+      "learning_rate": 3.335196309116534e-06,
+      "loss": 0.1868,
+      "step": 12884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5038307443785937,
+      "learning_rate": 3.3338090721613547e-06,
+      "loss": 0.2664,
+      "step": 12885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5952356862239482,
+      "learning_rate": 3.332422066056079e-06,
+      "loss": 0.3309,
+      "step": 12886
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2904352494406116,
+      "learning_rate": 3.3310352908487387e-06,
+      "loss": 0.1837,
+      "step": 12887
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5566544726291275,
+      "learning_rate": 3.3296487465873617e-06,
+      "loss": 0.3727,
+      "step": 12888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.338582329143749,
+      "learning_rate": 3.328262433319962e-06,
+      "loss": 0.7545,
+      "step": 12889
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18720255509199032,
+      "learning_rate": 3.3268763510945477e-06,
+      "loss": 0.1371,
+      "step": 12890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3551963786547063,
+      "learning_rate": 3.325490499959114e-06,
+      "loss": 0.2981,
+      "step": 12891
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6958333900137992,
+      "learning_rate": 3.3241048799616616e-06,
+      "loss": 0.4138,
+      "step": 12892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3327651923692982,
+      "learning_rate": 3.3227194911501705e-06,
+      "loss": 0.267,
+      "step": 12893
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2262784695854978,
+      "learning_rate": 3.3213343335726157e-06,
+      "loss": 0.3285,
+      "step": 12894
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.35721902129772304,
+      "learning_rate": 3.3199494072769657e-06,
+      "loss": 0.3184,
+      "step": 12895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34515026641036906,
+      "learning_rate": 3.3185647123111776e-06,
+      "loss": 0.2565,
+      "step": 12896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4102070107736664,
+      "learning_rate": 3.3171802487232087e-06,
+      "loss": 0.1409,
+      "step": 12897
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6009426131881008,
+      "learning_rate": 3.3157960165610035e-06,
+      "loss": 0.333,
+      "step": 12898
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26019221615739213,
+      "learning_rate": 3.314412015872489e-06,
+      "loss": 0.253,
+      "step": 12899
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.239600152449063,
+      "learning_rate": 3.313028246705603e-06,
+      "loss": 0.3466,
+      "step": 12900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4657976696878919,
+      "learning_rate": 3.3116447091082593e-06,
+      "loss": 0.3098,
+      "step": 12901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3989409195000362,
+      "learning_rate": 3.310261403128373e-06,
+      "loss": 0.2568,
+      "step": 12902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28897844817397594,
+      "learning_rate": 3.3088783288138436e-06,
+      "loss": 0.2548,
+      "step": 12903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4652505440048004,
+      "learning_rate": 3.307495486212572e-06,
+      "loss": 0.2714,
+      "step": 12904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.388391044884601,
+      "learning_rate": 3.306112875372445e-06,
+      "loss": 0.2605,
+      "step": 12905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5730762996066526,
+      "learning_rate": 3.3047304963413407e-06,
+      "loss": 0.3871,
+      "step": 12906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3801978509249738,
+      "learning_rate": 3.3033483491671316e-06,
+      "loss": 0.2487,
+      "step": 12907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4299270873716904,
+      "learning_rate": 3.3019664338976787e-06,
+      "loss": 0.2827,
+      "step": 12908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.463795814167745,
+      "learning_rate": 3.300584750580842e-06,
+      "loss": 0.3093,
+      "step": 12909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2269061260272887,
+      "learning_rate": 3.2992032992644686e-06,
+      "loss": 0.1444,
+      "step": 12910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2822473309634225,
+      "learning_rate": 3.2978220799963955e-06,
+      "loss": 0.2462,
+      "step": 12911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.558699342627396,
+      "learning_rate": 3.2964410928244526e-06,
+      "loss": 0.739,
+      "step": 12912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8887185637630161,
+      "learning_rate": 3.2950603377964706e-06,
+      "loss": 0.3034,
+      "step": 12913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3173946544253534,
+      "learning_rate": 3.29367981496026e-06,
+      "loss": 0.2621,
+      "step": 12914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3141577245945441,
+      "learning_rate": 3.29229952436363e-06,
+      "loss": 0.2978,
+      "step": 12915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3278313577965663,
+      "learning_rate": 3.2909194660543742e-06,
+      "loss": 0.1864,
+      "step": 12916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37208100190566185,
+      "learning_rate": 3.289539640080294e-06,
+      "loss": 0.1933,
+      "step": 12917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3163936504011386,
+      "learning_rate": 3.288160046489166e-06,
+      "loss": 0.8189,
+      "step": 12918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3683715557013782,
+      "learning_rate": 3.2867806853287675e-06,
+      "loss": 0.2944,
+      "step": 12919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37802142040575226,
+      "learning_rate": 3.2854015566468643e-06,
+      "loss": 0.1896,
+      "step": 12920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6399011890671984,
+      "learning_rate": 3.284022660491214e-06,
+      "loss": 0.3545,
+      "step": 12921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22873817154729156,
+      "learning_rate": 3.2826439969095737e-06,
+      "loss": 0.2049,
+      "step": 12922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3336483088508467,
+      "learning_rate": 3.281265565949683e-06,
+      "loss": 0.1561,
+      "step": 12923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5578981476572488,
+      "learning_rate": 3.2798873676592755e-06,
+      "loss": 0.4109,
+      "step": 12924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6593018831054235,
+      "learning_rate": 3.2785094020860777e-06,
+      "loss": 0.3806,
+      "step": 12925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36905433775237373,
+      "learning_rate": 3.277131669277813e-06,
+      "loss": 0.2017,
+      "step": 12926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3000027094105889,
+      "learning_rate": 3.275754169282189e-06,
+      "loss": 0.2773,
+      "step": 12927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42583234394958763,
+      "learning_rate": 3.2743769021469096e-06,
+      "loss": 0.2012,
+      "step": 12928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2963137515632942,
+      "learning_rate": 3.2729998679196663e-06,
+      "loss": 0.1866,
+      "step": 12929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2859908008194674,
+      "learning_rate": 3.2716230666481506e-06,
+      "loss": 0.2344,
+      "step": 12930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.458580153617439,
+      "learning_rate": 3.2702464983800386e-06,
+      "loss": 0.3144,
+      "step": 12931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3739316025400197,
+      "learning_rate": 3.2688701631630047e-06,
+      "loss": 0.2944,
+      "step": 12932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5003894526988127,
+      "learning_rate": 3.2674940610447005e-06,
+      "loss": 0.2209,
+      "step": 12933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2815996714466963,
+      "learning_rate": 3.2661181920727913e-06,
+      "loss": 0.2559,
+      "step": 12934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31949144720216227,
+      "learning_rate": 3.2647425562949196e-06,
+      "loss": 0.2264,
+      "step": 12935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43227858764097016,
+      "learning_rate": 3.263367153758723e-06,
+      "loss": 0.2649,
+      "step": 12936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6654739405314112,
+      "learning_rate": 3.26199198451183e-06,
+      "loss": 0.3732,
+      "step": 12937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4519785519202604,
+      "learning_rate": 3.2606170486018662e-06,
+      "loss": 0.312,
+      "step": 12938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29017049811434276,
+      "learning_rate": 3.2592423460764457e-06,
+      "loss": 0.2275,
+      "step": 12939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36334147109219417,
+      "learning_rate": 3.257867876983173e-06,
+      "loss": 0.1923,
+      "step": 12940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5450956167249345,
+      "learning_rate": 3.256493641369641e-06,
+      "loss": 0.315,
+      "step": 12941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4020490717875685,
+      "learning_rate": 3.2551196392834496e-06,
+      "loss": 0.3194,
+      "step": 12942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32817468224538004,
+      "learning_rate": 3.2537458707721735e-06,
+      "loss": 0.2755,
+      "step": 12943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5816663916046452,
+      "learning_rate": 3.252372335883388e-06,
+      "loss": 0.3265,
+      "step": 12944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3815486034011904,
+      "learning_rate": 3.250999034664659e-06,
+      "loss": 0.2811,
+      "step": 12945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3526554552274751,
+      "learning_rate": 3.24962596716354e-06,
+      "loss": 0.2452,
+      "step": 12946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24509700046609045,
+      "learning_rate": 3.2482531334275856e-06,
+      "loss": 0.1689,
+      "step": 12947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3338020441543314,
+      "learning_rate": 3.2468805335043363e-06,
+      "loss": 0.3091,
+      "step": 12948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9020955309045062,
+      "learning_rate": 3.2455081674413226e-06,
+      "loss": 0.3683,
+      "step": 12949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3050120934170287,
+      "learning_rate": 3.2441360352860675e-06,
+      "loss": 0.2703,
+      "step": 12950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.41887257623271296,
+      "learning_rate": 3.2427641370860953e-06,
+      "loss": 0.273,
+      "step": 12951
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42530719140404794,
+      "learning_rate": 3.241392472888909e-06,
+      "loss": 0.1314,
+      "step": 12952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38376438245705596,
+      "learning_rate": 3.240021042742012e-06,
+      "loss": 0.2803,
+      "step": 12953
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47996045275196836,
+      "learning_rate": 3.2386498466928916e-06,
+      "loss": 0.3495,
+      "step": 12954
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3492844356549713,
+      "learning_rate": 3.237278884789039e-06,
+      "loss": 0.3225,
+      "step": 12955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21621202245976492,
+      "learning_rate": 3.235908157077929e-06,
+      "loss": 0.0953,
+      "step": 12956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37093987547274837,
+      "learning_rate": 3.234537663607028e-06,
+      "loss": 0.2762,
+      "step": 12957
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33780033787498953,
+      "learning_rate": 3.233167404423797e-06,
+      "loss": 0.2939,
+      "step": 12958
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8582542909743728,
+      "learning_rate": 3.231797379575684e-06,
+      "loss": 0.2335,
+      "step": 12959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34765139094691233,
+      "learning_rate": 3.230427589110141e-06,
+      "loss": 0.2791,
+      "step": 12960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1022929147980127,
+      "learning_rate": 3.229058033074599e-06,
+      "loss": 0.7693,
+      "step": 12961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2048980960713727,
+      "learning_rate": 3.227688711516486e-06,
+      "loss": 0.18,
+      "step": 12962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3108323466882076,
+      "learning_rate": 3.2263196244832183e-06,
+      "loss": 0.243,
+      "step": 12963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.351882544006569,
+      "learning_rate": 3.224950772022214e-06,
+      "loss": 0.5611,
+      "step": 12964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.633610003120019,
+      "learning_rate": 3.223582154180873e-06,
+      "loss": 0.2477,
+      "step": 12965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.293079591213699,
+      "learning_rate": 3.2222137710065915e-06,
+      "loss": 0.2608,
+      "step": 12966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2513694259705406,
+      "learning_rate": 3.2208456225467554e-06,
+      "loss": 0.748,
+      "step": 12967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2295037376326925,
+      "learning_rate": 3.219477708848743e-06,
+      "loss": 0.1619,
+      "step": 12968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3458317104854778,
+      "learning_rate": 3.2181100299599268e-06,
+      "loss": 0.2019,
+      "step": 12969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3743554598909583,
+      "learning_rate": 3.2167425859276678e-06,
+      "loss": 0.3023,
+      "step": 12970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5944735307244252,
+      "learning_rate": 3.215375376799319e-06,
+      "loss": 0.3297,
+      "step": 12971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3758967024239241,
+      "learning_rate": 3.214008402622232e-06,
+      "loss": 0.2157,
+      "step": 12972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0402206032842936,
+      "learning_rate": 3.2126416634437428e-06,
+      "loss": 0.6525,
+      "step": 12973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2605416145760141,
+      "learning_rate": 3.2112751593111803e-06,
+      "loss": 0.2228,
+      "step": 12974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2663291579048332,
+      "learning_rate": 3.2099088902718635e-06,
+      "loss": 0.1584,
+      "step": 12975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7818741848797203,
+      "learning_rate": 3.2085428563731137e-06,
+      "loss": 0.4055,
+      "step": 12976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0304535099556142,
+      "learning_rate": 3.207177057662233e-06,
+      "loss": 0.4526,
+      "step": 12977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23643939698926889,
+      "learning_rate": 3.205811494186518e-06,
+      "loss": 0.2101,
+      "step": 12978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1088209673773897,
+      "learning_rate": 3.2044461659932557e-06,
+      "loss": 0.7036,
+      "step": 12979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.286646221853249,
+      "learning_rate": 3.2030810731297334e-06,
+      "loss": 0.1906,
+      "step": 12980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3675670697198052,
+      "learning_rate": 3.2017162156432222e-06,
+      "loss": 0.2737,
+      "step": 12981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42808874432795885,
+      "learning_rate": 3.2003515935809858e-06,
+      "loss": 0.248,
+      "step": 12982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9420101336635036,
+      "learning_rate": 3.1989872069902804e-06,
+      "loss": 0.3599,
+      "step": 12983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3487176167214536,
+      "learning_rate": 3.197623055918354e-06,
+      "loss": 0.2668,
+      "step": 12984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4368984598467063,
+      "learning_rate": 3.196259140412451e-06,
+      "loss": 0.4066,
+      "step": 12985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2809539051773061,
+      "learning_rate": 3.1948954605198014e-06,
+      "loss": 0.2199,
+      "step": 12986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3451613359950105,
+      "learning_rate": 3.193532016287629e-06,
+      "loss": 0.2908,
+      "step": 12987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5062091570808256,
+      "learning_rate": 3.1921688077631476e-06,
+      "loss": 0.2377,
+      "step": 12988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.39386024451333884,
+      "learning_rate": 3.19080583499357e-06,
+      "loss": 0.2658,
+      "step": 12989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3344590966805647,
+      "learning_rate": 3.189443098026094e-06,
+      "loss": 0.2519,
+      "step": 12990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41114489266247534,
+      "learning_rate": 3.188080596907911e-06,
+      "loss": 0.2871,
+      "step": 12991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3871947212487157,
+      "learning_rate": 3.1867183316862005e-06,
+      "loss": 0.2061,
+      "step": 12992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3329933325144057,
+      "learning_rate": 3.1853563024081446e-06,
+      "loss": 0.2933,
+      "step": 12993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3416280510762696,
+      "learning_rate": 3.183994509120907e-06,
+      "loss": 0.3046,
+      "step": 12994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5561745393411215,
+      "learning_rate": 3.182632951871646e-06,
+      "loss": 0.1458,
+      "step": 12995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32559993723873837,
+      "learning_rate": 3.18127163070751e-06,
+      "loss": 0.2581,
+      "step": 12996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5157714793821341,
+      "learning_rate": 3.1799105456756463e-06,
+      "loss": 0.3919,
+      "step": 12997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.45047672562360686,
+      "learning_rate": 3.1785496968231877e-06,
+      "loss": 0.2458,
+      "step": 12998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3044374430749099,
+      "learning_rate": 3.1771890841972643e-06,
+      "loss": 0.2535,
+      "step": 12999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4077212584271979,
+      "learning_rate": 3.1758287078449812e-06,
+      "loss": 0.2976,
+      "step": 13000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2829686186815421,
+      "learning_rate": 3.174468567813461e-06,
+      "loss": 0.1982,
+      "step": 13001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3156669845287359,
+      "learning_rate": 3.1731086641497997e-06,
+      "loss": 0.2602,
+      "step": 13002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0682277454930609,
+      "learning_rate": 3.171748996901093e-06,
+      "loss": 0.7463,
+      "step": 13003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9602481865968809,
+      "learning_rate": 3.1703895661144213e-06,
+      "loss": 0.3494,
+      "step": 13004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2875938503824787,
+      "learning_rate": 3.1690303718368675e-06,
+      "loss": 0.1851,
+      "step": 13005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3280084081282049,
+      "learning_rate": 3.1676714141154998e-06,
+      "loss": 0.2983,
+      "step": 13006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28378818890057067,
+      "learning_rate": 3.1663126929973766e-06,
+      "loss": 0.1949,
+      "step": 13007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29700206228395926,
+      "learning_rate": 3.1649542085295503e-06,
+      "loss": 0.1941,
+      "step": 13008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4696627068472638,
+      "learning_rate": 3.163595960759063e-06,
+      "loss": 0.3623,
+      "step": 13009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.47099670577215524,
+      "learning_rate": 3.162237949732957e-06,
+      "loss": 0.3376,
+      "step": 13010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2863904511902697,
+      "learning_rate": 3.1608801754982564e-06,
+      "loss": 0.1883,
+      "step": 13011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33279925478274053,
+      "learning_rate": 3.1595226381019817e-06,
+      "loss": 0.1941,
+      "step": 13012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5530973813912018,
+      "learning_rate": 3.15816533759114e-06,
+      "loss": 0.3521,
+      "step": 13013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23414052147715458,
+      "learning_rate": 3.1568082740127425e-06,
+      "loss": 0.2081,
+      "step": 13014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9721095642376917,
+      "learning_rate": 3.1554514474137797e-06,
+      "loss": 0.5654,
+      "step": 13015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5991458451958498,
+      "learning_rate": 3.154094857841239e-06,
+      "loss": 0.397,
+      "step": 13016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31615058172984073,
+      "learning_rate": 3.152738505342097e-06,
+      "loss": 0.2359,
+      "step": 13017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36932228495972225,
+      "learning_rate": 3.1513823899633276e-06,
+      "loss": 0.2451,
+      "step": 13018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2985834736662441,
+      "learning_rate": 3.1500265117518926e-06,
+      "loss": 0.2106,
+      "step": 13019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31142451629486045,
+      "learning_rate": 3.148670870754744e-06,
+      "loss": 0.2453,
+      "step": 13020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7538480855399459,
+      "learning_rate": 3.1473154670188255e-06,
+      "loss": 0.3306,
+      "step": 13021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3409916931889605,
+      "learning_rate": 3.145960300591081e-06,
+      "loss": 0.2929,
+      "step": 13022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3780237167598833,
+      "learning_rate": 3.1446053715184367e-06,
+      "loss": 0.2681,
+      "step": 13023
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5302115363553981,
+      "learning_rate": 3.1432506798478134e-06,
+      "loss": 0.2266,
+      "step": 13024
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20758940749083343,
+      "learning_rate": 3.1418962256261256e-06,
+      "loss": 0.2004,
+      "step": 13025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3761950098958392,
+      "learning_rate": 3.1405420089002713e-06,
+      "loss": 0.2534,
+      "step": 13026
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6177190474844302,
+      "learning_rate": 3.1391880297171574e-06,
+      "loss": 0.2827,
+      "step": 13027
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.686885988108865,
+      "learning_rate": 3.1378342881236657e-06,
+      "loss": 0.3982,
+      "step": 13028
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42586124579859064,
+      "learning_rate": 3.1364807841666776e-06,
+      "loss": 0.2896,
+      "step": 13029
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2834899923338788,
+      "learning_rate": 3.1351275178930616e-06,
+      "loss": 0.2551,
+      "step": 13030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17052564170743853,
+      "learning_rate": 3.133774489349688e-06,
+      "loss": 0.0897,
+      "step": 13031
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3672592846418563,
+      "learning_rate": 3.1324216985834088e-06,
+      "loss": 0.2728,
+      "step": 13032
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41687149656453454,
+      "learning_rate": 3.1310691456410703e-06,
+      "loss": 0.3083,
+      "step": 13033
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5042667901150454,
+      "learning_rate": 3.1297168305695125e-06,
+      "loss": 0.2768,
+      "step": 13034
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3323112138388973,
+      "learning_rate": 3.128364753415565e-06,
+      "loss": 0.2576,
+      "step": 13035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2650359396451536,
+      "learning_rate": 3.127012914226051e-06,
+      "loss": 0.4992,
+      "step": 13036
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.256286588856045,
+      "learning_rate": 3.125661313047783e-06,
+      "loss": 0.2135,
+      "step": 13037
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30430831676033854,
+      "learning_rate": 3.1243099499275666e-06,
+      "loss": 0.2453,
+      "step": 13038
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7948641796041248,
+      "learning_rate": 3.1229588249122034e-06,
+      "loss": 0.4419,
+      "step": 13039
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.46946317228487927,
+      "learning_rate": 3.12160793804848e-06,
+      "loss": 0.2173,
+      "step": 13040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34864930416461026,
+      "learning_rate": 3.120257289383178e-06,
+      "loss": 0.2764,
+      "step": 13041
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4017789633656141,
+      "learning_rate": 3.1189068789630672e-06,
+      "loss": 0.2943,
+      "step": 13042
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4361244313994006,
+      "learning_rate": 3.117556706834919e-06,
+      "loss": 0.2576,
+      "step": 13043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.727950689186441,
+      "learning_rate": 3.116206773045486e-06,
+      "loss": 0.212,
+      "step": 13044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33404541531367393,
+      "learning_rate": 3.1148570776415153e-06,
+      "loss": 0.2929,
+      "step": 13045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3258378879850681,
+      "learning_rate": 3.1135076206697456e-06,
+      "loss": 0.2331,
+      "step": 13046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3613632019699203,
+      "learning_rate": 3.112158402176915e-06,
+      "loss": 0.1676,
+      "step": 13047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5140961520189742,
+      "learning_rate": 3.110809422209742e-06,
+      "loss": 0.353,
+      "step": 13048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34242223527610316,
+      "learning_rate": 3.109460680814942e-06,
+      "loss": 0.2875,
+      "step": 13049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4790141871496615,
+      "learning_rate": 3.108112178039222e-06,
+      "loss": 0.2166,
+      "step": 13050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5128694160336205,
+      "learning_rate": 3.106763913929278e-06,
+      "loss": 0.3545,
+      "step": 13051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25136946231934676,
+      "learning_rate": 3.1054158885318075e-06,
+      "loss": 0.1828,
+      "step": 13052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27467417356526297,
+      "learning_rate": 3.104068101893487e-06,
+      "loss": 0.232,
+      "step": 13053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5452312702861496,
+      "learning_rate": 3.102720554060993e-06,
+      "loss": 0.3419,
+      "step": 13054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.770701859262707,
+      "learning_rate": 3.101373245080985e-06,
+      "loss": 0.4232,
+      "step": 13055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3705818865576376,
+      "learning_rate": 3.100026175000128e-06,
+      "loss": 0.2777,
+      "step": 13056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3255903926001107,
+      "learning_rate": 3.0986793438650686e-06,
+      "loss": 0.2491,
+      "step": 13057
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29833688651425433,
+      "learning_rate": 3.097332751722447e-06,
+      "loss": 0.265,
+      "step": 13058
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.280230038707804,
+      "learning_rate": 3.095986398618892e-06,
+      "loss": 0.1964,
+      "step": 13059
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5305521770532992,
+      "learning_rate": 3.094640284601034e-06,
+      "loss": 0.2259,
+      "step": 13060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3592098971698116,
+      "learning_rate": 3.093294409715486e-06,
+      "loss": 0.3027,
+      "step": 13061
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7451073577020624,
+      "learning_rate": 3.0919487740088563e-06,
+      "loss": 0.3797,
+      "step": 13062
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3298793993154702,
+      "learning_rate": 3.090603377527742e-06,
+      "loss": 0.2196,
+      "step": 13063
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23678147433076321,
+      "learning_rate": 3.0892582203187337e-06,
+      "loss": 0.184,
+      "step": 13064
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36824332059382947,
+      "learning_rate": 3.087913302428419e-06,
+      "loss": 0.2789,
+      "step": 13065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3972684606203196,
+      "learning_rate": 3.0865686239033687e-06,
+      "loss": 0.2218,
+      "step": 13066
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7531669824152212,
+      "learning_rate": 3.085224184790151e-06,
+      "loss": 0.3647,
+      "step": 13067
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5677699683736633,
+      "learning_rate": 3.083879985135322e-06,
+      "loss": 0.3255,
+      "step": 13068
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25851571654009503,
+      "learning_rate": 3.082536024985431e-06,
+      "loss": 0.2738,
+      "step": 13069
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9987692197544271,
+      "learning_rate": 3.0811923043870206e-06,
+      "loss": 0.4699,
+      "step": 13070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2454516431873957,
+      "learning_rate": 3.0798488233866196e-06,
+      "loss": 0.1554,
+      "step": 13071
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3892847958002378,
+      "learning_rate": 3.0785055820307595e-06,
+      "loss": 0.2817,
+      "step": 13072
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3326192433773473,
+      "learning_rate": 3.077162580365953e-06,
+      "loss": 0.2553,
+      "step": 13073
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6536009982343787,
+      "learning_rate": 3.07581981843871e-06,
+      "loss": 0.3113,
+      "step": 13074
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36056431727318683,
+      "learning_rate": 3.0744772962955283e-06,
+      "loss": 0.2842,
+      "step": 13075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26550009542834724,
+      "learning_rate": 3.0731350139828963e-06,
+      "loss": 0.2026,
+      "step": 13076
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23690984887180597,
+      "learning_rate": 3.071792971547305e-06,
+      "loss": 0.2057,
+      "step": 13077
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5352459519460655,
+      "learning_rate": 3.0704511690352246e-06,
+      "loss": 0.3448,
+      "step": 13078
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7568358868361343,
+      "learning_rate": 3.0691096064931226e-06,
+      "loss": 0.3506,
+      "step": 13079
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8462214742386458,
+      "learning_rate": 3.0677682839674526e-06,
+      "loss": 0.2057,
+      "step": 13080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2447690890806674,
+      "learning_rate": 3.0664272015046735e-06,
+      "loss": 0.2524,
+      "step": 13081
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2331150403392583,
+      "learning_rate": 3.0650863591512215e-06,
+      "loss": 0.6398,
+      "step": 13082
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32002625401700546,
+      "learning_rate": 3.063745756953531e-06,
+      "loss": 0.1126,
+      "step": 13083
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38562212812440144,
+      "learning_rate": 3.062405394958022e-06,
+      "loss": 0.2825,
+      "step": 13084
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4755195083621667,
+      "learning_rate": 3.061065273211121e-06,
+      "loss": 0.2858,
+      "step": 13085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5813518421794341,
+      "learning_rate": 3.0597253917592308e-06,
+      "loss": 0.1095,
+      "step": 13086
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38714185646623833,
+      "learning_rate": 3.0583857506487514e-06,
+      "loss": 0.33,
+      "step": 13087
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2848101672311991,
+      "learning_rate": 3.057046349926075e-06,
+      "loss": 0.7524,
+      "step": 13088
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2332084154602774,
+      "learning_rate": 3.0557071896375824e-06,
+      "loss": 0.1926,
+      "step": 13089
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3755476958237447,
+      "learning_rate": 3.054368269829654e-06,
+      "loss": 0.2989,
+      "step": 13090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.46469134330943285,
+      "learning_rate": 3.0530295905486527e-06,
+      "loss": 0.2978,
+      "step": 13091
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4148719623586969,
+      "learning_rate": 3.0516911518409387e-06,
+      "loss": 0.3245,
+      "step": 13092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28103354751520443,
+      "learning_rate": 3.0503529537528585e-06,
+      "loss": 0.2178,
+      "step": 13093
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1477225314022288,
+      "learning_rate": 3.04901499633076e-06,
+      "loss": 0.7313,
+      "step": 13094
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6204403347702722,
+      "learning_rate": 3.047677279620973e-06,
+      "loss": 0.3122,
+      "step": 13095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3277311313482158,
+      "learning_rate": 3.0463398036698222e-06,
+      "loss": 0.2242,
+      "step": 13096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24065037007139628,
+      "learning_rate": 3.0450025685236227e-06,
+      "loss": 0.2127,
+      "step": 13097
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3160438437281266,
+      "learning_rate": 3.043665574228688e-06,
+      "loss": 0.677,
+      "step": 13098
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.305915681969301,
+      "learning_rate": 3.042328820831315e-06,
+      "loss": 0.2111,
+      "step": 13099
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.47529902211212993,
+      "learning_rate": 3.040992308377796e-06,
+      "loss": 0.3485,
+      "step": 13100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4365889963384517,
+      "learning_rate": 3.0396560369144145e-06,
+      "loss": 0.3292,
+      "step": 13101
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3014203717649501,
+      "learning_rate": 3.038320006487445e-06,
+      "loss": 0.2112,
+      "step": 13102
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2850257890850647,
+      "learning_rate": 3.036984217143154e-06,
+      "loss": 0.1755,
+      "step": 13103
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3429237306537405,
+      "learning_rate": 3.0356486689278e-06,
+      "loss": 0.3056,
+      "step": 13104
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3024864803605085,
+      "learning_rate": 3.034313361887631e-06,
+      "loss": 0.242,
+      "step": 13105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8906428296674906,
+      "learning_rate": 3.0329782960688926e-06,
+      "loss": 0.3281,
+      "step": 13106
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5836982031317877,
+      "learning_rate": 3.031643471517817e-06,
+      "loss": 0.3389,
+      "step": 13107
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3551032461668801,
+      "learning_rate": 3.0303088882806276e-06,
+      "loss": 0.2684,
+      "step": 13108
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2158504718862656,
+      "learning_rate": 3.028974546403539e-06,
+      "loss": 0.1674,
+      "step": 13109
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7293506132465545,
+      "learning_rate": 3.027640445932766e-06,
+      "loss": 0.4317,
+      "step": 13110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36427266642415257,
+      "learning_rate": 3.0263065869145035e-06,
+      "loss": 0.26,
+      "step": 13111
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5653448581871906,
+      "learning_rate": 3.024972969394944e-06,
+      "loss": 0.2791,
+      "step": 13112
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4739343898820716,
+      "learning_rate": 3.023639593420271e-06,
+      "loss": 0.3312,
+      "step": 13113
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36681901328564676,
+      "learning_rate": 3.022306459036656e-06,
+      "loss": 0.2701,
+      "step": 13114
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29844337218234673,
+      "learning_rate": 3.0209735662902706e-06,
+      "loss": 0.1129,
+      "step": 13115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.35840107821791856,
+      "learning_rate": 3.019640915227271e-06,
+      "loss": 0.2941,
+      "step": 13116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3529763050954908,
+      "learning_rate": 3.0183085058938068e-06,
+      "loss": 0.2667,
+      "step": 13117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8960601500229003,
+      "learning_rate": 3.016976338336015e-06,
+      "loss": 0.3917,
+      "step": 13118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8699740633932406,
+      "learning_rate": 3.015644412600036e-06,
+      "loss": 0.2787,
+      "step": 13119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.316723083079564,
+      "learning_rate": 3.0143127287319895e-06,
+      "loss": 0.2392,
+      "step": 13120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2449187149952753,
+      "learning_rate": 3.012981286777994e-06,
+      "loss": 0.2233,
+      "step": 13121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1979911990400887,
+      "learning_rate": 3.0116500867841525e-06,
+      "loss": 0.4074,
+      "step": 13122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3560474829287622,
+      "learning_rate": 3.0103191287965715e-06,
+      "loss": 0.2798,
+      "step": 13123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9490243159614561,
+      "learning_rate": 3.008988412861338e-06,
+      "loss": 0.3888,
+      "step": 13124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34330007122672573,
+      "learning_rate": 3.007657939024535e-06,
+      "loss": 0.2458,
+      "step": 13125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31736090452794763,
+      "learning_rate": 3.006327707332235e-06,
+      "loss": 0.2582,
+      "step": 13126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4724601701873738,
+      "learning_rate": 3.004997717830508e-06,
+      "loss": 0.2203,
+      "step": 13127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2326895772207429,
+      "learning_rate": 3.003667970565409e-06,
+      "loss": 0.1987,
+      "step": 13128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3605528897187082,
+      "learning_rate": 3.002338465582988e-06,
+      "loss": 0.2826,
+      "step": 13129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9734472280455081,
+      "learning_rate": 3.0010092029292835e-06,
+      "loss": 0.3781,
+      "step": 13130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0805909522215216,
+      "learning_rate": 2.9996801826503275e-06,
+      "loss": 0.5344,
+      "step": 13131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2773513817656313,
+      "learning_rate": 2.9983514047921493e-06,
+      "loss": 0.1889,
+      "step": 13132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.35829546695635645,
+      "learning_rate": 2.9970228694007598e-06,
+      "loss": 0.3274,
+      "step": 13133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4805023126371464,
+      "learning_rate": 2.995694576522168e-06,
+      "loss": 0.2938,
+      "step": 13134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32261178589923567,
+      "learning_rate": 2.9943665262023714e-06,
+      "loss": 0.2114,
+      "step": 13135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36506769087255786,
+      "learning_rate": 2.993038718487361e-06,
+      "loss": 0.3249,
+      "step": 13136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3643438925093802,
+      "learning_rate": 2.991711153423118e-06,
+      "loss": 0.215,
+      "step": 13137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33487426259480946,
+      "learning_rate": 2.9903838310556133e-06,
+      "loss": 0.1942,
+      "step": 13138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4488417837759212,
+      "learning_rate": 2.989056751430819e-06,
+      "loss": 0.2472,
+      "step": 13139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3362232531745725,
+      "learning_rate": 2.987729914594687e-06,
+      "loss": 0.3241,
+      "step": 13140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29966928249023517,
+      "learning_rate": 2.9864033205931675e-06,
+      "loss": 0.2217,
+      "step": 13141
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.863128083275095,
+      "learning_rate": 2.9850769694721982e-06,
+      "loss": 0.5081,
+      "step": 13142
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39978155517146674,
+      "learning_rate": 2.9837508612777087e-06,
+      "loss": 0.2601,
+      "step": 13143
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2719235906065331,
+      "learning_rate": 2.9824249960556294e-06,
+      "loss": 0.2627,
+      "step": 13144
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5339971689860459,
+      "learning_rate": 2.9810993738518702e-06,
+      "loss": 0.2528,
+      "step": 13145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7843067034359767,
+      "learning_rate": 2.9797739947123383e-06,
+      "loss": 0.4102,
+      "step": 13146
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3894676336923343,
+      "learning_rate": 2.9784488586829272e-06,
+      "loss": 0.2814,
+      "step": 13147
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2722933405316551,
+      "learning_rate": 2.9771239658095342e-06,
+      "loss": 0.249,
+      "step": 13148
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2812989717087355,
+      "learning_rate": 2.975799316138035e-06,
+      "loss": 0.187,
+      "step": 13149
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39908579003212574,
+      "learning_rate": 2.9744749097143046e-06,
+      "loss": 0.2796,
+      "step": 13150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6090914983354809,
+      "learning_rate": 2.9731507465842025e-06,
+      "loss": 0.2293,
+      "step": 13151
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4057610007364994,
+      "learning_rate": 2.97182682679359e-06,
+      "loss": 0.304,
+      "step": 13152
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4026742727902014,
+      "learning_rate": 2.970503150388313e-06,
+      "loss": 0.3015,
+      "step": 13153
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29655660277141555,
+      "learning_rate": 2.96917971741421e-06,
+      "loss": 0.1956,
+      "step": 13154
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4737388090525767,
+      "learning_rate": 2.9678565279171113e-06,
+      "loss": 0.2305,
+      "step": 13155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25693414461923525,
+      "learning_rate": 2.9665335819428354e-06,
+      "loss": 0.2463,
+      "step": 13156
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5539166843080455,
+      "learning_rate": 2.9652108795372016e-06,
+      "loss": 0.323,
+      "step": 13157
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6583760802800553,
+      "learning_rate": 2.963888420746013e-06,
+      "loss": 0.2997,
+      "step": 13158
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3184635760317434,
+      "learning_rate": 2.962566205615065e-06,
+      "loss": 0.2428,
+      "step": 13159
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.45352708893623567,
+      "learning_rate": 2.9612442341901448e-06,
+      "loss": 0.3603,
+      "step": 13160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22507968518561244,
+      "learning_rate": 2.9599225065170356e-06,
+      "loss": 0.1509,
+      "step": 13161
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32301562997476935,
+      "learning_rate": 2.9586010226415085e-06,
+      "loss": 0.2435,
+      "step": 13162
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9494394947854288,
+      "learning_rate": 2.9572797826093256e-06,
+      "loss": 0.4594,
+      "step": 13163
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36294253708825114,
+      "learning_rate": 2.9559587864662365e-06,
+      "loss": 0.2539,
+      "step": 13164
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6077925906601935,
+      "learning_rate": 2.9546380342579962e-06,
+      "loss": 0.3771,
+      "step": 13165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3826961493987988,
+      "learning_rate": 2.953317526030337e-06,
+      "loss": 0.3284,
+      "step": 13166
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26520872395312056,
+      "learning_rate": 2.9519972618289894e-06,
+      "loss": 0.2096,
+      "step": 13167
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3666621982903667,
+      "learning_rate": 2.9506772416996732e-06,
+      "loss": 0.1855,
+      "step": 13168
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3930965638463575,
+      "learning_rate": 2.9493574656881006e-06,
+      "loss": 0.29,
+      "step": 13169
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8084312110357765,
+      "learning_rate": 2.9480379338399757e-06,
+      "loss": 0.3901,
+      "step": 13170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3213576415786909,
+      "learning_rate": 2.9467186462009943e-06,
+      "loss": 0.1748,
+      "step": 13171
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2886036323658359,
+      "learning_rate": 2.94539960281684e-06,
+      "loss": 0.2953,
+      "step": 13172
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44952697496270905,
+      "learning_rate": 2.944080803733197e-06,
+      "loss": 0.2489,
+      "step": 13173
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24913717295057006,
+      "learning_rate": 2.942762248995733e-06,
+      "loss": 0.1602,
+      "step": 13174
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5314385966365966,
+      "learning_rate": 2.9414439386501082e-06,
+      "loss": 0.3329,
+      "step": 13175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3573835288969693,
+      "learning_rate": 2.9401258727419723e-06,
+      "loss": 0.3176,
+      "step": 13176
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31895803358139446,
+      "learning_rate": 2.938808051316978e-06,
+      "loss": 0.1936,
+      "step": 13177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5156004074194172,
+      "learning_rate": 2.937490474420758e-06,
+      "loss": 0.3777,
+      "step": 13178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3266568107674821,
+      "learning_rate": 2.9361731420989382e-06,
+      "loss": 0.1728,
+      "step": 13179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2562266580469743,
+      "learning_rate": 2.9348560543971383e-06,
+      "loss": 0.2493,
+      "step": 13180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4862114642170611,
+      "learning_rate": 2.933539211360966e-06,
+      "loss": 0.2606,
+      "step": 13181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8436076426893645,
+      "learning_rate": 2.932222613036032e-06,
+      "loss": 0.4518,
+      "step": 13182
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8093127081004269,
+      "learning_rate": 2.930906259467924e-06,
+      "loss": 0.3676,
+      "step": 13183
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24389401278509923,
+      "learning_rate": 2.9295901507022275e-06,
+      "loss": 0.2305,
+      "step": 13184
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5109376582936659,
+      "learning_rate": 2.928274286784517e-06,
+      "loss": 0.2928,
+      "step": 13185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6477468458115577,
+      "learning_rate": 2.9269586677603677e-06,
+      "loss": 0.3776,
+      "step": 13186
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28910265146819936,
+      "learning_rate": 2.9256432936753354e-06,
+      "loss": 0.1995,
+      "step": 13187
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.350700075545137,
+      "learning_rate": 2.924328164574972e-06,
+      "loss": 0.3169,
+      "step": 13188
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7686568143004432,
+      "learning_rate": 2.923013280504816e-06,
+      "loss": 0.4884,
+      "step": 13189
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3587136303027198,
+      "learning_rate": 2.9216986415104097e-06,
+      "loss": 0.2156,
+      "step": 13190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7374740757701493,
+      "learning_rate": 2.9203842476372747e-06,
+      "loss": 0.3926,
+      "step": 13191
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2838512447883044,
+      "learning_rate": 2.9190700989309285e-06,
+      "loss": 0.2682,
+      "step": 13192
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.290622311729339,
+      "learning_rate": 2.9177561954368804e-06,
+      "loss": 0.2525,
+      "step": 13193
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5359851867679772,
+      "learning_rate": 2.916442537200629e-06,
+      "loss": 0.1526,
+      "step": 13194
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37080513662796255,
+      "learning_rate": 2.9151291242676692e-06,
+      "loss": 0.2884,
+      "step": 13195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30793946557758933,
+      "learning_rate": 2.9138159566834834e-06,
+      "loss": 0.2773,
+      "step": 13196
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8039012607707117,
+      "learning_rate": 2.912503034493547e-06,
+      "loss": 0.2705,
+      "step": 13197
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5254037076250225,
+      "learning_rate": 2.911190357743322e-06,
+      "loss": 0.3449,
+      "step": 13198
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.387990735879651,
+      "learning_rate": 2.909877926478274e-06,
+      "loss": 0.3074,
+      "step": 13199
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22161277090349127,
+      "learning_rate": 2.9085657407438485e-06,
+      "loss": 0.1872,
+      "step": 13200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5582834510824618,
+      "learning_rate": 2.9072538005854855e-06,
+      "loss": 0.3282,
+      "step": 13201
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40851544103697307,
+      "learning_rate": 2.9059421060486193e-06,
+      "loss": 0.3156,
+      "step": 13202
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.35263245779401653,
+      "learning_rate": 2.904630657178672e-06,
+      "loss": 0.2656,
+      "step": 13203
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6917076508307758,
+      "learning_rate": 2.903319454021061e-06,
+      "loss": 0.3521,
+      "step": 13204
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34619152479873244,
+      "learning_rate": 2.9020084966211913e-06,
+      "loss": 0.2813,
+      "step": 13205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2818608501668898,
+      "learning_rate": 2.900697785024459e-06,
+      "loss": 0.1854,
+      "step": 13206
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2772904606806348,
+      "learning_rate": 2.89938731927626e-06,
+      "loss": 0.2237,
+      "step": 13207
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3456570326485675,
+      "learning_rate": 2.8980770994219743e-06,
+      "loss": 0.2759,
+      "step": 13208
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6722228995159082,
+      "learning_rate": 2.8967671255069717e-06,
+      "loss": 0.3808,
+      "step": 13209
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5544282846657669,
+      "learning_rate": 2.8954573975766156e-06,
+      "loss": 0.1777,
+      "step": 13210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33520447957453975,
+      "learning_rate": 2.8941479156762675e-06,
+      "loss": 0.2917,
+      "step": 13211
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23424922457156772,
+      "learning_rate": 2.892838679851272e-06,
+      "loss": 0.1917,
+      "step": 13212
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3888829363306553,
+      "learning_rate": 2.891529690146966e-06,
+      "loss": 0.2193,
+      "step": 13213
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39275318593597286,
+      "learning_rate": 2.8902209466086794e-06,
+      "loss": 0.3079,
+      "step": 13214
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6074206260942908,
+      "learning_rate": 2.8889124492817377e-06,
+      "loss": 0.3461,
+      "step": 13215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4121332734309223,
+      "learning_rate": 2.887604198211453e-06,
+      "loss": 0.2094,
+      "step": 13216
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4086837165721845,
+      "learning_rate": 2.886296193443129e-06,
+      "loss": 0.2966,
+      "step": 13217
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31241744718481923,
+      "learning_rate": 2.8849884350220614e-06,
+      "loss": 0.2266,
+      "step": 13218
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3312463483694761,
+      "learning_rate": 2.883680922993536e-06,
+      "loss": 0.2859,
+      "step": 13219
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3602296698578734,
+      "learning_rate": 2.882373657402836e-06,
+      "loss": 0.2236,
+      "step": 13220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6188669949771659,
+      "learning_rate": 2.8810666382952314e-06,
+      "loss": 0.3728,
+      "step": 13221
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3765443199588423,
+      "learning_rate": 2.879759865715982e-06,
+      "loss": 0.5777,
+      "step": 13222
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24491184844570574,
+      "learning_rate": 2.87845333971034e-06,
+      "loss": 0.2082,
+      "step": 13223
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31949829603425767,
+      "learning_rate": 2.877147060323555e-06,
+      "loss": 0.2706,
+      "step": 13224
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8517481211049055,
+      "learning_rate": 2.875841027600862e-06,
+      "loss": 0.4386,
+      "step": 13225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2866011692470818,
+      "learning_rate": 2.8745352415874872e-06,
+      "loss": 0.2287,
+      "step": 13226
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3154714117522481,
+      "learning_rate": 2.873229702328647e-06,
+      "loss": 0.251,
+      "step": 13227
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.47557481422269005,
+      "learning_rate": 2.8719244098695597e-06,
+      "loss": 0.3048,
+      "step": 13228
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28793328291945003,
+      "learning_rate": 2.8706193642554237e-06,
+      "loss": 0.1866,
+      "step": 13229
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1838012367861193,
+      "learning_rate": 2.8693145655314327e-06,
+      "loss": 0.6212,
+      "step": 13230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3389412390749291,
+      "learning_rate": 2.86801001374277e-06,
+      "loss": 0.284,
+      "step": 13231
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3654452826305494,
+      "learning_rate": 2.8667057089346127e-06,
+      "loss": 0.2889,
+      "step": 13232
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.14945398474158922,
+      "learning_rate": 2.865401651152132e-06,
+      "loss": 0.0971,
+      "step": 13233
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8662144879164575,
+      "learning_rate": 2.864097840440485e-06,
+      "loss": 0.3408,
+      "step": 13234
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33595790732355874,
+      "learning_rate": 2.8627942768448234e-06,
+      "loss": 0.2453,
+      "step": 13235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34996311712384964,
+      "learning_rate": 2.861490960410289e-06,
+      "loss": 0.2498,
+      "step": 13236
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4919206950943628,
+      "learning_rate": 2.8601878911820168e-06,
+      "loss": 0.3234,
+      "step": 13237
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3451700720073584,
+      "learning_rate": 2.8588850692051296e-06,
+      "loss": 0.2718,
+      "step": 13238
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1936044629355527,
+      "learning_rate": 2.857582494524742e-06,
+      "loss": 0.1738,
+      "step": 13239
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.196637382273675,
+      "learning_rate": 2.8562801671859697e-06,
+      "loss": 0.5177,
+      "step": 13240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3138689667270411,
+      "learning_rate": 2.8549780872339073e-06,
+      "loss": 0.2528,
+      "step": 13241
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7183731018697147,
+      "learning_rate": 2.8536762547136464e-06,
+      "loss": 0.3813,
+      "step": 13242
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3149057673584101,
+      "learning_rate": 2.85237466967027e-06,
+      "loss": 0.2506,
+      "step": 13243
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31630528529321505,
+      "learning_rate": 2.851073332148848e-06,
+      "loss": 0.2517,
+      "step": 13244
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4393966081472135,
+      "learning_rate": 2.849772242194453e-06,
+      "loss": 0.2605,
+      "step": 13245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3954421837785276,
+      "learning_rate": 2.8484713998521364e-06,
+      "loss": 0.1002,
+      "step": 13246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25876846616548765,
+      "learning_rate": 2.847170805166949e-06,
+      "loss": 0.2416,
+      "step": 13247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.47757103299646225,
+      "learning_rate": 2.8458704581839247e-06,
+      "loss": 0.3262,
+      "step": 13248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9076141858611277,
+      "learning_rate": 2.844570358948103e-06,
+      "loss": 0.3371,
+      "step": 13249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3070964861074088,
+      "learning_rate": 2.843270507504502e-06,
+      "loss": 0.2474,
+      "step": 13250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37473028523472246,
+      "learning_rate": 2.8419709038981345e-06,
+      "loss": 0.2904,
+      "step": 13251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1318432605062941,
+      "learning_rate": 2.840671548174004e-06,
+      "loss": 0.0704,
+      "step": 13252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39967919119913864,
+      "learning_rate": 2.8393724403771137e-06,
+      "loss": 0.2531,
+      "step": 13253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0046598432534393,
+      "learning_rate": 2.8380735805524475e-06,
+      "loss": 0.3925,
+      "step": 13254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3777098827872954,
+      "learning_rate": 2.8367749687449853e-06,
+      "loss": 0.334,
+      "step": 13255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3147497604198872,
+      "learning_rate": 2.835476604999695e-06,
+      "loss": 0.1906,
+      "step": 13256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4294056542690755,
+      "learning_rate": 2.8341784893615443e-06,
+      "loss": 0.2689,
+      "step": 13257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.43842450172880576,
+      "learning_rate": 2.8328806218754855e-06,
+      "loss": 0.2174,
+      "step": 13258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25153714414036416,
+      "learning_rate": 2.831583002586461e-06,
+      "loss": 0.2005,
+      "step": 13259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546683602382102,
+      "learning_rate": 2.83028563153941e-06,
+      "loss": 0.3272,
+      "step": 13260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3431270553899364,
+      "learning_rate": 2.8289885087792557e-06,
+      "loss": 0.6881,
+      "step": 13261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2682221801490365,
+      "learning_rate": 2.827691634350924e-06,
+      "loss": 0.2002,
+      "step": 13262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3618717416925984,
+      "learning_rate": 2.826395008299323e-06,
+      "loss": 0.29,
+      "step": 13263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3130034952304038,
+      "learning_rate": 2.8250986306693553e-06,
+      "loss": 0.1539,
+      "step": 13264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3121843996017311,
+      "learning_rate": 2.823802501505909e-06,
+      "loss": 0.2223,
+      "step": 13265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7111069282689343,
+      "learning_rate": 2.8225066208538765e-06,
+      "loss": 0.4252,
+      "step": 13266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3587339369972459,
+      "learning_rate": 2.821210988758132e-06,
+      "loss": 0.3243,
+      "step": 13267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31053986252393684,
+      "learning_rate": 2.8199156052635412e-06,
+      "loss": 0.2829,
+      "step": 13268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3607437392178636,
+      "learning_rate": 2.8186204704149643e-06,
+      "loss": 0.2799,
+      "step": 13269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.259545723920443,
+      "learning_rate": 2.817325584257252e-06,
+      "loss": 0.2053,
+      "step": 13270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36841259735356224,
+      "learning_rate": 2.8160309468352465e-06,
+      "loss": 0.3017,
+      "step": 13271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38905719201812056,
+      "learning_rate": 2.81473655819378e-06,
+      "loss": 0.2515,
+      "step": 13272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0396985241949528,
+      "learning_rate": 2.813442418377674e-06,
+      "loss": 0.7514,
+      "step": 13273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39836820205825924,
+      "learning_rate": 2.812148527431752e-06,
+      "loss": 0.2428,
+      "step": 13274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.305449725807176,
+      "learning_rate": 2.8108548854008166e-06,
+      "loss": 0.242,
+      "step": 13275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44681122919529676,
+      "learning_rate": 2.8095614923296676e-06,
+      "loss": 0.2259,
+      "step": 13276
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3816614658218707,
+      "learning_rate": 2.8082683482630912e-06,
+      "loss": 0.2823,
+      "step": 13277
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2946987437084205,
+      "learning_rate": 2.806975453245877e-06,
+      "loss": 0.2036,
+      "step": 13278
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49942919812183445,
+      "learning_rate": 2.8056828073227925e-06,
+      "loss": 0.3453,
+      "step": 13279
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40850999837166446,
+      "learning_rate": 2.804390410538603e-06,
+      "loss": 0.3256,
+      "step": 13280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5696317559539408,
+      "learning_rate": 2.803098262938062e-06,
+      "loss": 0.3323,
+      "step": 13281
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4290707314341114,
+      "learning_rate": 2.801806364565921e-06,
+      "loss": 0.2762,
+      "step": 13282
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26581716980266745,
+      "learning_rate": 2.8005147154669166e-06,
+      "loss": 0.2421,
+      "step": 13283
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.299496740652704,
+      "learning_rate": 2.7992233156857784e-06,
+      "loss": 0.1897,
+      "step": 13284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0457263308117586,
+      "learning_rate": 2.7979321652672266e-06,
+      "loss": 0.5193,
+      "step": 13285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3313512064073587,
+      "learning_rate": 2.79664126425597e-06,
+      "loss": 0.2552,
+      "step": 13286
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3630065685033784,
+      "learning_rate": 2.795350612696721e-06,
+      "loss": 0.276,
+      "step": 13287
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6448201348910892,
+      "learning_rate": 2.794060210634171e-06,
+      "loss": 0.2862,
+      "step": 13288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6742071686259051,
+      "learning_rate": 2.7927700581130046e-06,
+      "loss": 0.3216,
+      "step": 13289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2330554908362521,
+      "learning_rate": 2.7914801551778994e-06,
+      "loss": 0.2058,
+      "step": 13290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31746453612708525,
+      "learning_rate": 2.7901905018735287e-06,
+      "loss": 0.2666,
+      "step": 13291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6076098034746343,
+      "learning_rate": 2.7889010982445508e-06,
+      "loss": 0.3607,
+      "step": 13292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3652652449404285,
+      "learning_rate": 2.7876119443356177e-06,
+      "loss": 0.3044,
+      "step": 13293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.441685084264975,
+      "learning_rate": 2.7863230401913698e-06,
+      "loss": 0.3271,
+      "step": 13294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28157585778291394,
+      "learning_rate": 2.7850343858564487e-06,
+      "loss": 0.1958,
+      "step": 13295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2538193604013508,
+      "learning_rate": 2.7837459813754765e-06,
+      "loss": 0.1991,
+      "step": 13296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1707178137167724,
+      "learning_rate": 2.782457826793069e-06,
+      "loss": 0.6914,
+      "step": 13297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4149696440496968,
+      "learning_rate": 2.781169922153838e-06,
+      "loss": 0.1949,
+      "step": 13298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2861185610671488,
+      "learning_rate": 2.7798822675023795e-06,
+      "loss": 0.2823,
+      "step": 13299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7178983508096588,
+      "learning_rate": 2.7785948628832904e-06,
+      "loss": 0.3858,
+      "step": 13300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4103924325196549,
+      "learning_rate": 2.7773077083411502e-06,
+      "loss": 0.1833,
+      "step": 13301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.244686535238677,
+      "learning_rate": 2.776020803920533e-06,
+      "loss": 0.1998,
+      "step": 13302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.355293710352619,
+      "learning_rate": 2.774734149666005e-06,
+      "loss": 0.3177,
+      "step": 13303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2862677955158507,
+      "learning_rate": 2.773447745622123e-06,
+      "loss": 0.2068,
+      "step": 13304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6852767845171588,
+      "learning_rate": 2.7721615918334355e-06,
+      "loss": 0.4025,
+      "step": 13305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36885958549727094,
+      "learning_rate": 2.7708756883444776e-06,
+      "loss": 0.297,
+      "step": 13306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5868284645412355,
+      "learning_rate": 2.7695900351997864e-06,
+      "loss": 0.3458,
+      "step": 13307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27339180895296544,
+      "learning_rate": 2.7683046324438822e-06,
+      "loss": 0.1766,
+      "step": 13308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36055577891955803,
+      "learning_rate": 2.7670194801212768e-06,
+      "loss": 0.2466,
+      "step": 13309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41817920256675367,
+      "learning_rate": 2.7657345782764765e-06,
+      "loss": 0.3027,
+      "step": 13310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2847934263341108,
+      "learning_rate": 2.7644499269539728e-06,
+      "loss": 0.242,
+      "step": 13311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9291208397012233,
+      "learning_rate": 2.7631655261982605e-06,
+      "loss": 0.4407,
+      "step": 13312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7871185563291784,
+      "learning_rate": 2.7618813760538145e-06,
+      "loss": 0.3596,
+      "step": 13313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2395530050943581,
+      "learning_rate": 2.7605974765651057e-06,
+      "loss": 0.2066,
+      "step": 13314
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43243853087168727,
+      "learning_rate": 2.759313827776592e-06,
+      "loss": 0.2994,
+      "step": 13315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.499861531513214,
+      "learning_rate": 2.758030429732732e-06,
+      "loss": 0.3264,
+      "step": 13316
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4019932945242539,
+      "learning_rate": 2.7567472824779663e-06,
+      "loss": 0.2627,
+      "step": 13317
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35582583706274945,
+      "learning_rate": 2.7554643860567308e-06,
+      "loss": 0.2845,
+      "step": 13318
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5481213981526601,
+      "learning_rate": 2.75418174051345e-06,
+      "loss": 0.2395,
+      "step": 13319
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41611532347043945,
+      "learning_rate": 2.7528993458925457e-06,
+      "loss": 0.3025,
+      "step": 13320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3799576843178077,
+      "learning_rate": 2.751617202238427e-06,
+      "loss": 0.1868,
+      "step": 13321
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2908916015181389,
+      "learning_rate": 2.750335309595491e-06,
+      "loss": 0.2636,
+      "step": 13322
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43567171245253916,
+      "learning_rate": 2.7490536680081325e-06,
+      "loss": 0.2941,
+      "step": 13323
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4234412711982623,
+      "learning_rate": 2.7477722775207303e-06,
+      "loss": 0.1738,
+      "step": 13324
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5576484292752899,
+      "learning_rate": 2.746491138177666e-06,
+      "loss": 0.2826,
+      "step": 13325
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3903393434473755,
+      "learning_rate": 2.745210250023301e-06,
+      "loss": 0.3149,
+      "step": 13326
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34900301912911175,
+      "learning_rate": 2.743929613101993e-06,
+      "loss": 0.2765,
+      "step": 13327
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8648744675226305,
+      "learning_rate": 2.7426492274580883e-06,
+      "loss": 0.3671,
+      "step": 13328
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3467725051148224,
+      "learning_rate": 2.7413690931359316e-06,
+      "loss": 0.3054,
+      "step": 13329
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22494972847791722,
+      "learning_rate": 2.7400892101798504e-06,
+      "loss": 0.2063,
+      "step": 13330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7541721363633993,
+      "learning_rate": 2.7388095786341682e-06,
+      "loss": 0.1194,
+      "step": 13331
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36934648699727984,
+      "learning_rate": 2.7375301985431947e-06,
+      "loss": 0.2673,
+      "step": 13332
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8084421696547506,
+      "learning_rate": 2.736251069951241e-06,
+      "loss": 0.4386,
+      "step": 13333
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29545160567203277,
+      "learning_rate": 2.734972192902601e-06,
+      "loss": 0.2284,
+      "step": 13334
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36633179660452714,
+      "learning_rate": 2.733693567441561e-06,
+      "loss": 0.3288,
+      "step": 13335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24642775837098382,
+      "learning_rate": 2.732415193612401e-06,
+      "loss": 0.1621,
+      "step": 13336
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5111131791764394,
+      "learning_rate": 2.73113707145939e-06,
+      "loss": 0.1199,
+      "step": 13337
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37770395293264974,
+      "learning_rate": 2.7298592010267887e-06,
+      "loss": 0.3027,
+      "step": 13338
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3905628714529268,
+      "learning_rate": 2.7285815823588513e-06,
+      "loss": 0.2994,
+      "step": 13339
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5113595561516364,
+      "learning_rate": 2.7273042154998188e-06,
+      "loss": 0.2487,
+      "step": 13340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.40329210177460667,
+      "learning_rate": 2.726027100493931e-06,
+      "loss": 0.3088,
+      "step": 13341
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2326471863617773,
+      "learning_rate": 2.724750237385412e-06,
+      "loss": 0.2229,
+      "step": 13342
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0041471856553938,
+      "learning_rate": 2.723473626218479e-06,
+      "loss": 0.4367,
+      "step": 13343
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3045613574682275,
+      "learning_rate": 2.722197267037339e-06,
+      "loss": 0.1968,
+      "step": 13344
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5701953085284389,
+      "learning_rate": 2.7209211598861975e-06,
+      "loss": 0.3772,
+      "step": 13345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3510620371677265,
+      "learning_rate": 2.719645304809242e-06,
+      "loss": 0.294,
+      "step": 13346
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2921036385897743,
+      "learning_rate": 2.7183697018506584e-06,
+      "loss": 0.2072,
+      "step": 13347
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2440216161862196,
+      "learning_rate": 2.7170943510546177e-06,
+      "loss": 0.1587,
+      "step": 13348
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.822130467598231,
+      "learning_rate": 2.715819252465284e-06,
+      "loss": 0.5094,
+      "step": 13349
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24005749395345125,
+      "learning_rate": 2.714544406126819e-06,
+      "loss": 0.2119,
+      "step": 13350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6209963089409438,
+      "learning_rate": 2.713269812083369e-06,
+      "loss": 0.3862,
+      "step": 13351
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0539666138783272,
+      "learning_rate": 2.711995470379071e-06,
+      "loss": 0.6506,
+      "step": 13352
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33989522147319307,
+      "learning_rate": 2.7107213810580536e-06,
+      "loss": 0.1925,
+      "step": 13353
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2547793738056594,
+      "learning_rate": 2.709447544164444e-06,
+      "loss": 0.2418,
+      "step": 13354
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3889272161112719,
+      "learning_rate": 2.708173959742353e-06,
+      "loss": 0.2009,
+      "step": 13355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36124339525567295,
+      "learning_rate": 2.7069006278358844e-06,
+      "loss": 0.2852,
+      "step": 13356
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.866977479010928,
+      "learning_rate": 2.70562754848913e-06,
+      "loss": 0.29,
+      "step": 13357
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32296997474230627,
+      "learning_rate": 2.704354721746183e-06,
+      "loss": 0.2969,
+      "step": 13358
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34896980548098877,
+      "learning_rate": 2.703082147651118e-06,
+      "loss": 0.2693,
+      "step": 13359
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2370600806689802,
+      "learning_rate": 2.7018098262480053e-06,
+      "loss": 0.103,
+      "step": 13360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5435084819332936,
+      "learning_rate": 2.700537757580901e-06,
+      "loss": 0.3418,
+      "step": 13361
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31758347380995683,
+      "learning_rate": 2.699265941693863e-06,
+      "loss": 0.2566,
+      "step": 13362
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4792746559962163,
+      "learning_rate": 2.6979943786309315e-06,
+      "loss": 0.2685,
+      "step": 13363
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2601019771202395,
+      "learning_rate": 2.6967230684361413e-06,
+      "loss": 0.7461,
+      "step": 13364
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3264807957462211,
+      "learning_rate": 2.6954520111535166e-06,
+      "loss": 0.251,
+      "step": 13365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23864470808248076,
+      "learning_rate": 2.694181206827071e-06,
+      "loss": 0.1862,
+      "step": 13366
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6936911184017519,
+      "learning_rate": 2.69291065550082e-06,
+      "loss": 0.3615,
+      "step": 13367
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.311294629785046,
+      "learning_rate": 2.691640357218759e-06,
+      "loss": 0.2626,
+      "step": 13368
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1184392357841415,
+      "learning_rate": 2.690370312024878e-06,
+      "loss": 0.4929,
+      "step": 13369
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32713334536191796,
+      "learning_rate": 2.6891005199631558e-06,
+      "loss": 0.2735,
+      "step": 13370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3248615650269381,
+      "learning_rate": 2.6878309810775738e-06,
+      "loss": 0.2617,
+      "step": 13371
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4369216717341033,
+      "learning_rate": 2.6865616954120878e-06,
+      "loss": 0.5723,
+      "step": 13372
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5156635728031603,
+      "learning_rate": 2.6852926630106558e-06,
+      "loss": 0.2641,
+      "step": 13373
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3373075279153563,
+      "learning_rate": 2.6840238839172206e-06,
+      "loss": 0.2585,
+      "step": 13374
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36128908717495617,
+      "learning_rate": 2.682755358175728e-06,
+      "loss": 0.1984,
+      "step": 13375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.71767892274232,
+      "learning_rate": 2.6814870858301013e-06,
+      "loss": 0.2892,
+      "step": 13376
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36937216939202855,
+      "learning_rate": 2.6802190669242634e-06,
+      "loss": 0.2787,
+      "step": 13377
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3642005381982544,
+      "learning_rate": 2.6789513015021207e-06,
+      "loss": 0.2875,
+      "step": 13378
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.44540040503421996,
+      "learning_rate": 2.6776837896075824e-06,
+      "loss": 0.2292,
+      "step": 13379
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33180184811006774,
+      "learning_rate": 2.6764165312845402e-06,
+      "loss": 0.2614,
+      "step": 13380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32777961922851084,
+      "learning_rate": 2.675149526576879e-06,
+      "loss": 0.2594,
+      "step": 13381
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4898636512120215,
+      "learning_rate": 2.67388277552847e-06,
+      "loss": 0.3877,
+      "step": 13382
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2721793650080163,
+      "learning_rate": 2.67261627818319e-06,
+      "loss": 0.197,
+      "step": 13383
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7355116149856278,
+      "learning_rate": 2.671350034584893e-06,
+      "loss": 0.3658,
+      "step": 13384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.47438055021689923,
+      "learning_rate": 2.670084044777429e-06,
+      "loss": 0.3412,
+      "step": 13385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2641302345028114,
+      "learning_rate": 2.668818308804636e-06,
+      "loss": 0.2146,
+      "step": 13386
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25613544160514196,
+      "learning_rate": 2.6675528267103534e-06,
+      "loss": 0.1642,
+      "step": 13387
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.011992324523561,
+      "learning_rate": 2.6662875985384007e-06,
+      "loss": 0.6767,
+      "step": 13388
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3127942387817402,
+      "learning_rate": 2.665022624332593e-06,
+      "loss": 0.2063,
+      "step": 13389
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3537706267658949,
+      "learning_rate": 2.6637579041367357e-06,
+      "loss": 0.2924,
+      "step": 13390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.71475916044793,
+      "learning_rate": 2.6624934379946243e-06,
+      "loss": 0.3884,
+      "step": 13391
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21694765840296173,
+      "learning_rate": 2.661229225950054e-06,
+      "loss": 0.1554,
+      "step": 13392
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35457329355787714,
+      "learning_rate": 2.659965268046798e-06,
+      "loss": 0.2834,
+      "step": 13393
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4947678076948455,
+      "learning_rate": 2.6587015643286295e-06,
+      "loss": 0.3978,
+      "step": 13394
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5838803034932754,
+      "learning_rate": 2.657438114839308e-06,
+      "loss": 0.3296,
+      "step": 13395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3963018840650698,
+      "learning_rate": 2.6561749196225915e-06,
+      "loss": 0.2608,
+      "step": 13396
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3411749326825493,
+      "learning_rate": 2.654911978722222e-06,
+      "loss": 0.2992,
+      "step": 13397
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4121763965343542,
+      "learning_rate": 2.6536492921819346e-06,
+      "loss": 0.2847,
+      "step": 13398
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23315859992533955,
+      "learning_rate": 2.6523868600454526e-06,
+      "loss": 0.143,
+      "step": 13399
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2290226217364666,
+      "learning_rate": 2.6511246823565016e-06,
+      "loss": 0.6549,
+      "step": 13400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.48684221439801073,
+      "learning_rate": 2.649862759158787e-06,
+      "loss": 0.2985,
+      "step": 13401
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27189307366780724,
+      "learning_rate": 2.648601090496008e-06,
+      "loss": 0.2455,
+      "step": 13402
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2851418916503219,
+      "learning_rate": 2.6473396764118575e-06,
+      "loss": 0.5728,
+      "step": 13403
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3704241967181858,
+      "learning_rate": 2.646078516950018e-06,
+      "loss": 0.2355,
+      "step": 13404
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32679798644361935,
+      "learning_rate": 2.6448176121541634e-06,
+      "loss": 0.2539,
+      "step": 13405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42009343086991197,
+      "learning_rate": 2.643556962067958e-06,
+      "loss": 0.2625,
+      "step": 13406
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3364463651284445,
+      "learning_rate": 2.6422965667350566e-06,
+      "loss": 0.2599,
+      "step": 13407
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4603678526160474,
+      "learning_rate": 2.6410364261991108e-06,
+      "loss": 0.2815,
+      "step": 13408
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34026034801807636,
+      "learning_rate": 2.6397765405037577e-06,
+      "loss": 0.2485,
+      "step": 13409
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32456868640721753,
+      "learning_rate": 2.6385169096926265e-06,
+      "loss": 0.2346,
+      "step": 13410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49966563733385355,
+      "learning_rate": 2.637257533809334e-06,
+      "loss": 0.2526,
+      "step": 13411
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6283074877544407,
+      "learning_rate": 2.6359984128975013e-06,
+      "loss": 0.33,
+      "step": 13412
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36892761734499885,
+      "learning_rate": 2.6347395470007254e-06,
+      "loss": 0.2784,
+      "step": 13413
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27568567981994324,
+      "learning_rate": 2.6334809361626034e-06,
+      "loss": 0.2485,
+      "step": 13414
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3477396319559145,
+      "learning_rate": 2.632222580426719e-06,
+      "loss": 0.2197,
+      "step": 13415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5419492347153053,
+      "learning_rate": 2.6309644798366474e-06,
+      "loss": 0.2762,
+      "step": 13416
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3174658239757126,
+      "learning_rate": 2.6297066344359612e-06,
+      "loss": 0.2882,
+      "step": 13417
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.47016789004810144,
+      "learning_rate": 2.6284490442682186e-06,
+      "loss": 0.3389,
+      "step": 13418
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3946797301594869,
+      "learning_rate": 2.6271917093769673e-06,
+      "loss": 0.2183,
+      "step": 13419
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23602412255899952,
+      "learning_rate": 2.6259346298057476e-06,
+      "loss": 0.1842,
+      "step": 13420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34069892961807774,
+      "learning_rate": 2.6246778055980983e-06,
+      "loss": 0.2678,
+      "step": 13421
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5221130511938119,
+      "learning_rate": 2.6234212367975375e-06,
+      "loss": 0.1991,
+      "step": 13422
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6077899342263603,
+      "learning_rate": 2.6221649234475845e-06,
+      "loss": 0.4096,
+      "step": 13423
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6059573421941735,
+      "learning_rate": 2.620908865591738e-06,
+      "loss": 0.3912,
+      "step": 13424
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25779870510643993,
+      "learning_rate": 2.619653063273504e-06,
+      "loss": 0.2187,
+      "step": 13425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3211535151394236,
+      "learning_rate": 2.618397516536367e-06,
+      "loss": 0.2455,
+      "step": 13426
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42875680879462036,
+      "learning_rate": 2.6171422254238067e-06,
+      "loss": 0.2418,
+      "step": 13427
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4063859681100399,
+      "learning_rate": 2.6158871899792927e-06,
+      "loss": 0.1835,
+      "step": 13428
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3628029769796745,
+      "learning_rate": 2.6146324102462862e-06,
+      "loss": 0.2992,
+      "step": 13429
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5550301775668914,
+      "learning_rate": 2.6133778862682433e-06,
+      "loss": 0.3215,
+      "step": 13430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7979129683971643,
+      "learning_rate": 2.612123618088608e-06,
+      "loss": 0.4975,
+      "step": 13431
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1949401131221945,
+      "learning_rate": 2.610869605750813e-06,
+      "loss": 0.138,
+      "step": 13432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35522383806764823,
+      "learning_rate": 2.6096158492982837e-06,
+      "loss": 0.2854,
+      "step": 13433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8215836187570505,
+      "learning_rate": 2.6083623487744423e-06,
+      "loss": 0.4089,
+      "step": 13434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32566860474390014,
+      "learning_rate": 2.6071091042226947e-06,
+      "loss": 0.2306,
+      "step": 13435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7199078363127668,
+      "learning_rate": 2.6058561156864415e-06,
+      "loss": 0.4138,
+      "step": 13436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3209376148836106,
+      "learning_rate": 2.60460338320907e-06,
+      "loss": 0.3103,
+      "step": 13437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25042851717939346,
+      "learning_rate": 2.603350906833971e-06,
+      "loss": 0.1709,
+      "step": 13438
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.482862170243484,
+      "learning_rate": 2.6020986866045085e-06,
+      "loss": 0.2197,
+      "step": 13439
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5706303570276031,
+      "learning_rate": 2.600846722564051e-06,
+      "loss": 0.2956,
+      "step": 13440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2703381171096475,
+      "learning_rate": 2.59959501475595e-06,
+      "loss": 0.233,
+      "step": 13441
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8217067922733584,
+      "learning_rate": 2.5983435632235586e-06,
+      "loss": 0.4331,
+      "step": 13442
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4817791637134085,
+      "learning_rate": 2.597092368010212e-06,
+      "loss": 0.2811,
+      "step": 13443
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42074169351000873,
+      "learning_rate": 2.5958414291592384e-06,
+      "loss": 0.3039,
+      "step": 13444
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22653893104331904,
+      "learning_rate": 2.594590746713953e-06,
+      "loss": 0.1724,
+      "step": 13445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5405885473781558,
+      "learning_rate": 2.5933403207176766e-06,
+      "loss": 0.2775,
+      "step": 13446
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3693178605084256,
+      "learning_rate": 2.5920901512137052e-06,
+      "loss": 0.296,
+      "step": 13447
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5107901097900251,
+      "learning_rate": 2.5908402382453337e-06,
+      "loss": 0.291,
+      "step": 13448
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.39069716983926167,
+      "learning_rate": 2.589590581855843e-06,
+      "loss": 0.2842,
+      "step": 13449
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3709867346737115,
+      "learning_rate": 2.588341182088514e-06,
+      "loss": 0.2768,
+      "step": 13450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30386942428806196,
+      "learning_rate": 2.587092038986613e-06,
+      "loss": 0.1273,
+      "step": 13451
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5989646935983985,
+      "learning_rate": 2.5858431525933946e-06,
+      "loss": 0.3982,
+      "step": 13452
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24380382171509266,
+      "learning_rate": 2.5845945229521095e-06,
+      "loss": 0.2705,
+      "step": 13453
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7642863991718087,
+      "learning_rate": 2.5833461501059933e-06,
+      "loss": 0.3247,
+      "step": 13454
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7580810104910045,
+      "learning_rate": 2.5820980340982847e-06,
+      "loss": 0.4048,
+      "step": 13455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30818453796959994,
+      "learning_rate": 2.5808501749722024e-06,
+      "loss": 0.2299,
+      "step": 13456
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27604039325913704,
+      "learning_rate": 2.5796025727709595e-06,
+      "loss": 0.2518,
+      "step": 13457
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8081791607239174,
+      "learning_rate": 2.5783552275377567e-06,
+      "loss": 0.2606,
+      "step": 13458
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.374970126576345,
+      "learning_rate": 2.577108139315797e-06,
+      "loss": 0.2767,
+      "step": 13459
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1719009812907486,
+      "learning_rate": 2.575861308148263e-06,
+      "loss": 0.748,
+      "step": 13460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28830064194617916,
+      "learning_rate": 2.574614734078332e-06,
+      "loss": 0.2245,
+      "step": 13461
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.40624340668557385,
+      "learning_rate": 2.5733684171491713e-06,
+      "loss": 0.3191,
+      "step": 13462
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31625190451738877,
+      "learning_rate": 2.5721223574039466e-06,
+      "loss": 0.1969,
+      "step": 13463
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42735286500023417,
+      "learning_rate": 2.570876554885804e-06,
+      "loss": 0.2341,
+      "step": 13464
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3077642913701503,
+      "learning_rate": 2.5696310096378875e-06,
+      "loss": 0.2794,
+      "step": 13465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1827315444704343,
+      "learning_rate": 2.568385721703329e-06,
+      "loss": 0.7581,
+      "step": 13466
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4823498393056806,
+      "learning_rate": 2.5671406911252506e-06,
+      "loss": 0.2698,
+      "step": 13467
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3486648926153057,
+      "learning_rate": 2.5658959179467734e-06,
+      "loss": 0.2548,
+      "step": 13468
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3394923711719963,
+      "learning_rate": 2.5646514022110013e-06,
+      "loss": 0.2921,
+      "step": 13469
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4458713311099046,
+      "learning_rate": 2.563407143961032e-06,
+      "loss": 0.2497,
+      "step": 13470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23033084641571222,
+      "learning_rate": 2.5621631432399496e-06,
+      "loss": 0.1549,
+      "step": 13471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.470600873287133,
+      "learning_rate": 2.5609194000908434e-06,
+      "loss": 0.3824,
+      "step": 13472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5083980754662374,
+      "learning_rate": 2.5596759145567763e-06,
+      "loss": 0.3503,
+      "step": 13473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3457918738633052,
+      "learning_rate": 2.5584326866808084e-06,
+      "loss": 0.1994,
+      "step": 13474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8561374677835701,
+      "learning_rate": 2.557189716506e-06,
+      "loss": 0.3989,
+      "step": 13475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3040416433764094,
+      "learning_rate": 2.555947004075392e-06,
+      "loss": 0.2597,
+      "step": 13476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22530902436018296,
+      "learning_rate": 2.5547045494320187e-06,
+      "loss": 0.1685,
+      "step": 13477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4681940576762507,
+      "learning_rate": 2.5534623526189075e-06,
+      "loss": 0.7422,
+      "step": 13478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8041062456322301,
+      "learning_rate": 2.5522204136790707e-06,
+      "loss": 0.4247,
+      "step": 13479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3134178270242085,
+      "learning_rate": 2.5509787326555245e-06,
+      "loss": 0.2055,
+      "step": 13480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37965284021285,
+      "learning_rate": 2.5497373095912638e-06,
+      "loss": 0.3178,
+      "step": 13481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3664967026991713,
+      "learning_rate": 2.5484961445292798e-06,
+      "loss": 0.2107,
+      "step": 13482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.348010313926068,
+      "learning_rate": 2.5472552375125514e-06,
+      "loss": 0.2591,
+      "step": 13483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27596022773032486,
+      "learning_rate": 2.546014588584057e-06,
+      "loss": 0.2083,
+      "step": 13484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5443468371920548,
+      "learning_rate": 2.5447741977867556e-06,
+      "loss": 0.3054,
+      "step": 13485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32076582873013293,
+      "learning_rate": 2.543534065163604e-06,
+      "loss": 0.2684,
+      "step": 13486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7357780117610295,
+      "learning_rate": 2.542294190757544e-06,
+      "loss": 0.3032,
+      "step": 13487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33348082391285816,
+      "learning_rate": 2.541054574611518e-06,
+      "loss": 0.2992,
+      "step": 13488
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24713867181114496,
+      "learning_rate": 2.539815216768452e-06,
+      "loss": 0.2075,
+      "step": 13489
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2824735073527296,
+      "learning_rate": 2.5385761172712642e-06,
+      "loss": 0.1629,
+      "step": 13490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7765884467550676,
+      "learning_rate": 2.537337276162861e-06,
+      "loss": 0.4139,
+      "step": 13491
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31663661173967594,
+      "learning_rate": 2.5360986934861507e-06,
+      "loss": 0.2695,
+      "step": 13492
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3654158248056001,
+      "learning_rate": 2.5348603692840214e-06,
+      "loss": 0.3131,
+      "step": 13493
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5894688790078348,
+      "learning_rate": 2.5336223035993566e-06,
+      "loss": 0.1376,
+      "step": 13494
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3125708863235182,
+      "learning_rate": 2.53238449647503e-06,
+      "loss": 0.2578,
+      "step": 13495
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1793852459395195,
+      "learning_rate": 2.5311469479539043e-06,
+      "loss": 0.7588,
+      "step": 13496
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2854328646753722,
+      "learning_rate": 2.5299096580788416e-06,
+      "loss": 0.2246,
+      "step": 13497
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.411565998534182,
+      "learning_rate": 2.5286726268926864e-06,
+      "loss": 0.299,
+      "step": 13498
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47853799166222705,
+      "learning_rate": 2.5274358544382773e-06,
+      "loss": 0.3148,
+      "step": 13499
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32126409083539537,
+      "learning_rate": 2.5261993407584394e-06,
+      "loss": 0.2218,
+      "step": 13500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35565643253633095,
+      "learning_rate": 2.5249630858960006e-06,
+      "loss": 0.2473,
+      "step": 13501
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0739827262463106,
+      "learning_rate": 2.5237270898937684e-06,
+      "loss": 0.7007,
+      "step": 13502
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4259404404044619,
+      "learning_rate": 2.522491352794545e-06,
+      "loss": 0.2166,
+      "step": 13503
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32529796843916775,
+      "learning_rate": 2.521255874641122e-06,
+      "loss": 0.2448,
+      "step": 13504
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24437696676073684,
+      "learning_rate": 2.5200206554762897e-06,
+      "loss": 0.2097,
+      "step": 13505
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2143331452246588,
+      "learning_rate": 2.5187856953428237e-06,
+      "loss": 0.4804,
+      "step": 13506
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31524982654847683,
+      "learning_rate": 2.5175509942834843e-06,
+      "loss": 0.2062,
+      "step": 13507
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3559843324635449,
+      "learning_rate": 2.516316552341028e-06,
+      "loss": 0.3224,
+      "step": 13508
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5991992837613603,
+      "learning_rate": 2.515082369558212e-06,
+      "loss": 0.3749,
+      "step": 13509
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24645117400291955,
+      "learning_rate": 2.513848445977771e-06,
+      "loss": 0.182,
+      "step": 13510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48245437149625137,
+      "learning_rate": 2.5126147816424364e-06,
+      "loss": 0.2498,
+      "step": 13511
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3648211240975149,
+      "learning_rate": 2.5113813765949267e-06,
+      "loss": 0.3079,
+      "step": 13512
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28731932512581504,
+      "learning_rate": 2.5101482308779625e-06,
+      "loss": 0.1939,
+      "step": 13513
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9704475832943743,
+      "learning_rate": 2.508915344534242e-06,
+      "loss": 0.4582,
+      "step": 13514
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6266296632397085,
+      "learning_rate": 2.50768271760646e-06,
+      "loss": 0.3931,
+      "step": 13515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27616660652349634,
+      "learning_rate": 2.5064503501373017e-06,
+      "loss": 0.2008,
+      "step": 13516
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2513559465711763,
+      "learning_rate": 2.505218242169448e-06,
+      "loss": 0.2077,
+      "step": 13517
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3205543433577367,
+      "learning_rate": 2.5039863937455645e-06,
+      "loss": 0.484,
+      "step": 13518
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6244415397855417,
+      "learning_rate": 2.5027548049083094e-06,
+      "loss": 0.3689,
+      "step": 13519
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3004803209648444,
+      "learning_rate": 2.5015234757003326e-06,
+      "loss": 0.2546,
+      "step": 13520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7853617262016618,
+      "learning_rate": 2.500292406164273e-06,
+      "loss": 0.3941,
+      "step": 13521
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3676844253504336,
+      "learning_rate": 2.4990615963427688e-06,
+      "loss": 0.2752,
+      "step": 13522
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22875365405484152,
+      "learning_rate": 2.4978310462784373e-06,
+      "loss": 0.1272,
+      "step": 13523
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35103757648372264,
+      "learning_rate": 2.496600756013895e-06,
+      "loss": 0.3124,
+      "step": 13524
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3743025090473022,
+      "learning_rate": 2.4953707255917426e-06,
+      "loss": 0.2631,
+      "step": 13525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4822393785473432,
+      "learning_rate": 2.4941409550545824e-06,
+      "loss": 0.2804,
+      "step": 13526
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1309185431941364,
+      "learning_rate": 2.492911444444999e-06,
+      "loss": 0.5591,
+      "step": 13527
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25582945570068455,
+      "learning_rate": 2.491682193805568e-06,
+      "loss": 0.2465,
+      "step": 13528
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24118639331888805,
+      "learning_rate": 2.4904532031788577e-06,
+      "loss": 0.1293,
+      "step": 13529
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2057509069942007,
+      "learning_rate": 2.489224472607432e-06,
+      "loss": 0.5991,
+      "step": 13530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4075751183842601,
+      "learning_rate": 2.487996002133841e-06,
+      "loss": 0.2751,
+      "step": 13531
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3821138342755005,
+      "learning_rate": 2.486767791800625e-06,
+      "loss": 0.2959,
+      "step": 13532
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4553999305831225,
+      "learning_rate": 2.4855398416503173e-06,
+      "loss": 0.3063,
+      "step": 13533
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39885559025142264,
+      "learning_rate": 2.4843121517254386e-06,
+      "loss": 0.2575,
+      "step": 13534
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2631468133379456,
+      "learning_rate": 2.4830847220685096e-06,
+      "loss": 0.2009,
+      "step": 13535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33484657861419903,
+      "learning_rate": 2.4818575527220347e-06,
+      "loss": 0.2423,
+      "step": 13536
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5743804627272125,
+      "learning_rate": 2.4806306437285075e-06,
+      "loss": 0.3096,
+      "step": 13537
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.36102512843751833,
+      "learning_rate": 2.479403995130416e-06,
+      "loss": 0.2962,
+      "step": 13538
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4681753138150729,
+      "learning_rate": 2.4781776069702446e-06,
+      "loss": 0.2943,
+      "step": 13539
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40481623216969953,
+      "learning_rate": 2.4769514792904603e-06,
+      "loss": 0.2884,
+      "step": 13540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2316043899089237,
+      "learning_rate": 2.4757256121335182e-06,
+      "loss": 0.195,
+      "step": 13541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8906998454080058,
+      "learning_rate": 2.4745000055418767e-06,
+      "loss": 0.2673,
+      "step": 13542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3756343888756985,
+      "learning_rate": 2.4732746595579772e-06,
+      "loss": 0.2692,
+      "step": 13543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.289010965803477,
+      "learning_rate": 2.4720495742242522e-06,
+      "loss": 0.2824,
+      "step": 13544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2453049542327923,
+      "learning_rate": 2.4708247495831263e-06,
+      "loss": 0.7404,
+      "step": 13545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.36664623886660264,
+      "learning_rate": 2.4696001856770137e-06,
+      "loss": 0.1901,
+      "step": 13546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32350893720100893,
+      "learning_rate": 2.468375882548325e-06,
+      "loss": 0.2296,
+      "step": 13547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28448727451578076,
+      "learning_rate": 2.4671518402394554e-06,
+      "loss": 0.2343,
+      "step": 13548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38941415911483,
+      "learning_rate": 2.4659280587927935e-06,
+      "loss": 0.2319,
+      "step": 13549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.540433968384648,
+      "learning_rate": 2.464704538250717e-06,
+      "loss": 0.2962,
+      "step": 13550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0027508666664908,
+      "learning_rate": 2.463481278655601e-06,
+      "loss": 0.6688,
+      "step": 13551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2635421406849195,
+      "learning_rate": 2.4622582800498042e-06,
+      "loss": 0.2202,
+      "step": 13552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3419175327603542,
+      "learning_rate": 2.4610355424756782e-06,
+      "loss": 0.2265,
+      "step": 13553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43706452382046834,
+      "learning_rate": 2.4598130659755647e-06,
+      "loss": 0.2644,
+      "step": 13554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47010643053265033,
+      "learning_rate": 2.4585908505918034e-06,
+      "loss": 0.1736,
+      "step": 13555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2701575469928234,
+      "learning_rate": 2.4573688963667176e-06,
+      "loss": 0.2637,
+      "step": 13556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1200006981818171,
+      "learning_rate": 2.4561472033426213e-06,
+      "loss": 0.8067,
+      "step": 13557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8991589933463543,
+      "learning_rate": 2.4549257715618234e-06,
+      "loss": 0.3886,
+      "step": 13558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2317283658849086,
+      "learning_rate": 2.4537046010666187e-06,
+      "loss": 0.1782,
+      "step": 13559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3356415177168732,
+      "learning_rate": 2.452483691899302e-06,
+      "loss": 0.2841,
+      "step": 13560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4554604862081215,
+      "learning_rate": 2.45126304410215e-06,
+      "loss": 0.2736,
+      "step": 13561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33245473228630973,
+      "learning_rate": 2.450042657717435e-06,
+      "loss": 0.2231,
+      "step": 13562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1404646769499847,
+      "learning_rate": 2.4488225327874147e-06,
+      "loss": 0.657,
+      "step": 13563
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33060524305523453,
+      "learning_rate": 2.4476026693543485e-06,
+      "loss": 0.2643,
+      "step": 13564
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4029416650173851,
+      "learning_rate": 2.4463830674604773e-06,
+      "loss": 0.2251,
+      "step": 13565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.532161244083854,
+      "learning_rate": 2.4451637271480357e-06,
+      "loss": 0.3595,
+      "step": 13566
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22270028050530902,
+      "learning_rate": 2.4439446484592466e-06,
+      "loss": 0.1721,
+      "step": 13567
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5652419091857782,
+      "learning_rate": 2.442725831436331e-06,
+      "loss": 0.2662,
+      "step": 13568
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8543186545019308,
+      "learning_rate": 2.4415072761214963e-06,
+      "loss": 0.5066,
+      "step": 13569
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5957280739948273,
+      "learning_rate": 2.4402889825569396e-06,
+      "loss": 0.3107,
+      "step": 13570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4228437651054475,
+      "learning_rate": 2.4390709507848497e-06,
+      "loss": 0.2888,
+      "step": 13571
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3197597496407284,
+      "learning_rate": 2.4378531808474048e-06,
+      "loss": 0.2583,
+      "step": 13572
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2274135399939389,
+      "learning_rate": 2.4366356727867847e-06,
+      "loss": 0.156,
+      "step": 13573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.45599695397853623,
+      "learning_rate": 2.435418426645144e-06,
+      "loss": 0.2762,
+      "step": 13574
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3557905396710193,
+      "learning_rate": 2.4342014424646343e-06,
+      "loss": 0.2661,
+      "step": 13575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7923026226601533,
+      "learning_rate": 2.4329847202874058e-06,
+      "loss": 0.3719,
+      "step": 13576
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7790919251786718,
+      "learning_rate": 2.4317682601555913e-06,
+      "loss": 0.2882,
+      "step": 13577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9627289488428679,
+      "learning_rate": 2.4305520621113175e-06,
+      "loss": 0.2932,
+      "step": 13578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2430678228878389,
+      "learning_rate": 2.4293361261966965e-06,
+      "loss": 0.2217,
+      "step": 13579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24952073494447208,
+      "learning_rate": 2.4281204524538425e-06,
+      "loss": 0.2056,
+      "step": 13580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1887519774137303,
+      "learning_rate": 2.426905040924853e-06,
+      "loss": 0.7538,
+      "step": 13581
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6848642705531716,
+      "learning_rate": 2.4256898916518145e-06,
+      "loss": 0.2607,
+      "step": 13582
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34377998627378237,
+      "learning_rate": 2.4244750046768105e-06,
+      "loss": 0.2902,
+      "step": 13583
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35628235811761977,
+      "learning_rate": 2.4232603800419087e-06,
+      "loss": 0.3136,
+      "step": 13584
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1818921796340036,
+      "learning_rate": 2.4220460177891757e-06,
+      "loss": 0.0816,
+      "step": 13585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3849502874315825,
+      "learning_rate": 2.4208319179606643e-06,
+      "loss": 0.3033,
+      "step": 13586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47780436818363103,
+      "learning_rate": 2.419618080598417e-06,
+      "loss": 0.3461,
+      "step": 13587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4695138835770297,
+      "learning_rate": 2.418404505744467e-06,
+      "loss": 0.222,
+      "step": 13588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5030258966165988,
+      "learning_rate": 2.4171911934408464e-06,
+      "loss": 0.2913,
+      "step": 13589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9470056441285403,
+      "learning_rate": 2.4159781437295684e-06,
+      "loss": 0.4908,
+      "step": 13590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2590739492221378,
+      "learning_rate": 2.414765356652641e-06,
+      "loss": 0.1915,
+      "step": 13591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2734733858017858,
+      "learning_rate": 2.4135528322520597e-06,
+      "loss": 0.2396,
+      "step": 13592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47194836624512676,
+      "learning_rate": 2.4123405705698213e-06,
+      "loss": 0.2789,
+      "step": 13593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7394128282774476,
+      "learning_rate": 2.4111285716479015e-06,
+      "loss": 0.3719,
+      "step": 13594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2604512556637114,
+      "learning_rate": 2.4099168355282743e-06,
+      "loss": 0.1993,
+      "step": 13595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49077072597667926,
+      "learning_rate": 2.4087053622529e-06,
+      "loss": 0.351,
+      "step": 13596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1799394846564877,
+      "learning_rate": 2.4074941518637295e-06,
+      "loss": 0.5699,
+      "step": 13597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3030824419466015,
+      "learning_rate": 2.406283204402714e-06,
+      "loss": 0.2124,
+      "step": 13598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46405446573905923,
+      "learning_rate": 2.405072519911783e-06,
+      "loss": 0.31,
+      "step": 13599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32419622170740764,
+      "learning_rate": 2.4038620984328655e-06,
+      "loss": 0.2435,
+      "step": 13600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21380776837026694,
+      "learning_rate": 2.4026519400078728e-06,
+      "loss": 0.1533,
+      "step": 13601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3300209163280692,
+      "learning_rate": 2.401442044678721e-06,
+      "loss": 0.5964,
+      "step": 13602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35887510821296326,
+      "learning_rate": 2.4002324124873033e-06,
+      "loss": 0.2981,
+      "step": 13603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2783468021005403,
+      "learning_rate": 2.3990230434755112e-06,
+      "loss": 0.1819,
+      "step": 13604
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7379956763914063,
+      "learning_rate": 2.3978139376852206e-06,
+      "loss": 0.4326,
+      "step": 13605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3247249377272244,
+      "learning_rate": 2.3966050951583096e-06,
+      "loss": 0.1835,
+      "step": 13606
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3669611320334294,
+      "learning_rate": 2.39539651593664e-06,
+      "loss": 0.2728,
+      "step": 13607
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35511621181556124,
+      "learning_rate": 2.3941882000620586e-06,
+      "loss": 0.2297,
+      "step": 13608
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2103759075595983,
+      "learning_rate": 2.3929801475764113e-06,
+      "loss": 0.498,
+      "step": 13609
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33080955912933513,
+      "learning_rate": 2.391772358521536e-06,
+      "loss": 0.2461,
+      "step": 13610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37747048144770934,
+      "learning_rate": 2.3905648329392574e-06,
+      "loss": 0.2751,
+      "step": 13611
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8894374852257119,
+      "learning_rate": 2.389357570871391e-06,
+      "loss": 0.4819,
+      "step": 13612
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24421514760385282,
+      "learning_rate": 2.3881505723597422e-06,
+      "loss": 0.202,
+      "step": 13613
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40684410356900697,
+      "learning_rate": 2.386943837446114e-06,
+      "loss": 0.1062,
+      "step": 13614
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3562939479119407,
+      "learning_rate": 2.385737366172294e-06,
+      "loss": 0.3223,
+      "step": 13615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34011491819599793,
+      "learning_rate": 2.3845311585800612e-06,
+      "loss": 0.2739,
+      "step": 13616
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7845879219744151,
+      "learning_rate": 2.3833252147111853e-06,
+      "loss": 0.2916,
+      "step": 13617
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.343820832697006,
+      "learning_rate": 2.382119534607431e-06,
+      "loss": 0.2475,
+      "step": 13618
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25729081529380954,
+      "learning_rate": 2.38091411831055e-06,
+      "loss": 0.2516,
+      "step": 13619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5159852122593639,
+      "learning_rate": 2.379708965862285e-06,
+      "loss": 0.2322,
+      "step": 13620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8836787867400838,
+      "learning_rate": 2.3785040773043686e-06,
+      "loss": 0.3155,
+      "step": 13621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4244576302501707,
+      "learning_rate": 2.3772994526785308e-06,
+      "loss": 0.2892,
+      "step": 13622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29146153581979123,
+      "learning_rate": 2.376095092026486e-06,
+      "loss": 0.289,
+      "step": 13623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6333592123809318,
+      "learning_rate": 2.37489099538994e-06,
+      "loss": 0.2585,
+      "step": 13624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3367141969825039,
+      "learning_rate": 2.3736871628105907e-06,
+      "loss": 0.2642,
+      "step": 13625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2573235517922094,
+      "learning_rate": 2.372483594330124e-06,
+      "loss": 0.1519,
+      "step": 13626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3441524091740002,
+      "learning_rate": 2.3712802899902256e-06,
+      "loss": 0.258,
+      "step": 13627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31942754964550046,
+      "learning_rate": 2.3700772498325617e-06,
+      "loss": 0.2458,
+      "step": 13628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7120882141052624,
+      "learning_rate": 2.3688744738987955e-06,
+      "loss": 0.3966,
+      "step": 13629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3333702259355196,
+      "learning_rate": 2.3676719622305754e-06,
+      "loss": 0.2666,
+      "step": 13630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2738749937750857,
+      "learning_rate": 2.3664697148695494e-06,
+      "loss": 0.2535,
+      "step": 13631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24507540728163543,
+      "learning_rate": 2.365267731857349e-06,
+      "loss": 0.1795,
+      "step": 13632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7427338888923851,
+      "learning_rate": 2.3640660132356e-06,
+      "loss": 0.3918,
+      "step": 13633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3615965217499706,
+      "learning_rate": 2.362864559045912e-06,
+      "loss": 0.2184,
+      "step": 13634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34559986606325616,
+      "learning_rate": 2.3616633693298996e-06,
+      "loss": 0.2857,
+      "step": 13635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.076346574627362,
+      "learning_rate": 2.360462444129156e-06,
+      "loss": 0.6005,
+      "step": 13636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2748717407836768,
+      "learning_rate": 2.3592617834852694e-06,
+      "loss": 0.1638,
+      "step": 13637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4605339516609135,
+      "learning_rate": 2.358061387439818e-06,
+      "loss": 0.3061,
+      "step": 13638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.36174894634119437,
+      "learning_rate": 2.356861256034371e-06,
+      "loss": 0.3071,
+      "step": 13639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40535929812470445,
+      "learning_rate": 2.355661389310492e-06,
+      "loss": 0.2054,
+      "step": 13640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5112923460526705,
+      "learning_rate": 2.354461787309733e-06,
+      "loss": 0.3737,
+      "step": 13641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5087292017954329,
+      "learning_rate": 2.353262450073628e-06,
+      "loss": 0.3251,
+      "step": 13642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29548431419215865,
+      "learning_rate": 2.3520633776437187e-06,
+      "loss": 0.199,
+      "step": 13643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26742981963554274,
+      "learning_rate": 2.3508645700615253e-06,
+      "loss": 0.1937,
+      "step": 13644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7458493819978979,
+      "learning_rate": 2.3496660273685633e-06,
+      "loss": 0.3586,
+      "step": 13645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.398854225716011,
+      "learning_rate": 2.348467749606335e-06,
+      "loss": 0.2928,
+      "step": 13646
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2920827396093392,
+      "learning_rate": 2.347269736816341e-06,
+      "loss": 0.2593,
+      "step": 13647
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4477857127683085,
+      "learning_rate": 2.3460719890400687e-06,
+      "loss": 0.748,
+      "step": 13648
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.45883127272812063,
+      "learning_rate": 2.3448745063189937e-06,
+      "loss": 0.2472,
+      "step": 13649
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2593031866666469,
+      "learning_rate": 2.3436772886945847e-06,
+      "loss": 0.1717,
+      "step": 13650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3450260325923007,
+      "learning_rate": 2.3424803362083005e-06,
+      "loss": 0.3068,
+      "step": 13651
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4138606948993418,
+      "learning_rate": 2.3412836489015945e-06,
+      "loss": 0.2737,
+      "step": 13652
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46342922046762425,
+      "learning_rate": 2.340087226815907e-06,
+      "loss": 0.2678,
+      "step": 13653
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5116239699745199,
+      "learning_rate": 2.338891069992669e-06,
+      "loss": 0.3142,
+      "step": 13654
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3018463186424405,
+      "learning_rate": 2.3376951784733014e-06,
+      "loss": 0.2575,
+      "step": 13655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5042604633777217,
+      "learning_rate": 2.336499552299223e-06,
+      "loss": 0.3364,
+      "step": 13656
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3450572288678373,
+      "learning_rate": 2.3353041915118357e-06,
+      "loss": 0.1711,
+      "step": 13657
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5783242085477613,
+      "learning_rate": 2.3341090961525347e-06,
+      "loss": 0.3407,
+      "step": 13658
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26315711604683223,
+      "learning_rate": 2.3329142662627026e-06,
+      "loss": 0.2855,
+      "step": 13659
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3037750222352267,
+      "learning_rate": 2.3317197018837233e-06,
+      "loss": 0.3251,
+      "step": 13660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6003607719198905,
+      "learning_rate": 2.330525403056961e-06,
+      "loss": 0.3302,
+      "step": 13661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.397572728143814,
+      "learning_rate": 2.329331369823774e-06,
+      "loss": 0.3095,
+      "step": 13662
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23107023596230547,
+      "learning_rate": 2.3281376022255107e-06,
+      "loss": 0.1822,
+      "step": 13663
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5463080263376949,
+      "learning_rate": 2.326944100303511e-06,
+      "loss": 0.323,
+      "step": 13664
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.389194949108078,
+      "learning_rate": 2.32575086409911e-06,
+      "loss": 0.3304,
+      "step": 13665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5020850413515092,
+      "learning_rate": 2.3245578936536263e-06,
+      "loss": 0.2403,
+      "step": 13666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3099713566620966,
+      "learning_rate": 2.323365189008372e-06,
+      "loss": 0.2437,
+      "step": 13667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4927324531266989,
+      "learning_rate": 2.3221727502046487e-06,
+      "loss": 0.3289,
+      "step": 13668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5103081683200503,
+      "learning_rate": 2.3209805772837557e-06,
+      "loss": 0.2563,
+      "step": 13669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23731099735038388,
+      "learning_rate": 2.3197886702869756e-06,
+      "loss": 0.1511,
+      "step": 13670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2998413324216504,
+      "learning_rate": 2.3185970292555827e-06,
+      "loss": 0.294,
+      "step": 13671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1268300194743812,
+      "learning_rate": 2.317405654230842e-06,
+      "loss": 0.5717,
+      "step": 13672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.42392272265792735,
+      "learning_rate": 2.3162145452540164e-06,
+      "loss": 0.2239,
+      "step": 13673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5320072799016502,
+      "learning_rate": 2.3150237023663503e-06,
+      "loss": 0.3257,
+      "step": 13674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3519616895462257,
+      "learning_rate": 2.3138331256090853e-06,
+      "loss": 0.2953,
+      "step": 13675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2006587422306295,
+      "learning_rate": 2.312642815023444e-06,
+      "loss": 0.1078,
+      "step": 13676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3494649945683863,
+      "learning_rate": 2.311452770650653e-06,
+      "loss": 0.2756,
+      "step": 13677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27624763479241404,
+      "learning_rate": 2.3102629925319233e-06,
+      "loss": 0.2441,
+      "step": 13678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5364171797371735,
+      "learning_rate": 2.3090734807084545e-06,
+      "loss": 0.2007,
+      "step": 13679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3835854642103055,
+      "learning_rate": 2.307884235221438e-06,
+      "loss": 0.2958,
+      "step": 13680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5006981875240444,
+      "learning_rate": 2.3066952561120616e-06,
+      "loss": 0.2308,
+      "step": 13681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3935079231579874,
+      "learning_rate": 2.3055065434214983e-06,
+      "loss": 0.2848,
+      "step": 13682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2531872170122483,
+      "learning_rate": 2.3043180971909128e-06,
+      "loss": 0.2199,
+      "step": 13683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.42624480063631087,
+      "learning_rate": 2.3031299174614572e-06,
+      "loss": 0.2557,
+      "step": 13684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6061150551108329,
+      "learning_rate": 2.3019420042742856e-06,
+      "loss": 0.3503,
+      "step": 13685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24096818349359286,
+      "learning_rate": 2.3007543576705303e-06,
+      "loss": 0.2254,
+      "step": 13686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5103883995679313,
+      "learning_rate": 2.299566977691321e-06,
+      "loss": 0.508,
+      "step": 13687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5508991805415692,
+      "learning_rate": 2.2983798643777755e-06,
+      "loss": 0.3538,
+      "step": 13688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3684502029643531,
+      "learning_rate": 2.297193017771002e-06,
+      "loss": 0.2364,
+      "step": 13689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28819135601703405,
+      "learning_rate": 2.296006437912106e-06,
+      "loss": 0.2523,
+      "step": 13690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.37804061739241884,
+      "learning_rate": 2.2948201248421754e-06,
+      "loss": 0.2241,
+      "step": 13691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.40514152634974704,
+      "learning_rate": 2.2936340786022926e-06,
+      "loss": 0.2199,
+      "step": 13692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.195907334850774,
+      "learning_rate": 2.2924482992335272e-06,
+      "loss": 0.4781,
+      "step": 13693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3126316257768985,
+      "learning_rate": 2.291262786776949e-06,
+      "loss": 0.2628,
+      "step": 13694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3317162176271463,
+      "learning_rate": 2.2900775412736086e-06,
+      "loss": 0.2689,
+      "step": 13695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.502522546503961,
+      "learning_rate": 2.288892562764552e-06,
+      "loss": 0.2287,
+      "step": 13696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4025773704640404,
+      "learning_rate": 2.28770785129081e-06,
+      "loss": 0.2352,
+      "step": 13697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3119324736608787,
+      "learning_rate": 2.286523406893418e-06,
+      "loss": 0.2736,
+      "step": 13698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46774707777608343,
+      "learning_rate": 2.285339229613388e-06,
+      "loss": 0.2402,
+      "step": 13699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7445728385464858,
+      "learning_rate": 2.2841553194917288e-06,
+      "loss": 0.5074,
+      "step": 13700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3503867920780646,
+      "learning_rate": 2.2829716765694397e-06,
+      "loss": 0.2881,
+      "step": 13701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31801113224197763,
+      "learning_rate": 2.2817883008875065e-06,
+      "loss": 0.2597,
+      "step": 13702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2962235175090054,
+      "learning_rate": 2.2806051924869144e-06,
+      "loss": 0.1875,
+      "step": 13703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3308268216739318,
+      "learning_rate": 2.2794223514086333e-06,
+      "loss": 0.25,
+      "step": 13704
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2976921395533858,
+      "learning_rate": 2.2782397776936237e-06,
+      "loss": 0.2965,
+      "step": 13705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3459148821492143,
+      "learning_rate": 2.277057471382836e-06,
+      "loss": 0.3103,
+      "step": 13706
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3780665332879215,
+      "learning_rate": 2.2758754325172194e-06,
+      "loss": 0.2764,
+      "step": 13707
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7272877027788468,
+      "learning_rate": 2.274693661137707e-06,
+      "loss": 0.3907,
+      "step": 13708
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.16950972666281172,
+      "learning_rate": 2.273512157285215e-06,
+      "loss": 0.0706,
+      "step": 13709
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.40555071835046586,
+      "learning_rate": 2.272330921000667e-06,
+      "loss": 0.2703,
+      "step": 13710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3774955734684031,
+      "learning_rate": 2.271149952324968e-06,
+      "loss": 0.3077,
+      "step": 13711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6892753385861591,
+      "learning_rate": 2.2699692512990135e-06,
+      "loss": 0.2721,
+      "step": 13712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.341879768930826,
+      "learning_rate": 2.268788817963692e-06,
+      "loss": 0.2822,
+      "step": 13713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36570922717372284,
+      "learning_rate": 2.2676086523598773e-06,
+      "loss": 0.3289,
+      "step": 13714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3003448965263176,
+      "learning_rate": 2.266428754528446e-06,
+      "loss": 0.0964,
+      "step": 13715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33720536672084483,
+      "learning_rate": 2.2652491245102537e-06,
+      "loss": 0.2491,
+      "step": 13716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.161787269663438,
+      "learning_rate": 2.2640697623461517e-06,
+      "loss": 0.5815,
+      "step": 13717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29078226584492334,
+      "learning_rate": 2.262890668076979e-06,
+      "loss": 0.2155,
+      "step": 13718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36384414626106976,
+      "learning_rate": 2.2617118417435725e-06,
+      "loss": 0.2685,
+      "step": 13719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9616109497695068,
+      "learning_rate": 2.260533283386751e-06,
+      "loss": 0.4991,
+      "step": 13720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30110611270374804,
+      "learning_rate": 2.25935499304733e-06,
+      "loss": 0.1999,
+      "step": 13721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23948729147878228,
+      "learning_rate": 2.2581769707661107e-06,
+      "loss": 0.2069,
+      "step": 13722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4181780988657122,
+      "learning_rate": 2.256999216583892e-06,
+      "loss": 0.4937,
+      "step": 13723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7846008419436977,
+      "learning_rate": 2.2558217305414564e-06,
+      "loss": 0.4049,
+      "step": 13724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2869992147760968,
+      "learning_rate": 2.2546445126795822e-06,
+      "loss": 0.1819,
+      "step": 13725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34869844428817903,
+      "learning_rate": 2.2534675630390366e-06,
+      "loss": 0.3234,
+      "step": 13726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3008860845713572,
+      "learning_rate": 2.2522908816605716e-06,
+      "loss": 0.1825,
+      "step": 13727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28978102239520515,
+      "learning_rate": 2.251114468584944e-06,
+      "loss": 0.1858,
+      "step": 13728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8083556054139469,
+      "learning_rate": 2.2499383238528894e-06,
+      "loss": 0.4195,
+      "step": 13729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3769533565114214,
+      "learning_rate": 2.2487624475051364e-06,
+      "loss": 0.2837,
+      "step": 13730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39109306811626965,
+      "learning_rate": 2.2475868395824043e-06,
+      "loss": 0.2734,
+      "step": 13731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7981003418145786,
+      "learning_rate": 2.2464115001254096e-06,
+      "loss": 0.3579,
+      "step": 13732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4237085330933973,
+      "learning_rate": 2.245236429174851e-06,
+      "loss": 0.2282,
+      "step": 13733
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26938040168964505,
+      "learning_rate": 2.244061626771421e-06,
+      "loss": 0.2416,
+      "step": 13734
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33512199491603817,
+      "learning_rate": 2.2428870929558012e-06,
+      "loss": 0.2038,
+      "step": 13735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7622171998110745,
+      "learning_rate": 2.2417128277686694e-06,
+      "loss": 0.4113,
+      "step": 13736
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34080168091289903,
+      "learning_rate": 2.2405388312506903e-06,
+      "loss": 0.2377,
+      "step": 13737
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34730323144893105,
+      "learning_rate": 2.239365103442517e-06,
+      "loss": 0.2583,
+      "step": 13738
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0072696237450594,
+      "learning_rate": 2.238191644384794e-06,
+      "loss": 0.5047,
+      "step": 13739
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24308109364594496,
+      "learning_rate": 2.237018454118163e-06,
+      "loss": 0.2012,
+      "step": 13740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5470089584947886,
+      "learning_rate": 2.2358455326832496e-06,
+      "loss": 0.2059,
+      "step": 13741
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34034621021093414,
+      "learning_rate": 2.234672880120674e-06,
+      "loss": 0.2917,
+      "step": 13742
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3159765482745945,
+      "learning_rate": 2.233500496471037e-06,
+      "loss": 0.2669,
+      "step": 13743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1613176017519513,
+      "learning_rate": 2.2323283817749463e-06,
+      "loss": 0.7208,
+      "step": 13744
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.338617416056063,
+      "learning_rate": 2.2311565360729903e-06,
+      "loss": 0.2283,
+      "step": 13745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3873334398545606,
+      "learning_rate": 2.2299849594057487e-06,
+      "loss": 0.2524,
+      "step": 13746
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2770440136651813,
+      "learning_rate": 2.2288136518137914e-06,
+      "loss": 0.2116,
+      "step": 13747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.678257433651019,
+      "learning_rate": 2.227642613337686e-06,
+      "loss": 0.2842,
+      "step": 13748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34415897142683244,
+      "learning_rate": 2.2264718440179835e-06,
+      "loss": 0.2728,
+      "step": 13749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3417732021973484,
+      "learning_rate": 2.2253013438952253e-06,
+      "loss": 0.2948,
+      "step": 13750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7331090075356483,
+      "learning_rate": 2.224131113009945e-06,
+      "loss": 0.1448,
+      "step": 13751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.35091864102953746,
+      "learning_rate": 2.222961151402674e-06,
+      "loss": 0.2522,
+      "step": 13752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44152654129417956,
+      "learning_rate": 2.2217914591139222e-06,
+      "loss": 0.3178,
+      "step": 13753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2511167710167237,
+      "learning_rate": 2.2206220361841978e-06,
+      "loss": 0.14,
+      "step": 13754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31288446501141626,
+      "learning_rate": 2.2194528826539984e-06,
+      "loss": 0.2369,
+      "step": 13755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.312231502877169,
+      "learning_rate": 2.218283998563808e-06,
+      "loss": 0.6555,
+      "step": 13756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5150153647059441,
+      "learning_rate": 2.2171153839541114e-06,
+      "loss": 0.3475,
+      "step": 13757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27675543088755755,
+      "learning_rate": 2.2159470388653737e-06,
+      "loss": 0.2179,
+      "step": 13758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8129461695250663,
+      "learning_rate": 2.2147789633380555e-06,
+      "loss": 0.401,
+      "step": 13759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3186014787104289,
+      "learning_rate": 2.213611157412605e-06,
+      "loss": 0.1735,
+      "step": 13760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2907238321322088,
+      "learning_rate": 2.2124436211294676e-06,
+      "loss": 0.1941,
+      "step": 13761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4140887679030627,
+      "learning_rate": 2.2112763545290728e-06,
+      "loss": 0.3082,
+      "step": 13762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1476560055281388,
+      "learning_rate": 2.2101093576518416e-06,
+      "loss": 0.3793,
+      "step": 13763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3858913981085996,
+      "learning_rate": 2.2089426305381865e-06,
+      "loss": 0.2011,
+      "step": 13764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39061485494929044,
+      "learning_rate": 2.2077761732285165e-06,
+      "loss": 0.3134,
+      "step": 13765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31571158619901163,
+      "learning_rate": 2.206609985763222e-06,
+      "loss": 0.2483,
+      "step": 13766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3321054971659981,
+      "learning_rate": 2.2054440681826896e-06,
+      "loss": 0.1801,
+      "step": 13767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34644794016271147,
+      "learning_rate": 2.2042784205272927e-06,
+      "loss": 0.2443,
+      "step": 13768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48747689057722515,
+      "learning_rate": 2.203113042837396e-06,
+      "loss": 0.324,
+      "step": 13769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3253743126407448,
+      "learning_rate": 2.2019479351533625e-06,
+      "loss": 0.2552,
+      "step": 13770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5069718524012903,
+      "learning_rate": 2.2007830975155366e-06,
+      "loss": 0.2568,
+      "step": 13771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1996810800114002,
+      "learning_rate": 2.199618529964257e-06,
+      "loss": 0.6718,
+      "step": 13772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3826603271787008,
+      "learning_rate": 2.198454232539848e-06,
+      "loss": 0.2895,
+      "step": 13773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19718297829960185,
+      "learning_rate": 2.1972902052826384e-06,
+      "loss": 0.1763,
+      "step": 13774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6762841450323746,
+      "learning_rate": 2.1961264482329326e-06,
+      "loss": 0.3604,
+      "step": 13775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.523526720693821,
+      "learning_rate": 2.194962961431032e-06,
+      "loss": 0.2805,
+      "step": 13776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.348195317352207,
+      "learning_rate": 2.1937997449172287e-06,
+      "loss": 0.2318,
+      "step": 13777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49968051927791735,
+      "learning_rate": 2.192636798731804e-06,
+      "loss": 0.3517,
+      "step": 13778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3862758209154229,
+      "learning_rate": 2.1914741229150315e-06,
+      "loss": 0.2706,
+      "step": 13779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20524174445611812,
+      "learning_rate": 2.1903117175071754e-06,
+      "loss": 0.1168,
+      "step": 13780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33607769190284076,
+      "learning_rate": 2.1891495825484856e-06,
+      "loss": 0.2942,
+      "step": 13781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.681396311351655,
+      "learning_rate": 2.1879877180792117e-06,
+      "loss": 0.3332,
+      "step": 13782
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3897404025541323,
+      "learning_rate": 2.186826124139587e-06,
+      "loss": 0.2887,
+      "step": 13783
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.199733904627748,
+      "learning_rate": 2.185664800769839e-06,
+      "loss": 0.3483,
+      "step": 13784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3624521652254083,
+      "learning_rate": 2.1845037480101793e-06,
+      "loss": 0.2673,
+      "step": 13785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23852031573910368,
+      "learning_rate": 2.183342965900821e-06,
+      "loss": 0.2303,
+      "step": 13786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6022579071701457,
+      "learning_rate": 2.18218245448196e-06,
+      "loss": 0.2888,
+      "step": 13787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29140171805652,
+      "learning_rate": 2.1810222137937855e-06,
+      "loss": 0.1838,
+      "step": 13788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3210911846698981,
+      "learning_rate": 2.1798622438764716e-06,
+      "loss": 0.2784,
+      "step": 13789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4872387173728299,
+      "learning_rate": 2.1787025447701947e-06,
+      "loss": 0.2583,
+      "step": 13790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3987621580560564,
+      "learning_rate": 2.177543116515113e-06,
+      "loss": 0.286,
+      "step": 13791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4894773404094698,
+      "learning_rate": 2.176383959151377e-06,
+      "loss": 0.3393,
+      "step": 13792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2736323672900733,
+      "learning_rate": 2.175225072719127e-06,
+      "loss": 0.2155,
+      "step": 13793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2835026201314597,
+      "learning_rate": 2.174066457258495e-06,
+      "loss": 0.1832,
+      "step": 13794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5998237761061286,
+      "learning_rate": 2.172908112809606e-06,
+      "loss": 0.3561,
+      "step": 13795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1528927782639422,
+      "learning_rate": 2.1717500394125735e-06,
+      "loss": 0.6232,
+      "step": 13796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2608272389082933,
+      "learning_rate": 2.1705922371075005e-06,
+      "loss": 0.2213,
+      "step": 13797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5306546366421715,
+      "learning_rate": 2.169434705934479e-06,
+      "loss": 0.308,
+      "step": 13798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4938087384739634,
+      "learning_rate": 2.1682774459335987e-06,
+      "loss": 0.2987,
+      "step": 13799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1622853869015634,
+      "learning_rate": 2.1671204571449345e-06,
+      "loss": 0.0694,
+      "step": 13800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3014907139564795,
+      "learning_rate": 2.165963739608552e-06,
+      "loss": 0.274,
+      "step": 13801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4934739195565253,
+      "learning_rate": 2.164807293364506e-06,
+      "loss": 0.3363,
+      "step": 13802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4667111156007383,
+      "learning_rate": 2.1636511184528484e-06,
+      "loss": 0.1946,
+      "step": 13803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4077960646746423,
+      "learning_rate": 2.162495214913616e-06,
+      "loss": 0.3125,
+      "step": 13804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32122695724294886,
+      "learning_rate": 2.1613395827868366e-06,
+      "loss": 0.2702,
+      "step": 13805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.14965512931372155,
+      "learning_rate": 2.160184222112531e-06,
+      "loss": 0.0691,
+      "step": 13806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4639504102647323,
+      "learning_rate": 2.159029132930707e-06,
+      "loss": 0.2847,
+      "step": 13807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.364104155468768,
+      "learning_rate": 2.1578743152813676e-06,
+      "loss": 0.7156,
+      "step": 13808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3139451351224941,
+      "learning_rate": 2.156719769204505e-06,
+      "loss": 0.2972,
+      "step": 13809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3411874232698844,
+      "learning_rate": 2.155565494740098e-06,
+      "loss": 0.2292,
+      "step": 13810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5518338776407492,
+      "learning_rate": 2.1544114919281223e-06,
+      "loss": 0.616,
+      "step": 13811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20569821108771164,
+      "learning_rate": 2.153257760808538e-06,
+      "loss": 0.146,
+      "step": 13812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2839430285999692,
+      "learning_rate": 2.152104301421302e-06,
+      "loss": 0.2202,
+      "step": 13813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4522044299979987,
+      "learning_rate": 2.150951113806351e-06,
+      "loss": 0.3257,
+      "step": 13814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4669588799542236,
+      "learning_rate": 2.1497981980036297e-06,
+      "loss": 0.3026,
+      "step": 13815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.37379373753426265,
+      "learning_rate": 2.1486455540530593e-06,
+      "loss": 0.2319,
+      "step": 13816
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3565690261654594,
+      "learning_rate": 2.1474931819945555e-06,
+      "loss": 0.3044,
+      "step": 13817
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27704374870348236,
+      "learning_rate": 2.1463410818680253e-06,
+      "loss": 0.1727,
+      "step": 13818
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3302566985170461,
+      "learning_rate": 2.1451892537133624e-06,
+      "loss": 0.2454,
+      "step": 13819
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.41198364612002364,
+      "learning_rate": 2.1440376975704614e-06,
+      "loss": 0.2576,
+      "step": 13820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.42590625607150334,
+      "learning_rate": 2.142886413479197e-06,
+      "loss": 0.3075,
+      "step": 13821
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3323355752642059,
+      "learning_rate": 2.1417354014794378e-06,
+      "loss": 0.2635,
+      "step": 13822
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.287156112327123,
+      "learning_rate": 2.1405846616110416e-06,
+      "loss": 0.3389,
+      "step": 13823
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27619945652710726,
+      "learning_rate": 2.1394341939138618e-06,
+      "loss": 0.149,
+      "step": 13824
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2555642275552039,
+      "learning_rate": 2.1382839984277395e-06,
+      "loss": 0.2574,
+      "step": 13825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.687890090018683,
+      "learning_rate": 2.137134075192504e-06,
+      "loss": 0.2607,
+      "step": 13826
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6251334392979514,
+      "learning_rate": 2.135984424247974e-06,
+      "loss": 0.3902,
+      "step": 13827
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.35576024720872323,
+      "learning_rate": 2.1348350456339684e-06,
+      "loss": 0.2901,
+      "step": 13828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33393058087833427,
+      "learning_rate": 2.1336859393902864e-06,
+      "loss": 0.2403,
+      "step": 13829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3272136365997171,
+      "learning_rate": 2.1325371055567236e-06,
+      "loss": 0.16,
+      "step": 13830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38288933342849923,
+      "learning_rate": 2.1313885441730607e-06,
+      "loss": 0.2884,
+      "step": 13831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6989908981679257,
+      "learning_rate": 2.1302402552790723e-06,
+      "loss": 0.3487,
+      "step": 13832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26166129345641775,
+      "learning_rate": 2.1290922389145284e-06,
+      "loss": 0.2073,
+      "step": 13833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38623162282324663,
+      "learning_rate": 2.1279444951191806e-06,
+      "loss": 0.3036,
+      "step": 13834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1632051177823823,
+      "learning_rate": 2.1267970239327773e-06,
+      "loss": 0.5774,
+      "step": 13835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27563291221288966,
+      "learning_rate": 2.1256498253950518e-06,
+      "loss": 0.1642,
+      "step": 13836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27018901283686697,
+      "learning_rate": 2.124502899545737e-06,
+      "loss": 0.2479,
+      "step": 13837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5005766607922495,
+      "learning_rate": 2.1233562464245483e-06,
+      "loss": 0.2709,
+      "step": 13838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.167362155705901,
+      "learning_rate": 2.122209866071194e-06,
+      "loss": 0.4289,
+      "step": 13839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3193880909220907,
+      "learning_rate": 2.12106375852537e-06,
+      "loss": 0.2474,
+      "step": 13840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35005318057106716,
+      "learning_rate": 2.119917923826773e-06,
+      "loss": 0.2889,
+      "step": 13841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33854876303697296,
+      "learning_rate": 2.118772362015078e-06,
+      "loss": 0.0908,
+      "step": 13842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3633576105580375,
+      "learning_rate": 2.117627073129961e-06,
+      "loss": 0.2801,
+      "step": 13843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.42219924579575663,
+      "learning_rate": 2.1164820572110734e-06,
+      "loss": 0.2845,
+      "step": 13844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35881211061754714,
+      "learning_rate": 2.115337314298077e-06,
+      "loss": 0.3175,
+      "step": 13845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2868125822886127,
+      "learning_rate": 2.1141928444306094e-06,
+      "loss": 0.2033,
+      "step": 13846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1661974844904361,
+      "learning_rate": 2.113048647648305e-06,
+      "loss": 0.4777,
+      "step": 13847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3302764354075652,
+      "learning_rate": 2.1119047239907833e-06,
+      "loss": 0.2756,
+      "step": 13848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29027735567796065,
+      "learning_rate": 2.110761073497665e-06,
+      "loss": 0.1903,
+      "step": 13849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48117388677328654,
+      "learning_rate": 2.1096176962085513e-06,
+      "loss": 0.262,
+      "step": 13850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4337568500734956,
+      "learning_rate": 2.1084745921630377e-06,
+      "loss": 0.3096,
+      "step": 13851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.284048549328376,
+      "learning_rate": 2.107331761400707e-06,
+      "loss": 0.1813,
+      "step": 13852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3678509849823864,
+      "learning_rate": 2.1061892039611407e-06,
+      "loss": 0.2978,
+      "step": 13853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9895870132408121,
+      "learning_rate": 2.105046919883903e-06,
+      "loss": 0.4123,
+      "step": 13854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30690631485293857,
+      "learning_rate": 2.1039049092085507e-06,
+      "loss": 0.2116,
+      "step": 13855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2885275065069769,
+      "learning_rate": 2.102763171974629e-06,
+      "loss": 0.232,
+      "step": 13856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39902371245566787,
+      "learning_rate": 2.1016217082216815e-06,
+      "loss": 0.2799,
+      "step": 13857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31971621759090113,
+      "learning_rate": 2.100480517989235e-06,
+      "loss": 0.259,
+      "step": 13858
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8751263325898722,
+      "learning_rate": 2.099339601316809e-06,
+      "loss": 0.2952,
+      "step": 13859
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34455456165955256,
+      "learning_rate": 2.098198958243911e-06,
+      "loss": 0.303,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3229178337704223,
+      "learning_rate": 2.0970585888100425e-06,
+      "loss": 0.266,
+      "step": 13861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8776049310548688,
+      "learning_rate": 2.0959184930546973e-06,
+      "loss": 0.2564,
+      "step": 13862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47569729397496585,
+      "learning_rate": 2.0947786710173545e-06,
+      "loss": 0.2821,
+      "step": 13863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2504288310318657,
+      "learning_rate": 2.0936391227374874e-06,
+      "loss": 0.2158,
+      "step": 13864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35032213602144774,
+      "learning_rate": 2.0924998482545535e-06,
+      "loss": 0.2503,
+      "step": 13865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6552856340397468,
+      "learning_rate": 2.0913608476080138e-06,
+      "loss": 0.3592,
+      "step": 13866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3133790814738462,
+      "learning_rate": 2.090222120837306e-06,
+      "loss": 0.2509,
+      "step": 13867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47362331697258986,
+      "learning_rate": 2.089083667981868e-06,
+      "loss": 0.2264,
+      "step": 13868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4945426270079345,
+      "learning_rate": 2.087945489081119e-06,
+      "loss": 0.4079,
+      "step": 13869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3993652479527239,
+      "learning_rate": 2.0868075841744795e-06,
+      "loss": 0.2696,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26067501072855986,
+      "learning_rate": 2.0856699533013535e-06,
+      "loss": 0.1919,
+      "step": 13871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3616882633023731,
+      "learning_rate": 2.0845325965011375e-06,
+      "loss": 0.264,
+      "step": 13872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3935744589887008,
+      "learning_rate": 2.083395513813217e-06,
+      "loss": 0.277,
+      "step": 13873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.662955242143687,
+      "learning_rate": 2.082258705276966e-06,
+      "loss": 0.3611,
+      "step": 13874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0458268635013401,
+      "learning_rate": 2.0811221709317587e-06,
+      "loss": 0.5303,
+      "step": 13875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23084717455090384,
+      "learning_rate": 2.0799859108169496e-06,
+      "loss": 0.2034,
+      "step": 13876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3093465003844103,
+      "learning_rate": 2.0788499249718887e-06,
+      "loss": 0.2502,
+      "step": 13877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0381983387677298,
+      "learning_rate": 2.077714213435914e-06,
+      "loss": 0.2885,
+      "step": 13878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32579748606769765,
+      "learning_rate": 2.0765787762483545e-06,
+      "loss": 0.2467,
+      "step": 13879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1957787010002299,
+      "learning_rate": 2.075443613448532e-06,
+      "loss": 0.4765,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33500602148308517,
+      "learning_rate": 2.0743087250757544e-06,
+      "loss": 0.2632,
+      "step": 13881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3294045154993524,
+      "learning_rate": 2.073174111169327e-06,
+      "loss": 0.2471,
+      "step": 13882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8524795933817103,
+      "learning_rate": 2.072039771768539e-06,
+      "loss": 0.3734,
+      "step": 13883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21219881196196438,
+      "learning_rate": 2.0709057069126726e-06,
+      "loss": 0.2046,
+      "step": 13884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3405533089230494,
+      "learning_rate": 2.0697719166410013e-06,
+      "loss": 0.2002,
+      "step": 13885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5915861638020246,
+      "learning_rate": 2.068638400992784e-06,
+      "loss": 0.757,
+      "step": 13886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2417936476128963,
+      "learning_rate": 2.0675051600072817e-06,
+      "loss": 0.7115,
+      "step": 13887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2680697096757186,
+      "learning_rate": 2.0663721937237334e-06,
+      "loss": 0.1963,
+      "step": 13888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3929872162313532,
+      "learning_rate": 2.0652395021813752e-06,
+      "loss": 0.3059,
+      "step": 13889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36593534101852726,
+      "learning_rate": 2.064107085419429e-06,
+      "loss": 0.1969,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32655675523124694,
+      "learning_rate": 2.062974943477116e-06,
+      "loss": 0.1665,
+      "step": 13891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29384082638874254,
+      "learning_rate": 2.0618430763936402e-06,
+      "loss": 0.2684,
+      "step": 13892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.209531727244767,
+      "learning_rate": 2.0607114842081966e-06,
+      "loss": 0.7586,
+      "step": 13893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3941791338393727,
+      "learning_rate": 2.0595801669599704e-06,
+      "loss": 0.2967,
+      "step": 13894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5237764192291912,
+      "learning_rate": 2.0584491246881443e-06,
+      "loss": 0.2825,
+      "step": 13895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2794762521852401,
+      "learning_rate": 2.0573183574318832e-06,
+      "loss": 0.249,
+      "step": 13896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2774111684414447,
+      "learning_rate": 2.0561878652303458e-06,
+      "loss": 0.1959,
+      "step": 13897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5438540602781272,
+      "learning_rate": 2.0550576481226814e-06,
+      "loss": 0.212,
+      "step": 13898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6511150660244753,
+      "learning_rate": 2.0539277061480256e-06,
+      "loss": 0.4511,
+      "step": 13899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2423417364066042,
+      "learning_rate": 2.0527980393455147e-06,
+      "loss": 0.2466,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47122913695901136,
+      "learning_rate": 2.051668647754267e-06,
+      "loss": 0.2733,
+      "step": 13901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27193278941468646,
+      "learning_rate": 2.0505395314133915e-06,
+      "loss": 0.156,
+      "step": 13902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5962094572756668,
+      "learning_rate": 2.049410690361987e-06,
+      "loss": 0.304,
+      "step": 13903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26728260818607263,
+      "learning_rate": 2.0482821246391515e-06,
+      "loss": 0.2332,
+      "step": 13904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7884511286921686,
+      "learning_rate": 2.0471538342839637e-06,
+      "loss": 0.4379,
+      "step": 13905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5466734579354183,
+      "learning_rate": 2.0460258193354963e-06,
+      "loss": 0.3473,
+      "step": 13906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39502877345893916,
+      "learning_rate": 2.0448980798328113e-06,
+      "loss": 0.3107,
+      "step": 13907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19179045423602922,
+      "learning_rate": 2.043770615814966e-06,
+      "loss": 0.1632,
+      "step": 13908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5438731088748233,
+      "learning_rate": 2.0426434273210016e-06,
+      "loss": 0.2428,
+      "step": 13909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3807065412872983,
+      "learning_rate": 2.041516514389954e-06,
+      "loss": 0.2902,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7299876932540644,
+      "learning_rate": 2.0403898770608466e-06,
+      "loss": 0.3083,
+      "step": 13911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3073229067569145,
+      "learning_rate": 2.0392635153726958e-06,
+      "loss": 0.2584,
+      "step": 13912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38588900994842523,
+      "learning_rate": 2.0381374293645072e-06,
+      "loss": 0.2999,
+      "step": 13913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47918109962924327,
+      "learning_rate": 2.0370116190752763e-06,
+      "loss": 0.1199,
+      "step": 13914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3032568846110088,
+      "learning_rate": 2.035886084543989e-06,
+      "loss": 0.19,
+      "step": 13915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3048366839804276,
+      "learning_rate": 2.0347608258096263e-06,
+      "loss": 0.2932,
+      "step": 13916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.860092659438108,
+      "learning_rate": 2.0336358429111534e-06,
+      "loss": 0.3054,
+      "step": 13917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.410237488915634,
+      "learning_rate": 2.0325111358875295e-06,
+      "loss": 0.2969,
+      "step": 13918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5844699302701855,
+      "learning_rate": 2.031386704777698e-06,
+      "loss": 0.3557,
+      "step": 13919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3807335069363862,
+      "learning_rate": 2.0302625496206065e-06,
+      "loss": 0.2834,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19674851030297955,
+      "learning_rate": 2.0291386704551795e-06,
+      "loss": 0.0822,
+      "step": 13921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4132502470459073,
+      "learning_rate": 2.028015067320338e-06,
+      "loss": 0.3011,
+      "step": 13922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6473087562464866,
+      "learning_rate": 2.0268917402549914e-06,
+      "loss": 0.3178,
+      "step": 13923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36673530265989734,
+      "learning_rate": 2.0257686892980387e-06,
+      "loss": 0.2243,
+      "step": 13924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36203099449274834,
+      "learning_rate": 2.0246459144883767e-06,
+      "loss": 0.2867,
+      "step": 13925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5054331029168437,
+      "learning_rate": 2.023523415864883e-06,
+      "loss": 0.2783,
+      "step": 13926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28828519889572235,
+      "learning_rate": 2.02240119346643e-06,
+      "loss": 0.0938,
+      "step": 13927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3016104369032031,
+      "learning_rate": 2.0212792473318788e-06,
+      "loss": 0.2868,
+      "step": 13928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6227814296163309,
+      "learning_rate": 2.020157577500086e-06,
+      "loss": 0.379,
+      "step": 13929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36615003349825465,
+      "learning_rate": 2.019036184009894e-06,
+      "loss": 0.2155,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40065400364964354,
+      "learning_rate": 2.0179150669001347e-06,
+      "loss": 0.3056,
+      "step": 13931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.359059617026297,
+      "learning_rate": 2.0167942262096317e-06,
+      "loss": 0.3046,
+      "step": 13932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3025116844483224,
+      "learning_rate": 2.0156736619772034e-06,
+      "loss": 0.1751,
+      "step": 13933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3910522596201756,
+      "learning_rate": 2.0145533742416536e-06,
+      "loss": 0.2408,
+      "step": 13934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8924655555675018,
+      "learning_rate": 2.013433363041777e-06,
+      "loss": 0.4662,
+      "step": 13935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3086428880041467,
+      "learning_rate": 2.012313628416359e-06,
+      "loss": 0.2813,
+      "step": 13936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3740806939413502,
+      "learning_rate": 2.0111941704041738e-06,
+      "loss": 0.2224,
+      "step": 13937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.312160171802165,
+      "learning_rate": 2.0100749890439943e-06,
+      "loss": 0.5937,
+      "step": 13938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39192858923914753,
+      "learning_rate": 2.0089560843745737e-06,
+      "loss": 0.2937,
+      "step": 13939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28806615219896364,
+      "learning_rate": 2.0078374564346605e-06,
+      "loss": 0.2294,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28823469322612705,
+      "learning_rate": 2.0067191052629897e-06,
+      "loss": 0.1929,
+      "step": 13941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1471591644672199,
+      "learning_rate": 2.0056010308982954e-06,
+      "loss": 0.6341,
+      "step": 13942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30332531713716976,
+      "learning_rate": 2.0044832333792942e-06,
+      "loss": 0.1903,
+      "step": 13943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37927445804426185,
+      "learning_rate": 2.003365712744694e-06,
+      "loss": 0.286,
+      "step": 13944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8720692601753824,
+      "learning_rate": 2.0022484690331957e-06,
+      "loss": 0.3939,
+      "step": 13945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.326152282045429,
+      "learning_rate": 2.0011315022834887e-06,
+      "loss": 0.2417,
+      "step": 13946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25737955441315835,
+      "learning_rate": 2.000014812534253e-06,
+      "loss": 0.121,
+      "step": 13947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3269429544155623,
+      "learning_rate": 1.9988983998241616e-06,
+      "loss": 0.2851,
+      "step": 13948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.262006983013765,
+      "learning_rate": 1.9977822641918722e-06,
+      "loss": 0.2173,
+      "step": 13949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0942238387697292,
+      "learning_rate": 1.996666405676041e-06,
+      "loss": 0.2634,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31107848659817255,
+      "learning_rate": 1.9955508243153075e-06,
+      "loss": 0.2901,
+      "step": 13951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36351275425339735,
+      "learning_rate": 1.9944355201483057e-06,
+      "loss": 0.2643,
+      "step": 13952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27098007640215643,
+      "learning_rate": 1.993320493213654e-06,
+      "loss": 0.1226,
+      "step": 13953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3990952691061257,
+      "learning_rate": 1.992205743549972e-06,
+      "loss": 0.2757,
+      "step": 13954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3607559119535387,
+      "learning_rate": 1.991091271195862e-06,
+      "loss": 0.2886,
+      "step": 13955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32360319442150737,
+      "learning_rate": 1.989977076189916e-06,
+      "loss": 0.254,
+      "step": 13956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5376372309933372,
+      "learning_rate": 1.9888631585707165e-06,
+      "loss": 0.3056,
+      "step": 13957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4290201609712683,
+      "learning_rate": 1.987749518376845e-06,
+      "loss": 0.2751,
+      "step": 13958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49871685728533977,
+      "learning_rate": 1.986636155646862e-06,
+      "loss": 0.3373,
+      "step": 13959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27146044669137814,
+      "learning_rate": 1.985523070419324e-06,
+      "loss": 0.1775,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2442071282951535,
+      "learning_rate": 1.984410262732779e-06,
+      "loss": 0.1945,
+      "step": 13961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6881344071763977,
+      "learning_rate": 1.9832977326257587e-06,
+      "loss": 0.3537,
+      "step": 13962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32311184584433045,
+      "learning_rate": 1.9821854801367947e-06,
+      "loss": 0.2321,
+      "step": 13963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29921551388686407,
+      "learning_rate": 1.981073505304404e-06,
+      "loss": 0.2423,
+      "step": 13964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1952034277542511,
+      "learning_rate": 1.9799618081670925e-06,
+      "loss": 0.6672,
+      "step": 13965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.16565970362132512,
+      "learning_rate": 1.978850388763356e-06,
+      "loss": 0.0987,
+      "step": 13966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28583421178478075,
+      "learning_rate": 1.977739247131688e-06,
+      "loss": 0.2643,
+      "step": 13967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45246208200159804,
+      "learning_rate": 1.976628383310566e-06,
+      "loss": 0.3104,
+      "step": 13968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7345956913222442,
+      "learning_rate": 1.9755177973384575e-06,
+      "loss": 0.2461,
+      "step": 13969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3468748142933442,
+      "learning_rate": 1.9744074892538203e-06,
+      "loss": 0.2622,
+      "step": 13970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4893550654394881,
+      "learning_rate": 1.9732974590951083e-06,
+      "loss": 0.3401,
+      "step": 13971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.313095481399203,
+      "learning_rate": 1.972187706900761e-06,
+      "loss": 0.256,
+      "step": 13972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28891551172507257,
+      "learning_rate": 1.9710782327092083e-06,
+      "loss": 0.1845,
+      "step": 13973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45100541255152954,
+      "learning_rate": 1.9699690365588674e-06,
+      "loss": 0.3064,
+      "step": 13974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.329722050488096,
+      "learning_rate": 1.9688601184881572e-06,
+      "loss": 0.2946,
+      "step": 13975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28484365336909273,
+      "learning_rate": 1.9677514785354747e-06,
+      "loss": 0.1848,
+      "step": 13976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0415921858447894,
+      "learning_rate": 1.966643116739214e-06,
+      "loss": 0.5841,
+      "step": 13977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.945207146977298,
+      "learning_rate": 1.9655350331377563e-06,
+      "loss": 0.4686,
+      "step": 13978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32279578400528663,
+      "learning_rate": 1.964427227769475e-06,
+      "loss": 0.1987,
+      "step": 13979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3187970274152468,
+      "learning_rate": 1.9633197006727333e-06,
+      "loss": 0.2472,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4706050178267718,
+      "learning_rate": 1.9622124518858855e-06,
+      "loss": 0.215,
+      "step": 13981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32904291628440924,
+      "learning_rate": 1.9611054814472707e-06,
+      "loss": 0.2531,
+      "step": 13982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3076825890196617,
+      "learning_rate": 1.959998789395231e-06,
+      "loss": 0.2521,
+      "step": 13983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.808660652729481,
+      "learning_rate": 1.9588923757680878e-06,
+      "loss": 0.496,
+      "step": 13984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3336398442274031,
+      "learning_rate": 1.9577862406041558e-06,
+      "loss": 0.2611,
+      "step": 13985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.741654787071884,
+      "learning_rate": 1.956680383941737e-06,
+      "loss": 0.1866,
+      "step": 13986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2751490630762147,
+      "learning_rate": 1.9555748058191337e-06,
+      "loss": 0.2453,
+      "step": 13987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3371685877572128,
+      "learning_rate": 1.9544695062746286e-06,
+      "loss": 0.2371,
+      "step": 13988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8680164276785397,
+      "learning_rate": 1.9533644853464996e-06,
+      "loss": 0.4721,
+      "step": 13989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47686024734591553,
+      "learning_rate": 1.952259743073012e-06,
+      "loss": 0.3312,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39545690708722075,
+      "learning_rate": 1.9511552794924194e-06,
+      "loss": 0.2362,
+      "step": 13991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36269709669257977,
+      "learning_rate": 1.9500510946429772e-06,
+      "loss": 0.2391,
+      "step": 13992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31630341720542904,
+      "learning_rate": 1.9489471885629196e-06,
+      "loss": 0.1847,
+      "step": 13993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40712626311337746,
+      "learning_rate": 1.9478435612904744e-06,
+      "loss": 0.2752,
+      "step": 13994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2942320985895319,
+      "learning_rate": 1.946740212863858e-06,
+      "loss": 0.288,
+      "step": 13995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7675447935881212,
+      "learning_rate": 1.945637143321284e-06,
+      "loss": 0.3028,
+      "step": 13996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.400006777628331,
+      "learning_rate": 1.9445343527009497e-06,
+      "loss": 0.2919,
+      "step": 13997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3556966067366299,
+      "learning_rate": 1.9434318410410435e-06,
+      "loss": 0.247,
+      "step": 13998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25185255386753236,
+      "learning_rate": 1.942329608379745e-06,
+      "loss": 0.183,
+      "step": 13999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31969515942588655,
+      "learning_rate": 1.9412276547552276e-06,
+      "loss": 0.2485,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9991830838318048,
+      "learning_rate": 1.9401259802056495e-06,
+      "loss": 0.5334,
+      "step": 14001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.42432293133392635,
+      "learning_rate": 1.9390245847691625e-06,
+      "loss": 0.2674,
+      "step": 14002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.257524155935746,
+      "learning_rate": 1.9379234684839075e-06,
+      "loss": 0.244,
+      "step": 14003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2512902112125146,
+      "learning_rate": 1.9368226313880134e-06,
+      "loss": 0.5929,
+      "step": 14004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2558079446835865,
+      "learning_rate": 1.935722073519608e-06,
+      "loss": 0.0865,
+      "step": 14005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30514259612396377,
+      "learning_rate": 1.9346217949168e-06,
+      "loss": 0.2563,
+      "step": 14006
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32676000041131353,
+      "learning_rate": 1.933521795617692e-06,
+      "loss": 0.2903,
+      "step": 14007
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6566724718751233,
+      "learning_rate": 1.932422075660376e-06,
+      "loss": 0.3683,
+      "step": 14008
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31808490481525514,
+      "learning_rate": 1.931322635082938e-06,
+      "loss": 0.175,
+      "step": 14009
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5415586763759216,
+      "learning_rate": 1.9302234739234507e-06,
+      "loss": 0.3217,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2155972087042718,
+      "learning_rate": 1.9291245922199776e-06,
+      "loss": 0.2108,
+      "step": 14011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40196080536904666,
+      "learning_rate": 1.9280259900105723e-06,
+      "loss": 0.1741,
+      "step": 14012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5085855651773016,
+      "learning_rate": 1.9269276673332806e-06,
+      "loss": 0.3481,
+      "step": 14013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6857508871116539,
+      "learning_rate": 1.9258296242261355e-06,
+      "loss": 0.4188,
+      "step": 14014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24640439130406608,
+      "learning_rate": 1.9247318607271637e-06,
+      "loss": 0.2135,
+      "step": 14015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5920314104837552,
+      "learning_rate": 1.923634376874378e-06,
+      "loss": 0.3719,
+      "step": 14016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25891151164406034,
+      "learning_rate": 1.9225371727057897e-06,
+      "loss": 0.1558,
+      "step": 14017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3841785028881913,
+      "learning_rate": 1.921440248259391e-06,
+      "loss": 0.2032,
+      "step": 14018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2969167120395358,
+      "learning_rate": 1.9203436035731694e-06,
+      "loss": 0.2832,
+      "step": 14019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8309038714782914,
+      "learning_rate": 1.919247238685098e-06,
+      "loss": 0.5118,
+      "step": 14020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7674720192887823,
+      "learning_rate": 1.918151153633151e-06,
+      "loss": 0.2878,
+      "step": 14021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.564361057127029,
+      "learning_rate": 1.917055348455281e-06,
+      "loss": 0.2283,
+      "step": 14022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2814577681747703,
+      "learning_rate": 1.9159598231894385e-06,
+      "loss": 0.2676,
+      "step": 14023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2937360059194529,
+      "learning_rate": 1.9148645778735555e-06,
+      "loss": 0.2098,
+      "step": 14024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5653575255098695,
+      "learning_rate": 1.9137696125455672e-06,
+      "loss": 0.2534,
+      "step": 14025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5171731933248757,
+      "learning_rate": 1.91267492724339e-06,
+      "loss": 0.3388,
+      "step": 14026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3752775546359505,
+      "learning_rate": 1.911580522004931e-06,
+      "loss": 0.2683,
+      "step": 14027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.368645831443498,
+      "learning_rate": 1.910486396868092e-06,
+      "loss": 0.2158,
+      "step": 14028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30316861367267955,
+      "learning_rate": 1.909392551870759e-06,
+      "loss": 0.1882,
+      "step": 14029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6934510092482077,
+      "learning_rate": 1.908298987050815e-06,
+      "loss": 0.3114,
+      "step": 14030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24347718869590526,
+      "learning_rate": 1.907205702446131e-06,
+      "loss": 0.2422,
+      "step": 14031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.259711777802538,
+      "learning_rate": 1.9061126980945644e-06,
+      "loss": 0.657,
+      "step": 14032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.572298799907,
+      "learning_rate": 1.9050199740339648e-06,
+      "loss": 0.3185,
+      "step": 14033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.42220044917716687,
+      "learning_rate": 1.9039275303021775e-06,
+      "loss": 0.2817,
+      "step": 14034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34749331845809894,
+      "learning_rate": 1.9028353669370315e-06,
+      "loss": 0.2275,
+      "step": 14035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49512336442006893,
+      "learning_rate": 1.9017434839763493e-06,
+      "loss": 0.3007,
+      "step": 14036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2254669338598251,
+      "learning_rate": 1.900651881457939e-06,
+      "loss": 0.1896,
+      "step": 14037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4076412787846548,
+      "learning_rate": 1.8995605594196086e-06,
+      "loss": 0.2732,
+      "step": 14038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32296030460251207,
+      "learning_rate": 1.8984695178991475e-06,
+      "loss": 0.2963,
+      "step": 14039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5481564846950949,
+      "learning_rate": 1.8973787569343394e-06,
+      "loss": 0.3206,
+      "step": 14040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6209723598460225,
+      "learning_rate": 1.8962882765629552e-06,
+      "loss": 0.2801,
+      "step": 14041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3055173237798795,
+      "learning_rate": 1.8951980768227586e-06,
+      "loss": 0.2729,
+      "step": 14042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.35224208256032374,
+      "learning_rate": 1.8941081577515053e-06,
+      "loss": 0.2993,
+      "step": 14043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28113135249311766,
+      "learning_rate": 1.8930185193869376e-06,
+      "loss": 0.1722,
+      "step": 14044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29298152458518495,
+      "learning_rate": 1.8919291617667912e-06,
+      "loss": 0.1652,
+      "step": 14045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.45232617515848045,
+      "learning_rate": 1.8908400849287889e-06,
+      "loss": 0.2993,
+      "step": 14046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37147528154819126,
+      "learning_rate": 1.8897512889106451e-06,
+      "loss": 0.3054,
+      "step": 14047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5620327131313082,
+      "learning_rate": 1.8886627737500663e-06,
+      "loss": 0.2291,
+      "step": 14048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26947274618159855,
+      "learning_rate": 1.8875745394847434e-06,
+      "loss": 0.2188,
+      "step": 14049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3342387244403811,
+      "learning_rate": 1.8864865861523684e-06,
+      "loss": 0.3209,
+      "step": 14050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.16985618467855765,
+      "learning_rate": 1.8853989137906137e-06,
+      "loss": 0.0886,
+      "step": 14051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.38349041687568947,
+      "learning_rate": 1.8843115224371467e-06,
+      "loss": 0.2769,
+      "step": 14052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6251738290863288,
+      "learning_rate": 1.8832244121296217e-06,
+      "loss": 0.3818,
+      "step": 14053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3763987776386038,
+      "learning_rate": 1.8821375829056842e-06,
+      "loss": 0.2174,
+      "step": 14054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31215780044633634,
+      "learning_rate": 1.8810510348029753e-06,
+      "loss": 0.2912,
+      "step": 14055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.225420066022846,
+      "learning_rate": 1.8799647678591203e-06,
+      "loss": 0.7848,
+      "step": 14056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2622067946912736,
+      "learning_rate": 1.878878782111736e-06,
+      "loss": 0.1747,
+      "step": 14057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.35062197492639324,
+      "learning_rate": 1.8777930775984277e-06,
+      "loss": 0.222,
+      "step": 14058
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36092983810301527,
+      "learning_rate": 1.8767076543567986e-06,
+      "loss": 0.305,
+      "step": 14059
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5759708595809394,
+      "learning_rate": 1.8756225124244332e-06,
+      "loss": 0.3023,
+      "step": 14060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3373531812343405,
+      "learning_rate": 1.8745376518389113e-06,
+      "loss": 0.2121,
+      "step": 14061
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3674552882376826,
+      "learning_rate": 1.8734530726377997e-06,
+      "loss": 0.3401,
+      "step": 14062
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20581830272049467,
+      "learning_rate": 1.8723687748586605e-06,
+      "loss": 0.1439,
+      "step": 14063
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3971041780692171,
+      "learning_rate": 1.8712847585390403e-06,
+      "loss": 0.2296,
+      "step": 14064
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7877507215245848,
+      "learning_rate": 1.8702010237164803e-06,
+      "loss": 0.3901,
+      "step": 14065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3564769039000346,
+      "learning_rate": 1.8691175704285091e-06,
+      "loss": 0.3047,
+      "step": 14066
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32404927429594854,
+      "learning_rate": 1.8680343987126448e-06,
+      "loss": 0.1954,
+      "step": 14067
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.125143640483104,
+      "learning_rate": 1.8669515086064006e-06,
+      "loss": 0.7126,
+      "step": 14068
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4371310503844042,
+      "learning_rate": 1.8658689001472775e-06,
+      "loss": 0.2545,
+      "step": 14069
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26684346084216576,
+      "learning_rate": 1.8647865733727644e-06,
+      "loss": 0.2519,
+      "step": 14070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30304389288387373,
+      "learning_rate": 1.8637045283203391e-06,
+      "loss": 0.1918,
+      "step": 14071
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9331663827577262,
+      "learning_rate": 1.8626227650274787e-06,
+      "loss": 0.4022,
+      "step": 14072
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3671040035406563,
+      "learning_rate": 1.8615412835316426e-06,
+      "loss": 0.2767,
+      "step": 14073
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33313642195654125,
+      "learning_rate": 1.8604600838702814e-06,
+      "loss": 0.2546,
+      "step": 14074
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6134663014323616,
+      "learning_rate": 1.8593791660808357e-06,
+      "loss": 0.3682,
+      "step": 14075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.35912139079324334,
+      "learning_rate": 1.8582985302007405e-06,
+      "loss": 0.2771,
+      "step": 14076
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3628918133959204,
+      "learning_rate": 1.8572181762674192e-06,
+      "loss": 0.1353,
+      "step": 14077
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31315581719439217,
+      "learning_rate": 1.8561381043182803e-06,
+      "loss": 0.2639,
+      "step": 14078
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3602637118248707,
+      "learning_rate": 1.8550583143907274e-06,
+      "loss": 0.2798,
+      "step": 14079
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7198841723664006,
+      "learning_rate": 1.8539788065221598e-06,
+      "loss": 0.351,
+      "step": 14080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6480899378091401,
+      "learning_rate": 1.8528995807499528e-06,
+      "loss": 0.3457,
+      "step": 14081
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2939543727068227,
+      "learning_rate": 1.8518206371114833e-06,
+      "loss": 0.264,
+      "step": 14082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2582146968956923,
+      "learning_rate": 1.8507419756441114e-06,
+      "loss": 0.193,
+      "step": 14083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4650946807988512,
+      "learning_rate": 1.8496635963851973e-06,
+      "loss": 0.1613,
+      "step": 14084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30943678757627774,
+      "learning_rate": 1.8485854993720831e-06,
+      "loss": 0.2519,
+      "step": 14085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33098435069101717,
+      "learning_rate": 1.8475076846421025e-06,
+      "loss": 0.303,
+      "step": 14086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8136153140125502,
+      "learning_rate": 1.8464301522325767e-06,
+      "loss": 0.3076,
+      "step": 14087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3524686554462998,
+      "learning_rate": 1.8453529021808282e-06,
+      "loss": 0.2519,
+      "step": 14088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4539966783959743,
+      "learning_rate": 1.8442759345241567e-06,
+      "loss": 0.2466,
+      "step": 14089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2479546022298384,
+      "learning_rate": 1.8431992492998595e-06,
+      "loss": 0.1885,
+      "step": 14090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2979881583987387,
+      "learning_rate": 1.8421228465452213e-06,
+      "loss": 0.2557,
+      "step": 14091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8722453179515083,
+      "learning_rate": 1.8410467262975152e-06,
+      "loss": 0.5817,
+      "step": 14092
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5745595604603996,
+      "learning_rate": 1.8399708885940136e-06,
+      "loss": 0.2642,
+      "step": 14093
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28733919145554143,
+      "learning_rate": 1.8388953334719684e-06,
+      "loss": 0.2482,
+      "step": 14094
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34111214746268476,
+      "learning_rate": 1.837820060968627e-06,
+      "loss": 0.2435,
+      "step": 14095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5477066304455874,
+      "learning_rate": 1.8367450711212232e-06,
+      "loss": 0.2656,
+      "step": 14096
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2817456644619849,
+      "learning_rate": 1.8356703639669904e-06,
+      "loss": 0.1833,
+      "step": 14097
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34795741382488465,
+      "learning_rate": 1.8345959395431401e-06,
+      "loss": 0.2963,
+      "step": 14098
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6400339452738504,
+      "learning_rate": 1.8335217978868825e-06,
+      "loss": 0.4564,
+      "step": 14099
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2979103780116249,
+      "learning_rate": 1.832447939035411e-06,
+      "loss": 0.1861,
+      "step": 14100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.375418491704937,
+      "learning_rate": 1.8313743630259184e-06,
+      "loss": 0.1906,
+      "step": 14101
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3613002064565976,
+      "learning_rate": 1.8303010698955803e-06,
+      "loss": 0.2885,
+      "step": 14102
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.331150447620705,
+      "learning_rate": 1.8292280596815649e-06,
+      "loss": 0.1993,
+      "step": 14103
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5396453246374567,
+      "learning_rate": 1.8281553324210278e-06,
+      "loss": 0.3703,
+      "step": 14104
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2441896449953618,
+      "learning_rate": 1.8270828881511238e-06,
+      "loss": 0.6757,
+      "step": 14105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22987355409113447,
+      "learning_rate": 1.8260107269089865e-06,
+      "loss": 0.206,
+      "step": 14106
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4125169977730276,
+      "learning_rate": 1.8249388487317465e-06,
+      "loss": 0.6422,
+      "step": 14107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3321933632573765,
+      "learning_rate": 1.823867253656524e-06,
+      "loss": 0.1871,
+      "step": 14108
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33090979186783415,
+      "learning_rate": 1.8227959417204222e-06,
+      "loss": 0.2569,
+      "step": 14109
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3767301347240297,
+      "learning_rate": 1.8217249129605496e-06,
+      "loss": 0.2513,
+      "step": 14110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9509296598908838,
+      "learning_rate": 1.820654167413991e-06,
+      "loss": 0.5372,
+      "step": 14111
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3118256550146787,
+      "learning_rate": 1.8195837051178267e-06,
+      "loss": 0.2466,
+      "step": 14112
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21152041292581364,
+      "learning_rate": 1.8185135261091247e-06,
+      "loss": 0.0666,
+      "step": 14113
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2945684219244831,
+      "learning_rate": 1.817443630424952e-06,
+      "loss": 0.2494,
+      "step": 14114
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3782111624907756,
+      "learning_rate": 1.8163740181023526e-06,
+      "loss": 0.2492,
+      "step": 14115
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4657972759725943,
+      "learning_rate": 1.8153046891783654e-06,
+      "loss": 0.2856,
+      "step": 14116
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.340304507867154,
+      "learning_rate": 1.8142356436900288e-06,
+      "loss": 0.3124,
+      "step": 14117
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3706279739554896,
+      "learning_rate": 1.8131668816743586e-06,
+      "loss": 0.2635,
+      "step": 14118
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6320484353599375,
+      "learning_rate": 1.8120984031683686e-06,
+      "loss": 0.2199,
+      "step": 14119
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3099710789466092,
+      "learning_rate": 1.811030208209058e-06,
+      "loss": 0.1903,
+      "step": 14120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4168781646528722,
+      "learning_rate": 1.8099622968334163e-06,
+      "loss": 0.2884,
+      "step": 14121
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28981006334400455,
+      "learning_rate": 1.8088946690784314e-06,
+      "loss": 0.2911,
+      "step": 14122
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0230460545064057,
+      "learning_rate": 1.8078273249810718e-06,
+      "loss": 0.4709,
+      "step": 14123
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36665359776477047,
+      "learning_rate": 1.806760264578299e-06,
+      "loss": 0.2622,
+      "step": 14124
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4271593257453827,
+      "learning_rate": 1.8056934879070642e-06,
+      "loss": 0.5857,
+      "step": 14125
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2838783227056397,
+      "learning_rate": 1.8046269950043138e-06,
+      "loss": 0.208,
+      "step": 14126
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3194121737248507,
+      "learning_rate": 1.803560785906977e-06,
+      "loss": 0.2564,
+      "step": 14127
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4439798130696629,
+      "learning_rate": 1.8024948606519787e-06,
+      "loss": 0.2703,
+      "step": 14128
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.301214908000095,
+      "learning_rate": 1.8014292192762285e-06,
+      "loss": 0.2513,
+      "step": 14129
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3262329360305483,
+      "learning_rate": 1.8003638618166342e-06,
+      "loss": 0.257,
+      "step": 14130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.184176185516564,
+      "learning_rate": 1.7992987883100877e-06,
+      "loss": 0.4787,
+      "step": 14131
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6445822576434146,
+      "learning_rate": 1.7982339987934705e-06,
+      "loss": 0.2847,
+      "step": 14132
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40948848681739713,
+      "learning_rate": 1.7971694933036576e-06,
+      "loss": 0.2681,
+      "step": 14133
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23710816941660431,
+      "learning_rate": 1.7961052718775096e-06,
+      "loss": 0.2382,
+      "step": 14134
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.47714786473531884,
+      "learning_rate": 1.7950413345518858e-06,
+      "loss": 0.2881,
+      "step": 14135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4062079094273239,
+      "learning_rate": 1.7939776813636278e-06,
+      "loss": 0.179,
+      "step": 14136
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3923200485078189,
+      "learning_rate": 1.7929143123495695e-06,
+      "loss": 0.2899,
+      "step": 14137
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4917783194741124,
+      "learning_rate": 1.7918512275465338e-06,
+      "loss": 0.3125,
+      "step": 14138
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3656188792929695,
+      "learning_rate": 1.790788426991339e-06,
+      "loss": 0.1853,
+      "step": 14139
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5413502468999466,
+      "learning_rate": 1.7897259107207888e-06,
+      "loss": 0.4162,
+      "step": 14140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32025923404560214,
+      "learning_rate": 1.7886636787716761e-06,
+      "loss": 0.2454,
+      "step": 14141
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2429282363963751,
+      "learning_rate": 1.787601731180786e-06,
+      "loss": 0.173,
+      "step": 14142
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5438855371114105,
+      "learning_rate": 1.7865400679848953e-06,
+      "loss": 0.3524,
+      "step": 14143
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6026762575249064,
+      "learning_rate": 1.7854786892207709e-06,
+      "loss": 0.4291,
+      "step": 14144
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.418008323429287,
+      "learning_rate": 1.7844175949251653e-06,
+      "loss": 0.2601,
+      "step": 14145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2651924993585563,
+      "learning_rate": 1.7833567851348254e-06,
+      "loss": 0.232,
+      "step": 14146
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2537043627469472,
+      "learning_rate": 1.7822962598864868e-06,
+      "loss": 0.1781,
+      "step": 14147
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37665750496002826,
+      "learning_rate": 1.7812360192168742e-06,
+      "loss": 0.2524,
+      "step": 14148
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.353511059581225,
+      "learning_rate": 1.7801760631627064e-06,
+      "loss": 0.2421,
+      "step": 14149
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7572573431074694,
+      "learning_rate": 1.7791163917606846e-06,
+      "loss": 0.3338,
+      "step": 14150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8686696236646662,
+      "learning_rate": 1.7780570050475122e-06,
+      "loss": 0.3671,
+      "step": 14151
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3290711277167447,
+      "learning_rate": 1.7769979030598706e-06,
+      "loss": 0.231,
+      "step": 14152
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26873437490942453,
+      "learning_rate": 1.7759390858344395e-06,
+      "loss": 0.2475,
+      "step": 14153
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3209443331607646,
+      "learning_rate": 1.7748805534078805e-06,
+      "loss": 0.1887,
+      "step": 14154
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.423738269074664,
+      "learning_rate": 1.773822305816857e-06,
+      "loss": 0.2495,
+      "step": 14155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8545198353982074,
+      "learning_rate": 1.7727643430980135e-06,
+      "loss": 0.411,
+      "step": 14156
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37808128113435296,
+      "learning_rate": 1.7717066652879877e-06,
+      "loss": 0.3064,
+      "step": 14157
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2977203707383954,
+      "learning_rate": 1.770649272423406e-06,
+      "loss": 0.276,
+      "step": 14158
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2657906575491931,
+      "learning_rate": 1.7695921645408832e-06,
+      "loss": 0.1227,
+      "step": 14159
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33109296938461874,
+      "learning_rate": 1.7685353416770322e-06,
+      "loss": 0.2161,
+      "step": 14160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24742271713944344,
+      "learning_rate": 1.7674788038684488e-06,
+      "loss": 0.258,
+      "step": 14161
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9296375061584038,
+      "learning_rate": 1.7664225511517196e-06,
+      "loss": 0.2892,
+      "step": 14162
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.427928964518346,
+      "learning_rate": 1.7653665835634214e-06,
+      "loss": 0.2758,
+      "step": 14163
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5014102721249964,
+      "learning_rate": 1.7643109011401272e-06,
+      "loss": 0.3853,
+      "step": 14164
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36698527266326064,
+      "learning_rate": 1.7632555039183918e-06,
+      "loss": 0.2581,
+      "step": 14165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40372847248005095,
+      "learning_rate": 1.762200391934764e-06,
+      "loss": 0.2874,
+      "step": 14166
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5512841267027948,
+      "learning_rate": 1.7611455652257802e-06,
+      "loss": 0.3146,
+      "step": 14167
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2772187100793543,
+      "learning_rate": 1.7600910238279745e-06,
+      "loss": 0.096,
+      "step": 14168
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32362380832747006,
+      "learning_rate": 1.7590367677778607e-06,
+      "loss": 0.2657,
+      "step": 14169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3396696390434374,
+      "learning_rate": 1.7579827971119501e-06,
+      "loss": 0.2754,
+      "step": 14170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9538742821374108,
+      "learning_rate": 1.756929111866741e-06,
+      "loss": 0.4811,
+      "step": 14171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6335048928071565,
+      "learning_rate": 1.7558757120787196e-06,
+      "loss": 0.2464,
+      "step": 14172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29133111716058585,
+      "learning_rate": 1.7548225977843703e-06,
+      "loss": 0.2619,
+      "step": 14173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23845768488865898,
+      "learning_rate": 1.7537697690201604e-06,
+      "loss": 0.1823,
+      "step": 14174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5880555726294913,
+      "learning_rate": 1.7527172258225479e-06,
+      "loss": 0.143,
+      "step": 14175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3607590540833608,
+      "learning_rate": 1.7516649682279807e-06,
+      "loss": 0.2992,
+      "step": 14176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3758488924017635,
+      "learning_rate": 1.7506129962729046e-06,
+      "loss": 0.2974,
+      "step": 14177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6228724462522425,
+      "learning_rate": 1.7495613099937447e-06,
+      "loss": 0.2398,
+      "step": 14178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28615182796675526,
+      "learning_rate": 1.748509909426922e-06,
+      "loss": 0.2267,
+      "step": 14179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5160793879630342,
+      "learning_rate": 1.747458794608844e-06,
+      "loss": 0.2544,
+      "step": 14180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28378991284710486,
+      "learning_rate": 1.7464079655759181e-06,
+      "loss": 0.2048,
+      "step": 14181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.39445167751605087,
+      "learning_rate": 1.7453574223645265e-06,
+      "loss": 0.3056,
+      "step": 14182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7048727007429841,
+      "learning_rate": 1.7443071650110532e-06,
+      "loss": 0.4061,
+      "step": 14183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.511720675101496,
+      "learning_rate": 1.743257193551865e-06,
+      "loss": 0.2747,
+      "step": 14184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27744333813057853,
+      "learning_rate": 1.742207508023327e-06,
+      "loss": 0.2208,
+      "step": 14185
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2413983160625698,
+      "learning_rate": 1.741158108461788e-06,
+      "loss": 0.1841,
+      "step": 14186
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9634224363895272,
+      "learning_rate": 1.7401089949035888e-06,
+      "loss": 0.4651,
+      "step": 14187
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3454464012521359,
+      "learning_rate": 1.7390601673850582e-06,
+      "loss": 0.2216,
+      "step": 14188
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36039692383388244,
+      "learning_rate": 1.7380116259425205e-06,
+      "loss": 0.2878,
+      "step": 14189
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0262897010065517,
+      "learning_rate": 1.7369633706122845e-06,
+      "loss": 0.457,
+      "step": 14190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32525366190097454,
+      "learning_rate": 1.7359154014306523e-06,
+      "loss": 0.1925,
+      "step": 14191
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25124789227221145,
+      "learning_rate": 1.7348677184339114e-06,
+      "loss": 0.173,
+      "step": 14192
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31799897298231583,
+      "learning_rate": 1.7338203216583493e-06,
+      "loss": 0.2934,
+      "step": 14193
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3329687176740427,
+      "learning_rate": 1.732773211140233e-06,
+      "loss": 0.2172,
+      "step": 14194
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6666999484070938,
+      "learning_rate": 1.7317263869158252e-06,
+      "loss": 0.3835,
+      "step": 14195
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.840595293551172,
+      "learning_rate": 1.7306798490213783e-06,
+      "loss": 0.4352,
+      "step": 14196
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24835261216243176,
+      "learning_rate": 1.729633597493129e-06,
+      "loss": 0.2628,
+      "step": 14197
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24067764811215303,
+      "learning_rate": 1.7285876323673144e-06,
+      "loss": 0.0807,
+      "step": 14198
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6601672537032992,
+      "learning_rate": 1.7275419536801552e-06,
+      "loss": 0.3598,
+      "step": 14199
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.366001352720004,
+      "learning_rate": 1.7264965614678631e-06,
+      "loss": 0.2781,
+      "step": 14200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3601613755013287,
+      "learning_rate": 1.7254514557666358e-06,
+      "loss": 0.2612,
+      "step": 14201
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.847066894311364,
+      "learning_rate": 1.7244066366126722e-06,
+      "loss": 0.3791,
+      "step": 14202
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35177083832531264,
+      "learning_rate": 1.72336210404215e-06,
+      "loss": 0.2582,
+      "step": 14203
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23594743243529212,
+      "learning_rate": 1.7223178580912426e-06,
+      "loss": 0.147,
+      "step": 14204
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41703845501159037,
+      "learning_rate": 1.7212738987961086e-06,
+      "loss": 0.3409,
+      "step": 14205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3426632852760692,
+      "learning_rate": 1.7202302261929071e-06,
+      "loss": 0.283,
+      "step": 14206
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7113318712831286,
+      "learning_rate": 1.7191868403177757e-06,
+      "loss": 0.3203,
+      "step": 14207
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5202922764450023,
+      "learning_rate": 1.7181437412068491e-06,
+      "loss": 0.3489,
+      "step": 14208
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29408898192812816,
+      "learning_rate": 1.717100928896246e-06,
+      "loss": 0.2608,
+      "step": 14209
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32304266882818133,
+      "learning_rate": 1.7160584034220828e-06,
+      "loss": 0.1571,
+      "step": 14210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.573358791143515,
+      "learning_rate": 1.7150161648204622e-06,
+      "loss": 0.2802,
+      "step": 14211
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30989601757233975,
+      "learning_rate": 1.713974213127475e-06,
+      "loss": 0.2593,
+      "step": 14212
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36938090057695033,
+      "learning_rate": 1.7129325483792048e-06,
+      "loss": 0.3223,
+      "step": 14213
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4585879963319681,
+      "learning_rate": 1.7118911706117213e-06,
+      "loss": 0.326,
+      "step": 14214
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32485245497389287,
+      "learning_rate": 1.710850079861095e-06,
+      "loss": 0.2697,
+      "step": 14215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44404969059636795,
+      "learning_rate": 1.7098092761633722e-06,
+      "loss": 0.247,
+      "step": 14216
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32738510941378074,
+      "learning_rate": 1.7087687595545943e-06,
+      "loss": 0.2433,
+      "step": 14217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34031209603664575,
+      "learning_rate": 1.7077285300708002e-06,
+      "loss": 0.2835,
+      "step": 14218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8347750409587513,
+      "learning_rate": 1.706688587748011e-06,
+      "loss": 0.5498,
+      "step": 14219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2600218973847508,
+      "learning_rate": 1.7056489326222392e-06,
+      "loss": 0.2242,
+      "step": 14220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2922422219720576,
+      "learning_rate": 1.7046095647294859e-06,
+      "loss": 0.1806,
+      "step": 14221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9252160238532011,
+      "learning_rate": 1.70357048410575e-06,
+      "loss": 0.3704,
+      "step": 14222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.46332376444231493,
+      "learning_rate": 1.7025316907870105e-06,
+      "loss": 0.2813,
+      "step": 14223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28819936990494277,
+      "learning_rate": 1.7014931848092409e-06,
+      "loss": 0.1896,
+      "step": 14224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32904385789124907,
+      "learning_rate": 1.700454966208407e-06,
+      "loss": 0.313,
+      "step": 14225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3804473251777745,
+      "learning_rate": 1.6994170350204576e-06,
+      "loss": 0.2184,
+      "step": 14226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31666399354697766,
+      "learning_rate": 1.6983793912813418e-06,
+      "loss": 0.1884,
+      "step": 14227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35107911647497736,
+      "learning_rate": 1.6973420350269909e-06,
+      "loss": 0.2859,
+      "step": 14228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.689539750224186,
+      "learning_rate": 1.6963049662933273e-06,
+      "loss": 0.3485,
+      "step": 14229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2906909140943072,
+      "learning_rate": 1.6952681851162644e-06,
+      "loss": 0.1952,
+      "step": 14230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1242614537154207,
+      "learning_rate": 1.6942316915317091e-06,
+      "loss": 0.7466,
+      "step": 14231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23330529085667315,
+      "learning_rate": 1.6931954855755527e-06,
+      "loss": 0.2129,
+      "step": 14232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3182776146060527,
+      "learning_rate": 1.6921595672836811e-06,
+      "loss": 0.235,
+      "step": 14233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1253697865393655,
+      "learning_rate": 1.6911239366919618e-06,
+      "loss": 0.2787,
+      "step": 14234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6584813922771027,
+      "learning_rate": 1.6900885938362677e-06,
+      "loss": 0.4179,
+      "step": 14235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31570298371256733,
+      "learning_rate": 1.6890535387524465e-06,
+      "loss": 0.2391,
+      "step": 14236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3460198335510046,
+      "learning_rate": 1.6880187714763453e-06,
+      "loss": 0.2658,
+      "step": 14237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3434216073340945,
+      "learning_rate": 1.6869842920437961e-06,
+      "loss": 0.1589,
+      "step": 14238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.37042829811528694,
+      "learning_rate": 1.6859501004906208e-06,
+      "loss": 0.2287,
+      "step": 14239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3684246063480538,
+      "learning_rate": 1.6849161968526384e-06,
+      "loss": 0.258,
+      "step": 14240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49687821321936904,
+      "learning_rate": 1.6838825811656512e-06,
+      "loss": 0.3434,
+      "step": 14241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4044736505228071,
+      "learning_rate": 1.6828492534654516e-06,
+      "loss": 0.2857,
+      "step": 14242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2720523490521247,
+      "learning_rate": 1.6818162137878224e-06,
+      "loss": 0.1785,
+      "step": 14243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2924842153658913,
+      "learning_rate": 1.6807834621685426e-06,
+      "loss": 0.2572,
+      "step": 14244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.42620199736044223,
+      "learning_rate": 1.6797509986433746e-06,
+      "loss": 0.2771,
+      "step": 14245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4717554256016822,
+      "learning_rate": 1.678718823248071e-06,
+      "loss": 0.3129,
+      "step": 14246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8840619730677773,
+      "learning_rate": 1.6776869360183746e-06,
+      "loss": 0.2366,
+      "step": 14247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3197993001659107,
+      "learning_rate": 1.6766553369900241e-06,
+      "loss": 0.2557,
+      "step": 14248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32803477233922695,
+      "learning_rate": 1.6756240261987434e-06,
+      "loss": 0.3214,
+      "step": 14249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3035876936412637,
+      "learning_rate": 1.6745930036802428e-06,
+      "loss": 0.1272,
+      "step": 14250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32877403922823634,
+      "learning_rate": 1.6735622694702259e-06,
+      "loss": 0.2405,
+      "step": 14251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3599618100090622,
+      "learning_rate": 1.6725318236043908e-06,
+      "loss": 0.3988,
+      "step": 14252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3396785730073297,
+      "learning_rate": 1.6715016661184225e-06,
+      "loss": 0.2526,
+      "step": 14253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3797148679905969,
+      "learning_rate": 1.6704717970479923e-06,
+      "loss": 0.2691,
+      "step": 14254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5481283074189417,
+      "learning_rate": 1.6694422164287627e-06,
+      "loss": 0.3733,
+      "step": 14255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26139269536422804,
+      "learning_rate": 1.6684129242963943e-06,
+      "loss": 0.2224,
+      "step": 14256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40947654935266997,
+      "learning_rate": 1.6673839206865283e-06,
+      "loss": 0.2732,
+      "step": 14257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3235516674745909,
+      "learning_rate": 1.6663552056347975e-06,
+      "loss": 0.2544,
+      "step": 14258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1695979780325698,
+      "learning_rate": 1.6653267791768258e-06,
+      "loss": 0.5732,
+      "step": 14259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2849039189767827,
+      "learning_rate": 1.6642986413482321e-06,
+      "loss": 0.1958,
+      "step": 14260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38973166723076286,
+      "learning_rate": 1.663270792184618e-06,
+      "loss": 0.2984,
+      "step": 14261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8836587032487782,
+      "learning_rate": 1.6622432317215776e-06,
+      "loss": 0.4592,
+      "step": 14262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40255447112780424,
+      "learning_rate": 1.6612159599946954e-06,
+      "loss": 0.2373,
+      "step": 14263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20061146179604905,
+      "learning_rate": 1.660188977039544e-06,
+      "loss": 0.2094,
+      "step": 14264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.583325395018498,
+      "learning_rate": 1.659162282891692e-06,
+      "loss": 0.7721,
+      "step": 14265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32561047876624966,
+      "learning_rate": 1.6581358775866907e-06,
+      "loss": 0.1869,
+      "step": 14266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5064695995116827,
+      "learning_rate": 1.6571097611600862e-06,
+      "loss": 0.3667,
+      "step": 14267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3729173917284346,
+      "learning_rate": 1.6560839336474088e-06,
+      "loss": 0.3192,
+      "step": 14268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3673667832560838,
+      "learning_rate": 1.6550583950841891e-06,
+      "loss": 0.2171,
+      "step": 14269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2352059959306334,
+      "learning_rate": 1.6540331455059377e-06,
+      "loss": 0.1389,
+      "step": 14270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2082826393706474,
+      "learning_rate": 1.6530081849481595e-06,
+      "loss": 0.5537,
+      "step": 14271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27082498922394227,
+      "learning_rate": 1.6519835134463468e-06,
+      "loss": 0.2635,
+      "step": 14272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4933710914894767,
+      "learning_rate": 1.6509591310359886e-06,
+      "loss": 0.2803,
+      "step": 14273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6557255957824724,
+      "learning_rate": 1.649935037752557e-06,
+      "loss": 0.3822,
+      "step": 14274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8281936757764693,
+      "learning_rate": 1.648911233631516e-06,
+      "loss": 0.3157,
+      "step": 14275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20637211972769265,
+      "learning_rate": 1.6478877187083187e-06,
+      "loss": 0.1921,
+      "step": 14276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4550300837789682,
+      "learning_rate": 1.6468644930184097e-06,
+      "loss": 0.2334,
+      "step": 14277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5805410905455298,
+      "learning_rate": 1.6458415565972253e-06,
+      "loss": 0.3295,
+      "step": 14278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43031247602796713,
+      "learning_rate": 1.6448189094801891e-06,
+      "loss": 0.2736,
+      "step": 14279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36717439759245385,
+      "learning_rate": 1.6437965517027143e-06,
+      "loss": 0.2979,
+      "step": 14280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6289420966546745,
+      "learning_rate": 1.6427744833002036e-06,
+      "loss": 0.3179,
+      "step": 14281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24428017486398262,
+      "learning_rate": 1.6417527043080583e-06,
+      "loss": 0.1558,
+      "step": 14282
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1756826326190717,
+      "learning_rate": 1.6407312147616539e-06,
+      "loss": 0.4512,
+      "step": 14283
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3007377368607552,
+      "learning_rate": 1.6397100146963662e-06,
+      "loss": 0.2467,
+      "step": 14284
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38786311492343356,
+      "learning_rate": 1.6386891041475639e-06,
+      "loss": 0.3136,
+      "step": 14285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.719273995861604,
+      "learning_rate": 1.6376684831505984e-06,
+      "loss": 0.2903,
+      "step": 14286
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.616064577176051,
+      "learning_rate": 1.636648151740814e-06,
+      "loss": 0.2994,
+      "step": 14287
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29643428083943496,
+      "learning_rate": 1.6356281099535432e-06,
+      "loss": 0.2724,
+      "step": 14288
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27973403877242464,
+      "learning_rate": 1.63460835782411e-06,
+      "loss": 0.1537,
+      "step": 14289
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3902255474494219,
+      "learning_rate": 1.633588895387832e-06,
+      "loss": 0.2725,
+      "step": 14290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6206832393439994,
+      "learning_rate": 1.6325697226800109e-06,
+      "loss": 0.3138,
+      "step": 14291
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32000206528220054,
+      "learning_rate": 1.6315508397359391e-06,
+      "loss": 0.2513,
+      "step": 14292
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9261649174707342,
+      "learning_rate": 1.6305322465909012e-06,
+      "loss": 0.3988,
+      "step": 14293
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3895447592390385,
+      "learning_rate": 1.6295139432801732e-06,
+      "loss": 0.2633,
+      "step": 14294
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20039587976898893,
+      "learning_rate": 1.628495929839018e-06,
+      "loss": 0.1685,
+      "step": 14295
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.678502925193134,
+      "learning_rate": 1.6274782063026883e-06,
+      "loss": 0.3645,
+      "step": 14296
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3716671212985712,
+      "learning_rate": 1.6264607727064253e-06,
+      "loss": 0.3074,
+      "step": 14297
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8235463841034898,
+      "learning_rate": 1.6254436290854691e-06,
+      "loss": 0.4441,
+      "step": 14298
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6843383792249415,
+      "learning_rate": 1.62442677547504e-06,
+      "loss": 0.1194,
+      "step": 14299
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25397188356352135,
+      "learning_rate": 1.62341021191035e-06,
+      "loss": 0.267,
+      "step": 14300
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32104682332962464,
+      "learning_rate": 1.6223939384266064e-06,
+      "loss": 0.1979,
+      "step": 14301
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6705088211808614,
+      "learning_rate": 1.6213779550589959e-06,
+      "loss": 0.2417,
+      "step": 14302
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41855789850269925,
+      "learning_rate": 1.6203622618427105e-06,
+      "loss": 0.2967,
+      "step": 14303
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3418585073461008,
+      "learning_rate": 1.6193468588129192e-06,
+      "loss": 0.311,
+      "step": 14304
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8013089892375115,
+      "learning_rate": 1.6183317460047853e-06,
+      "loss": 0.1193,
+      "step": 14305
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4221278782395689,
+      "learning_rate": 1.6173169234534602e-06,
+      "loss": 0.2809,
+      "step": 14306
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.286743200585038,
+      "learning_rate": 1.6163023911940923e-06,
+      "loss": 0.2143,
+      "step": 14307
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4101065643986548,
+      "learning_rate": 1.6152881492618123e-06,
+      "loss": 0.2727,
+      "step": 14308
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3389018916742762,
+      "learning_rate": 1.614274197691743e-06,
+      "loss": 0.2422,
+      "step": 14309
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.02805515707719,
+      "learning_rate": 1.6132605365189945e-06,
+      "loss": 0.6135,
+      "step": 14310
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.46498023154122176,
+      "learning_rate": 1.6122471657786764e-06,
+      "loss": 0.336,
+      "step": 14311
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24162416824879168,
+      "learning_rate": 1.6112340855058784e-06,
+      "loss": 0.199,
+      "step": 14312
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4745324261022142,
+      "learning_rate": 1.6102212957356821e-06,
+      "loss": 0.2715,
+      "step": 14313
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8109722485644276,
+      "learning_rate": 1.6092087965031623e-06,
+      "loss": 0.4024,
+      "step": 14314
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29032677627833714,
+      "learning_rate": 1.6081965878433781e-06,
+      "loss": 0.1993,
+      "step": 14315
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2764951495219204,
+      "learning_rate": 1.6071846697913907e-06,
+      "loss": 0.2629,
+      "step": 14316
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0850913168148355,
+      "learning_rate": 1.6061730423822353e-06,
+      "loss": 0.4369,
+      "step": 14317
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3368554761044434,
+      "learning_rate": 1.6051617056509427e-06,
+      "loss": 0.1969,
+      "step": 14318
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6791350794668766,
+      "learning_rate": 1.604150659632543e-06,
+      "loss": 0.3668,
+      "step": 14319
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2824110904939739,
+      "learning_rate": 1.6031399043620444e-06,
+      "loss": 0.2334,
+      "step": 14320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3433389634591825,
+      "learning_rate": 1.6021294398744491e-06,
+      "loss": 0.2899,
+      "step": 14321
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27915034093289076,
+      "learning_rate": 1.6011192662047493e-06,
+      "loss": 0.1297,
+      "step": 14322
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36788278715290085,
+      "learning_rate": 1.6001093833879288e-06,
+      "loss": 0.3075,
+      "step": 14323
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3496665619411987,
+      "learning_rate": 1.5990997914589602e-06,
+      "loss": 0.2855,
+      "step": 14324
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5939639419698534,
+      "learning_rate": 1.598090490452805e-06,
+      "loss": 0.2828,
+      "step": 14325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7697198306103441,
+      "learning_rate": 1.5970814804044143e-06,
+      "loss": 0.4741,
+      "step": 14326
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3498249717359082,
+      "learning_rate": 1.5960727613487282e-06,
+      "loss": 0.2813,
+      "step": 14327
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2114390332677578,
+      "learning_rate": 1.5950643333206827e-06,
+      "loss": 0.1838,
+      "step": 14328
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.312899098888681,
+      "learning_rate": 1.5940561963551982e-06,
+      "loss": 0.3905,
+      "step": 14329
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39726885707439447,
+      "learning_rate": 1.5930483504871863e-06,
+      "loss": 0.2603,
+      "step": 14330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3441962419375129,
+      "learning_rate": 1.5920407957515472e-06,
+      "loss": 0.2621,
+      "step": 14331
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.244026243921077,
+      "learning_rate": 1.5910335321831749e-06,
+      "loss": 0.7247,
+      "step": 14332
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32768146343407234,
+      "learning_rate": 1.5900265598169507e-06,
+      "loss": 0.2683,
+      "step": 14333
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2208843892022444,
+      "learning_rate": 1.5890198786877442e-06,
+      "loss": 0.1797,
+      "step": 14334
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3084248363985326,
+      "learning_rate": 1.5880134888304155e-06,
+      "loss": 0.2296,
+      "step": 14335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3367945172727081,
+      "learning_rate": 1.58700739027982e-06,
+      "loss": 0.2464,
+      "step": 14336
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6608801653776333,
+      "learning_rate": 1.5860015830707976e-06,
+      "loss": 0.3797,
+      "step": 14337
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5594061912195747,
+      "learning_rate": 1.5849960672381781e-06,
+      "loss": 0.3649,
+      "step": 14338
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26050111086502553,
+      "learning_rate": 1.5839908428167806e-06,
+      "loss": 0.2343,
+      "step": 14339
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.537863403344283,
+      "learning_rate": 1.5829859098414202e-06,
+      "loss": 0.3252,
+      "step": 14340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3168205101844687,
+      "learning_rate": 1.5819812683468971e-06,
+      "loss": 0.1417,
+      "step": 14341
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36311874865663823,
+      "learning_rate": 1.5809769183680001e-06,
+      "loss": 0.2854,
+      "step": 14342
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44223521127969234,
+      "learning_rate": 1.5799728599395093e-06,
+      "loss": 0.3553,
+      "step": 14343
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40043078151688744,
+      "learning_rate": 1.5789690930961955e-06,
+      "loss": 0.2374,
+      "step": 14344
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3187473275956427,
+      "learning_rate": 1.577965617872821e-06,
+      "loss": 0.2568,
+      "step": 14345
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40970369506063714,
+      "learning_rate": 1.5769624343041356e-06,
+      "loss": 0.2775,
+      "step": 14346
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27328936575328555,
+      "learning_rate": 1.5759595424248798e-06,
+      "loss": 0.2324,
+      "step": 14347
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3200545522278074,
+      "learning_rate": 1.5749569422697786e-06,
+      "loss": 0.2029,
+      "step": 14348
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7813268400056336,
+      "learning_rate": 1.57395463387356e-06,
+      "loss": 0.4164,
+      "step": 14349
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.362908227303668,
+      "learning_rate": 1.572952617270932e-06,
+      "loss": 0.7034,
+      "step": 14350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2630012527604643,
+      "learning_rate": 1.5719508924965876e-06,
+      "loss": 0.2133,
+      "step": 14351
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5084932701306548,
+      "learning_rate": 1.5709494595852238e-06,
+      "loss": 0.3867,
+      "step": 14352
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4391672927929206,
+      "learning_rate": 1.569948318571517e-06,
+      "loss": 0.2645,
+      "step": 14353
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24787677119492113,
+      "learning_rate": 1.5689474694901386e-06,
+      "loss": 0.1697,
+      "step": 14354
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5045982631903566,
+      "learning_rate": 1.5679469123757463e-06,
+      "loss": 0.331,
+      "step": 14355
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5158521633631827,
+      "learning_rate": 1.566946647262988e-06,
+      "loss": 0.3973,
+      "step": 14356
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2975250660129979,
+      "learning_rate": 1.5659466741865059e-06,
+      "loss": 0.2084,
+      "step": 14357
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.882686185713709,
+      "learning_rate": 1.5649469931809291e-06,
+      "loss": 0.4715,
+      "step": 14358
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4060897628478708,
+      "learning_rate": 1.5639476042808743e-06,
+      "loss": 0.3158,
+      "step": 14359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37822526003360196,
+      "learning_rate": 1.5629485075209494e-06,
+      "loss": 0.2885,
+      "step": 14360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26592851936967965,
+      "learning_rate": 1.5619497029357566e-06,
+      "loss": 0.1386,
+      "step": 14361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48297293822615317,
+      "learning_rate": 1.5609511905598828e-06,
+      "loss": 0.3686,
+      "step": 14362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3980489462284402,
+      "learning_rate": 1.559952970427907e-06,
+      "loss": 0.2771,
+      "step": 14363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3248249460067187,
+      "learning_rate": 1.5589550425743938e-06,
+      "loss": 0.2326,
+      "step": 14364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7106496612344982,
+      "learning_rate": 1.5579574070339077e-06,
+      "loss": 0.397,
+      "step": 14365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25155760454843434,
+      "learning_rate": 1.5569600638409931e-06,
+      "loss": 0.179,
+      "step": 14366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23556004098954053,
+      "learning_rate": 1.5559630130301885e-06,
+      "loss": 0.1974,
+      "step": 14367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3610236878509954,
+      "learning_rate": 1.5549662546360223e-06,
+      "loss": 0.7535,
+      "step": 14368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4102886902865449,
+      "learning_rate": 1.5539697886930082e-06,
+      "loss": 0.3094,
+      "step": 14369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4806336247620259,
+      "learning_rate": 1.5529736152356601e-06,
+      "loss": 0.2572,
+      "step": 14370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3447873590811517,
+      "learning_rate": 1.551977734298472e-06,
+      "loss": 0.3017,
+      "step": 14371
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3259821776760638,
+      "learning_rate": 1.5509821459159312e-06,
+      "loss": 0.2352,
+      "step": 14372
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.35766169837779577,
+      "learning_rate": 1.5499868501225135e-06,
+      "loss": 0.2043,
+      "step": 14373
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.45759062221042746,
+      "learning_rate": 1.548991846952691e-06,
+      "loss": 0.2734,
+      "step": 14374
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26891305999108633,
+      "learning_rate": 1.5479971364409163e-06,
+      "loss": 0.2624,
+      "step": 14375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6950747504087013,
+      "learning_rate": 1.5470027186216386e-06,
+      "loss": 0.3777,
+      "step": 14376
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8274621163981223,
+      "learning_rate": 1.5460085935292902e-06,
+      "loss": 0.2935,
+      "step": 14377
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5470660206047755,
+      "learning_rate": 1.5450147611983024e-06,
+      "loss": 0.2792,
+      "step": 14378
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22596268465440378,
+      "learning_rate": 1.5440212216630902e-06,
+      "loss": 0.2247,
+      "step": 14379
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4563189462879652,
+      "learning_rate": 1.54302797495806e-06,
+      "loss": 0.2061,
+      "step": 14380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.408530641518608,
+      "learning_rate": 1.5420350211176072e-06,
+      "loss": 0.2611,
+      "step": 14381
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4614799902770901,
+      "learning_rate": 1.541042360176115e-06,
+      "loss": 0.3243,
+      "step": 14382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3767321614513136,
+      "learning_rate": 1.5400499921679647e-06,
+      "loss": 0.2978,
+      "step": 14383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32608810726150617,
+      "learning_rate": 1.5390579171275222e-06,
+      "loss": 0.1694,
+      "step": 14384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2786817942246057,
+      "learning_rate": 1.5380661350891346e-06,
+      "loss": 0.1812,
+      "step": 14385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2423112132808052,
+      "learning_rate": 1.5370746460871555e-06,
+      "loss": 0.7655,
+      "step": 14386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24830800114573304,
+      "learning_rate": 1.5360834501559185e-06,
+      "loss": 0.2066,
+      "step": 14387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6941033661044869,
+      "learning_rate": 1.5350925473297462e-06,
+      "loss": 0.3225,
+      "step": 14388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3297493516790682,
+      "learning_rate": 1.5341019376429533e-06,
+      "loss": 0.4718,
+      "step": 14389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33027397591322355,
+      "learning_rate": 1.5331116211298492e-06,
+      "loss": 0.1663,
+      "step": 14390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3162742331726112,
+      "learning_rate": 1.532121597824725e-06,
+      "loss": 0.2882,
+      "step": 14391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3078259870411867,
+      "learning_rate": 1.5311318677618658e-06,
+      "loss": 0.1998,
+      "step": 14392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3203383686794931,
+      "learning_rate": 1.5301424309755464e-06,
+      "loss": 0.2119,
+      "step": 14393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9170395837276937,
+      "learning_rate": 1.529153287500027e-06,
+      "loss": 0.5176,
+      "step": 14394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3675141498763632,
+      "learning_rate": 1.5281644373695682e-06,
+      "loss": 0.2876,
+      "step": 14395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6819063271267362,
+      "learning_rate": 1.52717588061841e-06,
+      "loss": 0.26,
+      "step": 14396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18812990097663482,
+      "learning_rate": 1.5261876172807865e-06,
+      "loss": 0.1458,
+      "step": 14397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4405755811998149,
+      "learning_rate": 1.5251996473909202e-06,
+      "loss": 0.3262,
+      "step": 14398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4274068515417533,
+      "learning_rate": 1.5242119709830272e-06,
+      "loss": 0.3208,
+      "step": 14399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30076883740926846,
+      "learning_rate": 1.5232245880913088e-06,
+      "loss": 0.2194,
+      "step": 14400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2838196412791696,
+      "learning_rate": 1.5222374987499588e-06,
+      "loss": 0.536,
+      "step": 14401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5728611563163679,
+      "learning_rate": 1.5212507029931578e-06,
+      "loss": 0.2791,
+      "step": 14402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22643317321614864,
+      "learning_rate": 1.5202642008550827e-06,
+      "loss": 0.2097,
+      "step": 14403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.781228803059026,
+      "learning_rate": 1.519277992369893e-06,
+      "loss": 0.3939,
+      "step": 14404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5420511061950547,
+      "learning_rate": 1.5182920775717425e-06,
+      "loss": 0.3284,
+      "step": 14405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23424207699255722,
+      "learning_rate": 1.5173064564947714e-06,
+      "loss": 0.2109,
+      "step": 14406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4855977165429987,
+      "learning_rate": 1.5163211291731116e-06,
+      "loss": 0.3538,
+      "step": 14407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7778771909877253,
+      "learning_rate": 1.5153360956408891e-06,
+      "loss": 0.3002,
+      "step": 14408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40367431465738035,
+      "learning_rate": 1.514351355932212e-06,
+      "loss": 0.2746,
+      "step": 14409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4594177145672264,
+      "learning_rate": 1.513366910081182e-06,
+      "loss": 0.2778,
+      "step": 14410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3612788772247959,
+      "learning_rate": 1.5123827581218898e-06,
+      "loss": 0.3111,
+      "step": 14411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3966527721965122,
+      "learning_rate": 1.5113989000884189e-06,
+      "loss": 0.303,
+      "step": 14412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17782252117943567,
+      "learning_rate": 1.51041533601484e-06,
+      "loss": 0.0841,
+      "step": 14413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3769117084741131,
+      "learning_rate": 1.5094320659352123e-06,
+      "loss": 0.2748,
+      "step": 14414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3041994326675657,
+      "learning_rate": 1.5084490898835857e-06,
+      "loss": 0.2886,
+      "step": 14415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6275402594647362,
+      "learning_rate": 1.5074664078940039e-06,
+      "loss": 0.2981,
+      "step": 14416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7745986505170465,
+      "learning_rate": 1.5064840200004972e-06,
+      "loss": 0.4663,
+      "step": 14417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26455906511621774,
+      "learning_rate": 1.5055019262370807e-06,
+      "loss": 0.2277,
+      "step": 14418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2744273616921841,
+      "learning_rate": 1.5045201266377662e-06,
+      "loss": 0.191,
+      "step": 14419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8188194127786869,
+      "learning_rate": 1.5035386212365554e-06,
+      "loss": 0.4115,
+      "step": 14420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3522078195909208,
+      "learning_rate": 1.502557410067438e-06,
+      "loss": 0.2702,
+      "step": 14421
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6756378638949322,
+      "learning_rate": 1.5015764931643916e-06,
+      "loss": 0.4208,
+      "step": 14422
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25982068134200115,
+      "learning_rate": 1.5005958705613833e-06,
+      "loss": 0.2264,
+      "step": 14423
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3699918270593306,
+      "learning_rate": 1.4996155422923764e-06,
+      "loss": 0.3032,
+      "step": 14424
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31773506752213887,
+      "learning_rate": 1.4986355083913184e-06,
+      "loss": 0.1678,
+      "step": 14425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28175897232877206,
+      "learning_rate": 1.4976557688921478e-06,
+      "loss": 0.2038,
+      "step": 14426
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3882773152820447,
+      "learning_rate": 1.4966763238287885e-06,
+      "loss": 0.2799,
+      "step": 14427
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6931088691932338,
+      "learning_rate": 1.4956971732351655e-06,
+      "loss": 0.3976,
+      "step": 14428
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3691483478603107,
+      "learning_rate": 1.4947183171451841e-06,
+      "loss": 0.2152,
+      "step": 14429
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38071812733088667,
+      "learning_rate": 1.4937397555927413e-06,
+      "loss": 0.2909,
+      "step": 14430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2094977356719134,
+      "learning_rate": 1.4927614886117248e-06,
+      "loss": 0.1925,
+      "step": 14431
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5582480502897131,
+      "learning_rate": 1.4917835162360107e-06,
+      "loss": 0.2479,
+      "step": 14432
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3547248993099505,
+      "learning_rate": 1.4908058384994684e-06,
+      "loss": 0.2633,
+      "step": 14433
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33898191546731055,
+      "learning_rate": 1.4898284554359555e-06,
+      "loss": 0.2929,
+      "step": 14434
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.20134478839783,
+      "learning_rate": 1.4888513670793159e-06,
+      "loss": 0.7416,
+      "step": 14435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31287229804617556,
+      "learning_rate": 1.4878745734633859e-06,
+      "loss": 0.1874,
+      "step": 14436
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28716920169026183,
+      "learning_rate": 1.4868980746219953e-06,
+      "loss": 0.1744,
+      "step": 14437
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3602326317426348,
+      "learning_rate": 1.485921870588959e-06,
+      "loss": 0.3254,
+      "step": 14438
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34644487846977096,
+      "learning_rate": 1.4849459613980821e-06,
+      "loss": 0.2226,
+      "step": 14439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7099936095796247,
+      "learning_rate": 1.4839703470831568e-06,
+      "loss": 0.6051,
+      "step": 14440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1426420559205335,
+      "learning_rate": 1.4829950276779759e-06,
+      "loss": 0.6309,
+      "step": 14441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21521110544361674,
+      "learning_rate": 1.4820200032163102e-06,
+      "loss": 0.2127,
+      "step": 14442
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5594517974144692,
+      "learning_rate": 1.481045273731926e-06,
+      "loss": 0.2786,
+      "step": 14443
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6070929376142952,
+      "learning_rate": 1.480070839258575e-06,
+      "loss": 0.3338,
+      "step": 14444
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2462055074718749,
+      "learning_rate": 1.479096699830007e-06,
+      "loss": 0.1646,
+      "step": 14445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3872583829909997,
+      "learning_rate": 1.4781228554799544e-06,
+      "loss": 0.3043,
+      "step": 14446
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7217473632643663,
+      "learning_rate": 1.4771493062421393e-06,
+      "loss": 0.5007,
+      "step": 14447
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3543168866164041,
+      "learning_rate": 1.4761760521502788e-06,
+      "loss": 0.272,
+      "step": 14448
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.694609659735516,
+      "learning_rate": 1.4752030932380723e-06,
+      "loss": 0.2679,
+      "step": 14449
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2668551292038593,
+      "learning_rate": 1.4742304295392173e-06,
+      "loss": 0.2415,
+      "step": 14450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.348151555424367,
+      "learning_rate": 1.4732580610873991e-06,
+      "loss": 0.2819,
+      "step": 14451
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4865809063782709,
+      "learning_rate": 1.4722859879162831e-06,
+      "loss": 0.164,
+      "step": 14452
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3427781510155108,
+      "learning_rate": 1.471314210059539e-06,
+      "loss": 0.8253,
+      "step": 14453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2627245576209838,
+      "learning_rate": 1.4703427275508175e-06,
+      "loss": 0.2502,
+      "step": 14454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38749675227945757,
+      "learning_rate": 1.4693715404237595e-06,
+      "loss": 0.2567,
+      "step": 14455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48887376832848684,
+      "learning_rate": 1.4684006487119996e-06,
+      "loss": 0.2749,
+      "step": 14456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23372320232067798,
+      "learning_rate": 1.4674300524491548e-06,
+      "loss": 0.1902,
+      "step": 14457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3504977034974888,
+      "learning_rate": 1.466459751668843e-06,
+      "loss": 0.2363,
+      "step": 14458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1512347307415474,
+      "learning_rate": 1.4654897464046624e-06,
+      "loss": 0.7318,
+      "step": 14459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33497458339458847,
+      "learning_rate": 1.4645200366902056e-06,
+      "loss": 0.2622,
+      "step": 14460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6891307734834573,
+      "learning_rate": 1.4635506225590511e-06,
+      "loss": 0.3649,
+      "step": 14461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24648296086863,
+      "learning_rate": 1.4625815040447733e-06,
+      "loss": 0.1953,
+      "step": 14462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3198739485686858,
+      "learning_rate": 1.4616126811809305e-06,
+      "loss": 0.2548,
+      "step": 14463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4785902437377649,
+      "learning_rate": 1.4606441540010742e-06,
+      "loss": 0.2262,
+      "step": 14464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42064172693904334,
+      "learning_rate": 1.4596759225387401e-06,
+      "loss": 0.2391,
+      "step": 14465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3519446079196051,
+      "learning_rate": 1.4587079868274644e-06,
+      "loss": 0.2952,
+      "step": 14466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4042615656613845,
+      "learning_rate": 1.4577403469007645e-06,
+      "loss": 0.3237,
+      "step": 14467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.35400671879270007,
+      "learning_rate": 1.4567730027921489e-06,
+      "loss": 0.0845,
+      "step": 14468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3576298559652004,
+      "learning_rate": 1.4558059545351144e-06,
+      "loss": 0.2974,
+      "step": 14469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2690274370266391,
+      "learning_rate": 1.4548392021631541e-06,
+      "loss": 0.264,
+      "step": 14470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.260476076309206,
+      "learning_rate": 1.4538727457097447e-06,
+      "loss": 0.7678,
+      "step": 14471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32094729111190423,
+      "learning_rate": 1.4529065852083557e-06,
+      "loss": 0.2062,
+      "step": 14472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.618603149339349,
+      "learning_rate": 1.451940720692443e-06,
+      "loss": 0.3733,
+      "step": 14473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3205669864381226,
+      "learning_rate": 1.450975152195454e-06,
+      "loss": 0.2918,
+      "step": 14474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3185380448055098,
+      "learning_rate": 1.4500098797508289e-06,
+      "loss": 0.1982,
+      "step": 14475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24713688538065756,
+      "learning_rate": 1.4490449033919952e-06,
+      "loss": 0.1745,
+      "step": 14476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7044439761827327,
+      "learning_rate": 1.4480802231523682e-06,
+      "loss": 0.4051,
+      "step": 14477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22536764749533747,
+      "learning_rate": 1.447115839065354e-06,
+      "loss": 0.2154,
+      "step": 14478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.754993666762245,
+      "learning_rate": 1.446151751164352e-06,
+      "loss": 0.4199,
+      "step": 14479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5464917011221442,
+      "learning_rate": 1.4451879594827467e-06,
+      "loss": 0.4158,
+      "step": 14480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22848279331701107,
+      "learning_rate": 1.444224464053916e-06,
+      "loss": 0.1559,
+      "step": 14481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26781759688718365,
+      "learning_rate": 1.44326126491122e-06,
+      "loss": 0.2418,
+      "step": 14482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7385480231144739,
+      "learning_rate": 1.4422983620880215e-06,
+      "loss": 0.4098,
+      "step": 14483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5600152349811515,
+      "learning_rate": 1.4413357556176633e-06,
+      "loss": 0.3144,
+      "step": 14484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38866276568802743,
+      "learning_rate": 1.4403734455334816e-06,
+      "loss": 0.2366,
+      "step": 14485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3608062477208995,
+      "learning_rate": 1.4394114318687947e-06,
+      "loss": 0.2916,
+      "step": 14486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39215757913532784,
+      "learning_rate": 1.4384497146569242e-06,
+      "loss": 0.2683,
+      "step": 14487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28871817789259424,
+      "learning_rate": 1.437488293931173e-06,
+      "loss": 0.1842,
+      "step": 14488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7837969352840899,
+      "learning_rate": 1.436527169724833e-06,
+      "loss": 0.3826,
+      "step": 14489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.252579482719496,
+      "learning_rate": 1.4355663420711863e-06,
+      "loss": 0.2509,
+      "step": 14490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38934719167353166,
+      "learning_rate": 1.434605811003511e-06,
+      "loss": 0.179,
+      "step": 14491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.327017593625259,
+      "learning_rate": 1.4336455765550684e-06,
+      "loss": 0.5127,
+      "step": 14492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3263390069030087,
+      "learning_rate": 1.4326856387591114e-06,
+      "loss": 0.1991,
+      "step": 14493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26938715579382766,
+      "learning_rate": 1.4317259976488806e-06,
+      "loss": 0.2504,
+      "step": 14494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.642656981752686,
+      "learning_rate": 1.4307666532576115e-06,
+      "loss": 0.3615,
+      "step": 14495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31108348734056135,
+      "learning_rate": 1.429807605618525e-06,
+      "loss": 0.2628,
+      "step": 14496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5051662006316499,
+      "learning_rate": 1.4288488547648328e-06,
+      "loss": 0.2379,
+      "step": 14497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37356454597055017,
+      "learning_rate": 1.4278904007297356e-06,
+      "loss": 0.2442,
+      "step": 14498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3072156818751481,
+      "learning_rate": 1.4269322435464229e-06,
+      "loss": 0.2397,
+      "step": 14499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9057485951009321,
+      "learning_rate": 1.425974383248081e-06,
+      "loss": 0.516,
+      "step": 14500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3275530243981576,
+      "learning_rate": 1.425016819867876e-06,
+      "loss": 0.2634,
+      "step": 14501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5692330083855973,
+      "learning_rate": 1.4240595534389712e-06,
+      "loss": 0.2918,
+      "step": 14502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22938802148421475,
+      "learning_rate": 1.4231025839945123e-06,
+      "loss": 0.1967,
+      "step": 14503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3925463484862073,
+      "learning_rate": 1.422145911567645e-06,
+      "loss": 0.192,
+      "step": 14504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.661586160240331,
+      "learning_rate": 1.4211895361914961e-06,
+      "loss": 0.2866,
+      "step": 14505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24761832146932847,
+      "learning_rate": 1.4202334578991838e-06,
+      "loss": 0.2535,
+      "step": 14506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6307606543288709,
+      "learning_rate": 1.419277676723816e-06,
+      "loss": 0.3035,
+      "step": 14507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3912702731263923,
+      "learning_rate": 1.4183221926984958e-06,
+      "loss": 0.294,
+      "step": 14508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21325530333087425,
+      "learning_rate": 1.4173670058563082e-06,
+      "loss": 0.2035,
+      "step": 14509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.46726091602695835,
+      "learning_rate": 1.4164121162303335e-06,
+      "loss": 0.2975,
+      "step": 14510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4739114950446037,
+      "learning_rate": 1.4154575238536373e-06,
+      "loss": 0.1686,
+      "step": 14511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4402434355675197,
+      "learning_rate": 1.4145032287592753e-06,
+      "loss": 0.3343,
+      "step": 14512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5244827607300405,
+      "learning_rate": 1.4135492309803e-06,
+      "loss": 0.3307,
+      "step": 14513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.277632422490554,
+      "learning_rate": 1.4125955305497453e-06,
+      "loss": 0.2066,
+      "step": 14514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32009804462674135,
+      "learning_rate": 1.4116421275006386e-06,
+      "loss": 0.2483,
+      "step": 14515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6266221865588981,
+      "learning_rate": 1.410689021865993e-06,
+      "loss": 0.2119,
+      "step": 14516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3426461166367405,
+      "learning_rate": 1.4097362136788196e-06,
+      "loss": 0.225,
+      "step": 14517
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32656014034310793,
+      "learning_rate": 1.408783702972112e-06,
+      "loss": 0.2931,
+      "step": 14518
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8290687136751642,
+      "learning_rate": 1.4078314897788558e-06,
+      "loss": 0.4557,
+      "step": 14519
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3310944158287448,
+      "learning_rate": 1.4068795741320241e-06,
+      "loss": 0.152,
+      "step": 14520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23025651377461948,
+      "learning_rate": 1.4059279560645845e-06,
+      "loss": 0.2156,
+      "step": 14521
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3482816224844952,
+      "learning_rate": 1.4049766356094897e-06,
+      "loss": 0.2382,
+      "step": 14522
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6896887320183515,
+      "learning_rate": 1.4040256127996842e-06,
+      "loss": 0.345,
+      "step": 14523
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.376874453876573,
+      "learning_rate": 1.403074887668101e-06,
+      "loss": 0.2253,
+      "step": 14524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34520731206085,
+      "learning_rate": 1.4021244602476658e-06,
+      "loss": 0.3182,
+      "step": 14525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.590729596317791,
+      "learning_rate": 1.401174330571291e-06,
+      "loss": 0.4018,
+      "step": 14526
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24916310579425358,
+      "learning_rate": 1.4002244986718793e-06,
+      "loss": 0.1518,
+      "step": 14527
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9781155735931566,
+      "learning_rate": 1.3992749645823224e-06,
+      "loss": 0.4291,
+      "step": 14528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37701695438860766,
+      "learning_rate": 1.3983257283355044e-06,
+      "loss": 0.2982,
+      "step": 14529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2968145176708719,
+      "learning_rate": 1.3973767899642976e-06,
+      "loss": 0.2351,
+      "step": 14530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2241899528039462,
+      "learning_rate": 1.396428149501562e-06,
+      "loss": 0.8343,
+      "step": 14531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6203633467767298,
+      "learning_rate": 1.3954798069801468e-06,
+      "loss": 0.4185,
+      "step": 14532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2830199485054341,
+      "learning_rate": 1.394531762432899e-06,
+      "loss": 0.2174,
+      "step": 14533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3082002175777337,
+      "learning_rate": 1.3935840158926461e-06,
+      "loss": 0.2218,
+      "step": 14534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5911035214641597,
+      "learning_rate": 1.3926365673922082e-06,
+      "loss": 0.3011,
+      "step": 14535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40215627658380904,
+      "learning_rate": 1.3916894169643969e-06,
+      "loss": 0.3002,
+      "step": 14536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3721036297309638,
+      "learning_rate": 1.390742564642007e-06,
+      "loss": 0.2617,
+      "step": 14537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6464102731862245,
+      "learning_rate": 1.3897960104578357e-06,
+      "loss": 0.3488,
+      "step": 14538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4497840173503336,
+      "learning_rate": 1.3888497544446578e-06,
+      "loss": 0.2986,
+      "step": 14539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3635716294730615,
+      "learning_rate": 1.3879037966352426e-06,
+      "loss": 0.192,
+      "step": 14540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31564871045658416,
+      "learning_rate": 1.3869581370623464e-06,
+      "loss": 0.2819,
+      "step": 14541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35301383748174164,
+      "learning_rate": 1.3860127757587215e-06,
+      "loss": 0.2734,
+      "step": 14542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9236845077921836,
+      "learning_rate": 1.3850677127571033e-06,
+      "loss": 0.444,
+      "step": 14543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8576242151750284,
+      "learning_rate": 1.3841229480902207e-06,
+      "loss": 0.449,
+      "step": 14544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.317992615475425,
+      "learning_rate": 1.3831784817907867e-06,
+      "loss": 0.27,
+      "step": 14545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4192884397471493,
+      "learning_rate": 1.382234313891515e-06,
+      "loss": 0.3332,
+      "step": 14546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.281125353055018,
+      "learning_rate": 1.3812904444250973e-06,
+      "loss": 0.1339,
+      "step": 14547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4240781429981899,
+      "learning_rate": 1.3803468734242208e-06,
+      "loss": 0.2947,
+      "step": 14548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34789522278090906,
+      "learning_rate": 1.3794036009215628e-06,
+      "loss": 0.3019,
+      "step": 14549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6447680621171787,
+      "learning_rate": 1.3784606269497835e-06,
+      "loss": 0.1991,
+      "step": 14550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.42031524895275557,
+      "learning_rate": 1.377517951541545e-06,
+      "loss": 0.2907,
+      "step": 14551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5888956245708498,
+      "learning_rate": 1.3765755747294906e-06,
+      "loss": 0.357,
+      "step": 14552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22556962698291796,
+      "learning_rate": 1.3756334965462502e-06,
+      "loss": 0.1829,
+      "step": 14553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2882609123569876,
+      "learning_rate": 1.3746917170244522e-06,
+      "loss": 0.2108,
+      "step": 14554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3360077461689062,
+      "learning_rate": 1.3737502361967092e-06,
+      "loss": 0.6013,
+      "step": 14555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8700106534469142,
+      "learning_rate": 1.3728090540956241e-06,
+      "loss": 0.2877,
+      "step": 14556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2599726917673043,
+      "learning_rate": 1.3718681707537895e-06,
+      "loss": 0.2489,
+      "step": 14557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4713611148460899,
+      "learning_rate": 1.3709275862037908e-06,
+      "loss": 0.3234,
+      "step": 14558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2975155229884535,
+      "learning_rate": 1.3699873004781983e-06,
+      "loss": 0.1767,
+      "step": 14559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31430545435525203,
+      "learning_rate": 1.369047313609575e-06,
+      "loss": 0.1938,
+      "step": 14560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3427307670134785,
+      "learning_rate": 1.3681076256304715e-06,
+      "loss": 0.3059,
+      "step": 14561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8615199044778867,
+      "learning_rate": 1.3671682365734273e-06,
+      "loss": 0.4229,
+      "step": 14562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31342070762916385,
+      "learning_rate": 1.3662291464709787e-06,
+      "loss": 0.2215,
+      "step": 14563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8984623397142127,
+      "learning_rate": 1.365290355355644e-06,
+      "loss": 0.4016,
+      "step": 14564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2759744627759715,
+      "learning_rate": 1.3643518632599317e-06,
+      "loss": 0.2307,
+      "step": 14565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2279301524681025,
+      "learning_rate": 1.3634136702163415e-06,
+      "loss": 0.1532,
+      "step": 14566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1406842605842138,
+      "learning_rate": 1.362475776257367e-06,
+      "loss": 0.7491,
+      "step": 14567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.586708889053408,
+      "learning_rate": 1.3615381814154848e-06,
+      "loss": 0.3272,
+      "step": 14568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28618586482601904,
+      "learning_rate": 1.3606008857231634e-06,
+      "loss": 0.222,
+      "step": 14569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5022053060065834,
+      "learning_rate": 1.3596638892128599e-06,
+      "loss": 0.3399,
+      "step": 14570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2986255975167238,
+      "learning_rate": 1.3587271919170276e-06,
+      "loss": 0.1707,
+      "step": 14571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35800026833904114,
+      "learning_rate": 1.3577907938681e-06,
+      "loss": 0.2807,
+      "step": 14572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.32434733804341365,
+      "learning_rate": 1.356854695098505e-06,
+      "loss": 0.256,
+      "step": 14573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7751751650216921,
+      "learning_rate": 1.3559188956406587e-06,
+      "loss": 0.3757,
+      "step": 14574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3422898662062519,
+      "learning_rate": 1.354983395526972e-06,
+      "loss": 0.2493,
+      "step": 14575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5236379453340966,
+      "learning_rate": 1.3540481947898377e-06,
+      "loss": 0.2442,
+      "step": 14576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4769062265048514,
+      "learning_rate": 1.3531132934616432e-06,
+      "loss": 0.3491,
+      "step": 14577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26125303282449164,
+      "learning_rate": 1.3521786915747636e-06,
+      "loss": 0.1996,
+      "step": 14578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4615285075486771,
+      "learning_rate": 1.3512443891615612e-06,
+      "loss": 0.2125,
+      "step": 14579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4793211131504592,
+      "learning_rate": 1.3503103862543964e-06,
+      "loss": 0.3389,
+      "step": 14580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27583841894197575,
+      "learning_rate": 1.3493766828856113e-06,
+      "loss": 0.2668,
+      "step": 14581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3888028705359416,
+      "learning_rate": 1.348443279087539e-06,
+      "loss": 0.247,
+      "step": 14582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5835514940606136,
+      "learning_rate": 1.3475101748925024e-06,
+      "loss": 0.2396,
+      "step": 14583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3265317907950052,
+      "learning_rate": 1.3465773703328177e-06,
+      "loss": 0.2626,
+      "step": 14584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34205468813191087,
+      "learning_rate": 1.3456448654407871e-06,
+      "loss": 0.2876,
+      "step": 14585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7978010123803017,
+      "learning_rate": 1.3447126602487026e-06,
+      "loss": 0.3144,
+      "step": 14586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31640964650076336,
+      "learning_rate": 1.343780754788847e-06,
+      "loss": 0.2539,
+      "step": 14587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35450717795972697,
+      "learning_rate": 1.3428491490934904e-06,
+      "loss": 0.1691,
+      "step": 14588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29752438649468915,
+      "learning_rate": 1.3419178431948964e-06,
+      "loss": 0.2152,
+      "step": 14589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34828109234051496,
+      "learning_rate": 1.3409868371253155e-06,
+      "loss": 0.266,
+      "step": 14590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.400501330494556,
+      "learning_rate": 1.3400561309169845e-06,
+      "loss": 0.2795,
+      "step": 14591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.32453042151607786,
+      "learning_rate": 1.3391257246021404e-06,
+      "loss": 0.2643,
+      "step": 14592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3316450046499373,
+      "learning_rate": 1.3381956182130008e-06,
+      "loss": 0.2635,
+      "step": 14593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5078673764787308,
+      "learning_rate": 1.3372658117817738e-06,
+      "loss": 0.2464,
+      "step": 14594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7956324440457225,
+      "learning_rate": 1.3363363053406564e-06,
+      "loss": 0.1765,
+      "step": 14595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.306535566421557,
+      "learning_rate": 1.3354070989218426e-06,
+      "loss": 0.2439,
+      "step": 14596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35559449451805636,
+      "learning_rate": 1.334478192557509e-06,
+      "loss": 0.3066,
+      "step": 14597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9039033717873965,
+      "learning_rate": 1.333549586279822e-06,
+      "loss": 0.4907,
+      "step": 14598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23362546523475522,
+      "learning_rate": 1.3326212801209392e-06,
+      "loss": 0.1655,
+      "step": 14599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41212155732044253,
+      "learning_rate": 1.3316932741130106e-06,
+      "loss": 0.2784,
+      "step": 14600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37043529185291274,
+      "learning_rate": 1.3307655682881704e-06,
+      "loss": 0.2799,
+      "step": 14601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3008023321460584,
+      "learning_rate": 1.3298381626785461e-06,
+      "loss": 0.2106,
+      "step": 14602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6480680200258,
+      "learning_rate": 1.3289110573162534e-06,
+      "loss": 0.3836,
+      "step": 14603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3419447209679332,
+      "learning_rate": 1.3279842522333964e-06,
+      "loss": 0.3303,
+      "step": 14604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28926904223399996,
+      "learning_rate": 1.3270577474620737e-06,
+      "loss": 0.1862,
+      "step": 14605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.333294240317376,
+      "learning_rate": 1.326131543034368e-06,
+      "loss": 0.1797,
+      "step": 14606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7753789409052013,
+      "learning_rate": 1.3252056389823542e-06,
+      "loss": 0.3864,
+      "step": 14607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4255672538071689,
+      "learning_rate": 1.3242800353380935e-06,
+      "loss": 0.2057,
+      "step": 14608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30257530347677447,
+      "learning_rate": 1.3233547321336449e-06,
+      "loss": 0.2747,
+      "step": 14609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1949937386415568,
+      "learning_rate": 1.322429729401048e-06,
+      "loss": 0.6453,
+      "step": 14610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39220038644195765,
+      "learning_rate": 1.3215050271723372e-06,
+      "loss": 0.2691,
+      "step": 14611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19771103947243748,
+      "learning_rate": 1.3205806254795316e-06,
+      "loss": 0.177,
+      "step": 14612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.340854636617087,
+      "learning_rate": 1.3196565243546477e-06,
+      "loss": 0.6778,
+      "step": 14613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40782738073281505,
+      "learning_rate": 1.3187327238296855e-06,
+      "loss": 0.2768,
+      "step": 14614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.48597633138533963,
+      "learning_rate": 1.3178092239366357e-06,
+      "loss": 0.2581,
+      "step": 14615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35702165673584346,
+      "learning_rate": 1.316886024707479e-06,
+      "loss": 0.298,
+      "step": 14616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43957135064561503,
+      "learning_rate": 1.3159631261741835e-06,
+      "loss": 0.2841,
+      "step": 14617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2303763190919167,
+      "learning_rate": 1.315040528368714e-06,
+      "loss": 0.1285,
+      "step": 14618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7590643493551357,
+      "learning_rate": 1.3141182313230173e-06,
+      "loss": 0.3696,
+      "step": 14619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3120861870291768,
+      "learning_rate": 1.313196235069033e-06,
+      "loss": 0.2635,
+      "step": 14620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3375885377689183,
+      "learning_rate": 1.3122745396386893e-06,
+      "loss": 0.2568,
+      "step": 14621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0596125325492274,
+      "learning_rate": 1.311353145063905e-06,
+      "loss": 0.6659,
+      "step": 14622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.32892295818439166,
+      "learning_rate": 1.3104320513765867e-06,
+      "loss": 0.2509,
+      "step": 14623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3057490849134694,
+      "learning_rate": 1.3095112586086322e-06,
+      "loss": 0.1777,
+      "step": 14624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35619765249654883,
+      "learning_rate": 1.3085907667919295e-06,
+      "loss": 0.2711,
+      "step": 14625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5930467136699953,
+      "learning_rate": 1.3076705759583562e-06,
+      "loss": 0.2686,
+      "step": 14626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4096537171452396,
+      "learning_rate": 1.3067506861397771e-06,
+      "loss": 0.3089,
+      "step": 14627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31363251366168626,
+      "learning_rate": 1.3058310973680478e-06,
+      "loss": 0.2472,
+      "step": 14628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6261934366587812,
+      "learning_rate": 1.3049118096750102e-06,
+      "loss": 0.322,
+      "step": 14629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23884814580511757,
+      "learning_rate": 1.3039928230925058e-06,
+      "loss": 0.2023,
+      "step": 14630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7587479042041084,
+      "learning_rate": 1.303074137652357e-06,
+      "loss": 0.2866,
+      "step": 14631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3200019958475434,
+      "learning_rate": 1.302155753386376e-06,
+      "loss": 0.2691,
+      "step": 14632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38354019565695807,
+      "learning_rate": 1.3012376703263652e-06,
+      "loss": 0.3087,
+      "step": 14633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2891788763038219,
+      "learning_rate": 1.3003198885041212e-06,
+      "loss": 0.757,
+      "step": 14634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3551321237848732,
+      "learning_rate": 1.2994024079514257e-06,
+      "loss": 0.161,
+      "step": 14635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2695399257198015,
+      "learning_rate": 1.2984852287000515e-06,
+      "loss": 0.2375,
+      "step": 14636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4797545726288128,
+      "learning_rate": 1.297568350781757e-06,
+      "loss": 0.3093,
+      "step": 14637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29754951174458355,
+      "learning_rate": 1.296651774228298e-06,
+      "loss": 0.1445,
+      "step": 14638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.573274589294748,
+      "learning_rate": 1.2957354990714145e-06,
+      "loss": 0.3267,
+      "step": 14639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3397778778825856,
+      "learning_rate": 1.2948195253428364e-06,
+      "loss": 0.3433,
+      "step": 14640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3975827403188928,
+      "learning_rate": 1.2939038530742832e-06,
+      "loss": 0.1751,
+      "step": 14641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5512201139354378,
+      "learning_rate": 1.2929884822974626e-06,
+      "loss": 0.3175,
+      "step": 14642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3345476677624379,
+      "learning_rate": 1.2920734130440793e-06,
+      "loss": 0.2317,
+      "step": 14643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27178999005541327,
+      "learning_rate": 1.2911586453458203e-06,
+      "loss": 0.1999,
+      "step": 14644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.393265473755049,
+      "learning_rate": 1.2902441792343611e-06,
+      "loss": 0.2757,
+      "step": 14645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0589891218703797,
+      "learning_rate": 1.2893300147413702e-06,
+      "loss": 0.5945,
+      "step": 14646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7805817018750784,
+      "learning_rate": 1.2884161518985083e-06,
+      "loss": 0.356,
+      "step": 14647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24815947867005486,
+      "learning_rate": 1.2875025907374206e-06,
+      "loss": 0.2283,
+      "step": 14648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7997982329037202,
+      "learning_rate": 1.2865893312897438e-06,
+      "loss": 0.3963,
+      "step": 14649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24864551993847825,
+      "learning_rate": 1.2856763735871003e-06,
+      "loss": 0.1527,
+      "step": 14650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3838687286845385,
+      "learning_rate": 1.2847637176611128e-06,
+      "loss": 0.2545,
+      "step": 14651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33946848678803093,
+      "learning_rate": 1.2838513635433824e-06,
+      "loss": 0.2888,
+      "step": 14652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5313711952676399,
+      "learning_rate": 1.2829393112655052e-06,
+      "loss": 0.3075,
+      "step": 14653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35368230488042623,
+      "learning_rate": 1.2820275608590638e-06,
+      "loss": 0.2496,
+      "step": 14654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5093236803162801,
+      "learning_rate": 1.2811161123556337e-06,
+      "loss": 0.2507,
+      "step": 14655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2429324623945723,
+      "learning_rate": 1.2802049657867777e-06,
+      "loss": 0.2158,
+      "step": 14656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4075378255586554,
+      "learning_rate": 1.2792941211840481e-06,
+      "loss": 0.2347,
+      "step": 14657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.813489974515319,
+      "learning_rate": 1.2783835785789867e-06,
+      "loss": 0.4763,
+      "step": 14658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5005999740093483,
+      "learning_rate": 1.277473338003129e-06,
+      "loss": 0.3164,
+      "step": 14659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3206738895171215,
+      "learning_rate": 1.2765633994879933e-06,
+      "loss": 0.2816,
+      "step": 14660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703995448615973,
+      "learning_rate": 1.2756537630650934e-06,
+      "loss": 0.2392,
+      "step": 14661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24137241230342263,
+      "learning_rate": 1.274744428765926e-06,
+      "loss": 0.1746,
+      "step": 14662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3367852646989017,
+      "learning_rate": 1.2738353966219863e-06,
+      "loss": 0.2836,
+      "step": 14663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3384807639190101,
+      "learning_rate": 1.2729266666647511e-06,
+      "loss": 0.264,
+      "step": 14664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6314320406801325,
+      "learning_rate": 1.2720182389256896e-06,
+      "loss": 0.362,
+      "step": 14665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3457808849956551,
+      "learning_rate": 1.2711101134362624e-06,
+      "loss": 0.2846,
+      "step": 14666
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5896610511400546,
+      "learning_rate": 1.2702022902279132e-06,
+      "loss": 0.1807,
+      "step": 14667
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20164738798995183,
+      "learning_rate": 1.2692947693320867e-06,
+      "loss": 0.1918,
+      "step": 14668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34464558067401213,
+      "learning_rate": 1.2683875507802058e-06,
+      "loss": 0.2875,
+      "step": 14669
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7667195525112961,
+      "learning_rate": 1.2674806346036895e-06,
+      "loss": 0.3144,
+      "step": 14670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.524432798062477,
+      "learning_rate": 1.2665740208339406e-06,
+      "loss": 0.3225,
+      "step": 14671
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25184485692333597,
+      "learning_rate": 1.2656677095023607e-06,
+      "loss": 0.2584,
+      "step": 14672
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4804813262002852,
+      "learning_rate": 1.2647617006403312e-06,
+      "loss": 0.4913,
+      "step": 14673
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.16779157793554556,
+      "learning_rate": 1.2638559942792294e-06,
+      "loss": 0.0875,
+      "step": 14674
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39519690266180457,
+      "learning_rate": 1.2629505904504158e-06,
+      "loss": 0.3048,
+      "step": 14675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3542800298985227,
+      "learning_rate": 1.2620454891852507e-06,
+      "loss": 0.3001,
+      "step": 14676
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38034372533991734,
+      "learning_rate": 1.2611406905150736e-06,
+      "loss": 0.1732,
+      "step": 14677
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38478117119739746,
+      "learning_rate": 1.2602361944712193e-06,
+      "loss": 0.2879,
+      "step": 14678
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5953891231925955,
+      "learning_rate": 1.2593320010850096e-06,
+      "loss": 0.329,
+      "step": 14679
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4313422987262344,
+      "learning_rate": 1.258428110387754e-06,
+      "loss": 0.2362,
+      "step": 14680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3322900970369777,
+      "learning_rate": 1.2575245224107602e-06,
+      "loss": 0.2568,
+      "step": 14681
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5778985999746848,
+      "learning_rate": 1.256621237185316e-06,
+      "loss": 0.388,
+      "step": 14682
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25766138829972673,
+      "learning_rate": 1.2557182547427016e-06,
+      "loss": 0.1669,
+      "step": 14683
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23468154882529396,
+      "learning_rate": 1.2548155751141867e-06,
+      "loss": 0.2158,
+      "step": 14684
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4762527134046515,
+      "learning_rate": 1.2539131983310349e-06,
+      "loss": 0.5962,
+      "step": 14685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3406802197347614,
+      "learning_rate": 1.2530111244244925e-06,
+      "loss": 0.6108,
+      "step": 14686
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2989282297239217,
+      "learning_rate": 1.2521093534257977e-06,
+      "loss": 0.2035,
+      "step": 14687
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.32967555204294186,
+      "learning_rate": 1.2512078853661813e-06,
+      "loss": 0.2942,
+      "step": 14688
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26575499268107783,
+      "learning_rate": 1.2503067202768592e-06,
+      "loss": 0.1926,
+      "step": 14689
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34793399148411513,
+      "learning_rate": 1.2494058581890388e-06,
+      "loss": 0.1929,
+      "step": 14690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.44843755470370344,
+      "learning_rate": 1.2485052991339174e-06,
+      "loss": 0.2978,
+      "step": 14691
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5016887458702256,
+      "learning_rate": 1.247605043142679e-06,
+      "loss": 0.2965,
+      "step": 14692
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35071326656103574,
+      "learning_rate": 1.2467050902465038e-06,
+      "loss": 0.2073,
+      "step": 14693
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7348697816631484,
+      "learning_rate": 1.2458054404765552e-06,
+      "loss": 0.4092,
+      "step": 14694
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3464063282503142,
+      "learning_rate": 1.2449060938639869e-06,
+      "loss": 0.3186,
+      "step": 14695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2497142048809561,
+      "learning_rate": 1.2440070504399426e-06,
+      "loss": 0.1682,
+      "step": 14696
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5003880498698998,
+      "learning_rate": 1.243108310235559e-06,
+      "loss": 0.2321,
+      "step": 14697
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6704924239750555,
+      "learning_rate": 1.2422098732819587e-06,
+      "loss": 0.3914,
+      "step": 14698
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30569771907205934,
+      "learning_rate": 1.2413117396102548e-06,
+      "loss": 0.2449,
+      "step": 14699
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3158934485374403,
+      "learning_rate": 1.2404139092515455e-06,
+      "loss": 0.2607,
+      "step": 14700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46902780635018154,
+      "learning_rate": 1.2395163822369283e-06,
+      "loss": 0.2458,
+      "step": 14701
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24793381285511187,
+      "learning_rate": 1.2386191585974815e-06,
+      "loss": 0.2093,
+      "step": 14702
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.568270893805069,
+      "learning_rate": 1.2377222383642773e-06,
+      "loss": 0.2455,
+      "step": 14703
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4898535112631793,
+      "learning_rate": 1.2368256215683727e-06,
+      "loss": 0.3084,
+      "step": 14704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3244437239492716,
+      "learning_rate": 1.235929308240822e-06,
+      "loss": 0.2521,
+      "step": 14705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8715598875641944,
+      "learning_rate": 1.2350332984126623e-06,
+      "loss": 0.3682,
+      "step": 14706
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.32979803200018204,
+      "learning_rate": 1.2341375921149224e-06,
+      "loss": 0.2899,
+      "step": 14707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2545636420010435,
+      "learning_rate": 1.2332421893786218e-06,
+      "loss": 0.2019,
+      "step": 14708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4792169662322209,
+      "learning_rate": 1.2323470902347645e-06,
+      "loss": 0.2032,
+      "step": 14709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5781818608555402,
+      "learning_rate": 1.2314522947143526e-06,
+      "loss": 0.2895,
+      "step": 14710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4198601434035437,
+      "learning_rate": 1.23055780284837e-06,
+      "loss": 0.27,
+      "step": 14711
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2760695555581516,
+      "learning_rate": 1.2296636146677942e-06,
+      "loss": 0.2711,
+      "step": 14712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8359298804618515,
+      "learning_rate": 1.2287697302035883e-06,
+      "loss": 0.2242,
+      "step": 14713
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35699267197134954,
+      "learning_rate": 1.227876149486712e-06,
+      "loss": 0.2875,
+      "step": 14714
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25071982956057953,
+      "learning_rate": 1.226982872548107e-06,
+      "loss": 0.1969,
+      "step": 14715
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43126805352484987,
+      "learning_rate": 1.2260898994187075e-06,
+      "loss": 0.2655,
+      "step": 14716
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3063248769232248,
+      "learning_rate": 1.2251972301294358e-06,
+      "loss": 0.2335,
+      "step": 14717
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0832029188795427,
+      "learning_rate": 1.2243048647112078e-06,
+      "loss": 0.7687,
+      "step": 14718
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3293628692072448,
+      "learning_rate": 1.2234128031949266e-06,
+      "loss": 0.2336,
+      "step": 14719
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3376833571189543,
+      "learning_rate": 1.222521045611481e-06,
+      "loss": 0.2651,
+      "step": 14720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37187860367678754,
+      "learning_rate": 1.2216295919917553e-06,
+      "loss": 0.2057,
+      "step": 14721
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8919618832579139,
+      "learning_rate": 1.220738442366619e-06,
+      "loss": 0.4321,
+      "step": 14722
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2582549238915909,
+      "learning_rate": 1.2198475967669333e-06,
+      "loss": 0.21,
+      "step": 14723
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4988948087103634,
+      "learning_rate": 1.2189570552235475e-06,
+      "loss": 0.3751,
+      "step": 14724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0828966430021156,
+      "learning_rate": 1.2180668177672984e-06,
+      "loss": 0.552,
+      "step": 14725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2535459801459226,
+      "learning_rate": 1.217176884429021e-06,
+      "loss": 0.1598,
+      "step": 14726
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4715556154628492,
+      "learning_rate": 1.21628725523953e-06,
+      "loss": 0.3407,
+      "step": 14727
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.333192726383167,
+      "learning_rate": 1.2153979302296338e-06,
+      "loss": 0.2542,
+      "step": 14728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3590585580216333,
+      "learning_rate": 1.2145089094301265e-06,
+      "loss": 0.1648,
+      "step": 14729
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4915144416030057,
+      "learning_rate": 1.2136201928718005e-06,
+      "loss": 0.3565,
+      "step": 14730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3515123559483987,
+      "learning_rate": 1.21273178058543e-06,
+      "loss": 0.2912,
+      "step": 14731
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.38349679742074594,
+      "learning_rate": 1.21184367260178e-06,
+      "loss": 0.1698,
+      "step": 14732
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40644587846781555,
+      "learning_rate": 1.2109558689516054e-06,
+      "loss": 0.3135,
+      "step": 14733
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3399566033908146,
+      "learning_rate": 1.210068369665649e-06,
+      "loss": 0.1875,
+      "step": 14734
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31943117545798166,
+      "learning_rate": 1.2091811747746484e-06,
+      "loss": 0.2578,
+      "step": 14735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3502609198287449,
+      "learning_rate": 1.208294284309327e-06,
+      "loss": 0.2722,
+      "step": 14736
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4569221299642199,
+      "learning_rate": 1.2074076983003956e-06,
+      "loss": 0.6047,
+      "step": 14737
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40102742209080916,
+      "learning_rate": 1.2065214167785554e-06,
+      "loss": 0.2702,
+      "step": 14738
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29010025220729196,
+      "learning_rate": 1.2056354397745029e-06,
+      "loss": 0.2478,
+      "step": 14739
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3096437615449813,
+      "learning_rate": 1.2047497673189169e-06,
+      "loss": 0.1753,
+      "step": 14740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3867388418154201,
+      "learning_rate": 1.2038643994424682e-06,
+      "loss": 0.2674,
+      "step": 14741
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4669314450298384,
+      "learning_rate": 1.2029793361758146e-06,
+      "loss": 0.2701,
+      "step": 14742
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36142721745288964,
+      "learning_rate": 1.2020945775496107e-06,
+      "loss": 0.2782,
+      "step": 14743
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2954197524558824,
+      "learning_rate": 1.201210123594494e-06,
+      "loss": 0.2351,
+      "step": 14744
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6366188960797053,
+      "learning_rate": 1.200325974341091e-06,
+      "loss": 0.3014,
+      "step": 14745
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23679746876696145,
+      "learning_rate": 1.199442129820022e-06,
+      "loss": 0.1938,
+      "step": 14746
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3110930947191409,
+      "learning_rate": 1.1985585900618912e-06,
+      "loss": 0.2664,
+      "step": 14747
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5009217757277309,
+      "learning_rate": 1.1976753550972998e-06,
+      "loss": 0.3774,
+      "step": 14748
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5908704134342048,
+      "learning_rate": 1.196792424956833e-06,
+      "loss": 0.2874,
+      "step": 14749
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.527797718182077,
+      "learning_rate": 1.1959097996710656e-06,
+      "loss": 0.2624,
+      "step": 14750
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25644427630796324,
+      "learning_rate": 1.1950274792705618e-06,
+      "loss": 0.2696,
+      "step": 14751
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.16068604115031754,
+      "learning_rate": 1.1941454637858784e-06,
+      "loss": 0.0712,
+      "step": 14752
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5669102028958434,
+      "learning_rate": 1.19326375324756e-06,
+      "loss": 0.3149,
+      "step": 14753
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37777598291555,
+      "learning_rate": 1.1923823476861395e-06,
+      "loss": 0.3254,
+      "step": 14754
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3227218948916543,
+      "learning_rate": 1.1915012471321385e-06,
+      "loss": 0.2587,
+      "step": 14755
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3981084157797156,
+      "learning_rate": 1.1906204516160713e-06,
+      "loss": 0.3059,
+      "step": 14756
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5493564974138063,
+      "learning_rate": 1.189739961168439e-06,
+      "loss": 0.3309,
+      "step": 14757
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31382816631129834,
+      "learning_rate": 1.1888597758197319e-06,
+      "loss": 0.1758,
+      "step": 14758
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25289083122443995,
+      "learning_rate": 1.1879798956004307e-06,
+      "loss": 0.2251,
+      "step": 14759
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5491193092701862,
+      "learning_rate": 1.1871003205410092e-06,
+      "loss": 0.4062,
+      "step": 14760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6820129227771387,
+      "learning_rate": 1.186221050671924e-06,
+      "loss": 0.3746,
+      "step": 14761
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.358774579788031,
+      "learning_rate": 1.1853420860236253e-06,
+      "loss": 0.1911,
+      "step": 14762
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30419959453244505,
+      "learning_rate": 1.1844634266265487e-06,
+      "loss": 0.266,
+      "step": 14763
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4638461851125854,
+      "learning_rate": 1.1835850725111264e-06,
+      "loss": 0.2343,
+      "step": 14764
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22267885583250036,
+      "learning_rate": 1.1827070237077743e-06,
+      "loss": 0.1363,
+      "step": 14765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5251370761115037,
+      "learning_rate": 1.1818292802468989e-06,
+      "loss": 0.3503,
+      "step": 14766
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4038183648154914,
+      "learning_rate": 1.1809518421588939e-06,
+      "loss": 0.2937,
+      "step": 14767
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5477079411295382,
+      "learning_rate": 1.1800747094741493e-06,
+      "loss": 0.1356,
+      "step": 14768
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.42781575837813424,
+      "learning_rate": 1.1791978822230388e-06,
+      "loss": 0.2874,
+      "step": 14769
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.38030178099469747,
+      "learning_rate": 1.1783213604359268e-06,
+      "loss": 0.2965,
+      "step": 14770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19456608161814554,
+      "learning_rate": 1.1774451441431655e-06,
+      "loss": 0.0866,
+      "step": 14771
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39569769749296096,
+      "learning_rate": 1.1765692333750977e-06,
+      "loss": 0.2675,
+      "step": 14772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6547391841835668,
+      "learning_rate": 1.17569362816206e-06,
+      "loss": 0.3861,
+      "step": 14773
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2570354731252588,
+      "learning_rate": 1.174818328534373e-06,
+      "loss": 0.1988,
+      "step": 14774
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27903556542289093,
+      "learning_rate": 1.1739433345223482e-06,
+      "loss": 0.2137,
+      "step": 14775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4667329346849802,
+      "learning_rate": 1.1730686461562835e-06,
+      "loss": 0.4638,
+      "step": 14776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5838213979491272,
+      "learning_rate": 1.172194263466474e-06,
+      "loss": 0.3362,
+      "step": 14777
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3959474359561361,
+      "learning_rate": 1.1713201864831968e-06,
+      "loss": 0.2403,
+      "step": 14778
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33580085060755466,
+      "learning_rate": 1.1704464152367234e-06,
+      "loss": 0.2984,
+      "step": 14779
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22515204910858588,
+      "learning_rate": 1.1695729497573082e-06,
+      "loss": 0.156,
+      "step": 14780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41575650449527546,
+      "learning_rate": 1.168699790075204e-06,
+      "loss": 0.2199,
+      "step": 14781
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5596586766328981,
+      "learning_rate": 1.1678269362206463e-06,
+      "loss": 0.3317,
+      "step": 14782
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3552412822573221,
+      "learning_rate": 1.166954388223862e-06,
+      "loss": 0.25,
+      "step": 14783
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36014971889060016,
+      "learning_rate": 1.1660821461150673e-06,
+      "loss": 0.2429,
+      "step": 14784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.42642570719810025,
+      "learning_rate": 1.1652102099244667e-06,
+      "loss": 0.2757,
+      "step": 14785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2994411664078834,
+      "learning_rate": 1.1643385796822582e-06,
+      "loss": 0.2318,
+      "step": 14786
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29557420750535496,
+      "learning_rate": 1.1634672554186243e-06,
+      "loss": 0.259,
+      "step": 14787
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.675274623281233,
+      "learning_rate": 1.16259623716374e-06,
+      "loss": 0.2517,
+      "step": 14788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6074414344124812,
+      "learning_rate": 1.1617255249477677e-06,
+      "loss": 0.316,
+      "step": 14789
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.360726252157882,
+      "learning_rate": 1.16085511880086e-06,
+      "loss": 0.2795,
+      "step": 14790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32731888538386056,
+      "learning_rate": 1.1599850187531603e-06,
+      "loss": 0.2495,
+      "step": 14791
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2693011419688268,
+      "learning_rate": 1.1591152248347959e-06,
+      "loss": 0.159,
+      "step": 14792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3694108954988957,
+      "learning_rate": 1.1582457370758948e-06,
+      "loss": 0.2588,
+      "step": 14793
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.42252599397338797,
+      "learning_rate": 1.157376555506562e-06,
+      "loss": 0.2219,
+      "step": 14794
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3670240559105416,
+      "learning_rate": 1.1565076801568997e-06,
+      "loss": 0.2801,
+      "step": 14795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4043612092318145,
+      "learning_rate": 1.1556391110569965e-06,
+      "loss": 0.2843,
+      "step": 14796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1551526431797676,
+      "learning_rate": 1.1547708482369279e-06,
+      "loss": 0.719,
+      "step": 14797
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22288561492500292,
+      "learning_rate": 1.1539028917267668e-06,
+      "loss": 0.1688,
+      "step": 14798
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2946033880771682,
+      "learning_rate": 1.1530352415565683e-06,
+      "loss": 0.2451,
+      "step": 14799
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9260136414904041,
+      "learning_rate": 1.152167897756379e-06,
+      "loss": 0.3756,
+      "step": 14800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.46156669228028563,
+      "learning_rate": 1.1513008603562327e-06,
+      "loss": 0.2187,
+      "step": 14801
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4098473307632537,
+      "learning_rate": 1.1504341293861588e-06,
+      "loss": 0.3288,
+      "step": 14802
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3391465615133372,
+      "learning_rate": 1.149567704876171e-06,
+      "loss": 0.3124,
+      "step": 14803
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24197185666517396,
+      "learning_rate": 1.1487015868562723e-06,
+      "loss": 0.0848,
+      "step": 14804
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30112775030509276,
+      "learning_rate": 1.147835775356455e-06,
+      "loss": 0.2428,
+      "step": 14805
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31423519705737857,
+      "learning_rate": 1.1469702704067064e-06,
+      "loss": 0.3014,
+      "step": 14806
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7919262173209054,
+      "learning_rate": 1.146105072036997e-06,
+      "loss": 0.2429,
+      "step": 14807
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35309423242059595,
+      "learning_rate": 1.1452401802772884e-06,
+      "loss": 0.2779,
+      "step": 14808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1765512888066414,
+      "learning_rate": 1.144375595157532e-06,
+      "loss": 0.7295,
+      "step": 14809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26690693768113766,
+      "learning_rate": 1.143511316707665e-06,
+      "loss": 0.2349,
+      "step": 14810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22741761840826571,
+      "learning_rate": 1.1426473449576225e-06,
+      "loss": 0.1552,
+      "step": 14811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8487452061726126,
+      "learning_rate": 1.1417836799373205e-06,
+      "loss": 0.3599,
+      "step": 14812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7051579592509625,
+      "learning_rate": 1.1409203216766706e-06,
+      "loss": 0.4184,
+      "step": 14813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24162397410920866,
+      "learning_rate": 1.1400572702055657e-06,
+      "loss": 0.1935,
+      "step": 14814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45583372486177876,
+      "learning_rate": 1.1391945255538994e-06,
+      "loss": 0.3813,
+      "step": 14815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4161358729060462,
+      "learning_rate": 1.1383320877515446e-06,
+      "loss": 0.2391,
+      "step": 14816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26035907038958234,
+      "learning_rate": 1.1374699568283698e-06,
+      "loss": 0.1717,
+      "step": 14817
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32519859938343304,
+      "learning_rate": 1.1366081328142264e-06,
+      "loss": 0.2887,
+      "step": 14818
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6847648618208944,
+      "learning_rate": 1.135746615738965e-06,
+      "loss": 0.3494,
+      "step": 14819
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32858251668966476,
+      "learning_rate": 1.1348854056324166e-06,
+      "loss": 0.212,
+      "step": 14820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1507405287489314,
+      "learning_rate": 1.1340245025244045e-06,
+      "loss": 0.7616,
+      "step": 14821
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36743606071499174,
+      "learning_rate": 1.133163906444742e-06,
+      "loss": 0.3013,
+      "step": 14822
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3275912685999875,
+      "learning_rate": 1.132303617423236e-06,
+      "loss": 0.2387,
+      "step": 14823
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34411631210529814,
+      "learning_rate": 1.131443635489672e-06,
+      "loss": 0.1776,
+      "step": 14824
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4080656699672398,
+      "learning_rate": 1.1305839606738334e-06,
+      "loss": 0.717,
+      "step": 14825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25755544714127737,
+      "learning_rate": 1.129724593005489e-06,
+      "loss": 0.2429,
+      "step": 14826
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4605922701134536,
+      "learning_rate": 1.1288655325144027e-06,
+      "loss": 0.3024,
+      "step": 14827
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8821629142995729,
+      "learning_rate": 1.1280067792303218e-06,
+      "loss": 0.4034,
+      "step": 14828
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3203927459270146,
+      "learning_rate": 1.1271483331829835e-06,
+      "loss": 0.2634,
+      "step": 14829
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33166642821629216,
+      "learning_rate": 1.1262901944021165e-06,
+      "loss": 0.245,
+      "step": 14830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3227943214683181,
+      "learning_rate": 1.125432362917439e-06,
+      "loss": 0.1672,
+      "step": 14831
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31125204319825606,
+      "learning_rate": 1.1245748387586575e-06,
+      "loss": 0.2647,
+      "step": 14832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8780581738774083,
+      "learning_rate": 1.123717621955468e-06,
+      "loss": 0.364,
+      "step": 14833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3650475322050246,
+      "learning_rate": 1.1228607125375534e-06,
+      "loss": 0.2866,
+      "step": 14834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40110708736483236,
+      "learning_rate": 1.1220041105345935e-06,
+      "loss": 0.2838,
+      "step": 14835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.560535169035649,
+      "learning_rate": 1.121147815976248e-06,
+      "loss": 0.3457,
+      "step": 14836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20911265248523273,
+      "learning_rate": 1.1202918288921727e-06,
+      "loss": 0.143,
+      "step": 14837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3534204337682163,
+      "learning_rate": 1.1194361493120099e-06,
+      "loss": 0.268,
+      "step": 14838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.47013886187275117,
+      "learning_rate": 1.118580777265388e-06,
+      "loss": 0.3605,
+      "step": 14839
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7626126324582352,
+      "learning_rate": 1.1177257127819353e-06,
+      "loss": 0.3003,
+      "step": 14840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31851659287056955,
+      "learning_rate": 1.1168709558912583e-06,
+      "loss": 0.2653,
+      "step": 14841
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3898838568972243,
+      "learning_rate": 1.116016506622959e-06,
+      "loss": 0.2784,
+      "step": 14842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.15368938464097603,
+      "learning_rate": 1.1151623650066224e-06,
+      "loss": 0.0877,
+      "step": 14843
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30142584311757015,
+      "learning_rate": 1.114308531071835e-06,
+      "loss": 0.238,
+      "step": 14844
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2307746884376518,
+      "learning_rate": 1.1134550048481596e-06,
+      "loss": 0.5617,
+      "step": 14845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3237698295002861,
+      "learning_rate": 1.1126017863651562e-06,
+      "loss": 0.2563,
+      "step": 14846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31133573341743925,
+      "learning_rate": 1.1117488756523677e-06,
+      "loss": 0.2625,
+      "step": 14847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3573196658449307,
+      "learning_rate": 1.1108962727393368e-06,
+      "loss": 0.5345,
+      "step": 14848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27633793321812017,
+      "learning_rate": 1.110043977655585e-06,
+      "loss": 0.1962,
+      "step": 14849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23377373058249695,
+      "learning_rate": 1.109191990430628e-06,
+      "loss": 0.2015,
+      "step": 14850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2731592859606273,
+      "learning_rate": 1.1083403110939695e-06,
+      "loss": 0.7351,
+      "step": 14851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.635556152341861,
+      "learning_rate": 1.107488939675102e-06,
+      "loss": 0.374,
+      "step": 14852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3848628568892182,
+      "learning_rate": 1.1066378762035125e-06,
+      "loss": 0.1684,
+      "step": 14853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3072768771702361,
+      "learning_rate": 1.1057871207086713e-06,
+      "loss": 0.2661,
+      "step": 14854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3946099595301403,
+      "learning_rate": 1.1049366732200383e-06,
+      "loss": 0.2413,
+      "step": 14855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3380880546143608,
+      "learning_rate": 1.104086533767067e-06,
+      "loss": 0.1868,
+      "step": 14856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4442492965914318,
+      "learning_rate": 1.1032367023791957e-06,
+      "loss": 0.3175,
+      "step": 14857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2528655784441553,
+      "learning_rate": 1.1023871790858553e-06,
+      "loss": 0.2477,
+      "step": 14858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28372712548736984,
+      "learning_rate": 1.1015379639164625e-06,
+      "loss": 0.1826,
+      "step": 14859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1834437417520254,
+      "learning_rate": 1.100689056900429e-06,
+      "loss": 0.5269,
+      "step": 14860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3866428294105834,
+      "learning_rate": 1.0998404580671507e-06,
+      "loss": 0.2489,
+      "step": 14861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25646265297625703,
+      "learning_rate": 1.0989921674460146e-06,
+      "loss": 0.2444,
+      "step": 14862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44426050608931333,
+      "learning_rate": 1.0981441850663976e-06,
+      "loss": 0.263,
+      "step": 14863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43875055594403534,
+      "learning_rate": 1.0972965109576628e-06,
+      "loss": 0.25,
+      "step": 14864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4141839426559247,
+      "learning_rate": 1.0964491451491677e-06,
+      "loss": 0.2651,
+      "step": 14865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2750628065825932,
+      "learning_rate": 1.0956020876702567e-06,
+      "loss": 0.2278,
+      "step": 14866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0146697997317078,
+      "learning_rate": 1.094755338550263e-06,
+      "loss": 0.5553,
+      "step": 14867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3163524879755541,
+      "learning_rate": 1.0939088978185053e-06,
+      "loss": 0.2456,
+      "step": 14868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6128769729330104,
+      "learning_rate": 1.0930627655043036e-06,
+      "loss": 0.2857,
+      "step": 14869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23060783739572102,
+      "learning_rate": 1.0922169416369531e-06,
+      "loss": 0.2126,
+      "step": 14870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6672465753746528,
+      "learning_rate": 1.0913714262457486e-06,
+      "loss": 0.3115,
+      "step": 14871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3828226055546082,
+      "learning_rate": 1.0905262193599665e-06,
+      "loss": 0.3032,
+      "step": 14872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30441911790214365,
+      "learning_rate": 1.0896813210088797e-06,
+      "loss": 0.2394,
+      "step": 14873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6542472129796099,
+      "learning_rate": 1.0888367312217452e-06,
+      "loss": 0.2881,
+      "step": 14874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37036987367157137,
+      "learning_rate": 1.0879924500278116e-06,
+      "loss": 0.2922,
+      "step": 14875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2711333428099278,
+      "learning_rate": 1.087148477456317e-06,
+      "loss": 0.1101,
+      "step": 14876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3898315605171779,
+      "learning_rate": 1.0863048135364851e-06,
+      "loss": 0.2556,
+      "step": 14877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32269214206454216,
+      "learning_rate": 1.0854614582975353e-06,
+      "loss": 0.2821,
+      "step": 14878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4122935758067834,
+      "learning_rate": 1.084618411768673e-06,
+      "loss": 0.3141,
+      "step": 14879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40451562657458473,
+      "learning_rate": 1.0837756739790916e-06,
+      "loss": 0.269,
+      "step": 14880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36283485462304216,
+      "learning_rate": 1.0829332449579732e-06,
+      "loss": 0.322,
+      "step": 14881
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28832612026905335,
+      "learning_rate": 1.0820911247344944e-06,
+      "loss": 0.1806,
+      "step": 14882
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26597397855279037,
+      "learning_rate": 1.0812493133378166e-06,
+      "loss": 0.1913,
+      "step": 14883
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5791713121666998,
+      "learning_rate": 1.0804078107970917e-06,
+      "loss": 0.3546,
+      "step": 14884
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5003590603484814,
+      "learning_rate": 1.0795666171414597e-06,
+      "loss": 0.3228,
+      "step": 14885
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2956405417591853,
+      "learning_rate": 1.0787257324000533e-06,
+      "loss": 0.2204,
+      "step": 14886
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5080100270148686,
+      "learning_rate": 1.077885156601991e-06,
+      "loss": 0.3585,
+      "step": 14887
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49071658954089276,
+      "learning_rate": 1.0770448897763818e-06,
+      "loss": 0.2817,
+      "step": 14888
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.15267670632605457,
+      "learning_rate": 1.0762049319523248e-06,
+      "loss": 0.0704,
+      "step": 14889
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2648087586985502,
+      "learning_rate": 1.075365283158908e-06,
+      "loss": 0.2688,
+      "step": 14890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6311471408354534,
+      "learning_rate": 1.0745259434252065e-06,
+      "loss": 0.3905,
+      "step": 14891
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.671787505272611,
+      "learning_rate": 1.0736869127802884e-06,
+      "loss": 0.1974,
+      "step": 14892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2957218483145482,
+      "learning_rate": 1.0728481912532062e-06,
+      "loss": 0.2763,
+      "step": 14893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32442168752672784,
+      "learning_rate": 1.07200977887301e-06,
+      "loss": 0.2365,
+      "step": 14894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3205645128811862,
+      "learning_rate": 1.0711716756687307e-06,
+      "loss": 0.0799,
+      "step": 14895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40600771972047073,
+      "learning_rate": 1.070333881669392e-06,
+      "loss": 0.2881,
+      "step": 14896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4958254382676697,
+      "learning_rate": 1.0694963969040062e-06,
+      "loss": 0.3264,
+      "step": 14897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40568174234009224,
+      "learning_rate": 1.0686592214015766e-06,
+      "loss": 0.2968,
+      "step": 14898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32870509799899306,
+      "learning_rate": 1.067822355191095e-06,
+      "loss": 0.235,
+      "step": 14899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0628361533545403,
+      "learning_rate": 1.0669857983015408e-06,
+      "loss": 0.4543,
+      "step": 14900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20065123792574188,
+      "learning_rate": 1.0661495507618845e-06,
+      "loss": 0.1679,
+      "step": 14901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29083032401850334,
+      "learning_rate": 1.0653136126010832e-06,
+      "loss": 0.223,
+      "step": 14902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5620063770534373,
+      "learning_rate": 1.06447798384809e-06,
+      "loss": 0.3486,
+      "step": 14903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.554067422855577,
+      "learning_rate": 1.0636426645318387e-06,
+      "loss": 0.3698,
+      "step": 14904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28406674047779545,
+      "learning_rate": 1.0628076546812583e-06,
+      "loss": 0.2318,
+      "step": 14905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5269874383643349,
+      "learning_rate": 1.0619729543252622e-06,
+      "loss": 0.3355,
+      "step": 14906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24121729656072172,
+      "learning_rate": 1.0611385634927607e-06,
+      "loss": 0.1645,
+      "step": 14907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.35797358852460026,
+      "learning_rate": 1.0603044822126463e-06,
+      "loss": 0.2266,
+      "step": 14908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3378453061765035,
+      "learning_rate": 1.0594707105138024e-06,
+      "loss": 0.2897,
+      "step": 14909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8746212764884232,
+      "learning_rate": 1.0586372484251018e-06,
+      "loss": 0.391,
+      "step": 14910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32721325459226824,
+      "learning_rate": 1.057804095975411e-06,
+      "loss": 0.2678,
+      "step": 14911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5419081597400264,
+      "learning_rate": 1.0569712531935805e-06,
+      "loss": 0.277,
+      "step": 14912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2279118449162286,
+      "learning_rate": 1.0561387201084494e-06,
+      "loss": 0.206,
+      "step": 14913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.288549030985663,
+      "learning_rate": 1.0553064967488514e-06,
+      "loss": 0.2256,
+      "step": 14914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6606990633306803,
+      "learning_rate": 1.054474583143602e-06,
+      "loss": 0.2891,
+      "step": 14915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6655593582916631,
+      "learning_rate": 1.0536429793215152e-06,
+      "loss": 0.399,
+      "step": 14916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24780553816203488,
+      "learning_rate": 1.0528116853113867e-06,
+      "loss": 0.2537,
+      "step": 14917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4479855245872628,
+      "learning_rate": 1.0519807011420057e-06,
+      "loss": 0.2714,
+      "step": 14918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6947298411391986,
+      "learning_rate": 1.051150026842146e-06,
+      "loss": 0.3632,
+      "step": 14919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3823419776922345,
+      "learning_rate": 1.0503196624405775e-06,
+      "loss": 0.259,
+      "step": 14920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2174358692451919,
+      "learning_rate": 1.0494896079660554e-06,
+      "loss": 0.1818,
+      "step": 14921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6206679548783893,
+      "learning_rate": 1.0486598634473221e-06,
+      "loss": 0.3312,
+      "step": 14922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40084767726771736,
+      "learning_rate": 1.0478304289131115e-06,
+      "loss": 0.3024,
+      "step": 14923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3130018162299788,
+      "learning_rate": 1.0470013043921523e-06,
+      "loss": 0.6565,
+      "step": 14924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.317664336895266,
+      "learning_rate": 1.046172489913151e-06,
+      "loss": 0.237,
+      "step": 14925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3425571725675746,
+      "learning_rate": 1.0453439855048108e-06,
+      "loss": 0.2797,
+      "step": 14926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30894367529548855,
+      "learning_rate": 1.0445157911958214e-06,
+      "loss": 0.1792,
+      "step": 14927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.014875460132905,
+      "learning_rate": 1.0436879070148675e-06,
+      "loss": 0.5139,
+      "step": 14928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26322226528383447,
+      "learning_rate": 1.042860332990615e-06,
+      "loss": 0.2515,
+      "step": 14929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5905553635015627,
+      "learning_rate": 1.0420330691517256e-06,
+      "loss": 0.3113,
+      "step": 14930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5689957343239428,
+      "learning_rate": 1.0412061155268428e-06,
+      "loss": 0.2489,
+      "step": 14931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3543218393322141,
+      "learning_rate": 1.0403794721446092e-06,
+      "loss": 0.257,
+      "step": 14932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.386805524160091,
+      "learning_rate": 1.03955313903365e-06,
+      "loss": 0.3118,
+      "step": 14933
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18668486006155857,
+      "learning_rate": 1.0387271162225787e-06,
+      "loss": 0.1076,
+      "step": 14934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.39926953655521297,
+      "learning_rate": 1.0379014037400014e-06,
+      "loss": 0.2844,
+      "step": 14935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8440325835444368,
+      "learning_rate": 1.0370760016145142e-06,
+      "loss": 0.4694,
+      "step": 14936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3958906252639677,
+      "learning_rate": 1.036250909874701e-06,
+      "loss": 0.3087,
+      "step": 14937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2876575469257746,
+      "learning_rate": 1.0354261285491319e-06,
+      "loss": 0.188,
+      "step": 14938
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24698100951360044,
+      "learning_rate": 1.0346016576663686e-06,
+      "loss": 0.1631,
+      "step": 14939
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8047880203921186,
+      "learning_rate": 1.0337774972549675e-06,
+      "loss": 0.5828,
+      "step": 14940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2618812698548808,
+      "learning_rate": 1.0329536473434653e-06,
+      "loss": 0.2086,
+      "step": 14941
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.513815456683671,
+      "learning_rate": 1.032130107960393e-06,
+      "loss": 0.3363,
+      "step": 14942
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6800722467991159,
+      "learning_rate": 1.0313068791342683e-06,
+      "loss": 0.3421,
+      "step": 14943
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2965021973445377,
+      "learning_rate": 1.0304839608936002e-06,
+      "loss": 0.197,
+      "step": 14944
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28748538312306504,
+      "learning_rate": 1.0296613532668875e-06,
+      "loss": 0.2301,
+      "step": 14945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1561225785911142,
+      "learning_rate": 1.0288390562826178e-06,
+      "loss": 0.697,
+      "step": 14946
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22153187673332958,
+      "learning_rate": 1.0280170699692648e-06,
+      "loss": 0.1684,
+      "step": 14947
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4287418102379653,
+      "learning_rate": 1.0271953943552938e-06,
+      "loss": 0.3314,
+      "step": 14948
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5123710533271378,
+      "learning_rate": 1.0263740294691615e-06,
+      "loss": 0.3298,
+      "step": 14949
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31713937270697534,
+      "learning_rate": 1.0255529753393112e-06,
+      "loss": 0.2379,
+      "step": 14950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6366738751403234,
+      "learning_rate": 1.0247322319941745e-06,
+      "loss": 0.2046,
+      "step": 14951
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.45705670046593916,
+      "learning_rate": 1.023911799462174e-06,
+      "loss": 0.3897,
+      "step": 14952
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27150260711708984,
+      "learning_rate": 1.0230916777717226e-06,
+      "loss": 0.2514,
+      "step": 14953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5578410636590693,
+      "learning_rate": 1.0222718669512211e-06,
+      "loss": 0.2062,
+      "step": 14954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5180872333426131,
+      "learning_rate": 1.0214523670290587e-06,
+      "loss": 0.2397,
+      "step": 14955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33763059935600714,
+      "learning_rate": 1.0206331780336154e-06,
+      "loss": 0.255,
+      "step": 14956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34511369249002094,
+      "learning_rate": 1.0198142999932559e-06,
+      "loss": 0.2375,
+      "step": 14957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8756372942785727,
+      "learning_rate": 1.0189957329363465e-06,
+      "loss": 0.5402,
+      "step": 14958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40461371137970736,
+      "learning_rate": 1.0181774768912255e-06,
+      "loss": 0.2893,
+      "step": 14959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38107539894620857,
+      "learning_rate": 1.0173595318862305e-06,
+      "loss": 0.2672,
+      "step": 14960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22422982508923442,
+      "learning_rate": 1.01654189794969e-06,
+      "loss": 0.1835,
+      "step": 14961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33600638655085446,
+      "learning_rate": 1.0157245751099188e-06,
+      "loss": 0.2601,
+      "step": 14962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4870843447136277,
+      "learning_rate": 1.0149075633952178e-06,
+      "loss": 0.5065,
+      "step": 14963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4281829927047752,
+      "learning_rate": 1.0140908628338796e-06,
+      "loss": 0.2577,
+      "step": 14964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2609028341284687,
+      "learning_rate": 1.013274473454191e-06,
+      "loss": 0.2483,
+      "step": 14965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6522575513963987,
+      "learning_rate": 1.0124583952844214e-06,
+      "loss": 0.3881,
+      "step": 14966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2859295010336732,
+      "learning_rate": 1.0116426283528301e-06,
+      "loss": 0.0838,
+      "step": 14967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32065626081543513,
+      "learning_rate": 1.0108271726876684e-06,
+      "loss": 0.239,
+      "step": 14968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34553309967875206,
+      "learning_rate": 1.0100120283171733e-06,
+      "loss": 0.2821,
+      "step": 14969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7021634157856494,
+      "learning_rate": 1.0091971952695768e-06,
+      "loss": 0.3325,
+      "step": 14970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3073095774730255,
+      "learning_rate": 1.008382673573095e-06,
+      "loss": 0.2581,
+      "step": 14971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8590640171628728,
+      "learning_rate": 1.0075684632559346e-06,
+      "loss": 0.4479,
+      "step": 14972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22537341360656152,
+      "learning_rate": 1.0067545643462895e-06,
+      "loss": 0.2082,
+      "step": 14973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3298064748892064,
+      "learning_rate": 1.0059409768723495e-06,
+      "loss": 0.1858,
+      "step": 14974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.147862075159621,
+      "learning_rate": 1.0051277008622861e-06,
+      "loss": 0.5509,
+      "step": 14975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31479894074709697,
+      "learning_rate": 1.004314736344264e-06,
+      "loss": 0.2775,
+      "step": 14976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3955090912581276,
+      "learning_rate": 1.0035020833464338e-06,
+      "loss": 0.1986,
+      "step": 14977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5281904876913323,
+      "learning_rate": 1.0026897418969417e-06,
+      "loss": 0.3507,
+      "step": 14978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30591248057503395,
+      "learning_rate": 1.0018777120239165e-06,
+      "loss": 0.1526,
+      "step": 14979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37828916984177724,
+      "learning_rate": 1.0010659937554789e-06,
+      "loss": 0.1737,
+      "step": 14980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31146064451236316,
+      "learning_rate": 1.00025458711974e-06,
+      "loss": 0.2923,
+      "step": 14981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7401631176839336,
+      "learning_rate": 9.99443492144795e-07,
+      "loss": 0.3723,
+      "step": 14982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29215619453143027,
+      "learning_rate": 9.986327088587378e-07,
+      "loss": 0.2067,
+      "step": 14983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34927543617015566,
+      "learning_rate": 9.978222372896417e-07,
+      "loss": 0.3224,
+      "step": 14984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.42741603016621244,
+      "learning_rate": 9.970120774655744e-07,
+      "loss": 0.2293,
+      "step": 14985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31526568971659064,
+      "learning_rate": 9.9620222941459e-07,
+      "loss": 0.2038,
+      "step": 14986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5250602102861858,
+      "learning_rate": 9.953926931647372e-07,
+      "loss": 0.2446,
+      "step": 14987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3332569485498697,
+      "learning_rate": 9.945834687440491e-07,
+      "loss": 0.297,
+      "step": 14988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40668710547493864,
+      "learning_rate": 9.937745561805478e-07,
+      "loss": 0.2977,
+      "step": 14989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5698087862797053,
+      "learning_rate": 9.92965955502244e-07,
+      "loss": 0.2558,
+      "step": 14990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.252459910213438,
+      "learning_rate": 9.921576667371458e-07,
+      "loss": 0.1633,
+      "step": 14991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3259181171702919,
+      "learning_rate": 9.91349689913238e-07,
+      "loss": 0.2698,
+      "step": 14992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33737889420877937,
+      "learning_rate": 9.90542025058503e-07,
+      "loss": 0.2462,
+      "step": 14993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6389429058168169,
+      "learning_rate": 9.897346722009095e-07,
+      "loss": 0.3715,
+      "step": 14994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6251259359040213,
+      "learning_rate": 9.889276313684171e-07,
+      "loss": 0.3601,
+      "step": 14995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2480426106504089,
+      "learning_rate": 9.88120902588975e-07,
+      "loss": 0.2357,
+      "step": 14996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3518028245212191,
+      "learning_rate": 9.87314485890517e-07,
+      "loss": 0.1729,
+      "step": 14997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6043858429194576,
+      "learning_rate": 9.8650838130097e-07,
+      "loss": 0.3186,
+      "step": 14998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37360874956552786,
+      "learning_rate": 9.857025888482518e-07,
+      "loss": 0.2994,
+      "step": 14999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33171762764892554,
+      "learning_rate": 9.848971085602655e-07,
+      "loss": 0.2454,
+      "step": 15000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.620980650429492,
+      "learning_rate": 9.84091940464904e-07,
+      "loss": 0.2839,
+      "step": 15001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4069755655302794,
+      "learning_rate": 9.832870845900488e-07,
+      "loss": 0.3088,
+      "step": 15002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.17326883779583802,
+      "learning_rate": 9.824825409635763e-07,
+      "loss": 0.0697,
+      "step": 15003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3107041090929514,
+      "learning_rate": 9.816783096133463e-07,
+      "loss": 0.2584,
+      "step": 15004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.354031848651155,
+      "learning_rate": 9.80874390567208e-07,
+      "loss": 0.2983,
+      "step": 15005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8439772860687007,
+      "learning_rate": 9.800707838530021e-07,
+      "loss": 0.312,
+      "step": 15006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41508853480396724,
+      "learning_rate": 9.792674894985553e-07,
+      "loss": 0.2951,
+      "step": 15007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5387336155751442,
+      "learning_rate": 9.78464507531689e-07,
+      "loss": 0.3999,
+      "step": 15008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30233725667514955,
+      "learning_rate": 9.776618379802093e-07,
+      "loss": 0.1861,
+      "step": 15009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4230746134209038,
+      "learning_rate": 9.768594808719113e-07,
+      "loss": 0.2633,
+      "step": 15010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48842838044318104,
+      "learning_rate": 9.76057436234581e-07,
+      "loss": 0.3434,
+      "step": 15011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.283475291759844,
+      "learning_rate": 9.752557040959943e-07,
+      "loss": 0.2384,
+      "step": 15012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32374263109755874,
+      "learning_rate": 9.744542844839145e-07,
+      "loss": 0.1523,
+      "step": 15013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3835245373644557,
+      "learning_rate": 9.736531774260948e-07,
+      "loss": 0.3108,
+      "step": 15014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8657684258513281,
+      "learning_rate": 9.728523829502768e-07,
+      "loss": 0.365,
+      "step": 15015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2924587753221655,
+      "learning_rate": 9.720519010841933e-07,
+      "loss": 0.2123,
+      "step": 15016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3928088182505235,
+      "learning_rate": 9.712517318555637e-07,
+      "loss": 0.3112,
+      "step": 15017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29042677444468945,
+      "learning_rate": 9.704518752921e-07,
+      "loss": 0.1606,
+      "step": 15018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3775481237884491,
+      "learning_rate": 9.696523314214978e-07,
+      "loss": 0.1905,
+      "step": 15019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2703635013510745,
+      "learning_rate": 9.688531002714464e-07,
+      "loss": 0.2672,
+      "step": 15020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2847216612253773,
+      "learning_rate": 9.680541818696254e-07,
+      "loss": 0.5317,
+      "step": 15021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5357320718859065,
+      "learning_rate": 9.672555762436997e-07,
+      "loss": 0.2516,
+      "step": 15022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29402400287893316,
+      "learning_rate": 9.66457283421325e-07,
+      "loss": 0.2488,
+      "step": 15023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2730938089664483,
+      "learning_rate": 9.656593034301432e-07,
+      "loss": 0.2212,
+      "step": 15024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5329117555031294,
+      "learning_rate": 9.648616362977959e-07,
+      "loss": 0.3298,
+      "step": 15025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3581964345211006,
+      "learning_rate": 9.640642820518997e-07,
+      "loss": 0.2328,
+      "step": 15026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6362827691871921,
+      "learning_rate": 9.63267240720067e-07,
+      "loss": 0.3489,
+      "step": 15027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29267465950559424,
+      "learning_rate": 9.62470512329904e-07,
+      "loss": 0.2516,
+      "step": 15028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36347141522829296,
+      "learning_rate": 9.616740969089967e-07,
+      "loss": 0.2125,
+      "step": 15029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2846777407910682,
+      "learning_rate": 9.608779944849278e-07,
+      "loss": 0.1992,
+      "step": 15030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2644418677329352,
+      "learning_rate": 9.600822050852654e-07,
+      "loss": 0.7545,
+      "step": 15031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21804919025258576,
+      "learning_rate": 9.592867287375652e-07,
+      "loss": 0.2162,
+      "step": 15032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.593650145047429,
+      "learning_rate": 9.584915654693782e-07,
+      "loss": 0.3514,
+      "step": 15033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9895576760248582,
+      "learning_rate": 9.576967153082406e-07,
+      "loss": 0.4155,
+      "step": 15034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23430158277619348,
+      "learning_rate": 9.569021782816767e-07,
+      "loss": 0.1612,
+      "step": 15035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27556174932560695,
+      "learning_rate": 9.561079544171992e-07,
+      "loss": 0.2646,
+      "step": 15036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.808225331928117,
+      "learning_rate": 9.553140437423157e-07,
+      "loss": 0.5553,
+      "step": 15037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37205200444797243,
+      "learning_rate": 9.545204462845192e-07,
+      "loss": 0.2542,
+      "step": 15038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6746390579793102,
+      "learning_rate": 9.537271620712896e-07,
+      "loss": 0.2945,
+      "step": 15039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3469141740881783,
+      "learning_rate": 9.529341911300982e-07,
+      "loss": 0.2678,
+      "step": 15040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3863948928230087,
+      "learning_rate": 9.521415334884088e-07,
+      "loss": 0.2668,
+      "step": 15041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34364358012632695,
+      "learning_rate": 9.513491891736681e-07,
+      "loss": 0.126,
+      "step": 15042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26545776152588485,
+      "learning_rate": 9.505571582133166e-07,
+      "loss": 0.2173,
+      "step": 15043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34794469679286133,
+      "learning_rate": 9.497654406347812e-07,
+      "loss": 0.2684,
+      "step": 15044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4445266623529254,
+      "learning_rate": 9.489740364654776e-07,
+      "loss": 0.2681,
+      "step": 15045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6025872062236165,
+      "learning_rate": 9.481829457328162e-07,
+      "loss": 0.3286,
+      "step": 15046
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37119355184782665,
+      "learning_rate": 9.473921684641896e-07,
+      "loss": 0.2974,
+      "step": 15047
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3547465998055965,
+      "learning_rate": 9.466017046869835e-07,
+      "loss": 0.3058,
+      "step": 15048
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8409961481560209,
+      "learning_rate": 9.458115544285684e-07,
+      "loss": 0.3469,
+      "step": 15049
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32572567903383914,
+      "learning_rate": 9.450217177163123e-07,
+      "loss": 0.2468,
+      "step": 15050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2476762696013451,
+      "learning_rate": 9.442321945775646e-07,
+      "loss": 0.206,
+      "step": 15051
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.638595287493113,
+      "learning_rate": 9.434429850396665e-07,
+      "loss": 0.2482,
+      "step": 15052
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33552988930315564,
+      "learning_rate": 9.426540891299463e-07,
+      "loss": 0.254,
+      "step": 15053
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.286349510665451,
+      "learning_rate": 9.418655068757276e-07,
+      "loss": 0.6941,
+      "step": 15054
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.39628036912695536,
+      "learning_rate": 9.410772383043176e-07,
+      "loss": 0.2364,
+      "step": 15055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2941849588804919,
+      "learning_rate": 9.402892834430122e-07,
+      "loss": 0.2882,
+      "step": 15056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42903190992151646,
+      "learning_rate": 9.395016423190984e-07,
+      "loss": 0.2431,
+      "step": 15057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4266723252802292,
+      "learning_rate": 9.387143149598543e-07,
+      "loss": 0.0998,
+      "step": 15058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3951244070626443,
+      "learning_rate": 9.379273013925449e-07,
+      "loss": 0.2723,
+      "step": 15059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3569081459857391,
+      "learning_rate": 9.371406016444229e-07,
+      "loss": 0.3066,
+      "step": 15060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7314097994167829,
+      "learning_rate": 9.363542157427297e-07,
+      "loss": 0.3921,
+      "step": 15061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2963625225107329,
+      "learning_rate": 9.355681437147024e-07,
+      "loss": 0.1872,
+      "step": 15062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3283347937303761,
+      "learning_rate": 9.347823855875604e-07,
+      "loss": 0.2963,
+      "step": 15063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.297389999754831,
+      "learning_rate": 9.339969413885142e-07,
+      "loss": 0.2066,
+      "step": 15064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35212332109937267,
+      "learning_rate": 9.33211811144763e-07,
+      "loss": 0.1968,
+      "step": 15065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8605096147432884,
+      "learning_rate": 9.324269948834985e-07,
+      "loss": 0.4698,
+      "step": 15066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30120778723725855,
+      "learning_rate": 9.316424926318967e-07,
+      "loss": 0.2745,
+      "step": 15067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3371975813552838,
+      "learning_rate": 9.30858304417126e-07,
+      "loss": 0.2051,
+      "step": 15068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2826029807524462,
+      "learning_rate": 9.300744302663401e-07,
+      "loss": 0.2156,
+      "step": 15069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4581018247895585,
+      "learning_rate": 9.292908702066883e-07,
+      "loss": 0.7467,
+      "step": 15070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3579852541537003,
+      "learning_rate": 9.285076242653035e-07,
+      "loss": 0.1911,
+      "step": 15071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34615695380006883,
+      "learning_rate": 9.277246924693106e-07,
+      "loss": 0.2995,
+      "step": 15072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6649173448975993,
+      "learning_rate": 9.269420748458202e-07,
+      "loss": 0.3692,
+      "step": 15073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3270716735112408,
+      "learning_rate": 9.261597714219351e-07,
+      "loss": 0.2589,
+      "step": 15074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5415825274888945,
+      "learning_rate": 9.253777822247479e-07,
+      "loss": 0.2645,
+      "step": 15075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28196842106823133,
+      "learning_rate": 9.24596107281338e-07,
+      "loss": 0.1962,
+      "step": 15076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.306635893181428,
+      "learning_rate": 9.238147466187742e-07,
+      "loss": 0.2415,
+      "step": 15077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7446253266487983,
+      "learning_rate": 9.230337002641144e-07,
+      "loss": 0.3045,
+      "step": 15078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3272308542677074,
+      "learning_rate": 9.222529682444081e-07,
+      "loss": 0.2904,
+      "step": 15079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3088128697895234,
+      "learning_rate": 9.214725505866929e-07,
+      "loss": 0.2339,
+      "step": 15080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9928672480486425,
+      "learning_rate": 9.206924473179913e-07,
+      "loss": 0.3694,
+      "step": 15081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3445131715535893,
+      "learning_rate": 9.199126584653184e-07,
+      "loss": 0.1636,
+      "step": 15082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3972284436044029,
+      "learning_rate": 9.191331840556816e-07,
+      "loss": 0.2828,
+      "step": 15083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27892382887335904,
+      "learning_rate": 9.183540241160715e-07,
+      "loss": 0.2442,
+      "step": 15084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0493720657729202,
+      "learning_rate": 9.175751786734722e-07,
+      "loss": 0.5299,
+      "step": 15085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.313880941212933,
+      "learning_rate": 9.167966477548529e-07,
+      "loss": 0.2344,
+      "step": 15086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5243978675286038,
+      "learning_rate": 9.160184313871745e-07,
+      "loss": 0.3414,
+      "step": 15087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23520310415924797,
+      "learning_rate": 9.152405295973877e-07,
+      "loss": 0.1365,
+      "step": 15088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.315451777087604,
+      "learning_rate": 9.144629424124318e-07,
+      "loss": 0.2393,
+      "step": 15089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6343131932703363,
+      "learning_rate": 9.136856698592323e-07,
+      "loss": 0.3756,
+      "step": 15090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33787984330910664,
+      "learning_rate": 9.129087119647062e-07,
+      "loss": 0.2509,
+      "step": 15091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31430488166281095,
+      "learning_rate": 9.121320687557622e-07,
+      "loss": 0.247,
+      "step": 15092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2083642377805037,
+      "learning_rate": 9.113557402592965e-07,
+      "loss": 0.5698,
+      "step": 15093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.272352156350627,
+      "learning_rate": 9.105797265021865e-07,
+      "loss": 0.123,
+      "step": 15094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26727226411819544,
+      "learning_rate": 9.098040275113118e-07,
+      "loss": 0.2503,
+      "step": 15095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4165842297184612,
+      "learning_rate": 9.09028643313532e-07,
+      "loss": 0.2908,
+      "step": 15096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8381958437662107,
+      "learning_rate": 9.082535739357001e-07,
+      "loss": 0.3653,
+      "step": 15097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3056167343849403,
+      "learning_rate": 9.074788194046557e-07,
+      "loss": 0.2505,
+      "step": 15098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5242561797322017,
+      "learning_rate": 9.067043797472264e-07,
+      "loss": 0.321,
+      "step": 15099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2684495155203418,
+      "learning_rate": 9.05930254990236e-07,
+      "loss": 0.2153,
+      "step": 15100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4350688261100347,
+      "learning_rate": 9.0515644516049e-07,
+      "loss": 0.188,
+      "step": 15101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5120139042621656,
+      "learning_rate": 9.043829502847845e-07,
+      "loss": 0.343,
+      "step": 15102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35141044392890897,
+      "learning_rate": 9.036097703899049e-07,
+      "loss": 0.329,
+      "step": 15103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3241126312677885,
+      "learning_rate": 9.028369055026287e-07,
+      "loss": 0.1678,
+      "step": 15104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6065691927483036,
+      "learning_rate": 9.020643556497211e-07,
+      "loss": 0.3254,
+      "step": 15105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7583614764645674,
+      "learning_rate": 9.012921208579317e-07,
+      "loss": 0.3871,
+      "step": 15106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25091863341591597,
+      "learning_rate": 9.005202011540037e-07,
+      "loss": 0.2057,
+      "step": 15107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.290779990812648,
+      "learning_rate": 8.997485965646724e-07,
+      "loss": 0.2231,
+      "step": 15108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2478961440750131,
+      "learning_rate": 8.989773071166552e-07,
+      "loss": 0.8038,
+      "step": 15109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2953342710533784,
+      "learning_rate": 8.982063328366631e-07,
+      "loss": 0.1998,
+      "step": 15110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5276949035579362,
+      "learning_rate": 8.974356737513934e-07,
+      "loss": 0.3769,
+      "step": 15111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4520637612174389,
+      "learning_rate": 8.966653298875339e-07,
+      "loss": 0.3065,
+      "step": 15112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3955320368936747,
+      "learning_rate": 8.958953012717641e-07,
+      "loss": 0.308,
+      "step": 15113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25203374514690974,
+      "learning_rate": 8.951255879307486e-07,
+      "loss": 0.1288,
+      "step": 15114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3478143118831568,
+      "learning_rate": 8.943561898911424e-07,
+      "loss": 0.3261,
+      "step": 15115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7966981365506505,
+      "learning_rate": 8.935871071795876e-07,
+      "loss": 0.3132,
+      "step": 15116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34943890063486505,
+      "learning_rate": 8.928183398227219e-07,
+      "loss": 0.2148,
+      "step": 15117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42402665162699893,
+      "learning_rate": 8.920498878471651e-07,
+      "loss": 0.2931,
+      "step": 15118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.374210011293103,
+      "learning_rate": 8.912817512795302e-07,
+      "loss": 0.2996,
+      "step": 15119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2495954819765893,
+      "learning_rate": 8.905139301464139e-07,
+      "loss": 0.1498,
+      "step": 15120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1354208775040406,
+      "learning_rate": 8.897464244744103e-07,
+      "loss": 0.808,
+      "step": 15121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.691843031359836,
+      "learning_rate": 8.88979234290096e-07,
+      "loss": 0.347,
+      "step": 15122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25810167310409116,
+      "learning_rate": 8.882123596200387e-07,
+      "loss": 0.2779,
+      "step": 15123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.664422180294338,
+      "learning_rate": 8.874458004907971e-07,
+      "loss": 0.2791,
+      "step": 15124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5426279817665275,
+      "learning_rate": 8.866795569289122e-07,
+      "loss": 0.266,
+      "step": 15125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31902793423548365,
+      "learning_rate": 8.859136289609272e-07,
+      "loss": 0.2559,
+      "step": 15126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3112222568220345,
+      "learning_rate": 8.851480166133586e-07,
+      "loss": 0.2472,
+      "step": 15127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2615724470027559,
+      "learning_rate": 8.843827199127208e-07,
+      "loss": 0.2101,
+      "step": 15128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.541148258436055,
+      "learning_rate": 8.836177388855183e-07,
+      "loss": 0.3172,
+      "step": 15129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4333991198695846,
+      "learning_rate": 8.82853073558243e-07,
+      "loss": 0.2728,
+      "step": 15130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3145467505444975,
+      "learning_rate": 8.820887239573728e-07,
+      "loss": 0.2714,
+      "step": 15131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5162011923967506,
+      "learning_rate": 8.813246901093763e-07,
+      "loss": 0.3445,
+      "step": 15132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1839520405237534,
+      "learning_rate": 8.80560972040716e-07,
+      "loss": 0.123,
+      "step": 15133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5364721309908936,
+      "learning_rate": 8.797975697778361e-07,
+      "loss": 0.3405,
+      "step": 15134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2625779164027942,
+      "learning_rate": 8.790344833471753e-07,
+      "loss": 0.2696,
+      "step": 15135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7388355872405135,
+      "learning_rate": 8.782717127751572e-07,
+      "loss": 0.3782,
+      "step": 15136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7507920836315224,
+      "learning_rate": 8.775092580881961e-07,
+      "loss": 0.1164,
+      "step": 15137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3979987615357961,
+      "learning_rate": 8.767471193126987e-07,
+      "loss": 0.2978,
+      "step": 15138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3314991768932052,
+      "learning_rate": 8.75985296475057e-07,
+      "loss": 0.2987,
+      "step": 15139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3167246892166629,
+      "learning_rate": 8.752237896016513e-07,
+      "loss": 0.1281,
+      "step": 15140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37153197907359864,
+      "learning_rate": 8.744625987188516e-07,
+      "loss": 0.3026,
+      "step": 15141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.50583112778751,
+      "learning_rate": 8.737017238530221e-07,
+      "loss": 0.2881,
+      "step": 15142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30099834150968574,
+      "learning_rate": 8.729411650305086e-07,
+      "loss": 0.2031,
+      "step": 15143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.40016934367057977,
+      "learning_rate": 8.721809222776512e-07,
+      "loss": 0.3229,
+      "step": 15144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1087364033476483,
+      "learning_rate": 8.71420995620773e-07,
+      "loss": 0.7046,
+      "step": 15145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24438582944496376,
+      "learning_rate": 8.706613850861955e-07,
+      "loss": 0.1804,
+      "step": 15146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2983420842661307,
+      "learning_rate": 8.69902090700222e-07,
+      "loss": 0.2841,
+      "step": 15147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.47140707285888844,
+      "learning_rate": 8.691431124891458e-07,
+      "loss": 0.259,
+      "step": 15148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5977151144145306,
+      "learning_rate": 8.683844504792516e-07,
+      "loss": 0.2461,
+      "step": 15149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3293070448411575,
+      "learning_rate": 8.676261046968082e-07,
+      "loss": 0.2188,
+      "step": 15150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35148732981494435,
+      "learning_rate": 8.668680751680836e-07,
+      "loss": 0.3271,
+      "step": 15151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5480960876299462,
+      "learning_rate": 8.661103619193235e-07,
+      "loss": 0.3355,
+      "step": 15152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3996302396384634,
+      "learning_rate": 8.653529649767689e-07,
+      "loss": 0.2401,
+      "step": 15153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28270246411719746,
+      "learning_rate": 8.64595884366648e-07,
+      "loss": 0.194,
+      "step": 15154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.453050144014881,
+      "learning_rate": 8.638391201151786e-07,
+      "loss": 0.2769,
+      "step": 15155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29278379222233086,
+      "learning_rate": 8.630826722485686e-07,
+      "loss": 0.1997,
+      "step": 15156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6741223471885249,
+      "learning_rate": 8.623265407930126e-07,
+      "loss": 0.4103,
+      "step": 15157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.46914591529300426,
+      "learning_rate": 8.615707257746942e-07,
+      "loss": 0.3219,
+      "step": 15158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2827772048682046,
+      "learning_rate": 8.608152272197901e-07,
+      "loss": 0.2253,
+      "step": 15159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31540560824285985,
+      "learning_rate": 8.600600451544638e-07,
+      "loss": 0.1769,
+      "step": 15160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0518770248703995,
+      "learning_rate": 8.593051796048623e-07,
+      "loss": 0.3286,
+      "step": 15161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3572068024893413,
+      "learning_rate": 8.58550630597128e-07,
+      "loss": 0.2964,
+      "step": 15162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30446816276194233,
+      "learning_rate": 8.577963981573944e-07,
+      "loss": 0.238,
+      "step": 15163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5673975171436122,
+      "learning_rate": 8.570424823117785e-07,
+      "loss": 0.3419,
+      "step": 15164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38060387474042917,
+      "learning_rate": 8.56288883086388e-07,
+      "loss": 0.2754,
+      "step": 15165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19743050333583234,
+      "learning_rate": 8.55535600507319e-07,
+      "loss": 0.1249,
+      "step": 15166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4663590562791462,
+      "learning_rate": 8.547826346006594e-07,
+      "loss": 0.2865,
+      "step": 15167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3707588567212866,
+      "learning_rate": 8.540299853924849e-07,
+      "loss": 0.3017,
+      "step": 15168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7253472188843394,
+      "learning_rate": 8.532776529088582e-07,
+      "loss": 0.304,
+      "step": 15169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27858164729369395,
+      "learning_rate": 8.525256371758317e-07,
+      "loss": 0.2592,
+      "step": 15170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38438389855409005,
+      "learning_rate": 8.517739382194512e-07,
+      "loss": 0.2976,
+      "step": 15171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35220266077902224,
+      "learning_rate": 8.510225560657459e-07,
+      "loss": 0.0923,
+      "step": 15172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3201247021161922,
+      "learning_rate": 8.50271490740735e-07,
+      "loss": 0.5593,
+      "step": 15173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32291369852792134,
+      "learning_rate": 8.495207422704299e-07,
+      "loss": 0.2695,
+      "step": 15174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34051148799422554,
+      "learning_rate": 8.487703106808254e-07,
+      "loss": 0.2965,
+      "step": 15175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5064423885781546,
+      "learning_rate": 8.480201959979139e-07,
+      "loss": 0.3937,
+      "step": 15176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31970419618362195,
+      "learning_rate": 8.472703982476694e-07,
+      "loss": 0.2563,
+      "step": 15177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23113554443336126,
+      "learning_rate": 8.465209174560574e-07,
+      "loss": 0.1821,
+      "step": 15178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4125928845847924,
+      "learning_rate": 8.457717536490307e-07,
+      "loss": 0.2227,
+      "step": 15179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33512692119389803,
+      "learning_rate": 8.450229068525351e-07,
+      "loss": 0.2999,
+      "step": 15180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.612047740136571,
+      "learning_rate": 8.442743770925044e-07,
+      "loss": 0.3659,
+      "step": 15181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32224924038452785,
+      "learning_rate": 8.435261643948567e-07,
+      "loss": 0.2515,
+      "step": 15182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4883487465666448,
+      "learning_rate": 8.427782687855035e-07,
+      "loss": 0.264,
+      "step": 15183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5277231776949652,
+      "learning_rate": 8.420306902903464e-07,
+      "loss": 0.217,
+      "step": 15184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3979449991906311,
+      "learning_rate": 8.412834289352734e-07,
+      "loss": 0.171,
+      "step": 15185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.36033874060148274,
+      "learning_rate": 8.405364847461606e-07,
+      "loss": 0.2843,
+      "step": 15186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3147674767160009,
+      "learning_rate": 8.397898577488739e-07,
+      "loss": 0.3099,
+      "step": 15187
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3165994656358568,
+      "learning_rate": 8.390435479692726e-07,
+      "loss": 0.7248,
+      "step": 15188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2795136011073454,
+      "learning_rate": 8.382975554331985e-07,
+      "loss": 0.1823,
+      "step": 15189
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33718281834093666,
+      "learning_rate": 8.375518801664873e-07,
+      "loss": 0.2743,
+      "step": 15190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31695562935276533,
+      "learning_rate": 8.368065221949595e-07,
+      "loss": 0.2208,
+      "step": 15191
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3135609234920389,
+      "learning_rate": 8.360614815444268e-07,
+      "loss": 0.2054,
+      "step": 15192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9047987658050868,
+      "learning_rate": 8.353167582406918e-07,
+      "loss": 0.4717,
+      "step": 15193
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32879063579291573,
+      "learning_rate": 8.345723523095462e-07,
+      "loss": 0.2928,
+      "step": 15194
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33331989963810826,
+      "learning_rate": 8.338282637767614e-07,
+      "loss": 0.1939,
+      "step": 15195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4331013516301031,
+      "learning_rate": 8.330844926681114e-07,
+      "loss": 0.2289,
+      "step": 15196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4826586936930948,
+      "learning_rate": 8.323410390093523e-07,
+      "loss": 0.2577,
+      "step": 15197
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20794647329271113,
+      "learning_rate": 8.315979028262277e-07,
+      "loss": 0.2144,
+      "step": 15198
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9167179052248758,
+      "learning_rate": 8.308550841444718e-07,
+      "loss": 0.493,
+      "step": 15199
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9684662978300949,
+      "learning_rate": 8.301125829898126e-07,
+      "loss": 0.6385,
+      "step": 15200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31667062423994347,
+      "learning_rate": 8.293703993879599e-07,
+      "loss": 0.262,
+      "step": 15201
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3174213408244066,
+      "learning_rate": 8.28628533364616e-07,
+      "loss": 0.2297,
+      "step": 15202
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6367147811096763,
+      "learning_rate": 8.278869849454718e-07,
+      "loss": 0.3327,
+      "step": 15203
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41740650979949356,
+      "learning_rate": 8.271457541562045e-07,
+      "loss": 0.2915,
+      "step": 15204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19492825352143867,
+      "learning_rate": 8.264048410224879e-07,
+      "loss": 0.15,
+      "step": 15205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34968830067182716,
+      "learning_rate": 8.25664245569976e-07,
+      "loss": 0.2954,
+      "step": 15206
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38126528864990045,
+      "learning_rate": 8.249239678243171e-07,
+      "loss": 0.2521,
+      "step": 15207
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4991098352243059,
+      "learning_rate": 8.241840078111452e-07,
+      "loss": 0.2371,
+      "step": 15208
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7523629811440098,
+      "learning_rate": 8.234443655560886e-07,
+      "loss": 0.3825,
+      "step": 15209
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27570811560535324,
+      "learning_rate": 8.227050410847592e-07,
+      "loss": 0.268,
+      "step": 15210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23902070137790352,
+      "learning_rate": 8.219660344227587e-07,
+      "loss": 0.2191,
+      "step": 15211
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.005631527918455,
+      "learning_rate": 8.212273455956787e-07,
+      "loss": 0.3593,
+      "step": 15212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31857938219013965,
+      "learning_rate": 8.204889746291022e-07,
+      "loss": 0.2688,
+      "step": 15213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38049878062907533,
+      "learning_rate": 8.197509215485988e-07,
+      "loss": 0.2892,
+      "step": 15214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7259267300163725,
+      "learning_rate": 8.190131863797246e-07,
+      "loss": 0.2979,
+      "step": 15215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3431148587829527,
+      "learning_rate": 8.182757691480303e-07,
+      "loss": 0.2404,
+      "step": 15216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4465896566952469,
+      "learning_rate": 8.175386698790489e-07,
+      "loss": 0.3118,
+      "step": 15217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27790041083033745,
+      "learning_rate": 8.168018885983109e-07,
+      "loss": 0.2146,
+      "step": 15218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3068585203990971,
+      "learning_rate": 8.160654253313282e-07,
+      "loss": 0.2547,
+      "step": 15219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7012299019399189,
+      "learning_rate": 8.15329280103605e-07,
+      "loss": 0.3996,
+      "step": 15220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3526300473721663,
+      "learning_rate": 8.145934529406319e-07,
+      "loss": 0.2639,
+      "step": 15221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4329912494767619,
+      "learning_rate": 8.13857943867894e-07,
+      "loss": 0.275,
+      "step": 15222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5705532148120219,
+      "learning_rate": 8.13122752910861e-07,
+      "loss": 0.3568,
+      "step": 15223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29636431949967834,
+      "learning_rate": 8.123878800949914e-07,
+      "loss": 0.2048,
+      "step": 15224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35704937882642657,
+      "learning_rate": 8.116533254457337e-07,
+      "loss": 0.186,
+      "step": 15225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3023244294858326,
+      "learning_rate": 8.109190889885277e-07,
+      "loss": 0.2889,
+      "step": 15226
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1384083091604984,
+      "learning_rate": 8.101851707487973e-07,
+      "loss": 0.4531,
+      "step": 15227
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.36463495630098414,
+      "learning_rate": 8.094515707519623e-07,
+      "loss": 0.1598,
+      "step": 15228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.544002008754593,
+      "learning_rate": 8.087182890234202e-07,
+      "loss": 0.3947,
+      "step": 15229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25996577539505555,
+      "learning_rate": 8.079853255885705e-07,
+      "loss": 0.2442,
+      "step": 15230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22928359992702024,
+      "learning_rate": 8.072526804727943e-07,
+      "loss": 0.1601,
+      "step": 15231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.548664404540612,
+      "learning_rate": 8.06520353701461e-07,
+      "loss": 0.369,
+      "step": 15232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4342997506272013,
+      "learning_rate": 8.057883452999316e-07,
+      "loss": 0.5444,
+      "step": 15233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22084093260537044,
+      "learning_rate": 8.050566552935579e-07,
+      "loss": 0.2115,
+      "step": 15234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2855665998775354,
+      "learning_rate": 8.043252837076776e-07,
+      "loss": 0.6016,
+      "step": 15235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43839777602461616,
+      "learning_rate": 8.03594230567617e-07,
+      "loss": 0.3084,
+      "step": 15236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3919309008857394,
+      "learning_rate": 8.028634958986903e-07,
+      "loss": 0.271,
+      "step": 15237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2905695696643427,
+      "learning_rate": 8.021330797262072e-07,
+      "loss": 0.2399,
+      "step": 15238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5220398686554123,
+      "learning_rate": 8.01402982075461e-07,
+      "loss": 0.2565,
+      "step": 15239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5903199356384773,
+      "learning_rate": 8.006732029717335e-07,
+      "loss": 0.2823,
+      "step": 15240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3598694129624296,
+      "learning_rate": 7.999437424402967e-07,
+      "loss": 0.2532,
+      "step": 15241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33400510821115537,
+      "learning_rate": 7.992146005064105e-07,
+      "loss": 0.3282,
+      "step": 15242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5293015346627737,
+      "learning_rate": 7.984857771953303e-07,
+      "loss": 0.2665,
+      "step": 15243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21466042777465733,
+      "learning_rate": 7.977572725322913e-07,
+      "loss": 0.1591,
+      "step": 15244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4726424277010858,
+      "learning_rate": 7.970290865425212e-07,
+      "loss": 0.3258,
+      "step": 15245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4202027933289129,
+      "learning_rate": 7.963012192512376e-07,
+      "loss": 0.2639,
+      "step": 15246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3945913782257165,
+      "learning_rate": 7.955736706836481e-07,
+      "loss": 0.2548,
+      "step": 15247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6633476242535613,
+      "learning_rate": 7.94846440864947e-07,
+      "loss": 0.3651,
+      "step": 15248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3693773855404697,
+      "learning_rate": 7.941195298203175e-07,
+      "loss": 0.2616,
+      "step": 15249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.260791519573847,
+      "learning_rate": 7.933929375749317e-07,
+      "loss": 0.2281,
+      "step": 15250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5155000696175669,
+      "learning_rate": 7.926666641539538e-07,
+      "loss": 0.1134,
+      "step": 15251
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3980620444030963,
+      "learning_rate": 7.919407095825337e-07,
+      "loss": 0.2641,
+      "step": 15252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5403914162100106,
+      "learning_rate": 7.912150738858104e-07,
+      "loss": 0.3727,
+      "step": 15253
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3470276410754074,
+      "learning_rate": 7.904897570889136e-07,
+      "loss": 0.2602,
+      "step": 15254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3700558072591282,
+      "learning_rate": 7.897647592169578e-07,
+      "loss": 0.2559,
+      "step": 15255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.531602700604024,
+      "learning_rate": 7.890400802950548e-07,
+      "loss": 0.3307,
+      "step": 15256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25900870477851295,
+      "learning_rate": 7.883157203482982e-07,
+      "loss": 0.1727,
+      "step": 15257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5904195157887109,
+      "learning_rate": 7.875916794017713e-07,
+      "loss": 0.2851,
+      "step": 15258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38738009696102815,
+      "learning_rate": 7.868679574805472e-07,
+      "loss": 0.3192,
+      "step": 15259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6042547956260029,
+      "learning_rate": 7.861445546096902e-07,
+      "loss": 0.3113,
+      "step": 15260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29172509317629647,
+      "learning_rate": 7.854214708142538e-07,
+      "loss": 0.2586,
+      "step": 15261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3643435241482438,
+      "learning_rate": 7.846987061192723e-07,
+      "loss": 0.2565,
+      "step": 15262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3800601781197826,
+      "learning_rate": 7.839762605497791e-07,
+      "loss": 0.1566,
+      "step": 15263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30248056133318446,
+      "learning_rate": 7.83254134130793e-07,
+      "loss": 0.1106,
+      "step": 15264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3365648909541551,
+      "learning_rate": 7.825323268873187e-07,
+      "loss": 0.2817,
+      "step": 15265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4406402840444798,
+      "learning_rate": 7.818108388443546e-07,
+      "loss": 0.3342,
+      "step": 15266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7090457854227306,
+      "learning_rate": 7.810896700268822e-07,
+      "loss": 0.223,
+      "step": 15267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2873213588453266,
+      "learning_rate": 7.803688204598803e-07,
+      "loss": 0.2282,
+      "step": 15268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36227730773662065,
+      "learning_rate": 7.796482901683089e-07,
+      "loss": 0.2392,
+      "step": 15269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3593253101252005,
+      "learning_rate": 7.789280791771214e-07,
+      "loss": 0.1979,
+      "step": 15270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3621820947758829,
+      "learning_rate": 7.782081875112568e-07,
+      "loss": 0.2827,
+      "step": 15271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6322813462372843,
+      "learning_rate": 7.774886151956473e-07,
+      "loss": 0.3907,
+      "step": 15272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2985759612155513,
+      "learning_rate": 7.767693622552097e-07,
+      "loss": 0.2031,
+      "step": 15273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40022176584749375,
+      "learning_rate": 7.76050428714854e-07,
+      "loss": 0.2846,
+      "step": 15274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3104592255829074,
+      "learning_rate": 7.753318145994727e-07,
+      "loss": 0.1448,
+      "step": 15275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49120492596421095,
+      "learning_rate": 7.746135199339556e-07,
+      "loss": 0.3065,
+      "step": 15276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28632281845942664,
+      "learning_rate": 7.738955447431762e-07,
+      "loss": 0.2348,
+      "step": 15277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4665069876855719,
+      "learning_rate": 7.731778890519969e-07,
+      "loss": 0.3292,
+      "step": 15278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4626098353168415,
+      "learning_rate": 7.724605528852702e-07,
+      "loss": 0.5868,
+      "step": 15279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35047118850359527,
+      "learning_rate": 7.717435362678361e-07,
+      "loss": 0.1945,
+      "step": 15280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23788708299943856,
+      "learning_rate": 7.71026839224529e-07,
+      "loss": 0.2045,
+      "step": 15281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6010442606371211,
+      "learning_rate": 7.703104617801649e-07,
+      "loss": 0.3379,
+      "step": 15282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32129184135251976,
+      "learning_rate": 7.695944039595526e-07,
+      "loss": 0.2314,
+      "step": 15283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2358162206069345,
+      "learning_rate": 7.688786657874881e-07,
+      "loss": 0.757,
+      "step": 15284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.302450566459852,
+      "learning_rate": 7.681632472887601e-07,
+      "loss": 0.2894,
+      "step": 15285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36525775072758804,
+      "learning_rate": 7.674481484881413e-07,
+      "loss": 0.2433,
+      "step": 15286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.12990322688717965,
+      "learning_rate": 7.667333694103962e-07,
+      "loss": 0.07,
+      "step": 15287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.530598155745847,
+      "learning_rate": 7.660189100802762e-07,
+      "loss": 0.3354,
+      "step": 15288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26335616434420217,
+      "learning_rate": 7.653047705225258e-07,
+      "loss": 0.2555,
+      "step": 15289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4268927274168955,
+      "learning_rate": 7.645909507618732e-07,
+      "loss": 0.2675,
+      "step": 15290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7647099515322016,
+      "learning_rate": 7.638774508230395e-07,
+      "loss": 0.4027,
+      "step": 15291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3398062115060478,
+      "learning_rate": 7.631642707307319e-07,
+      "loss": 0.2611,
+      "step": 15292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2679660624561897,
+      "learning_rate": 7.624514105096492e-07,
+      "loss": 0.2,
+      "step": 15293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6026562844344492,
+      "learning_rate": 7.617388701844764e-07,
+      "loss": 0.3465,
+      "step": 15294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25505961167579344,
+      "learning_rate": 7.610266497798913e-07,
+      "loss": 0.2145,
+      "step": 15295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.200358480855169,
+      "learning_rate": 7.603147493205531e-07,
+      "loss": 0.3905,
+      "step": 15296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3565559048959195,
+      "learning_rate": 7.59603168831119e-07,
+      "loss": 0.2969,
+      "step": 15297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3149087956944596,
+      "learning_rate": 7.588919083362301e-07,
+      "loss": 0.2396,
+      "step": 15298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6796516619216567,
+      "learning_rate": 7.581809678605167e-07,
+      "loss": 0.3689,
+      "step": 15299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34306570273858433,
+      "learning_rate": 7.574703474285971e-07,
+      "loss": 0.2018,
+      "step": 15300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.253075691212108,
+      "learning_rate": 7.567600470650849e-07,
+      "loss": 0.2466,
+      "step": 15301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41639118533895647,
+      "learning_rate": 7.560500667945736e-07,
+      "loss": 0.2633,
+      "step": 15302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9729597607745787,
+      "learning_rate": 7.553404066416514e-07,
+      "loss": 0.2259,
+      "step": 15303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36267534769957355,
+      "learning_rate": 7.546310666308909e-07,
+      "loss": 0.2609,
+      "step": 15304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3762109035053571,
+      "learning_rate": 7.539220467868613e-07,
+      "loss": 0.2912,
+      "step": 15305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37831387911988446,
+      "learning_rate": 7.532133471341141e-07,
+      "loss": 0.1914,
+      "step": 15306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2679166493911752,
+      "learning_rate": 7.525049676971907e-07,
+      "loss": 0.2282,
+      "step": 15307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1468120655160323,
+      "learning_rate": 7.517969085006227e-07,
+      "loss": 0.7836,
+      "step": 15308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3255850537957562,
+      "learning_rate": 7.510891695689282e-07,
+      "loss": 0.2195,
+      "step": 15309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35726564990652665,
+      "learning_rate": 7.503817509266198e-07,
+      "loss": 0.2898,
+      "step": 15310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6503155915878299,
+      "learning_rate": 7.496746525981935e-07,
+      "loss": 0.3677,
+      "step": 15311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37840816000222754,
+      "learning_rate": 7.489678746081364e-07,
+      "loss": 0.2404,
+      "step": 15312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29412005921689116,
+      "learning_rate": 7.482614169809222e-07,
+      "loss": 0.1814,
+      "step": 15313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4733546443356241,
+      "learning_rate": 7.475552797410191e-07,
+      "loss": 0.396,
+      "step": 15314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44759966457147815,
+      "learning_rate": 7.468494629128786e-07,
+      "loss": 0.2897,
+      "step": 15315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32367457949655865,
+      "learning_rate": 7.461439665209435e-07,
+      "loss": 0.2157,
+      "step": 15316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4080924876089095,
+      "learning_rate": 7.454387905896432e-07,
+      "loss": 0.306,
+      "step": 15317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2171090379852552,
+      "learning_rate": 7.447339351434013e-07,
+      "loss": 0.4721,
+      "step": 15318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2858970690734332,
+      "learning_rate": 7.44029400206625e-07,
+      "loss": 0.1938,
+      "step": 15319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2425084966006583,
+      "learning_rate": 7.433251858037127e-07,
+      "loss": 0.7118,
+      "step": 15320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2196747731628239,
+      "learning_rate": 7.426212919590503e-07,
+      "loss": 0.2056,
+      "step": 15321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29882954862740696,
+      "learning_rate": 7.419177186970139e-07,
+      "loss": 0.2113,
+      "step": 15322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8049704432913872,
+      "learning_rate": 7.412144660419706e-07,
+      "loss": 0.4391,
+      "step": 15323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5968983898094012,
+      "learning_rate": 7.405115340182723e-07,
+      "loss": 0.3162,
+      "step": 15324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28930754291418537,
+      "learning_rate": 7.398089226502603e-07,
+      "loss": 0.2629,
+      "step": 15325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1015836359844582,
+      "learning_rate": 7.391066319622664e-07,
+      "loss": 0.5367,
+      "step": 15326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3365458212983297,
+      "learning_rate": 7.384046619786123e-07,
+      "loss": 0.2162,
+      "step": 15327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44176905324324517,
+      "learning_rate": 7.377030127236073e-07,
+      "loss": 0.2576,
+      "step": 15328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2811618795819501,
+      "learning_rate": 7.370016842215488e-07,
+      "loss": 0.2339,
+      "step": 15329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.281599200325273,
+      "learning_rate": 7.363006764967228e-07,
+      "loss": 0.6169,
+      "step": 15330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3883951981613511,
+      "learning_rate": 7.355999895734067e-07,
+      "loss": 0.2458,
+      "step": 15331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35906664183412457,
+      "learning_rate": 7.348996234758643e-07,
+      "loss": 0.2571,
+      "step": 15332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4223644010226172,
+      "learning_rate": 7.341995782283506e-07,
+      "loss": 0.2908,
+      "step": 15333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39088380406576523,
+      "learning_rate": 7.334998538551042e-07,
+      "loss": 0.273,
+      "step": 15334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3214635287901621,
+      "learning_rate": 7.328004503803609e-07,
+      "loss": 0.1508,
+      "step": 15335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5600623509613721,
+      "learning_rate": 7.321013678283407e-07,
+      "loss": 0.3081,
+      "step": 15336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.272040944032163,
+      "learning_rate": 7.314026062232504e-07,
+      "loss": 0.244,
+      "step": 15337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.168178953668496,
+      "learning_rate": 7.307041655892877e-07,
+      "loss": 0.7761,
+      "step": 15338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5828528196274684,
+      "learning_rate": 7.300060459506431e-07,
+      "loss": 0.3033,
+      "step": 15339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25286178191941955,
+      "learning_rate": 7.293082473314905e-07,
+      "loss": 0.2107,
+      "step": 15340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2580447143807208,
+      "learning_rate": 7.286107697559952e-07,
+      "loss": 0.2347,
+      "step": 15341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3155362942915954,
+      "learning_rate": 7.279136132483078e-07,
+      "loss": 0.2337,
+      "step": 15342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3069915058726905,
+      "learning_rate": 7.272167778325756e-07,
+      "loss": 0.2612,
+      "step": 15343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.872607402488223,
+      "learning_rate": 7.265202635329272e-07,
+      "loss": 0.5209,
+      "step": 15344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33572217596314796,
+      "learning_rate": 7.258240703734832e-07,
+      "loss": 0.2534,
+      "step": 15345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36957113937852054,
+      "learning_rate": 7.251281983783532e-07,
+      "loss": 0.2541,
+      "step": 15346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25658032069472986,
+      "learning_rate": 7.244326475716323e-07,
+      "loss": 0.1934,
+      "step": 15347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36518152555370853,
+      "learning_rate": 7.237374179774125e-07,
+      "loss": 0.2461,
+      "step": 15348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3860018083993852,
+      "learning_rate": 7.230425096197669e-07,
+      "loss": 0.2512,
+      "step": 15349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4407144779168672,
+      "learning_rate": 7.223479225227603e-07,
+      "loss": 0.3225,
+      "step": 15350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8099477334987211,
+      "learning_rate": 7.216536567104449e-07,
+      "loss": 0.4107,
+      "step": 15351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3715733078958953,
+      "learning_rate": 7.209597122068657e-07,
+      "loss": 0.1808,
+      "step": 15352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2045474274396735,
+      "learning_rate": 7.202660890360524e-07,
+      "loss": 0.2173,
+      "step": 15353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.600767535273116,
+      "learning_rate": 7.195727872220248e-07,
+      "loss": 0.7834,
+      "step": 15354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2723621726814691,
+      "learning_rate": 7.188798067887926e-07,
+      "loss": 0.1995,
+      "step": 15355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4313450825328893,
+      "learning_rate": 7.181871477603542e-07,
+      "loss": 0.3322,
+      "step": 15356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43795447255344067,
+      "learning_rate": 7.174948101606949e-07,
+      "loss": 0.3239,
+      "step": 15357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3489367764018725,
+      "learning_rate": 7.168027940137923e-07,
+      "loss": 0.1882,
+      "step": 15358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2626421008762479,
+      "learning_rate": 7.161110993436093e-07,
+      "loss": 0.1515,
+      "step": 15359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3860502345972539,
+      "learning_rate": 7.15419726174098e-07,
+      "loss": 0.2858,
+      "step": 15360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36349650078670603,
+      "learning_rate": 7.147286745292049e-07,
+      "loss": 0.1971,
+      "step": 15361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44592750277136384,
+      "learning_rate": 7.140379444328571e-07,
+      "loss": 0.3327,
+      "step": 15362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0136525434025818,
+      "learning_rate": 7.13347535908977e-07,
+      "loss": 0.5487,
+      "step": 15363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.405578078064654,
+      "learning_rate": 7.126574489814719e-07,
+      "loss": 0.2968,
+      "step": 15364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2241058466957011,
+      "learning_rate": 7.119676836742407e-07,
+      "loss": 0.1817,
+      "step": 15365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5356592030749281,
+      "learning_rate": 7.112782400111684e-07,
+      "loss": 0.2669,
+      "step": 15366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6030656386460477,
+      "learning_rate": 7.105891180161306e-07,
+      "loss": 0.3748,
+      "step": 15367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24970782740649353,
+      "learning_rate": 7.099003177129926e-07,
+      "loss": 0.2436,
+      "step": 15368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3865134035245326,
+      "learning_rate": 7.092118391256076e-07,
+      "loss": 0.6055,
+      "step": 15369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7176124074822394,
+      "learning_rate": 7.085236822778174e-07,
+      "loss": 0.2688,
+      "step": 15370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2343011862368419,
+      "learning_rate": 7.078358471934521e-07,
+      "loss": 0.1487,
+      "step": 15371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.378432979445646,
+      "learning_rate": 7.071483338963303e-07,
+      "loss": 0.3006,
+      "step": 15372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48086836045064396,
+      "learning_rate": 7.064611424102641e-07,
+      "loss": 0.2961,
+      "step": 15373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37437277656526946,
+      "learning_rate": 7.057742727590478e-07,
+      "loss": 0.3221,
+      "step": 15374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9436427928262109,
+      "learning_rate": 7.050877249664701e-07,
+      "loss": 0.3674,
+      "step": 15375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2964656910026224,
+      "learning_rate": 7.04401499056302e-07,
+      "loss": 0.2569,
+      "step": 15376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3690822885989384,
+      "learning_rate": 7.037155950523123e-07,
+      "loss": 0.307,
+      "step": 15377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26884006202752997,
+      "learning_rate": 7.030300129782519e-07,
+      "loss": 0.131,
+      "step": 15378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40264916010642365,
+      "learning_rate": 7.023447528578631e-07,
+      "loss": 0.2621,
+      "step": 15379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35794945612338336,
+      "learning_rate": 7.016598147148735e-07,
+      "loss": 0.2974,
+      "step": 15380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5066463148507542,
+      "learning_rate": 7.009751985730062e-07,
+      "loss": 0.2751,
+      "step": 15381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8604249835369419,
+      "learning_rate": 7.00290904455968e-07,
+      "loss": 0.3744,
+      "step": 15382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29377213308048267,
+      "learning_rate": 6.996069323874555e-07,
+      "loss": 0.2061,
+      "step": 15383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30323452832902037,
+      "learning_rate": 6.989232823911551e-07,
+      "loss": 0.2308,
+      "step": 15384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7647519096570284,
+      "learning_rate": 6.982399544907403e-07,
+      "loss": 0.3747,
+      "step": 15385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36798900089852343,
+      "learning_rate": 6.975569487098766e-07,
+      "loss": 0.3002,
+      "step": 15386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3360407176215778,
+      "learning_rate": 6.968742650722172e-07,
+      "loss": 0.7616,
+      "step": 15387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2787925854981777,
+      "learning_rate": 6.961919036014009e-07,
+      "loss": 0.2048,
+      "step": 15388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4689798303850397,
+      "learning_rate": 6.955098643210578e-07,
+      "loss": 0.2941,
+      "step": 15389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.641487238466754,
+      "learning_rate": 6.94828147254809e-07,
+      "loss": 0.3726,
+      "step": 15390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2461102831506282,
+      "learning_rate": 6.941467524262613e-07,
+      "loss": 0.1119,
+      "step": 15391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25551098270018807,
+      "learning_rate": 6.934656798590122e-07,
+      "loss": 0.281,
+      "step": 15392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1202496115317333,
+      "learning_rate": 6.927849295766442e-07,
+      "loss": 0.5997,
+      "step": 15393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6445255900823864,
+      "learning_rate": 6.92104501602735e-07,
+      "loss": 0.0993,
+      "step": 15394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40685012455812186,
+      "learning_rate": 6.91424395960848e-07,
+      "loss": 0.295,
+      "step": 15395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3592962271783745,
+      "learning_rate": 6.907446126745332e-07,
+      "loss": 0.3047,
+      "step": 15396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2749244052812137,
+      "learning_rate": 6.900651517673318e-07,
+      "loss": 0.1568,
+      "step": 15397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6044148011731777,
+      "learning_rate": 6.893860132627739e-07,
+      "loss": 0.3646,
+      "step": 15398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49796483369447875,
+      "learning_rate": 6.887071971843783e-07,
+      "loss": 0.2891,
+      "step": 15399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33803796040103634,
+      "learning_rate": 6.880287035556521e-07,
+      "loss": 0.2728,
+      "step": 15400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.343782499945726,
+      "learning_rate": 6.873505324000895e-07,
+      "loss": 0.2097,
+      "step": 15401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6284160615889932,
+      "learning_rate": 6.866726837411797e-07,
+      "loss": 0.3594,
+      "step": 15402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3844903560072926,
+      "learning_rate": 6.859951576023937e-07,
+      "loss": 0.2492,
+      "step": 15403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22190970897036372,
+      "learning_rate": 6.853179540071963e-07,
+      "loss": 0.2188,
+      "step": 15404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.46152722721730727,
+      "learning_rate": 6.846410729790342e-07,
+      "loss": 0.2762,
+      "step": 15405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6960344758067857,
+      "learning_rate": 6.839645145413543e-07,
+      "loss": 0.3359,
+      "step": 15406
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3315547930125616,
+      "learning_rate": 6.832882787175809e-07,
+      "loss": 0.2416,
+      "step": 15407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34418183719622625,
+      "learning_rate": 6.826123655311356e-07,
+      "loss": 0.2972,
+      "step": 15408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7768735537156152,
+      "learning_rate": 6.819367750054217e-07,
+      "loss": 0.4253,
+      "step": 15409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3022317412638133,
+      "learning_rate": 6.812615071638363e-07,
+      "loss": 0.2187,
+      "step": 15410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30097106667699985,
+      "learning_rate": 6.805865620297659e-07,
+      "loss": 0.1357,
+      "step": 15411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3699431546810101,
+      "learning_rate": 6.799119396265807e-07,
+      "loss": 0.2815,
+      "step": 15412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.36422006459287853,
+      "learning_rate": 6.792376399776457e-07,
+      "loss": 0.2794,
+      "step": 15413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7376350234779959,
+      "learning_rate": 6.785636631063075e-07,
+      "loss": 0.2866,
+      "step": 15414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4651567840856827,
+      "learning_rate": 6.778900090359119e-07,
+      "loss": 0.3252,
+      "step": 15415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3479007390754298,
+      "learning_rate": 6.772166777897838e-07,
+      "loss": 0.2525,
+      "step": 15416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20645592583378722,
+      "learning_rate": 6.765436693912408e-07,
+      "loss": 0.1553,
+      "step": 15417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5925146855129958,
+      "learning_rate": 6.758709838635879e-07,
+      "loss": 0.3312,
+      "step": 15418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.36018604511633256,
+      "learning_rate": 6.751986212301242e-07,
+      "loss": 0.2976,
+      "step": 15419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34434040713944425,
+      "learning_rate": 6.745265815141311e-07,
+      "loss": 0.2374,
+      "step": 15420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2688158338463227,
+      "learning_rate": 6.73854864738881e-07,
+      "loss": 0.5905,
+      "step": 15421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3130992444743742,
+      "learning_rate": 6.731834709276353e-07,
+      "loss": 0.2281,
+      "step": 15422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.373687076761885,
+      "learning_rate": 6.725124001036454e-07,
+      "loss": 0.2071,
+      "step": 15423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49870489961270104,
+      "learning_rate": 6.718416522901506e-07,
+      "loss": 0.3269,
+      "step": 15424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23765638274613635,
+      "learning_rate": 6.711712275103776e-07,
+      "loss": 0.221,
+      "step": 15425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3561355023422061,
+      "learning_rate": 6.705011257875449e-07,
+      "loss": 0.5169,
+      "step": 15426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3737441156518405,
+      "learning_rate": 6.698313471448547e-07,
+      "loss": 0.2337,
+      "step": 15427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28412374842415977,
+      "learning_rate": 6.691618916055053e-07,
+      "loss": 0.2634,
+      "step": 15428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.260617821916191,
+      "learning_rate": 6.684927591926793e-07,
+      "loss": 0.6065,
+      "step": 15429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.416719995810165,
+      "learning_rate": 6.678239499295469e-07,
+      "loss": 0.1985,
+      "step": 15430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3731138532303315,
+      "learning_rate": 6.671554638392696e-07,
+      "loss": 0.2707,
+      "step": 15431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26428445121349753,
+      "learning_rate": 6.664873009449979e-07,
+      "loss": 0.2391,
+      "step": 15432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4735796217441745,
+      "learning_rate": 6.658194612698687e-07,
+      "loss": 0.2467,
+      "step": 15433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3392053344013183,
+      "learning_rate": 6.651519448370092e-07,
+      "loss": 0.2764,
+      "step": 15434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4614700653109635,
+      "learning_rate": 6.644847516695385e-07,
+      "loss": 0.3329,
+      "step": 15435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3974557145871179,
+      "learning_rate": 6.638178817905594e-07,
+      "loss": 0.2162,
+      "step": 15436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31729301878064864,
+      "learning_rate": 6.631513352231644e-07,
+      "loss": 0.2351,
+      "step": 15437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6014970938156791,
+      "learning_rate": 6.624851119904385e-07,
+      "loss": 0.2425,
+      "step": 15438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.46471253231576026,
+      "learning_rate": 6.618192121154488e-07,
+      "loss": 0.3678,
+      "step": 15439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3236888175086646,
+      "learning_rate": 6.611536356212612e-07,
+      "loss": 0.2154,
+      "step": 15440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5426087946165549,
+      "learning_rate": 6.604883825309205e-07,
+      "loss": 0.3937,
+      "step": 15441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5414638027935905,
+      "learning_rate": 6.598234528674663e-07,
+      "loss": 0.6158,
+      "step": 15442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18180585805377175,
+      "learning_rate": 6.591588466539222e-07,
+      "loss": 0.1348,
+      "step": 15443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33920005873970366,
+      "learning_rate": 6.584945639133067e-07,
+      "loss": 0.299,
+      "step": 15444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3235160572056892,
+      "learning_rate": 6.578306046686234e-07,
+      "loss": 0.6165,
+      "step": 15445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31586260327256654,
+      "learning_rate": 6.57166968942865e-07,
+      "loss": 0.2081,
+      "step": 15446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4470041807775742,
+      "learning_rate": 6.565036567590099e-07,
+      "loss": 0.3292,
+      "step": 15447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.50546493909433,
+      "learning_rate": 6.558406681400342e-07,
+      "loss": 0.3378,
+      "step": 15448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33419202257022956,
+      "learning_rate": 6.55178003108894e-07,
+      "loss": 0.2611,
+      "step": 15449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16134420138718436,
+      "learning_rate": 6.545156616885373e-07,
+      "loss": 0.071,
+      "step": 15450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32891260695083896,
+      "learning_rate": 6.538536439019016e-07,
+      "loss": 0.3021,
+      "step": 15451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37903738502174683,
+      "learning_rate": 6.531919497719097e-07,
+      "loss": 0.2468,
+      "step": 15452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4409361141174233,
+      "learning_rate": 6.52530579321482e-07,
+      "loss": 0.2674,
+      "step": 15453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5984776888889717,
+      "learning_rate": 6.51869532573517e-07,
+      "loss": 0.2518,
+      "step": 15454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37448200080698674,
+      "learning_rate": 6.512088095509095e-07,
+      "loss": 0.2645,
+      "step": 15455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24355986018680098,
+      "learning_rate": 6.505484102765358e-07,
+      "loss": 0.1973,
+      "step": 15456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8738256093138846,
+      "learning_rate": 6.498883347732709e-07,
+      "loss": 0.4653,
+      "step": 15457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39720267810411064,
+      "learning_rate": 6.492285830639711e-07,
+      "loss": 0.2898,
+      "step": 15458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.284470585352059,
+      "learning_rate": 6.485691551714835e-07,
+      "loss": 0.2527,
+      "step": 15459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3712482470378675,
+      "learning_rate": 6.47910051118642e-07,
+      "loss": 0.5142,
+      "step": 15460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3337385065605379,
+      "learning_rate": 6.472512709282752e-07,
+      "loss": 0.2579,
+      "step": 15461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2548757339390069,
+      "learning_rate": 6.465928146231937e-07,
+      "loss": 0.1738,
+      "step": 15462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33354702595338487,
+      "learning_rate": 6.459346822262014e-07,
+      "loss": 0.2561,
+      "step": 15463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31942036759526615,
+      "learning_rate": 6.45276873760089e-07,
+      "loss": 0.2559,
+      "step": 15464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8281221679177381,
+      "learning_rate": 6.44619389247636e-07,
+      "loss": 0.4434,
+      "step": 15465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4513030484506004,
+      "learning_rate": 6.439622287116121e-07,
+      "loss": 0.2286,
+      "step": 15466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2902777682810745,
+      "learning_rate": 6.433053921747734e-07,
+      "loss": 0.2554,
+      "step": 15467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22307685318256087,
+      "learning_rate": 6.42648879659864e-07,
+      "loss": 0.1937,
+      "step": 15468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6450982193385442,
+      "learning_rate": 6.419926911896246e-07,
+      "loss": 0.2968,
+      "step": 15469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40342553770057044,
+      "learning_rate": 6.413368267867748e-07,
+      "loss": 0.2651,
+      "step": 15470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3625069981846449,
+      "learning_rate": 6.406812864740286e-07,
+      "loss": 0.3225,
+      "step": 15471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5406013839654554,
+      "learning_rate": 6.400260702740857e-07,
+      "loss": 0.2295,
+      "step": 15472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3813327291140086,
+      "learning_rate": 6.39371178209639e-07,
+      "loss": 0.2555,
+      "step": 15473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.248468868214602,
+      "learning_rate": 6.387166103033659e-07,
+      "loss": 0.1838,
+      "step": 15474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3369238767144438,
+      "learning_rate": 6.38062366577934e-07,
+      "loss": 0.2943,
+      "step": 15475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49041949518461936,
+      "learning_rate": 6.374084470559993e-07,
+      "loss": 0.2287,
+      "step": 15476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5345573593990525,
+      "learning_rate": 6.367548517602062e-07,
+      "loss": 0.3918,
+      "step": 15477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.267321575127367,
+      "learning_rate": 6.36101580713191e-07,
+      "loss": 0.4503,
+      "step": 15478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24989474901885625,
+      "learning_rate": 6.354486339375765e-07,
+      "loss": 0.2039,
+      "step": 15479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30215440765844515,
+      "learning_rate": 6.347960114559726e-07,
+      "loss": 0.2355,
+      "step": 15480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47055963128076644,
+      "learning_rate": 6.341437132909778e-07,
+      "loss": 0.2552,
+      "step": 15481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28025144025055243,
+      "learning_rate": 6.334917394651863e-07,
+      "loss": 0.1913,
+      "step": 15482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47105549542319564,
+      "learning_rate": 6.328400900011722e-07,
+      "loss": 0.3661,
+      "step": 15483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5301274686641892,
+      "learning_rate": 6.321887649215031e-07,
+      "loss": 0.3488,
+      "step": 15484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34412718839375167,
+      "learning_rate": 6.31537764248733e-07,
+      "loss": 0.1598,
+      "step": 15485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49793609859674237,
+      "learning_rate": 6.308870880054085e-07,
+      "loss": 0.3045,
+      "step": 15486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2893282719469128,
+      "learning_rate": 6.302367362140616e-07,
+      "loss": 0.2616,
+      "step": 15487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34383649815869993,
+      "learning_rate": 6.295867088972141e-07,
+      "loss": 0.2231,
+      "step": 15488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3281453009232224,
+      "learning_rate": 6.289370060773748e-07,
+      "loss": 0.2295,
+      "step": 15489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.536206100347878,
+      "learning_rate": 6.282876277770433e-07,
+      "loss": 0.2901,
+      "step": 15490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3532249976082039,
+      "learning_rate": 6.276385740187097e-07,
+      "loss": 0.258,
+      "step": 15491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3603329578354436,
+      "learning_rate": 6.26989844824849e-07,
+      "loss": 0.2514,
+      "step": 15492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9432767223998106,
+      "learning_rate": 6.263414402179269e-07,
+      "loss": 0.402,
+      "step": 15493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30261911972964056,
+      "learning_rate": 6.256933602203963e-07,
+      "loss": 0.2115,
+      "step": 15494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23543770198075897,
+      "learning_rate": 6.250456048547027e-07,
+      "loss": 0.2335,
+      "step": 15495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4415768113353055,
+      "learning_rate": 6.243981741432769e-07,
+      "loss": 0.2614,
+      "step": 15496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5812569473819925,
+      "learning_rate": 6.23751068108539e-07,
+      "loss": 0.3034,
+      "step": 15497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.36478079392776935,
+      "learning_rate": 6.231042867728987e-07,
+      "loss": 0.2275,
+      "step": 15498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3750949965379845,
+      "learning_rate": 6.224578301587536e-07,
+      "loss": 0.29,
+      "step": 15499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.266345866146398,
+      "learning_rate": 6.218116982884903e-07,
+      "loss": 0.2056,
+      "step": 15500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5182817369574994,
+      "learning_rate": 6.211658911844854e-07,
+      "loss": 0.3603,
+      "step": 15501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32107800900252353,
+      "learning_rate": 6.205204088690997e-07,
+      "loss": 0.1834,
+      "step": 15502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3540271412785568,
+      "learning_rate": 6.198752513646911e-07,
+      "loss": 0.2723,
+      "step": 15503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5177309374569469,
+      "learning_rate": 6.192304186935993e-07,
+      "loss": 0.3293,
+      "step": 15504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8842981293585628,
+      "learning_rate": 6.185859108781544e-07,
+      "loss": 0.2621,
+      "step": 15505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8063127372023199,
+      "learning_rate": 6.179417279406752e-07,
+      "loss": 0.5123,
+      "step": 15506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2330689085737917,
+      "learning_rate": 6.172978699034715e-07,
+      "loss": 0.2458,
+      "step": 15507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2998705740226622,
+      "learning_rate": 6.166543367888389e-07,
+      "loss": 0.1521,
+      "step": 15508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5991385238658149,
+      "learning_rate": 6.160111286190629e-07,
+      "loss": 0.3678,
+      "step": 15509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.41892626382614,
+      "learning_rate": 6.153682454164167e-07,
+      "loss": 0.2752,
+      "step": 15510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.38775365797584355,
+      "learning_rate": 6.14725687203167e-07,
+      "loss": 0.2397,
+      "step": 15511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6237671895525037,
+      "learning_rate": 6.140834540015617e-07,
+      "loss": 0.3729,
+      "step": 15512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29434266789911306,
+      "learning_rate": 6.134415458338439e-07,
+      "loss": 0.2485,
+      "step": 15513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44565125934806743,
+      "learning_rate": 6.127999627222414e-07,
+      "loss": 0.238,
+      "step": 15514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28957696302459607,
+      "learning_rate": 6.121587046889709e-07,
+      "loss": 0.2293,
+      "step": 15515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.381380545026879,
+      "learning_rate": 6.115177717562426e-07,
+      "loss": 0.3005,
+      "step": 15516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.280503866463015,
+      "learning_rate": 6.108771639462496e-07,
+      "loss": 0.4066,
+      "step": 15517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31632331249443907,
+      "learning_rate": 6.102368812811776e-07,
+      "loss": 0.2172,
+      "step": 15518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30631108778593535,
+      "learning_rate": 6.095969237831956e-07,
+      "loss": 0.2671,
+      "step": 15519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4657106174751502,
+      "learning_rate": 6.089572914744712e-07,
+      "loss": 0.2943,
+      "step": 15520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3192673232648922,
+      "learning_rate": 6.083179843771513e-07,
+      "loss": 0.1749,
+      "step": 15521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5648607556412087,
+      "learning_rate": 6.076790025133761e-07,
+      "loss": 0.3423,
+      "step": 15522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.38998891302247396,
+      "learning_rate": 6.070403459052721e-07,
+      "loss": 0.3006,
+      "step": 15523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47612449309706734,
+      "learning_rate": 6.064020145749572e-07,
+      "loss": 0.2506,
+      "step": 15524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3699572894444798,
+      "learning_rate": 6.057640085445371e-07,
+      "loss": 0.2902,
+      "step": 15525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33072600556263054,
+      "learning_rate": 6.051263278361064e-07,
+      "loss": 0.2912,
+      "step": 15526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5229226557439318,
+      "learning_rate": 6.04488972471744e-07,
+      "loss": 0.2665,
+      "step": 15527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23899799542644962,
+      "learning_rate": 6.038519424735268e-07,
+      "loss": 0.1533,
+      "step": 15528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2521803906393414,
+      "learning_rate": 6.032152378635125e-07,
+      "loss": 0.71,
+      "step": 15529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.332009032357602,
+      "learning_rate": 6.025788586637516e-07,
+      "loss": 0.3173,
+      "step": 15530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3054581468802576,
+      "learning_rate": 6.019428048962794e-07,
+      "loss": 0.2127,
+      "step": 15531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6988792384416537,
+      "learning_rate": 6.013070765831242e-07,
+      "loss": 0.3557,
+      "step": 15532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2612692825758358,
+      "learning_rate": 6.006716737463003e-07,
+      "loss": 0.1574,
+      "step": 15533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.35768416541218084,
+      "learning_rate": 6.000365964078125e-07,
+      "loss": 0.2071,
+      "step": 15534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.360522914540225,
+      "learning_rate": 5.99401844589651e-07,
+      "loss": 0.2844,
+      "step": 15535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7207181920540063,
+      "learning_rate": 5.987674183138015e-07,
+      "loss": 0.4223,
+      "step": 15536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34250474149267907,
+      "learning_rate": 5.98133317602233e-07,
+      "loss": 0.2512,
+      "step": 15537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42762939818118384,
+      "learning_rate": 5.974995424769026e-07,
+      "loss": 0.2639,
+      "step": 15538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28191737906201586,
+      "learning_rate": 5.968660929597581e-07,
+      "loss": 0.2076,
+      "step": 15539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3401995555671143,
+      "learning_rate": 5.962329690727353e-07,
+      "loss": 0.2638,
+      "step": 15540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5155883682044767,
+      "learning_rate": 5.956001708377623e-07,
+      "loss": 0.1423,
+      "step": 15541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3712414578625411,
+      "learning_rate": 5.949676982767505e-07,
+      "loss": 0.3232,
+      "step": 15542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32353924162903186,
+      "learning_rate": 5.943355514116033e-07,
+      "loss": 0.2585,
+      "step": 15543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7305438124068102,
+      "learning_rate": 5.937037302642101e-07,
+      "loss": 0.3036,
+      "step": 15544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29228028451163485,
+      "learning_rate": 5.930722348564533e-07,
+      "loss": 0.1928,
+      "step": 15545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34765287698989866,
+      "learning_rate": 5.924410652102009e-07,
+      "loss": 0.255,
+      "step": 15546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3763887260898906,
+      "learning_rate": 5.918102213473087e-07,
+      "loss": 0.2487,
+      "step": 15547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6285264044725666,
+      "learning_rate": 5.911797032896239e-07,
+      "loss": 0.3656,
+      "step": 15548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33477221937072554,
+      "learning_rate": 5.905495110589821e-07,
+      "loss": 0.2613,
+      "step": 15549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.252759326879859,
+      "learning_rate": 5.89919644677206e-07,
+      "loss": 0.8002,
+      "step": 15550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3344029733336263,
+      "learning_rate": 5.892901041661092e-07,
+      "loss": 0.2161,
+      "step": 15551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20909466560736703,
+      "learning_rate": 5.886608895474888e-07,
+      "loss": 0.1845,
+      "step": 15552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1886244262668104,
+      "learning_rate": 5.880320008431384e-07,
+      "loss": 0.6513,
+      "step": 15553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29300410427650325,
+      "learning_rate": 5.874034380748362e-07,
+      "loss": 0.2342,
+      "step": 15554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34717227520428,
+      "learning_rate": 5.867752012643469e-07,
+      "loss": 0.285,
+      "step": 15555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7185220879859632,
+      "learning_rate": 5.861472904334287e-07,
+      "loss": 0.3823,
+      "step": 15556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48824641014498565,
+      "learning_rate": 5.855197056038231e-07,
+      "loss": 0.1113,
+      "step": 15557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34978211070504545,
+      "learning_rate": 5.848924467972661e-07,
+      "loss": 0.2534,
+      "step": 15558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.257049533946427,
+      "learning_rate": 5.842655140354791e-07,
+      "loss": 0.2372,
+      "step": 15559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.596660439530161,
+      "learning_rate": 5.836389073401727e-07,
+      "loss": 0.283,
+      "step": 15560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31315917280327593,
+      "learning_rate": 5.830126267330449e-07,
+      "loss": 0.2764,
+      "step": 15561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.35240639619956515,
+      "learning_rate": 5.823866722357863e-07,
+      "loss": 0.3343,
+      "step": 15562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.620476298451587,
+      "learning_rate": 5.817610438700716e-07,
+      "loss": 0.5078,
+      "step": 15563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2704294455548528,
+      "learning_rate": 5.811357416575681e-07,
+      "loss": 0.1573,
+      "step": 15564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4428142185190504,
+      "learning_rate": 5.805107656199272e-07,
+      "loss": 0.2494,
+      "step": 15565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3292874439000769,
+      "learning_rate": 5.79886115778795e-07,
+      "loss": 0.2965,
+      "step": 15566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30624266788213034,
+      "learning_rate": 5.792617921558008e-07,
+      "loss": 0.1981,
+      "step": 15567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2556421070753658,
+      "learning_rate": 5.786377947725652e-07,
+      "loss": 0.736,
+      "step": 15568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5193881640144948,
+      "learning_rate": 5.780141236506975e-07,
+      "loss": 0.567,
+      "step": 15569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22623774423910054,
+      "learning_rate": 5.77390778811796e-07,
+      "loss": 0.2157,
+      "step": 15570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28603043110314136,
+      "learning_rate": 5.767677602774469e-07,
+      "loss": 0.1825,
+      "step": 15571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5968504572934171,
+      "learning_rate": 5.761450680692249e-07,
+      "loss": 0.3951,
+      "step": 15572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2821106282963114,
+      "learning_rate": 5.755227022086918e-07,
+      "loss": 0.1834,
+      "step": 15573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3511480956550723,
+      "learning_rate": 5.749006627174048e-07,
+      "loss": 0.3039,
+      "step": 15574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2740202637398323,
+      "learning_rate": 5.742789496169021e-07,
+      "loss": 0.4536,
+      "step": 15575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.36967814771582713,
+      "learning_rate": 5.736575629287145e-07,
+      "loss": 0.2536,
+      "step": 15576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3182282890886471,
+      "learning_rate": 5.730365026743579e-07,
+      "loss": 0.2127,
+      "step": 15577
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2597618110444984,
+      "learning_rate": 5.72415768875344e-07,
+      "loss": 0.2431,
+      "step": 15578
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40873608355276525,
+      "learning_rate": 5.717953615531668e-07,
+      "loss": 0.2492,
+      "step": 15579
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5488195331582982,
+      "learning_rate": 5.711752807293102e-07,
+      "loss": 0.2588,
+      "step": 15580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.40896979648835,
+      "learning_rate": 5.705555264252483e-07,
+      "loss": 0.5422,
+      "step": 15581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2511592502937873,
+      "learning_rate": 5.699360986624414e-07,
+      "loss": 0.2355,
+      "step": 15582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5226165719071535,
+      "learning_rate": 5.693169974623435e-07,
+      "loss": 0.2665,
+      "step": 15583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4341487348411721,
+      "learning_rate": 5.686982228463933e-07,
+      "loss": 0.2977,
+      "step": 15584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2937040743182103,
+      "learning_rate": 5.680797748360168e-07,
+      "loss": 0.2368,
+      "step": 15585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24353668748816057,
+      "learning_rate": 5.674616534526312e-07,
+      "loss": 0.1934,
+      "step": 15586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.674275057185165,
+      "learning_rate": 5.66843858717645e-07,
+      "loss": 0.4936,
+      "step": 15587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33392941409136523,
+      "learning_rate": 5.66226390652449e-07,
+      "loss": 0.2589,
+      "step": 15588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5883240032566877,
+      "learning_rate": 5.656092492784282e-07,
+      "loss": 0.3758,
+      "step": 15589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3373756640161482,
+      "learning_rate": 5.649924346169522e-07,
+      "loss": 0.2541,
+      "step": 15590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5502096680976466,
+      "learning_rate": 5.643759466893839e-07,
+      "loss": 0.3176,
+      "step": 15591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23347398396256974,
+      "learning_rate": 5.637597855170707e-07,
+      "loss": 0.1993,
+      "step": 15592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3502673113280859,
+      "learning_rate": 5.631439511213499e-07,
+      "loss": 0.2337,
+      "step": 15593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3869961913722,
+      "learning_rate": 5.625284435235478e-07,
+      "loss": 0.2529,
+      "step": 15594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.676089532643328,
+      "learning_rate": 5.619132627449797e-07,
+      "loss": 0.3139,
+      "step": 15595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3271539215905461,
+      "learning_rate": 5.612984088069507e-07,
+      "loss": 0.3681,
+      "step": 15596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41527056555619096,
+      "learning_rate": 5.606838817307514e-07,
+      "loss": 0.2618,
+      "step": 15597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20296961040711048,
+      "learning_rate": 5.600696815376639e-07,
+      "loss": 0.207,
+      "step": 15598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7358714300620751,
+      "learning_rate": 5.594558082489565e-07,
+      "loss": 0.2504,
+      "step": 15599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37675397571544944,
+      "learning_rate": 5.58842261885889e-07,
+      "loss": 0.277,
+      "step": 15600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6241152115560328,
+      "learning_rate": 5.582290424697078e-07,
+      "loss": 0.2907,
+      "step": 15601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3432574165161779,
+      "learning_rate": 5.576161500216481e-07,
+      "loss": 0.3207,
+      "step": 15602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32805857188412496,
+      "learning_rate": 5.570035845629362e-07,
+      "loss": 0.207,
+      "step": 15603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3566758030118385,
+      "learning_rate": 5.563913461147841e-07,
+      "loss": 0.2272,
+      "step": 15604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33717272175655655,
+      "learning_rate": 5.557794346983936e-07,
+      "loss": 0.2613,
+      "step": 15605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25959042748500355,
+      "learning_rate": 5.551678503349545e-07,
+      "loss": 0.2154,
+      "step": 15606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8584149753122465,
+      "learning_rate": 5.545565930456464e-07,
+      "loss": 0.4831,
+      "step": 15607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2299027222034258,
+      "learning_rate": 5.539456628516382e-07,
+      "loss": 0.7181,
+      "step": 15608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7561475327791286,
+      "learning_rate": 5.53335059774085e-07,
+      "loss": 0.1436,
+      "step": 15609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2262751318435284,
+      "learning_rate": 5.527247838341332e-07,
+      "loss": 0.2338,
+      "step": 15610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40729868452013945,
+      "learning_rate": 5.521148350529137e-07,
+      "loss": 0.2774,
+      "step": 15611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8980545590794237,
+      "learning_rate": 5.51505213451553e-07,
+      "loss": 0.3379,
+      "step": 15612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3863824149031958,
+      "learning_rate": 5.508959190511609e-07,
+      "loss": 0.2108,
+      "step": 15613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33532397215502635,
+      "learning_rate": 5.502869518728359e-07,
+      "loss": 0.3238,
+      "step": 15614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5569176551550571,
+      "learning_rate": 5.49678311937667e-07,
+      "loss": 0.3709,
+      "step": 15615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.36933007186294964,
+      "learning_rate": 5.490699992667326e-07,
+      "loss": 0.2373,
+      "step": 15616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45251797358873475,
+      "learning_rate": 5.48462013881097e-07,
+      "loss": 0.2836,
+      "step": 15617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23540557748561147,
+      "learning_rate": 5.478543558018167e-07,
+      "loss": 0.2141,
+      "step": 15618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4132520532778912,
+      "learning_rate": 5.472470250499328e-07,
+      "loss": 0.2118,
+      "step": 15619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2284634022109895,
+      "learning_rate": 5.466400216464774e-07,
+      "loss": 0.7019,
+      "step": 15620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3909616326979364,
+      "learning_rate": 5.460333456124722e-07,
+      "loss": 0.2444,
+      "step": 15621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.280909579303553,
+      "learning_rate": 5.454269969689252e-07,
+      "loss": 0.2283,
+      "step": 15622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6943885695673067,
+      "learning_rate": 5.448209757368361e-07,
+      "loss": 0.3467,
+      "step": 15623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.214468125396571,
+      "learning_rate": 5.442152819371882e-07,
+      "loss": 0.1708,
+      "step": 15624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5495278130056973,
+      "learning_rate": 5.436099155909592e-07,
+      "loss": 0.3312,
+      "step": 15625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.314657048380129,
+      "learning_rate": 5.430048767191121e-07,
+      "loss": 0.2532,
+      "step": 15626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5585893096454551,
+      "learning_rate": 5.424001653426003e-07,
+      "loss": 0.3362,
+      "step": 15627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.36554119112794337,
+      "learning_rate": 5.417957814823627e-07,
+      "loss": 0.2746,
+      "step": 15628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3853663692179569,
+      "learning_rate": 5.411917251593313e-07,
+      "loss": 0.2478,
+      "step": 15629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21621671494700923,
+      "learning_rate": 5.405879963944238e-07,
+      "loss": 0.1392,
+      "step": 15630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4678193698155111,
+      "learning_rate": 5.39984595208547e-07,
+      "loss": 0.2854,
+      "step": 15631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8065746449842706,
+      "learning_rate": 5.393815216225972e-07,
+      "loss": 0.3653,
+      "step": 15632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3145087220616073,
+      "learning_rate": 5.387787756574592e-07,
+      "loss": 0.2735,
+      "step": 15633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3790211085952469,
+      "learning_rate": 5.381763573340049e-07,
+      "loss": 0.2876,
+      "step": 15634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9396952767327181,
+      "learning_rate": 5.375742666730955e-07,
+      "loss": 0.2666,
+      "step": 15635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23651179482052748,
+      "learning_rate": 5.36972503695582e-07,
+      "loss": 0.1632,
+      "step": 15636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31046373469158267,
+      "learning_rate": 5.363710684223045e-07,
+      "loss": 0.264,
+      "step": 15637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48144032175554063,
+      "learning_rate": 5.357699608740907e-07,
+      "loss": 0.3497,
+      "step": 15638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.538184412368237,
+      "learning_rate": 5.351691810717552e-07,
+      "loss": 0.2247,
+      "step": 15639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42659928158946236,
+      "learning_rate": 5.345687290361035e-07,
+      "loss": 0.3106,
+      "step": 15640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6164374849438661,
+      "learning_rate": 5.339686047879311e-07,
+      "loss": 0.3437,
+      "step": 15641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21254629178819517,
+      "learning_rate": 5.333688083480182e-07,
+      "loss": 0.1303,
+      "step": 15642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4149495701684598,
+      "learning_rate": 5.327693397371369e-07,
+      "loss": 0.2945,
+      "step": 15643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8395100725171276,
+      "learning_rate": 5.321701989760452e-07,
+      "loss": 0.4472,
+      "step": 15644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2746630354052698,
+      "learning_rate": 5.315713860854921e-07,
+      "loss": 0.2244,
+      "step": 15645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3804893461734729,
+      "learning_rate": 5.309729010862163e-07,
+      "loss": 0.3088,
+      "step": 15646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3065235001536424,
+      "learning_rate": 5.303747439989415e-07,
+      "loss": 0.4975,
+      "step": 15647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1564879560582931,
+      "learning_rate": 5.29776914844382e-07,
+      "loss": 0.0707,
+      "step": 15648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.285253213852128,
+      "learning_rate": 5.291794136432393e-07,
+      "loss": 0.2486,
+      "step": 15649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5030656323905838,
+      "learning_rate": 5.285822404162066e-07,
+      "loss": 0.3373,
+      "step": 15650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7274055892952281,
+      "learning_rate": 5.279853951839653e-07,
+      "loss": 0.4009,
+      "step": 15651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3379741609185097,
+      "learning_rate": 5.27388877967181e-07,
+      "loss": 0.2087,
+      "step": 15652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5153461591870591,
+      "learning_rate": 5.267926887865127e-07,
+      "loss": 0.2932,
+      "step": 15653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33147772787988616,
+      "learning_rate": 5.261968276626062e-07,
+      "loss": 0.2518,
+      "step": 15654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2169728639824801,
+      "learning_rate": 5.256012946160971e-07,
+      "loss": 0.1588,
+      "step": 15655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7769396824390693,
+      "learning_rate": 5.250060896676068e-07,
+      "loss": 0.4372,
+      "step": 15656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3540478962758659,
+      "learning_rate": 5.244112128377477e-07,
+      "loss": 0.2981,
+      "step": 15657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31449744620930736,
+      "learning_rate": 5.238166641471221e-07,
+      "loss": 0.2103,
+      "step": 15658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3570129545424443,
+      "learning_rate": 5.23222443616318e-07,
+      "loss": 0.4924,
+      "step": 15659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28912954651128253,
+      "learning_rate": 5.226285512659123e-07,
+      "loss": 0.1437,
+      "step": 15660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2773332070677685,
+      "learning_rate": 5.220349871164732e-07,
+      "loss": 0.2187,
+      "step": 15661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4015892649555536,
+      "learning_rate": 5.214417511885539e-07,
+      "loss": 0.308,
+      "step": 15662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6413421071522666,
+      "learning_rate": 5.208488435026992e-07,
+      "loss": 0.4395,
+      "step": 15663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4767384537292822,
+      "learning_rate": 5.202562640794429e-07,
+      "loss": 0.2826,
+      "step": 15664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38488084441750964,
+      "learning_rate": 5.196640129393038e-07,
+      "loss": 0.2458,
+      "step": 15665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3259987938419826,
+      "learning_rate": 5.190720901027901e-07,
+      "loss": 0.5462,
+      "step": 15666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2478349774975706,
+      "learning_rate": 5.184804955904066e-07,
+      "loss": 0.2062,
+      "step": 15667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5459979152515864,
+      "learning_rate": 5.178892294226334e-07,
+      "loss": 0.3124,
+      "step": 15668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35238957217762756,
+      "learning_rate": 5.172982916199465e-07,
+      "loss": 0.3126,
+      "step": 15669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3040986720543595,
+      "learning_rate": 5.167076822028149e-07,
+      "loss": 0.2522,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23105199235660073,
+      "learning_rate": 5.16117401191687e-07,
+      "loss": 0.0663,
+      "step": 15671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4992153585528372,
+      "learning_rate": 5.155274486070072e-07,
+      "loss": 0.3346,
+      "step": 15672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.261839200482087,
+      "learning_rate": 5.149378244692027e-07,
+      "loss": 0.2322,
+      "step": 15673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5089769058171006,
+      "learning_rate": 5.143485287986927e-07,
+      "loss": 0.2294,
+      "step": 15674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3794901247148647,
+      "learning_rate": 5.137595616158863e-07,
+      "loss": 0.2656,
+      "step": 15675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33643466430036256,
+      "learning_rate": 5.131709229411785e-07,
+      "loss": 0.2438,
+      "step": 15676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3647097642493535,
+      "learning_rate": 5.12582612794954e-07,
+      "loss": 0.282,
+      "step": 15677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7738986859788042,
+      "learning_rate": 5.119946311975843e-07,
+      "loss": 0.2076,
+      "step": 15678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34867075246133533,
+      "learning_rate": 5.114069781694331e-07,
+      "loss": 0.2548,
+      "step": 15679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6216173151236968,
+      "learning_rate": 5.108196537308507e-07,
+      "loss": 0.3724,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2894107217697507,
+      "learning_rate": 5.102326579021754e-07,
+      "loss": 0.2367,
+      "step": 15681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33805317461218487,
+      "learning_rate": 5.096459907037344e-07,
+      "loss": 0.2521,
+      "step": 15682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30537643202596476,
+      "learning_rate": 5.090596521558455e-07,
+      "loss": 0.1848,
+      "step": 15683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.44721678914316987,
+      "learning_rate": 5.084736422788128e-07,
+      "loss": 0.2464,
+      "step": 15684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25725829442115,
+      "learning_rate": 5.078879610929299e-07,
+      "loss": 0.2437,
+      "step": 15685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6813280238885259,
+      "learning_rate": 5.073026086184785e-07,
+      "loss": 0.4465,
+      "step": 15686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22755669708281842,
+      "learning_rate": 5.067175848757288e-07,
+      "loss": 0.1386,
+      "step": 15687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23563842364379814,
+      "learning_rate": 5.061328898849416e-07,
+      "loss": 0.1959,
+      "step": 15688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3534225275012728,
+      "learning_rate": 5.055485236663638e-07,
+      "loss": 0.2719,
+      "step": 15689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6615023017392966,
+      "learning_rate": 5.049644862402336e-07,
+      "loss": 0.3489,
+      "step": 15690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30468397890498616,
+      "learning_rate": 5.043807776267729e-07,
+      "loss": 0.203,
+      "step": 15691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2458868096231006,
+      "learning_rate": 5.037973978461985e-07,
+      "loss": 0.6232,
+      "step": 15692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3412161058565019,
+      "learning_rate": 5.032143469187123e-07,
+      "loss": 0.3043,
+      "step": 15693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2687736892933335,
+      "learning_rate": 5.026316248645047e-07,
+      "loss": 0.1856,
+      "step": 15694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2705685324835451,
+      "learning_rate": 5.020492317037539e-07,
+      "loss": 0.1572,
+      "step": 15695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4139012116678503,
+      "learning_rate": 5.014671674566319e-07,
+      "loss": 0.3283,
+      "step": 15696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.304048179285271,
+      "learning_rate": 5.008854321432932e-07,
+      "loss": 0.1848,
+      "step": 15697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5339104269414403,
+      "learning_rate": 5.003040257838831e-07,
+      "loss": 0.3313,
+      "step": 15698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1617194889579199,
+      "learning_rate": 4.997229483985366e-07,
+      "loss": 0.4965,
+      "step": 15699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4026837415595859,
+      "learning_rate": 4.991422000073753e-07,
+      "loss": 0.2646,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21280087447699006,
+      "learning_rate": 4.985617806305121e-07,
+      "loss": 0.1822,
+      "step": 15701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7633811247449117,
+      "learning_rate": 4.979816902880441e-07,
+      "loss": 0.3467,
+      "step": 15702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3855766888818858,
+      "learning_rate": 4.97401929000062e-07,
+      "loss": 0.2693,
+      "step": 15703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34135917725265313,
+      "learning_rate": 4.968224967866431e-07,
+      "loss": 0.2474,
+      "step": 15704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4960999651515975,
+      "learning_rate": 4.962433936678523e-07,
+      "loss": 0.3536,
+      "step": 15705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4109252916237451,
+      "learning_rate": 4.956646196637438e-07,
+      "loss": 0.2418,
+      "step": 15706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3962821887700003,
+      "learning_rate": 4.950861747943603e-07,
+      "loss": 0.203,
+      "step": 15707
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32311958760614523,
+      "learning_rate": 4.945080590797346e-07,
+      "loss": 0.2377,
+      "step": 15708
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2812146473652669,
+      "learning_rate": 4.939302725398865e-07,
+      "loss": 0.2372,
+      "step": 15709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.489036443344534,
+      "learning_rate": 4.933528151948241e-07,
+      "loss": 0.2296,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1842550860891952,
+      "learning_rate": 4.92775687064545e-07,
+      "loss": 0.8144,
+      "step": 15711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3325210959542368,
+      "learning_rate": 4.921988881690332e-07,
+      "loss": 0.2314,
+      "step": 15712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4057706489840598,
+      "learning_rate": 4.91622418528267e-07,
+      "loss": 0.3007,
+      "step": 15713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33307672193546267,
+      "learning_rate": 4.910462781622072e-07,
+      "loss": 0.1754,
+      "step": 15714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6227748994436654,
+      "learning_rate": 4.904704670908067e-07,
+      "loss": 0.3603,
+      "step": 15715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42078372345346,
+      "learning_rate": 4.89894985334003e-07,
+      "loss": 0.2779,
+      "step": 15716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30774036410484645,
+      "learning_rate": 4.893198329117277e-07,
+      "loss": 0.2554,
+      "step": 15717
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41423426847798006,
+      "learning_rate": 4.887450098438984e-07,
+      "loss": 0.2504,
+      "step": 15718
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4670151159835837,
+      "learning_rate": 4.881705161504202e-07,
+      "loss": 0.3101,
+      "step": 15719
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21198178399623216,
+      "learning_rate": 4.875963518511872e-07,
+      "loss": 0.1375,
+      "step": 15720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3436783552350103,
+      "learning_rate": 4.870225169660836e-07,
+      "loss": 0.2452,
+      "step": 15721
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6392564449176125,
+      "learning_rate": 4.864490115149823e-07,
+      "loss": 0.3374,
+      "step": 15722
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.781527813043474,
+      "learning_rate": 4.858758355177418e-07,
+      "loss": 0.337,
+      "step": 15723
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3130251635811382,
+      "learning_rate": 4.853029889942129e-07,
+      "loss": 0.2628,
+      "step": 15724
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35452990662259,
+      "learning_rate": 4.84730471964231e-07,
+      "loss": 0.2988,
+      "step": 15725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41564857624579266,
+      "learning_rate": 4.841582844476244e-07,
+      "loss": 0.2544,
+      "step": 15726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2698526571349199,
+      "learning_rate": 4.835864264642076e-07,
+      "loss": 0.1338,
+      "step": 15727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5453059779722633,
+      "learning_rate": 4.830148980337834e-07,
+      "loss": 0.3428,
+      "step": 15728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3471356881056745,
+      "learning_rate": 4.824436991761428e-07,
+      "loss": 0.3259,
+      "step": 15729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34684817296127346,
+      "learning_rate": 4.818728299110686e-07,
+      "loss": 0.2164,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8953070877213622,
+      "learning_rate": 4.813022902583286e-07,
+      "loss": 0.4185,
+      "step": 15731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41813356179096717,
+      "learning_rate": 4.807320802376824e-07,
+      "loss": 0.2948,
+      "step": 15732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1447814308629961,
+      "learning_rate": 4.801621998688722e-07,
+      "loss": 0.0716,
+      "step": 15733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3883904878065946,
+      "learning_rate": 4.795926491716396e-07,
+      "loss": 0.3139,
+      "step": 15734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6516174961256956,
+      "learning_rate": 4.790234281657025e-07,
+      "loss": 0.3845,
+      "step": 15735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3672484706266206,
+      "learning_rate": 4.784545368707738e-07,
+      "loss": 0.204,
+      "step": 15736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32729802797217145,
+      "learning_rate": 4.778859753065545e-07,
+      "loss": 0.2943,
+      "step": 15737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4014564746228084,
+      "learning_rate": 4.773177434927356e-07,
+      "loss": 0.1903,
+      "step": 15738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42003454062639795,
+      "learning_rate": 4.767498414489935e-07,
+      "loss": 0.2329,
+      "step": 15739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25069726353847577,
+      "learning_rate": 4.7618226919499465e-07,
+      "loss": 0.233,
+      "step": 15740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7447684355120814,
+      "learning_rate": 4.756150267503934e-07,
+      "loss": 0.4077,
+      "step": 15741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5399344736483198,
+      "learning_rate": 4.750481141348362e-07,
+      "loss": 0.3129,
+      "step": 15742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34196695458758986,
+      "learning_rate": 4.7448153136795185e-07,
+      "loss": 0.2277,
+      "step": 15743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38306387196871694,
+      "learning_rate": 4.739152784693635e-07,
+      "loss": 0.2897,
+      "step": 15744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2161941592051655,
+      "learning_rate": 4.733493554586777e-07,
+      "loss": 0.1393,
+      "step": 15745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.359758705296901,
+      "learning_rate": 4.727837623554954e-07,
+      "loss": 0.2245,
+      "step": 15746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6364944282856332,
+      "learning_rate": 4.722184991794021e-07,
+      "loss": 0.368,
+      "step": 15747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29869548642001714,
+      "learning_rate": 4.7165356594997215e-07,
+      "loss": 0.2794,
+      "step": 15748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3514622933369766,
+      "learning_rate": 4.710889626867687e-07,
+      "loss": 0.2274,
+      "step": 15749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.534507202273457,
+      "learning_rate": 4.7052468940934405e-07,
+      "loss": 0.2445,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40165702496329014,
+      "learning_rate": 4.699607461372413e-07,
+      "loss": 0.2209,
+      "step": 15751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29933132482387254,
+      "learning_rate": 4.6939713288998824e-07,
+      "loss": 0.2888,
+      "step": 15752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39195373443976395,
+      "learning_rate": 4.6883384968710146e-07,
+      "loss": 0.275,
+      "step": 15753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.804794875047898,
+      "learning_rate": 4.682708965480887e-07,
+      "loss": 0.447,
+      "step": 15754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.38284619581894447,
+      "learning_rate": 4.677082734924454e-07,
+      "loss": 0.2555,
+      "step": 15755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34044252869584235,
+      "learning_rate": 4.67145980539655e-07,
+      "loss": 0.2351,
+      "step": 15756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2561805020204338,
+      "learning_rate": 4.665840177091885e-07,
+      "loss": 0.1603,
+      "step": 15757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3286363690720469,
+      "learning_rate": 4.66022385020507e-07,
+      "loss": 0.2712,
+      "step": 15758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7834022205802545,
+      "learning_rate": 4.6546108249306163e-07,
+      "loss": 0.3245,
+      "step": 15759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2849494240717211,
+      "learning_rate": 4.649001101462891e-07,
+      "loss": 0.2749,
+      "step": 15760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3713613181819242,
+      "learning_rate": 4.6433946799961605e-07,
+      "loss": 0.2917,
+      "step": 15761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3943161939139344,
+      "learning_rate": 4.6377915607245583e-07,
+      "loss": 0.218,
+      "step": 15762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35360704239027846,
+      "learning_rate": 4.6321917438421294e-07,
+      "loss": 0.2162,
+      "step": 15763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29520125023410765,
+      "learning_rate": 4.626595229542818e-07,
+      "loss": 0.2779,
+      "step": 15764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40025625010104504,
+      "learning_rate": 4.621002018020404e-07,
+      "loss": 0.3336,
+      "step": 15765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18183834726311107,
+      "learning_rate": 4.615412109468587e-07,
+      "loss": 0.1206,
+      "step": 15766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3861098147618549,
+      "learning_rate": 4.6098255040809447e-07,
+      "loss": 0.275,
+      "step": 15767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39267454900302395,
+      "learning_rate": 4.604242202050957e-07,
+      "loss": 0.2835,
+      "step": 15768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8010348120564672,
+      "learning_rate": 4.5986622035719575e-07,
+      "loss": 0.1828,
+      "step": 15769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35292055566037195,
+      "learning_rate": 4.59308550883717e-07,
+      "loss": 0.2744,
+      "step": 15770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44243941838987,
+      "learning_rate": 4.5875121180397276e-07,
+      "loss": 0.2978,
+      "step": 15771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29214452693872744,
+      "learning_rate": 4.581942031372655e-07,
+      "loss": 0.2196,
+      "step": 15772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3078542986697614,
+      "learning_rate": 4.5763752490288194e-07,
+      "loss": 0.242,
+      "step": 15773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8844199684923539,
+      "learning_rate": 4.570811771201e-07,
+      "loss": 0.4156,
+      "step": 15774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6335765581868591,
+      "learning_rate": 4.5652515980818546e-07,
+      "loss": 0.3745,
+      "step": 15775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22674350740370733,
+      "learning_rate": 4.5596947298639614e-07,
+      "loss": 0.212,
+      "step": 15776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7659474533173485,
+      "learning_rate": 4.554141166739734e-07,
+      "loss": 0.5087,
+      "step": 15777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2534645385419527,
+      "learning_rate": 4.548590908901496e-07,
+      "loss": 0.1883,
+      "step": 15778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.283779004129321,
+      "learning_rate": 4.5430439565414263e-07,
+      "loss": 0.1839,
+      "step": 15779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3476013206398067,
+      "learning_rate": 4.5375003098516613e-07,
+      "loss": 0.2717,
+      "step": 15780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5541988538944898,
+      "learning_rate": 4.5319599690241576e-07,
+      "loss": 0.3387,
+      "step": 15781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34841705919786675,
+      "learning_rate": 4.5264229342507736e-07,
+      "loss": 0.2278,
+      "step": 15782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2826263441085395,
+      "learning_rate": 4.5208892057232446e-07,
+      "loss": 0.6443,
+      "step": 15783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25060351873355025,
+      "learning_rate": 4.515358783633228e-07,
+      "loss": 0.238,
+      "step": 15784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2457619049739712,
+      "learning_rate": 4.5098316681722266e-07,
+      "loss": 0.1596,
+      "step": 15785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7165815667752252,
+      "learning_rate": 4.5043078595316536e-07,
+      "loss": 0.3654,
+      "step": 15786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45627630512144096,
+      "learning_rate": 4.4987873579027784e-07,
+      "loss": 0.3349,
+      "step": 15787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3075011835404249,
+      "learning_rate": 4.493270163476804e-07,
+      "loss": 0.2568,
+      "step": 15788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48719044332200573,
+      "learning_rate": 4.4877562764447766e-07,
+      "loss": 0.2597,
+      "step": 15789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5000417018144054,
+      "learning_rate": 4.4822456969976444e-07,
+      "loss": 0.2893,
+      "step": 15790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2556509931355065,
+      "learning_rate": 4.4767384253262326e-07,
+      "loss": 0.2022,
+      "step": 15791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3334183800000595,
+      "learning_rate": 4.4712344616212433e-07,
+      "loss": 0.246,
+      "step": 15792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.978515196075592,
+      "learning_rate": 4.4657338060733246e-07,
+      "loss": 0.4858,
+      "step": 15793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3399680398954725,
+      "learning_rate": 4.4602364588729243e-07,
+      "loss": 0.2275,
+      "step": 15794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5585273086353948,
+      "learning_rate": 4.454742420210434e-07,
+      "loss": 0.3749,
+      "step": 15795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3308699135535816,
+      "learning_rate": 4.44925169027608e-07,
+      "loss": 0.3088,
+      "step": 15796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24890530106124995,
+      "learning_rate": 4.4437642692600534e-07,
+      "loss": 0.1998,
+      "step": 15797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42886569435392907,
+      "learning_rate": 4.4382801573523595e-07,
+      "loss": 0.1926,
+      "step": 15798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3308619773627193,
+      "learning_rate": 4.4327993547429225e-07,
+      "loss": 0.2951,
+      "step": 15799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32885301500455993,
+      "learning_rate": 4.427321861621514e-07,
+      "loss": 0.2736,
+      "step": 15800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1618546884805907,
+      "learning_rate": 4.4218476781778483e-07,
+      "loss": 0.5503,
+      "step": 15801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35167729216980037,
+      "learning_rate": 4.416376804601508e-07,
+      "loss": 0.1287,
+      "step": 15802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3502201067992776,
+      "learning_rate": 4.410909241081918e-07,
+      "loss": 0.2668,
+      "step": 15803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3340832982352592,
+      "learning_rate": 4.405444987808405e-07,
+      "loss": 0.2879,
+      "step": 15804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.43546774821008033,
+      "learning_rate": 4.39998404497024e-07,
+      "loss": 0.1536,
+      "step": 15805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3112353537779798,
+      "learning_rate": 4.3945264127565166e-07,
+      "loss": 0.2528,
+      "step": 15806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.551693833136559,
+      "learning_rate": 4.389072091356239e-07,
+      "loss": 0.3786,
+      "step": 15807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.49880144907478224,
+      "learning_rate": 4.383621080958267e-07,
+      "loss": 0.2335,
+      "step": 15808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3142694304193792,
+      "learning_rate": 4.378173381751394e-07,
+      "loss": 0.2585,
+      "step": 15809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40112648470346424,
+      "learning_rate": 4.372728993924269e-07,
+      "loss": 0.266,
+      "step": 15810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2930212171494269,
+      "learning_rate": 4.3672879176654303e-07,
+      "loss": 0.1907,
+      "step": 15811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34939185440615567,
+      "learning_rate": 4.3618501531632717e-07,
+      "loss": 0.282,
+      "step": 15812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1777285605456114,
+      "learning_rate": 4.3564157006061535e-07,
+      "loss": 0.685,
+      "step": 15813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7188027543719689,
+      "learning_rate": 4.3509845601822474e-07,
+      "loss": 0.4602,
+      "step": 15814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30789053770206504,
+      "learning_rate": 4.3455567320796366e-07,
+      "loss": 0.216,
+      "step": 15815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37512687043786214,
+      "learning_rate": 4.34013221648627e-07,
+      "loss": 0.3312,
+      "step": 15816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3616392023467653,
+      "learning_rate": 4.3347110135900094e-07,
+      "loss": 0.1928,
+      "step": 15817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31199271737258344,
+      "learning_rate": 4.329293123578604e-07,
+      "loss": 0.1853,
+      "step": 15818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4895600138237852,
+      "learning_rate": 4.3238785466396596e-07,
+      "loss": 0.3757,
+      "step": 15819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45924672686202866,
+      "learning_rate": 4.318467282960681e-07,
+      "loss": 0.3507,
+      "step": 15820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3120934451623118,
+      "learning_rate": 4.3130593327290637e-07,
+      "loss": 0.1888,
+      "step": 15821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.770298360679297,
+      "learning_rate": 4.307654696132102e-07,
+      "loss": 0.3835,
+      "step": 15822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21490372065099173,
+      "learning_rate": 4.302253373356935e-07,
+      "loss": 0.2062,
+      "step": 15823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2986581235557719,
+      "learning_rate": 4.296855364590624e-07,
+      "loss": 0.1844,
+      "step": 15824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.069549265915451,
+      "learning_rate": 4.2914606700200755e-07,
+      "loss": 0.7423,
+      "step": 15825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6599039669777897,
+      "learning_rate": 4.286069289832151e-07,
+      "loss": 0.3809,
+      "step": 15826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.419239808764103,
+      "learning_rate": 4.280681224213523e-07,
+      "loss": 0.2807,
+      "step": 15827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31727020613520757,
+      "learning_rate": 4.2752964733507984e-07,
+      "loss": 0.2382,
+      "step": 15828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29397819184063156,
+      "learning_rate": 4.2699150374304275e-07,
+      "loss": 0.1889,
+      "step": 15829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4446204152271284,
+      "learning_rate": 4.2645369166387727e-07,
+      "loss": 0.2613,
+      "step": 15830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39269983702966865,
+      "learning_rate": 4.259162111162107e-07,
+      "loss": 0.2553,
+      "step": 15831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45515828284942017,
+      "learning_rate": 4.2537906211865375e-07,
+      "loss": 0.3144,
+      "step": 15832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3112784755800918,
+      "learning_rate": 4.2484224468980815e-07,
+      "loss": 0.2632,
+      "step": 15833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8587599044679906,
+      "learning_rate": 4.243057588482624e-07,
+      "loss": 0.2075,
+      "step": 15834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23261596393289166,
+      "learning_rate": 4.237696046125994e-07,
+      "loss": 0.2007,
+      "step": 15835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4171560765483998,
+      "learning_rate": 4.232337820013821e-07,
+      "loss": 0.2908,
+      "step": 15836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6894450007615852,
+      "learning_rate": 4.226982910331656e-07,
+      "loss": 0.2665,
+      "step": 15837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.631586559127634,
+      "learning_rate": 4.2216313172649623e-07,
+      "loss": 0.3694,
+      "step": 15838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3701344195302594,
+      "learning_rate": 4.2162830409990583e-07,
+      "loss": 0.2664,
+      "step": 15839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32238862938044927,
+      "learning_rate": 4.2109380817191626e-07,
+      "loss": 0.2694,
+      "step": 15840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36861815680269605,
+      "learning_rate": 4.205596439610349e-07,
+      "loss": 0.1159,
+      "step": 15841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3877743415272305,
+      "learning_rate": 4.2002581148576136e-07,
+      "loss": 0.2594,
+      "step": 15842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35220346095956107,
+      "learning_rate": 4.194923107645821e-07,
+      "loss": 0.3235,
+      "step": 15843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5239046538366025,
+      "learning_rate": 4.189591418159722e-07,
+      "loss": 0.2708,
+      "step": 15844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39195351825581,
+      "learning_rate": 4.1842630465839586e-07,
+      "loss": 0.2784,
+      "step": 15845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5495615089399805,
+      "learning_rate": 4.178937993103027e-07,
+      "loss": 0.3197,
+      "step": 15846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27379809113010045,
+      "learning_rate": 4.1736162579013694e-07,
+      "loss": 0.1906,
+      "step": 15847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2684468644693002,
+      "learning_rate": 4.16829784116326e-07,
+      "loss": 0.2042,
+      "step": 15848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4441215646728875,
+      "learning_rate": 4.1629827430728743e-07,
+      "loss": 0.2991,
+      "step": 15849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7808231682802714,
+      "learning_rate": 4.157670963814264e-07,
+      "loss": 0.3301,
+      "step": 15850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2581293927684206,
+      "learning_rate": 4.1523625035713943e-07,
+      "loss": 0.2583,
+      "step": 15851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6004003120556647,
+      "learning_rate": 4.147057362528095e-07,
+      "loss": 0.3379,
+      "step": 15852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5608318585976009,
+      "learning_rate": 4.141755540868075e-07,
+      "loss": 0.2723,
+      "step": 15853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3815204124123482,
+      "learning_rate": 4.1364570387749324e-07,
+      "loss": 0.213,
+      "step": 15854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3536760296474735,
+      "learning_rate": 4.1311618564321534e-07,
+      "loss": 0.2938,
+      "step": 15855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4356058557524649,
+      "learning_rate": 4.1258699940231353e-07,
+      "loss": 0.355,
+      "step": 15856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18578229436004529,
+      "learning_rate": 4.120581451731109e-07,
+      "loss": 0.0713,
+      "step": 15857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4081226800618796,
+      "learning_rate": 4.1152962297392297e-07,
+      "loss": 0.273,
+      "step": 15858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37698659558030323,
+      "learning_rate": 4.110014328230505e-07,
+      "loss": 0.3109,
+      "step": 15859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5250626358401506,
+      "learning_rate": 4.104735747387867e-07,
+      "loss": 0.2035,
+      "step": 15860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3725837020347488,
+      "learning_rate": 4.099460487394114e-07,
+      "loss": 0.3016,
+      "step": 15861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2275656487782942,
+      "learning_rate": 4.09418854843191e-07,
+      "loss": 0.7537,
+      "step": 15862
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2383154149448102,
+      "learning_rate": 4.0889199306838323e-07,
+      "loss": 0.2219,
+      "step": 15863
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3139643851031367,
+      "learning_rate": 4.083654634332335e-07,
+      "loss": 0.1762,
+      "step": 15864
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9530704152814369,
+      "learning_rate": 4.07839265955976e-07,
+      "loss": 0.4449,
+      "step": 15865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5492062970741735,
+      "learning_rate": 4.073134006548318e-07,
+      "loss": 0.2962,
+      "step": 15866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25208480398575406,
+      "learning_rate": 4.06787867548013e-07,
+      "loss": 0.2353,
+      "step": 15867
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1304469249273643,
+      "learning_rate": 4.062626666537162e-07,
+      "loss": 0.8001,
+      "step": 15868
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19056843516548483,
+      "learning_rate": 4.0573779799013226e-07,
+      "loss": 0.1272,
+      "step": 15869
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3996361879336976,
+      "learning_rate": 4.0521326157543563e-07,
+      "loss": 0.2277,
+      "step": 15870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3637325778437535,
+      "learning_rate": 4.046890574277895e-07,
+      "loss": 0.2886,
+      "step": 15871
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5237881779689043,
+      "learning_rate": 4.0416518556534944e-07,
+      "loss": 0.3128,
+      "step": 15872
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3676819772907118,
+      "learning_rate": 4.0364164600625753e-07,
+      "loss": 0.2235,
+      "step": 15873
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.47406659233561166,
+      "learning_rate": 4.0311843876864155e-07,
+      "loss": 0.3483,
+      "step": 15874
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2780747777411099,
+      "learning_rate": 4.025955638706203e-07,
+      "loss": 0.2054,
+      "step": 15875
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28877513130409865,
+      "learning_rate": 4.020730213303037e-07,
+      "loss": 0.2163,
+      "step": 15876
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6012066973017609,
+      "learning_rate": 4.015508111657862e-07,
+      "loss": 0.2896,
+      "step": 15877
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.77096149159783,
+      "learning_rate": 4.0102893339515e-07,
+      "loss": 0.3788,
+      "step": 15878
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24983255991293724,
+      "learning_rate": 4.005073880364696e-07,
+      "loss": 0.2577,
+      "step": 15879
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8911082860363854,
+      "learning_rate": 3.999861751078049e-07,
+      "loss": 0.5244,
+      "step": 15880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21183230763143193,
+      "learning_rate": 3.994652946272071e-07,
+      "loss": 0.1423,
+      "step": 15881
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39131821841894676,
+      "learning_rate": 3.989447466127128e-07,
+      "loss": 0.2829,
+      "step": 15882
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32030932741388807,
+      "learning_rate": 3.984245310823498e-07,
+      "loss": 0.2383,
+      "step": 15883
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5922727533151578,
+      "learning_rate": 3.9790464805413044e-07,
+      "loss": 0.3336,
+      "step": 15884
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3930583813091873,
+      "learning_rate": 3.973850975460614e-07,
+      "loss": 0.33,
+      "step": 15885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4788539037882134,
+      "learning_rate": 3.9686587957613377e-07,
+      "loss": 0.2603,
+      "step": 15886
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2870566855876852,
+      "learning_rate": 3.963469941623288e-07,
+      "loss": 0.2128,
+      "step": 15887
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2673197346089193,
+      "learning_rate": 3.958284413226121e-07,
+      "loss": 0.2114,
+      "step": 15888
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5354140355201578,
+      "learning_rate": 3.9531022107494486e-07,
+      "loss": 0.3267,
+      "step": 15889
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35721314864795733,
+      "learning_rate": 3.9479233343727165e-07,
+      "loss": 0.1804,
+      "step": 15890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28916473867304665,
+      "learning_rate": 3.9427477842752693e-07,
+      "loss": 0.2645,
+      "step": 15891
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0814982506679522,
+      "learning_rate": 3.9375755606363306e-07,
+      "loss": 0.6821,
+      "step": 15892
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4496409680340063,
+      "learning_rate": 3.9324066636350136e-07,
+      "loss": 0.1017,
+      "step": 15893
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3384135104379383,
+      "learning_rate": 3.92724109345034e-07,
+      "loss": 0.2612,
+      "step": 15894
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2919069332707678,
+      "learning_rate": 3.922078850261168e-07,
+      "loss": 0.2677,
+      "step": 15895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.075665763003214,
+      "learning_rate": 3.9169199342462774e-07,
+      "loss": 0.1625,
+      "step": 15896
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3534241016716571,
+      "learning_rate": 3.9117643455843016e-07,
+      "loss": 0.2953,
+      "step": 15897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5174200198693831,
+      "learning_rate": 3.906612084453809e-07,
+      "loss": 0.3611,
+      "step": 15898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.408525886252803,
+      "learning_rate": 3.9014631510332135e-07,
+      "loss": 0.2211,
+      "step": 15899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.396189831826831,
+      "learning_rate": 3.896317545500805e-07,
+      "loss": 0.2782,
+      "step": 15900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29171815179801414,
+      "learning_rate": 3.8911752680347857e-07,
+      "loss": 0.1854,
+      "step": 15901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3603406957800757,
+      "learning_rate": 3.8860363188132356e-07,
+      "loss": 0.2891,
+      "step": 15902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2890968051915757,
+      "learning_rate": 3.880900698014134e-07,
+      "loss": 0.1957,
+      "step": 15903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2876865594992615,
+      "learning_rate": 3.8757684058152947e-07,
+      "loss": 0.7166,
+      "step": 15904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.602658793155093,
+      "learning_rate": 3.8706394423944524e-07,
+      "loss": 0.3449,
+      "step": 15905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34185570155092393,
+      "learning_rate": 3.8655138079292444e-07,
+      "loss": 0.2216,
+      "step": 15906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2891611853218796,
+      "learning_rate": 3.8603915025971605e-07,
+      "loss": 0.2511,
+      "step": 15907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3595831908817427,
+      "learning_rate": 3.855272526575582e-07,
+      "loss": 0.2175,
+      "step": 15908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31805676469838995,
+      "learning_rate": 3.8501568800417663e-07,
+      "loss": 0.2128,
+      "step": 15909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3700899135798883,
+      "learning_rate": 3.845044563172895e-07,
+      "loss": 0.3155,
+      "step": 15910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6586535137821252,
+      "learning_rate": 3.8399355761460036e-07,
+      "loss": 0.3655,
+      "step": 15911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3068097109261226,
+      "learning_rate": 3.8348299191380057e-07,
+      "loss": 0.2049,
+      "step": 15912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.288649576944417,
+      "learning_rate": 3.8297275923256936e-07,
+      "loss": 0.1484,
+      "step": 15913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3339644155133282,
+      "learning_rate": 3.824628595885793e-07,
+      "loss": 0.2961,
+      "step": 15914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3166963614134166,
+      "learning_rate": 3.8195329299948737e-07,
+      "loss": 0.2589,
+      "step": 15915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8291946937549082,
+      "learning_rate": 3.814440594829394e-07,
+      "loss": 0.3325,
+      "step": 15916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6188592393447846,
+      "learning_rate": 3.8093515905656797e-07,
+      "loss": 0.3916,
+      "step": 15917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3410052252876215,
+      "learning_rate": 3.804265917380001e-07,
+      "loss": 0.2626,
+      "step": 15918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3440402045586406,
+      "learning_rate": 3.7991835754484616e-07,
+      "loss": 0.224,
+      "step": 15919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2800763142332023,
+      "learning_rate": 3.794104564947054e-07,
+      "loss": 0.1761,
+      "step": 15920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31545824296470104,
+      "learning_rate": 3.789028886051671e-07,
+      "loss": 0.2512,
+      "step": 15921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5130003457284722,
+      "learning_rate": 3.7839565389380606e-07,
+      "loss": 0.2595,
+      "step": 15922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4206333487665775,
+      "learning_rate": 3.7788875237819156e-07,
+      "loss": 0.3264,
+      "step": 15923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4075238314358624,
+      "learning_rate": 3.7738218407587514e-07,
+      "loss": 0.2339,
+      "step": 15924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40120391979014103,
+      "learning_rate": 3.768759490044005e-07,
+      "loss": 0.1618,
+      "step": 15925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24959491992631752,
+      "learning_rate": 3.76370047181297e-07,
+      "loss": 0.2329,
+      "step": 15926
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.35098746815589077,
+      "learning_rate": 3.7586447862408617e-07,
+      "loss": 0.2625,
+      "step": 15927
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6981973004763733,
+      "learning_rate": 3.7535924335027396e-07,
+      "loss": 0.4138,
+      "step": 15928
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7501774342098396,
+      "learning_rate": 3.7485434137735754e-07,
+      "loss": 0.2783,
+      "step": 15929
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3149510966890482,
+      "learning_rate": 3.743497727228207e-07,
+      "loss": 0.2552,
+      "step": 15930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3666851503761762,
+      "learning_rate": 3.738455374041372e-07,
+      "loss": 0.3152,
+      "step": 15931
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1810162314400088,
+      "learning_rate": 3.7334163543876977e-07,
+      "loss": 0.098,
+      "step": 15932
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.304726241285453,
+      "learning_rate": 3.7283806684416777e-07,
+      "loss": 0.259,
+      "step": 15933
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3382470948180131,
+      "learning_rate": 3.723348316377695e-07,
+      "loss": 0.2932,
+      "step": 15934
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8426507956023479,
+      "learning_rate": 3.718319298369999e-07,
+      "loss": 0.3081,
+      "step": 15935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.35343634631778686,
+      "learning_rate": 3.7132936145927835e-07,
+      "loss": 0.2407,
+      "step": 15936
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.545706171785692,
+      "learning_rate": 3.708271265220087e-07,
+      "loss": 0.5846,
+      "step": 15937
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2584584444218997,
+      "learning_rate": 3.703252250425782e-07,
+      "loss": 0.2462,
+      "step": 15938
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28528019760679263,
+      "learning_rate": 3.6982365703837286e-07,
+      "loss": 0.2056,
+      "step": 15939
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6252324364531351,
+      "learning_rate": 3.6932242252675997e-07,
+      "loss": 0.3615,
+      "step": 15940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5203778132134718,
+      "learning_rate": 3.6882152152509674e-07,
+      "loss": 0.2695,
+      "step": 15941
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28609701313106567,
+      "learning_rate": 3.683209540507304e-07,
+      "loss": 0.1995,
+      "step": 15942
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4592887764684259,
+      "learning_rate": 3.678207201209949e-07,
+      "loss": 0.2873,
+      "step": 15943
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4645147472870708,
+      "learning_rate": 3.673208197532152e-07,
+      "loss": 0.2766,
+      "step": 15944
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34419848756733823,
+      "learning_rate": 3.6682125296469973e-07,
+      "loss": 0.1808,
+      "step": 15945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3378083035715725,
+      "learning_rate": 3.6632201977275126e-07,
+      "loss": 0.3049,
+      "step": 15946
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.45811410825743665,
+      "learning_rate": 3.658231201946549e-07,
+      "loss": 0.2619,
+      "step": 15947
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28931462287240073,
+      "learning_rate": 3.6532455424769133e-07,
+      "loss": 0.1857,
+      "step": 15948
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.285888454753432,
+      "learning_rate": 3.6482632194912436e-07,
+      "loss": 0.4671,
+      "step": 15949
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.36370191667119234,
+      "learning_rate": 3.64328423316207e-07,
+      "loss": 0.3186,
+      "step": 15950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.39727285401672247,
+      "learning_rate": 3.638308583661809e-07,
+      "loss": 0.2611,
+      "step": 15951
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42335563840042184,
+      "learning_rate": 3.633336271162791e-07,
+      "loss": 0.2897,
+      "step": 15952
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32788117129585664,
+      "learning_rate": 3.6283672958371987e-07,
+      "loss": 0.1735,
+      "step": 15953
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2670674166828153,
+      "learning_rate": 3.623401657857095e-07,
+      "loss": 0.2516,
+      "step": 15954
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5172019125312697,
+      "learning_rate": 3.618439357394443e-07,
+      "loss": 0.2023,
+      "step": 15955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.765943640524781,
+      "learning_rate": 3.613480394621094e-07,
+      "loss": 0.3744,
+      "step": 15956
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3254959834274066,
+      "learning_rate": 3.608524769708788e-07,
+      "loss": 0.2652,
+      "step": 15957
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29855816184102585,
+      "learning_rate": 3.6035724828291096e-07,
+      "loss": 0.2446,
+      "step": 15958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28100532394478656,
+      "learning_rate": 3.598623534153578e-07,
+      "loss": 0.2126,
+      "step": 15959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42355858312063266,
+      "learning_rate": 3.593677923853556e-07,
+      "loss": 0.3086,
+      "step": 15960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5485094544149791,
+      "learning_rate": 3.5887356521003283e-07,
+      "loss": 0.2255,
+      "step": 15961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3308684754469159,
+      "learning_rate": 3.583796719065047e-07,
+      "loss": 0.3017,
+      "step": 15962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5477406568092534,
+      "learning_rate": 3.578861124918731e-07,
+      "loss": 0.3118,
+      "step": 15963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.37672975801509384,
+      "learning_rate": 3.5739288698323107e-07,
+      "loss": 0.3116,
+      "step": 15964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.20393972677598093,
+      "learning_rate": 3.568999953976582e-07,
+      "loss": 0.154,
+      "step": 15965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3064175262555501,
+      "learning_rate": 3.564074377522253e-07,
+      "loss": 0.2573,
+      "step": 15966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5742413490053281,
+      "learning_rate": 3.5591521406398654e-07,
+      "loss": 0.3097,
+      "step": 15967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5365842583125046,
+      "learning_rate": 3.554233243499894e-07,
+      "loss": 0.2825,
+      "step": 15968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43114460143022276,
+      "learning_rate": 3.5493176862726794e-07,
+      "loss": 0.253,
+      "step": 15969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2924940242715081,
+      "learning_rate": 3.5444054691284535e-07,
+      "loss": 0.2877,
+      "step": 15970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2374504232369525,
+      "learning_rate": 3.539496592237335e-07,
+      "loss": 0.1154,
+      "step": 15971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2552973385212853,
+      "learning_rate": 3.5345910557692655e-07,
+      "loss": 0.1876,
+      "step": 15972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4709639877224451,
+      "learning_rate": 3.529688859894176e-07,
+      "loss": 0.3019,
+      "step": 15973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3310973746232077,
+      "learning_rate": 3.5247900047818193e-07,
+      "loss": 0.2604,
+      "step": 15974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43855069686197606,
+      "learning_rate": 3.5198944906018273e-07,
+      "loss": 0.269,
+      "step": 15975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5369183319927688,
+      "learning_rate": 3.5150023175237303e-07,
+      "loss": 0.3707,
+      "step": 15976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3176682852295683,
+      "learning_rate": 3.5101134857169704e-07,
+      "loss": 0.2594,
+      "step": 15977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24078063210633607,
+      "learning_rate": 3.505227995350824e-07,
+      "loss": 0.1773,
+      "step": 15978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5265718725264137,
+      "learning_rate": 3.5003458465944884e-07,
+      "loss": 0.3234,
+      "step": 15979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.667710297637546,
+      "learning_rate": 3.495467039617018e-07,
+      "loss": 0.3737,
+      "step": 15980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7002359606242801,
+      "learning_rate": 3.4905915745873763e-07,
+      "loss": 0.1409,
+      "step": 15981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2570986693620255,
+      "learning_rate": 3.4857194516744075e-07,
+      "loss": 0.2817,
+      "step": 15982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4987963183833602,
+      "learning_rate": 3.4808506710468204e-07,
+      "loss": 0.2641,
+      "step": 15983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28035991687745376,
+      "learning_rate": 3.4759852328732136e-07,
+      "loss": 0.1183,
+      "step": 15984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3520688668657019,
+      "learning_rate": 3.4711231373220854e-07,
+      "loss": 0.2765,
+      "step": 15985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3869463660741132,
+      "learning_rate": 3.466264384561824e-07,
+      "loss": 0.2959,
+      "step": 15986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41454146732657704,
+      "learning_rate": 3.461408974760683e-07,
+      "loss": 0.1616,
+      "step": 15987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5291939781843723,
+      "learning_rate": 3.456556908086783e-07,
+      "loss": 0.4009,
+      "step": 15988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48867694572595083,
+      "learning_rate": 3.4517081847081693e-07,
+      "loss": 0.3361,
+      "step": 15989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30895000784527976,
+      "learning_rate": 3.44686280479275e-07,
+      "loss": 0.2123,
+      "step": 15990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4070864437798445,
+      "learning_rate": 3.442020768508325e-07,
+      "loss": 0.2457,
+      "step": 15991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.525896819446171,
+      "learning_rate": 3.4371820760225606e-07,
+      "loss": 0.2261,
+      "step": 15992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.389551843659952,
+      "learning_rate": 3.432346727503033e-07,
+      "loss": 0.2679,
+      "step": 15993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29140564211839626,
+      "learning_rate": 3.427514723117187e-07,
+      "loss": 0.2335,
+      "step": 15994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1747614721852366,
+      "learning_rate": 3.4226860630323545e-07,
+      "loss": 0.604,
+      "step": 15995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4861490590325455,
+      "learning_rate": 3.4178607474157464e-07,
+      "loss": 0.3133,
+      "step": 15996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2647163166220459,
+      "learning_rate": 3.413038776434474e-07,
+      "loss": 0.1841,
+      "step": 15997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30848881097191544,
+      "learning_rate": 3.408220150255492e-07,
+      "loss": 0.243,
+      "step": 15998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7268327398882253,
+      "learning_rate": 3.403404869045712e-07,
+      "loss": 0.3952,
+      "step": 15999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3471361648604528,
+      "learning_rate": 3.398592932971878e-07,
+      "loss": 0.2387,
+      "step": 16000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4117894608929416,
+      "learning_rate": 3.393784342200601e-07,
+      "loss": 0.2862,
+      "step": 16001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7838377566107934,
+      "learning_rate": 3.388979096898415e-07,
+      "loss": 0.3532,
+      "step": 16002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34177113425113487,
+      "learning_rate": 3.3841771972317414e-07,
+      "loss": 0.269,
+      "step": 16003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.301355411444743,
+      "learning_rate": 3.3793786433668596e-07,
+      "loss": 0.0814,
+      "step": 16004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30844575612555714,
+      "learning_rate": 3.3745834354699247e-07,
+      "loss": 0.262,
+      "step": 16005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3747679957008508,
+      "learning_rate": 3.3697915737070154e-07,
+      "loss": 0.306,
+      "step": 16006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7646152394209245,
+      "learning_rate": 3.365003058244076e-07,
+      "loss": 0.3003,
+      "step": 16007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.536131589767793,
+      "learning_rate": 3.3602178892469193e-07,
+      "loss": 0.3079,
+      "step": 16008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3947217103658755,
+      "learning_rate": 3.355436066881268e-07,
+      "loss": 0.283,
+      "step": 16009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22066852034689402,
+      "learning_rate": 3.3506575913127006e-07,
+      "loss": 0.1689,
+      "step": 16010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5902292519591517,
+      "learning_rate": 3.3458824627067067e-07,
+      "loss": 0.3642,
+      "step": 16011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.381931762583577,
+      "learning_rate": 3.3411106812286544e-07,
+      "loss": 0.3037,
+      "step": 16012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33341807366384135,
+      "learning_rate": 3.336342247043778e-07,
+      "loss": 0.2449,
+      "step": 16013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5143593485060982,
+      "learning_rate": 3.331577160317201e-07,
+      "loss": 0.294,
+      "step": 16014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4661460986029127,
+      "learning_rate": 3.3268154212139583e-07,
+      "loss": 0.2655,
+      "step": 16015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3431806133508932,
+      "learning_rate": 3.3220570298989507e-07,
+      "loss": 0.1848,
+      "step": 16016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30176925676400573,
+      "learning_rate": 3.317301986536947e-07,
+      "loss": 0.2456,
+      "step": 16017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3353558657145527,
+      "learning_rate": 3.3125502912926044e-07,
+      "loss": 0.2543,
+      "step": 16018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5655804810882348,
+      "learning_rate": 3.307801944330491e-07,
+      "loss": 0.3556,
+      "step": 16019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33186886555261347,
+      "learning_rate": 3.303056945815053e-07,
+      "loss": 0.0592,
+      "step": 16020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27348568532345147,
+      "learning_rate": 3.2983152959105924e-07,
+      "loss": 0.2488,
+      "step": 16021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25027602062501897,
+      "learning_rate": 3.2935769947813003e-07,
+      "loss": 0.2053,
+      "step": 16022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.810186576406879,
+      "learning_rate": 3.2888420425912783e-07,
+      "loss": 0.2417,
+      "step": 16023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.347940790008678,
+      "learning_rate": 3.2841104395045174e-07,
+      "loss": 0.2546,
+      "step": 16024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43409577516476583,
+      "learning_rate": 3.279382185684843e-07,
+      "loss": 0.332,
+      "step": 16025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5257069977199804,
+      "learning_rate": 3.274657281296001e-07,
+      "loss": 0.317,
+      "step": 16026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3115977415753507,
+      "learning_rate": 3.269935726501616e-07,
+      "loss": 0.1908,
+      "step": 16027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31145969908220766,
+      "learning_rate": 3.265217521465203e-07,
+      "loss": 0.1627,
+      "step": 16028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.304969715955574,
+      "learning_rate": 3.2605026663501403e-07,
+      "loss": 0.271,
+      "step": 16029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3198391503543351,
+      "learning_rate": 3.2557911613197213e-07,
+      "loss": 0.2093,
+      "step": 16030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5495654930323556,
+      "learning_rate": 3.251083006537081e-07,
+      "loss": 0.3726,
+      "step": 16031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9376399677814717,
+      "learning_rate": 3.24637820216529e-07,
+      "loss": 0.4465,
+      "step": 16032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27483983067880535,
+      "learning_rate": 3.241676748367251e-07,
+      "loss": 0.2189,
+      "step": 16033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3926888204978273,
+      "learning_rate": 3.2369786453057996e-07,
+      "loss": 0.2467,
+      "step": 16034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8656459700472929,
+      "learning_rate": 3.232283893143595e-07,
+      "loss": 0.4894,
+      "step": 16035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.347362517901405,
+      "learning_rate": 3.2275924920432525e-07,
+      "loss": 0.2104,
+      "step": 16036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3434981831692561,
+      "learning_rate": 3.2229044421672183e-07,
+      "loss": 0.3017,
+      "step": 16037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6020647893512497,
+      "learning_rate": 3.2182197436778527e-07,
+      "loss": 0.212,
+      "step": 16038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33810505101496374,
+      "learning_rate": 3.213538396737359e-07,
+      "loss": 0.2483,
+      "step": 16039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31886277314169437,
+      "learning_rate": 3.2088604015078737e-07,
+      "loss": 0.0928,
+      "step": 16040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3547676364178774,
+      "learning_rate": 3.20418575815139e-07,
+      "loss": 0.2903,
+      "step": 16041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.330343531602598,
+      "learning_rate": 3.1995144668298004e-07,
+      "loss": 0.2459,
+      "step": 16042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8111748137958689,
+      "learning_rate": 3.1948465277048424e-07,
+      "loss": 0.3584,
+      "step": 16043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.35521287147768266,
+      "learning_rate": 3.190181940938197e-07,
+      "loss": 0.2285,
+      "step": 16044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2691973622179739,
+      "learning_rate": 3.185520706691392e-07,
+      "loss": 0.2408,
+      "step": 16045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3561969579912896,
+      "learning_rate": 3.1808628251258413e-07,
+      "loss": 0.2051,
+      "step": 16046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6097716949414717,
+      "learning_rate": 3.1762082964028385e-07,
+      "loss": 0.3688,
+      "step": 16047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.37746575318449793,
+      "learning_rate": 3.171557120683588e-07,
+      "loss": 0.2568,
+      "step": 16048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28599774774097053,
+      "learning_rate": 3.166909298129139e-07,
+      "loss": 0.2567,
+      "step": 16049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2873137393901833,
+      "learning_rate": 3.162264828900474e-07,
+      "loss": 0.1658,
+      "step": 16050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3359843271780522,
+      "learning_rate": 3.1576237131584084e-07,
+      "loss": 0.2672,
+      "step": 16051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8905987280455674,
+      "learning_rate": 3.1529859510636585e-07,
+      "loss": 0.4539,
+      "step": 16052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3531928411300171,
+      "learning_rate": 3.1483515427768506e-07,
+      "loss": 0.2499,
+      "step": 16053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32954114792842953,
+      "learning_rate": 3.143720488458457e-07,
+      "loss": 0.2502,
+      "step": 16054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3633234307925469,
+      "learning_rate": 3.139092788268872e-07,
+      "loss": 0.2719,
+      "step": 16055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29719624462330824,
+      "learning_rate": 3.1344684423683214e-07,
+      "loss": 0.1561,
+      "step": 16056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27943018114548945,
+      "learning_rate": 3.1298474509169676e-07,
+      "loss": 0.2495,
+      "step": 16057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8756150438352185,
+      "learning_rate": 3.1252298140748374e-07,
+      "loss": 0.3744,
+      "step": 16058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.719734852026446,
+      "learning_rate": 3.120615532001836e-07,
+      "loss": 0.2874,
+      "step": 16059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3029303968028634,
+      "learning_rate": 3.1160046048577365e-07,
+      "loss": 0.2542,
+      "step": 16060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3691633990742467,
+      "learning_rate": 3.1113970328022433e-07,
+      "loss": 0.3045,
+      "step": 16061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3109473557320388,
+      "learning_rate": 3.106792815994908e-07,
+      "loss": 0.1067,
+      "step": 16062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.332168543197011,
+      "learning_rate": 3.1021919545951683e-07,
+      "loss": 0.2444,
+      "step": 16063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8139045767802737,
+      "learning_rate": 3.0975944487623534e-07,
+      "loss": 0.3702,
+      "step": 16064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33745728439735845,
+      "learning_rate": 3.093000298655668e-07,
+      "loss": 0.296,
+      "step": 16065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32467491783135444,
+      "learning_rate": 3.0884095044342197e-07,
+      "loss": 0.18,
+      "step": 16066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5142360344956464,
+      "learning_rate": 3.0838220662569807e-07,
+      "loss": 0.3945,
+      "step": 16067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23229234122279033,
+      "learning_rate": 3.0792379842828234e-07,
+      "loss": 0.2061,
+      "step": 16068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4327552771597201,
+      "learning_rate": 3.074657258670477e-07,
+      "loss": 0.1711,
+      "step": 16069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4375268155712646,
+      "learning_rate": 3.070079889578592e-07,
+      "loss": 0.3044,
+      "step": 16070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9263858301221611,
+      "learning_rate": 3.0655058771656755e-07,
+      "loss": 0.4369,
+      "step": 16071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2924196438960813,
+      "learning_rate": 3.060935221590111e-07,
+      "loss": 0.1915,
+      "step": 16072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3187869889458041,
+      "learning_rate": 3.056367923010195e-07,
+      "loss": 0.2959,
+      "step": 16073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2271257776392587,
+      "learning_rate": 3.0518039815841004e-07,
+      "loss": 0.13,
+      "step": 16074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3284032596538808,
+      "learning_rate": 3.0472433974698566e-07,
+      "loss": 0.2056,
+      "step": 16075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7342704365795844,
+      "learning_rate": 3.042686170825404e-07,
+      "loss": 0.3854,
+      "step": 16076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34661328109772904,
+      "learning_rate": 3.0381323018085496e-07,
+      "loss": 0.3171,
+      "step": 16077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.38368727591985285,
+      "learning_rate": 3.0335817905770115e-07,
+      "loss": 0.2618,
+      "step": 16078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5057502110746248,
+      "learning_rate": 3.029034637288375e-07,
+      "loss": 0.2623,
+      "step": 16079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2947538085012821,
+      "learning_rate": 3.024490842100092e-07,
+      "loss": 0.2066,
+      "step": 16080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26156274907930127,
+      "learning_rate": 3.019950405169514e-07,
+      "loss": 0.2431,
+      "step": 16081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5329165928014303,
+      "learning_rate": 3.015413326653893e-07,
+      "loss": 0.2872,
+      "step": 16082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.347908220713831,
+      "learning_rate": 3.0108796067103376e-07,
+      "loss": 0.5954,
+      "step": 16083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5714727630722087,
+      "learning_rate": 3.0063492454958434e-07,
+      "loss": 0.3317,
+      "step": 16084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2701745129360195,
+      "learning_rate": 3.001822243167307e-07,
+      "loss": 0.2408,
+      "step": 16085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43758200115949175,
+      "learning_rate": 2.997298599881493e-07,
+      "loss": 0.2901,
+      "step": 16086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.576522748883847,
+      "learning_rate": 2.992778315795064e-07,
+      "loss": 0.3509,
+      "step": 16087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23680880305406177,
+      "learning_rate": 2.988261391064551e-07,
+      "loss": 0.2101,
+      "step": 16088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4501988796850427,
+      "learning_rate": 2.9837478258463725e-07,
+      "loss": 0.324,
+      "step": 16089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5710226158588361,
+      "learning_rate": 2.9792376202968263e-07,
+      "loss": 0.3863,
+      "step": 16090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41143826535651795,
+      "learning_rate": 2.974730774572121e-07,
+      "loss": 0.2862,
+      "step": 16091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.45703676201711263,
+      "learning_rate": 2.970227288828309e-07,
+      "loss": 0.2569,
+      "step": 16092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30212625877525195,
+      "learning_rate": 2.965727163221366e-07,
+      "loss": 0.2634,
+      "step": 16093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2798986506628534,
+      "learning_rate": 2.9612303979071e-07,
+      "loss": 0.2054,
+      "step": 16094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.641861194235436,
+      "learning_rate": 2.9567369930412646e-07,
+      "loss": 0.3078,
+      "step": 16095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30785882672148424,
+      "learning_rate": 2.9522469487794467e-07,
+      "loss": 0.2824,
+      "step": 16096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.37031077069707474,
+      "learning_rate": 2.9477602652771554e-07,
+      "loss": 0.318,
+      "step": 16097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6730152886289593,
+      "learning_rate": 2.943276942689732e-07,
+      "loss": 0.2634,
+      "step": 16098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4854497688120956,
+      "learning_rate": 2.9387969811724757e-07,
+      "loss": 0.2925,
+      "step": 16099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35689041820120426,
+      "learning_rate": 2.9343203808804953e-07,
+      "loss": 0.252,
+      "step": 16100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29101679336873626,
+      "learning_rate": 2.9298471419688335e-07,
+      "loss": 0.2409,
+      "step": 16101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33551396449091,
+      "learning_rate": 2.925377264592388e-07,
+      "loss": 0.1526,
+      "step": 16102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.403795727148337,
+      "learning_rate": 2.9209107489059474e-07,
+      "loss": 0.2942,
+      "step": 16103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35408713603736885,
+      "learning_rate": 2.916447595064198e-07,
+      "loss": 0.2958,
+      "step": 16104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6365072948732012,
+      "learning_rate": 2.9119878032216944e-07,
+      "loss": 0.1797,
+      "step": 16105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2916939704666153,
+      "learning_rate": 2.90753137353289e-07,
+      "loss": 0.2252,
+      "step": 16106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.46903057760889666,
+      "learning_rate": 2.903078306152085e-07,
+      "loss": 0.2796,
+      "step": 16107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30356242185573823,
+      "learning_rate": 2.8986286012335105e-07,
+      "loss": 0.2131,
+      "step": 16108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3753404509166921,
+      "learning_rate": 2.8941822589312443e-07,
+      "loss": 0.3212,
+      "step": 16109
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6360470819564403,
+      "learning_rate": 2.889739279399262e-07,
+      "loss": 0.3638,
+      "step": 16110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33098743442000983,
+      "learning_rate": 2.885299662791452e-07,
+      "loss": 0.1529,
+      "step": 16111
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3058883589226098,
+      "learning_rate": 2.880863409261525e-07,
+      "loss": 0.28,
+      "step": 16112
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25707865836058724,
+      "learning_rate": 2.876430518963136e-07,
+      "loss": 0.1953,
+      "step": 16113
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6049082040792976,
+      "learning_rate": 2.872000992049773e-07,
+      "loss": 0.3962,
+      "step": 16114
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36179534195311464,
+      "learning_rate": 2.867574828674824e-07,
+      "loss": 0.2377,
+      "step": 16115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35134754663657647,
+      "learning_rate": 2.8631520289916004e-07,
+      "loss": 0.305,
+      "step": 16116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6873273078284741,
+      "learning_rate": 2.858732593153246e-07,
+      "loss": 0.2425,
+      "step": 16117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4355865580407592,
+      "learning_rate": 2.8543165213128057e-07,
+      "loss": 0.2198,
+      "step": 16118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2721525238168374,
+      "learning_rate": 2.8499038136231894e-07,
+      "loss": 0.2223,
+      "step": 16119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3305990370089696,
+      "learning_rate": 2.845494470237242e-07,
+      "loss": 0.2855,
+      "step": 16120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3257204072520625,
+      "learning_rate": 2.841088491307642e-07,
+      "loss": 0.2172,
+      "step": 16121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7299514903666677,
+      "learning_rate": 2.8366858769869663e-07,
+      "loss": 0.4215,
+      "step": 16122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0502896884292527,
+      "learning_rate": 2.8322866274276715e-07,
+      "loss": 0.3683,
+      "step": 16123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2414259325709617,
+      "learning_rate": 2.8278907427821245e-07,
+      "loss": 0.223,
+      "step": 16124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4580905634662621,
+      "learning_rate": 2.8234982232025365e-07,
+      "loss": 0.2462,
+      "step": 16125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6134757196152566,
+      "learning_rate": 2.8191090688410305e-07,
+      "loss": 0.3354,
+      "step": 16126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34813149354241363,
+      "learning_rate": 2.8147232798496074e-07,
+      "loss": 0.2758,
+      "step": 16127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2445101180031528,
+      "learning_rate": 2.810340856380112e-07,
+      "loss": 0.1985,
+      "step": 16128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9321604201835165,
+      "learning_rate": 2.8059617985843557e-07,
+      "loss": 0.3523,
+      "step": 16129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36475217838282614,
+      "learning_rate": 2.801586106613963e-07,
+      "loss": 0.2633,
+      "step": 16130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9028096980700738,
+      "learning_rate": 2.797213780620456e-07,
+      "loss": 0.2922,
+      "step": 16131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2847058382844327,
+      "learning_rate": 2.7928448207552474e-07,
+      "loss": 0.2463,
+      "step": 16132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38924969775553325,
+      "learning_rate": 2.7884792271696603e-07,
+      "loss": 0.3169,
+      "step": 16133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24599208379016946,
+      "learning_rate": 2.784117000014852e-07,
+      "loss": 0.1124,
+      "step": 16134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6790767935493142,
+      "learning_rate": 2.7797581394418907e-07,
+      "loss": 0.3211,
+      "step": 16135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.308392724398636,
+      "learning_rate": 2.7754026456017104e-07,
+      "loss": 0.2932,
+      "step": 16136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4612725789887395,
+      "learning_rate": 2.7710505186451684e-07,
+      "loss": 0.2619,
+      "step": 16137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6119165445617544,
+      "learning_rate": 2.766701758722967e-07,
+      "loss": 0.3625,
+      "step": 16138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3401083159351843,
+      "learning_rate": 2.7623563659857186e-07,
+      "loss": 0.3047,
+      "step": 16139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22356893956156418,
+      "learning_rate": 2.758014340583859e-07,
+      "loss": 0.216,
+      "step": 16140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9203710855895133,
+      "learning_rate": 2.7536756826678e-07,
+      "loss": 0.199,
+      "step": 16141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3429263728146882,
+      "learning_rate": 2.749340392387767e-07,
+      "loss": 0.2606,
+      "step": 16142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6493811930991484,
+      "learning_rate": 2.745008469893884e-07,
+      "loss": 0.3888,
+      "step": 16143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3245298437449268,
+      "learning_rate": 2.740679915336175e-07,
+      "loss": 0.2407,
+      "step": 16144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3407142144619474,
+      "learning_rate": 2.7363547288645544e-07,
+      "loss": 0.3001,
+      "step": 16145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4127223640520308,
+      "learning_rate": 2.732032910628779e-07,
+      "loss": 0.2387,
+      "step": 16146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4410479453956271,
+      "learning_rate": 2.727714460778519e-07,
+      "loss": 0.0986,
+      "step": 16147
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23114652387431825,
+      "learning_rate": 2.7233993794633205e-07,
+      "loss": 0.2363,
+      "step": 16148
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7707876240880364,
+      "learning_rate": 2.7190876668326207e-07,
+      "loss": 0.4322,
+      "step": 16149
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5681333562696601,
+      "learning_rate": 2.7147793230357434e-07,
+      "loss": 0.3007,
+      "step": 16150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.334019256141992,
+      "learning_rate": 2.71047434822187e-07,
+      "loss": 0.2636,
+      "step": 16151
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25145171079300954,
+      "learning_rate": 2.7061727425400695e-07,
+      "loss": 0.2368,
+      "step": 16152
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37370438219164975,
+      "learning_rate": 2.701874506139335e-07,
+      "loss": 0.2354,
+      "step": 16153
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32680459189413763,
+      "learning_rate": 2.6975796391685125e-07,
+      "loss": 0.1946,
+      "step": 16154
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5853293839882767,
+      "learning_rate": 2.6932881417763067e-07,
+      "loss": 0.342,
+      "step": 16155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3247975509729384,
+      "learning_rate": 2.689000014111365e-07,
+      "loss": 0.3094,
+      "step": 16156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33160684037157295,
+      "learning_rate": 2.684715256322146e-07,
+      "loss": 0.2061,
+      "step": 16157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36609071143521194,
+      "learning_rate": 2.680433868557064e-07,
+      "loss": 0.2004,
+      "step": 16158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2305571895568488,
+      "learning_rate": 2.676155850964379e-07,
+      "loss": 0.4733,
+      "step": 16159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22411989695587825,
+      "learning_rate": 2.6718812036922283e-07,
+      "loss": 0.2119,
+      "step": 16160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6334435102540285,
+      "learning_rate": 2.6676099268886367e-07,
+      "loss": 0.3685,
+      "step": 16161
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2434403340381204,
+      "learning_rate": 2.6633420207015426e-07,
+      "loss": 0.8358,
+      "step": 16162
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2768553025982602,
+      "learning_rate": 2.659077485278716e-07,
+      "loss": 0.1891,
+      "step": 16163
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8087700309256254,
+      "learning_rate": 2.6548163207678614e-07,
+      "loss": 0.3039,
+      "step": 16164
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3263779764663397,
+      "learning_rate": 2.6505585273165156e-07,
+      "loss": 0.1664,
+      "step": 16165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31180764115083154,
+      "learning_rate": 2.6463041050721615e-07,
+      "loss": 0.2553,
+      "step": 16166
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8241373179238439,
+      "learning_rate": 2.642053054182103e-07,
+      "loss": 0.3087,
+      "step": 16167
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3446713828640923,
+      "learning_rate": 2.637805374793556e-07,
+      "loss": 0.3118,
+      "step": 16168
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32013571828818826,
+      "learning_rate": 2.633561067053625e-07,
+      "loss": 0.2598,
+      "step": 16169
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4038815496361895,
+      "learning_rate": 2.629320131109281e-07,
+      "loss": 0.1137,
+      "step": 16170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2753926358926855,
+      "learning_rate": 2.6250825671074065e-07,
+      "loss": 0.2389,
+      "step": 16171
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4119719794526924,
+      "learning_rate": 2.6208483751947286e-07,
+      "loss": 0.2821,
+      "step": 16172
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47065265482665175,
+      "learning_rate": 2.6166175555178864e-07,
+      "loss": 0.2604,
+      "step": 16173
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7827261475775145,
+      "learning_rate": 2.612390108223395e-07,
+      "loss": 0.579,
+      "step": 16174
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33742011781888565,
+      "learning_rate": 2.6081660334576376e-07,
+      "loss": 0.2539,
+      "step": 16175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35489690817213293,
+      "learning_rate": 2.6039453313669084e-07,
+      "loss": 0.224,
+      "step": 16176
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40363854862203224,
+      "learning_rate": 2.599728002097346e-07,
+      "loss": 0.2276,
+      "step": 16177
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2998525771419882,
+      "learning_rate": 2.5955140457950334e-07,
+      "loss": 0.254,
+      "step": 16178
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3225328457185091,
+      "learning_rate": 2.591303462605876e-07,
+      "loss": 0.2568,
+      "step": 16179
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40344121357429547,
+      "learning_rate": 2.587096252675703e-07,
+      "loss": 0.2584,
+      "step": 16180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4726299423223656,
+      "learning_rate": 2.5828924161501847e-07,
+      "loss": 0.2424,
+      "step": 16181
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2427679629599802,
+      "learning_rate": 2.5786919531749056e-07,
+      "loss": 0.5494,
+      "step": 16182
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3302371945217039,
+      "learning_rate": 2.5744948638953495e-07,
+      "loss": 0.2585,
+      "step": 16183
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2960406165691513,
+      "learning_rate": 2.570301148456833e-07,
+      "loss": 0.246,
+      "step": 16184
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22857777782773264,
+      "learning_rate": 2.566110807004618e-07,
+      "loss": 0.1342,
+      "step": 16185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9455741587639533,
+      "learning_rate": 2.5619238396837665e-07,
+      "loss": 0.5122,
+      "step": 16186
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3025016716582683,
+      "learning_rate": 2.5577402466393286e-07,
+      "loss": 0.238,
+      "step": 16187
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39578317038127414,
+      "learning_rate": 2.5535600280161444e-07,
+      "loss": 0.3052,
+      "step": 16188
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5815913779636789,
+      "learning_rate": 2.549383183958998e-07,
+      "loss": 0.3511,
+      "step": 16189
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38518298786026156,
+      "learning_rate": 2.5452097146125063e-07,
+      "loss": 0.1683,
+      "step": 16190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22205133621720435,
+      "learning_rate": 2.5410396201212105e-07,
+      "loss": 0.2007,
+      "step": 16191
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5057323622641279,
+      "learning_rate": 2.536872900629539e-07,
+      "loss": 0.3944,
+      "step": 16192
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3595273845184508,
+      "learning_rate": 2.5327095562817537e-07,
+      "loss": 0.1673,
+      "step": 16193
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5360787331974713,
+      "learning_rate": 2.528549587222051e-07,
+      "loss": 0.3074,
+      "step": 16194
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.442058729074706,
+      "learning_rate": 2.524392993594482e-07,
+      "loss": 0.3311,
+      "step": 16195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2855255735605113,
+      "learning_rate": 2.5202397755429876e-07,
+      "loss": 0.2012,
+      "step": 16196
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23458277362119945,
+      "learning_rate": 2.5160899332114075e-07,
+      "loss": 0.1566,
+      "step": 16197
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.307433304730941,
+      "learning_rate": 2.5119434667434384e-07,
+      "loss": 0.7377,
+      "step": 16198
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3418249915562307,
+      "learning_rate": 2.507800376282665e-07,
+      "loss": 0.1928,
+      "step": 16199
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3322723765898175,
+      "learning_rate": 2.5036606619725847e-07,
+      "loss": 0.3077,
+      "step": 16200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7679882980968665,
+      "learning_rate": 2.499524323956537e-07,
+      "loss": 0.4053,
+      "step": 16201
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5105729769391152,
+      "learning_rate": 2.495391362377775e-07,
+      "loss": 0.2702,
+      "step": 16202
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20248897495158377,
+      "learning_rate": 2.4912617773794057e-07,
+      "loss": 0.16,
+      "step": 16203
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45523324068273235,
+      "learning_rate": 2.4871355691044595e-07,
+      "loss": 0.3446,
+      "step": 16204
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31116609474106843,
+      "learning_rate": 2.4830127376958113e-07,
+      "loss": 0.2482,
+      "step": 16205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7147870871308801,
+      "learning_rate": 2.4788932832962465e-07,
+      "loss": 0.2833,
+      "step": 16206
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36457125106980254,
+      "learning_rate": 2.4747772060484064e-07,
+      "loss": 0.2912,
+      "step": 16207
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5240461723284267,
+      "learning_rate": 2.470664506094844e-07,
+      "loss": 0.2765,
+      "step": 16208
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22721409677761747,
+      "learning_rate": 2.466555183577968e-07,
+      "loss": 0.146,
+      "step": 16209
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8567658532302702,
+      "learning_rate": 2.462449238640097e-07,
+      "loss": 0.5097,
+      "step": 16210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37293244224659416,
+      "learning_rate": 2.458346671423406e-07,
+      "loss": 0.2645,
+      "step": 16211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27145054299585214,
+      "learning_rate": 2.4542474820699823e-07,
+      "loss": 0.244,
+      "step": 16212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3981516737715696,
+      "learning_rate": 2.450151670721768e-07,
+      "loss": 0.4875,
+      "step": 16213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5986289510427129,
+      "learning_rate": 2.446059237520615e-07,
+      "loss": 0.2993,
+      "step": 16214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22481400806917337,
+      "learning_rate": 2.441970182608222e-07,
+      "loss": 0.2347,
+      "step": 16215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30778683391514594,
+      "learning_rate": 2.43788450612622e-07,
+      "loss": 0.1411,
+      "step": 16216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.41000123055651855,
+      "learning_rate": 2.433802208216085e-07,
+      "loss": 0.2763,
+      "step": 16217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4744266292202893,
+      "learning_rate": 2.42972328901917e-07,
+      "loss": 0.3694,
+      "step": 16218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3541890062856887,
+      "learning_rate": 2.4256477486767517e-07,
+      "loss": 0.2324,
+      "step": 16219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.41346592434653073,
+      "learning_rate": 2.4215755873299497e-07,
+      "loss": 0.2446,
+      "step": 16220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5884279680128669,
+      "learning_rate": 2.4175068051197957e-07,
+      "loss": 0.3041,
+      "step": 16221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4402183600759972,
+      "learning_rate": 2.413441402187178e-07,
+      "loss": 0.2166,
+      "step": 16222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2384371145087381,
+      "learning_rate": 2.4093793786728934e-07,
+      "loss": 0.2403,
+      "step": 16223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4502437190455561,
+      "learning_rate": 2.4053207347175976e-07,
+      "loss": 0.3311,
+      "step": 16224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4684769525518414,
+      "learning_rate": 2.401265470461855e-07,
+      "loss": 0.12,
+      "step": 16225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5547322450352692,
+      "learning_rate": 2.397213586046099e-07,
+      "loss": 0.2999,
+      "step": 16226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2818633700967168,
+      "learning_rate": 2.3931650816106267e-07,
+      "loss": 0.2502,
+      "step": 16227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4363782611754942,
+      "learning_rate": 2.3891199572956493e-07,
+      "loss": 0.3209,
+      "step": 16228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47684193737555375,
+      "learning_rate": 2.3850782132412544e-07,
+      "loss": 0.2063,
+      "step": 16229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3005978735016527,
+      "learning_rate": 2.3810398495874076e-07,
+      "loss": 0.2733,
+      "step": 16230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2758614772557928,
+      "learning_rate": 2.377004866473953e-07,
+      "loss": 0.252,
+      "step": 16231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5232494032312904,
+      "learning_rate": 2.3729732640406233e-07,
+      "loss": 0.0864,
+      "step": 16232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43196860650976154,
+      "learning_rate": 2.3689450424270179e-07,
+      "loss": 0.2845,
+      "step": 16233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6300939297391976,
+      "learning_rate": 2.364920201772658e-07,
+      "loss": 0.3533,
+      "step": 16234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24702854532908614,
+      "learning_rate": 2.3608987422169106e-07,
+      "loss": 0.2239,
+      "step": 16235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5147166917845664,
+      "learning_rate": 2.3568806638990527e-07,
+      "loss": 0.384,
+      "step": 16236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34917477325603674,
+      "learning_rate": 2.3528659669581954e-07,
+      "loss": 0.1735,
+      "step": 16237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35170576502926393,
+      "learning_rate": 2.3488546515334054e-07,
+      "loss": 0.1596,
+      "step": 16238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35057358099260666,
+      "learning_rate": 2.3448467177635826e-07,
+      "loss": 0.291,
+      "step": 16239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43437427364444037,
+      "learning_rate": 2.3408421657875158e-07,
+      "loss": 0.3281,
+      "step": 16240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4994355097839041,
+      "learning_rate": 2.3368409957438832e-07,
+      "loss": 0.3334,
+      "step": 16241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3989427545060095,
+      "learning_rate": 2.3328432077712516e-07,
+      "loss": 0.2308,
+      "step": 16242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22736027420925461,
+      "learning_rate": 2.3288488020080546e-07,
+      "loss": 0.2094,
+      "step": 16243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.355549695082054,
+      "learning_rate": 2.3248577785926264e-07,
+      "loss": 0.3942,
+      "step": 16244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3061093657317648,
+      "learning_rate": 2.3208701376631672e-07,
+      "loss": 0.2012,
+      "step": 16245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4283494643377407,
+      "learning_rate": 2.3168858793577774e-07,
+      "loss": 0.3167,
+      "step": 16246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40586491113206713,
+      "learning_rate": 2.3129050038144362e-07,
+      "loss": 0.3005,
+      "step": 16247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30763307469813905,
+      "learning_rate": 2.3089275111709886e-07,
+      "loss": 0.2203,
+      "step": 16248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2669467232153082,
+      "learning_rate": 2.3049534015651686e-07,
+      "loss": 0.1807,
+      "step": 16249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7019907586813147,
+      "learning_rate": 2.3009826751346332e-07,
+      "loss": 0.3552,
+      "step": 16250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2421896836266599,
+      "learning_rate": 2.2970153320168498e-07,
+      "loss": 0.2243,
+      "step": 16251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8531496453021868,
+      "learning_rate": 2.293051372349231e-07,
+      "loss": 0.404,
+      "step": 16252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7552706084359576,
+      "learning_rate": 2.2890907962690335e-07,
+      "loss": 0.548,
+      "step": 16253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3945319567289995,
+      "learning_rate": 2.2851336039134363e-07,
+      "loss": 0.2455,
+      "step": 16254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21129773765826465,
+      "learning_rate": 2.2811797954194527e-07,
+      "loss": 0.172,
+      "step": 16255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5848773278732337,
+      "learning_rate": 2.2772293709240168e-07,
+      "loss": 0.3105,
+      "step": 16256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39209953000005016,
+      "learning_rate": 2.2732823305639197e-07,
+      "loss": 0.3028,
+      "step": 16257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45565939341834677,
+      "learning_rate": 2.2693386744758405e-07,
+      "loss": 0.2577,
+      "step": 16258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3638561329873051,
+      "learning_rate": 2.2653984027963704e-07,
+      "loss": 0.2771,
+      "step": 16259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3977592178558374,
+      "learning_rate": 2.2614615156619556e-07,
+      "loss": 0.3164,
+      "step": 16260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2486224244899392,
+      "learning_rate": 2.2575280132089315e-07,
+      "loss": 0.0902,
+      "step": 16261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.41724306546277456,
+      "learning_rate": 2.2535978955734895e-07,
+      "loss": 0.2994,
+      "step": 16262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26274641689573147,
+      "learning_rate": 2.2496711628917644e-07,
+      "loss": 0.2548,
+      "step": 16263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7297326479778752,
+      "learning_rate": 2.2457478152997148e-07,
+      "loss": 0.4205,
+      "step": 16264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0018169427080976,
+      "learning_rate": 2.2418278529332203e-07,
+      "loss": 0.4632,
+      "step": 16265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3327648178772939,
+      "learning_rate": 2.2379112759280176e-07,
+      "loss": 0.2609,
+      "step": 16266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23721807132324804,
+      "learning_rate": 2.2339980844197527e-07,
+      "loss": 0.2063,
+      "step": 16267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5633019302848433,
+      "learning_rate": 2.2300882785439183e-07,
+      "loss": 0.2385,
+      "step": 16268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39111215606894656,
+      "learning_rate": 2.2261818584359386e-07,
+      "loss": 0.2894,
+      "step": 16269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4038427231112138,
+      "learning_rate": 2.222278824231061e-07,
+      "loss": 0.4999,
+      "step": 16270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2967884736462052,
+      "learning_rate": 2.2183791760644668e-07,
+      "loss": 0.2486,
+      "step": 16271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34349186837418483,
+      "learning_rate": 2.214482914071203e-07,
+      "loss": 0.2819,
+      "step": 16272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8920189124308012,
+      "learning_rate": 2.2105900383861956e-07,
+      "loss": 0.4225,
+      "step": 16273
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28684107801549114,
+      "learning_rate": 2.2067005491442362e-07,
+      "loss": 0.244,
+      "step": 16274
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3494953028521881,
+      "learning_rate": 2.2028144464800393e-07,
+      "loss": 0.275,
+      "step": 16275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35670384309614495,
+      "learning_rate": 2.1989317305281755e-07,
+      "loss": 0.179,
+      "step": 16276
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.996066078136281,
+      "learning_rate": 2.1950524014231033e-07,
+      "loss": 0.4171,
+      "step": 16277
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.347523126063495,
+      "learning_rate": 2.191176459299138e-07,
+      "loss": 0.1977,
+      "step": 16278
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32388881265928143,
+      "learning_rate": 2.1873039042905497e-07,
+      "loss": 0.2703,
+      "step": 16279
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39745643594948593,
+      "learning_rate": 2.183434736531409e-07,
+      "loss": 0.2683,
+      "step": 16280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22006091226414445,
+      "learning_rate": 2.1795689561557308e-07,
+      "loss": 0.152,
+      "step": 16281
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34373104169964486,
+      "learning_rate": 2.175706563297353e-07,
+      "loss": 0.2868,
+      "step": 16282
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.263619920512286,
+      "learning_rate": 2.1718475580900567e-07,
+      "loss": 0.7482,
+      "step": 16283
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29616054478209414,
+      "learning_rate": 2.16799194066748e-07,
+      "loss": 0.1906,
+      "step": 16284
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5626197863716071,
+      "learning_rate": 2.164139711163138e-07,
+      "loss": 0.3607,
+      "step": 16285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3590447014920199,
+      "learning_rate": 2.160290869710424e-07,
+      "loss": 0.3023,
+      "step": 16286
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22648512679657767,
+      "learning_rate": 2.1564454164426207e-07,
+      "loss": 0.1571,
+      "step": 16287
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4333570135949622,
+      "learning_rate": 2.1526033514929213e-07,
+      "loss": 0.2604,
+      "step": 16288
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.220905292397416,
+      "learning_rate": 2.1487646749943524e-07,
+      "loss": 0.5481,
+      "step": 16289
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32203788282666024,
+      "learning_rate": 2.1449293870798637e-07,
+      "loss": 0.2688,
+      "step": 16290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34693965055254655,
+      "learning_rate": 2.1410974878822487e-07,
+      "loss": 0.2552,
+      "step": 16291
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.551396319559656,
+      "learning_rate": 2.1372689775342238e-07,
+      "loss": 0.2369,
+      "step": 16292
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30887122673256734,
+      "learning_rate": 2.1334438561683713e-07,
+      "loss": 0.2373,
+      "step": 16293
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22820054330620598,
+      "learning_rate": 2.1296221239171523e-07,
+      "loss": 0.1915,
+      "step": 16294
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.137231060487121,
+      "learning_rate": 2.125803780912894e-07,
+      "loss": 0.7172,
+      "step": 16295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.40545905296169865,
+      "learning_rate": 2.1219888272878575e-07,
+      "loss": 0.2602,
+      "step": 16296
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4730775018614647,
+      "learning_rate": 2.118177263174137e-07,
+      "loss": 0.2668,
+      "step": 16297
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34205215126746363,
+      "learning_rate": 2.114369088703727e-07,
+      "loss": 0.2772,
+      "step": 16298
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32729918038453004,
+      "learning_rate": 2.1105643040085112e-07,
+      "loss": 0.2464,
+      "step": 16299
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24639056641925142,
+      "learning_rate": 2.106762909220228e-07,
+      "loss": 0.1184,
+      "step": 16300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.99073060626058,
+      "learning_rate": 2.10296490447055e-07,
+      "loss": 0.4329,
+      "step": 16301
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32121737540817974,
+      "learning_rate": 2.0991702898909838e-07,
+      "loss": 0.2518,
+      "step": 16302
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3301117105900323,
+      "learning_rate": 2.0953790656129457e-07,
+      "loss": 0.3058,
+      "step": 16303
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7456922022694272,
+      "learning_rate": 2.091591231767709e-07,
+      "loss": 0.1374,
+      "step": 16304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32467499870042676,
+      "learning_rate": 2.0878067884864683e-07,
+      "loss": 0.2509,
+      "step": 16305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39956950076075687,
+      "learning_rate": 2.0840257359002635e-07,
+      "loss": 0.2391,
+      "step": 16306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3293116288624257,
+      "learning_rate": 2.0802480741400456e-07,
+      "loss": 0.2718,
+      "step": 16307
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3131425030863296,
+      "learning_rate": 2.0764738033366095e-07,
+      "loss": 0.2571,
+      "step": 16308
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44822453408391827,
+      "learning_rate": 2.0727029236206953e-07,
+      "loss": 0.237,
+      "step": 16309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3492819711372927,
+      "learning_rate": 2.068935435122854e-07,
+      "loss": 0.2403,
+      "step": 16310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3019190826036198,
+      "learning_rate": 2.0651713379735706e-07,
+      "loss": 0.2593,
+      "step": 16311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4584884405827178,
+      "learning_rate": 2.0614106323031846e-07,
+      "loss": 0.2696,
+      "step": 16312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.411254505830708,
+      "learning_rate": 2.0576533182419477e-07,
+      "loss": 0.2811,
+      "step": 16313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3902520439694165,
+      "learning_rate": 2.053899395919956e-07,
+      "loss": 0.2648,
+      "step": 16314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.353453516685148,
+      "learning_rate": 2.0501488654672276e-07,
+      "loss": 0.2453,
+      "step": 16315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1755006016717067,
+      "learning_rate": 2.0464017270136139e-07,
+      "loss": 0.4489,
+      "step": 16316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36115634958306614,
+      "learning_rate": 2.0426579806889114e-07,
+      "loss": 0.1607,
+      "step": 16317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2925206750954721,
+      "learning_rate": 2.0389176266227494e-07,
+      "loss": 0.3033,
+      "step": 16318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7559070874858043,
+      "learning_rate": 2.0351806649446582e-07,
+      "loss": 0.4248,
+      "step": 16319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3267369295851818,
+      "learning_rate": 2.0314470957840337e-07,
+      "loss": 0.2155,
+      "step": 16320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27187878348807465,
+      "learning_rate": 2.0277169192701951e-07,
+      "loss": 0.1862,
+      "step": 16321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3450218801283784,
+      "learning_rate": 2.0239901355323166e-07,
+      "loss": 0.2994,
+      "step": 16322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33760499589787035,
+      "learning_rate": 2.0202667446994396e-07,
+      "loss": 0.1877,
+      "step": 16323
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3079162281268204,
+      "learning_rate": 2.016546746900505e-07,
+      "loss": 0.5858,
+      "step": 16324
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5508684131706696,
+      "learning_rate": 2.0128301422643437e-07,
+      "loss": 0.3595,
+      "step": 16325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22361769791366404,
+      "learning_rate": 2.0091169309196635e-07,
+      "loss": 0.219,
+      "step": 16326
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27103442373090697,
+      "learning_rate": 2.0054071129950503e-07,
+      "loss": 0.1882,
+      "step": 16327
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4862555952033187,
+      "learning_rate": 2.0017006886189793e-07,
+      "loss": 0.5859,
+      "step": 16328
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43694139275442706,
+      "learning_rate": 1.9979976579197924e-07,
+      "loss": 0.2516,
+      "step": 16329
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2777763051503804,
+      "learning_rate": 1.9942980210257313e-07,
+      "loss": 0.2499,
+      "step": 16330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5401113539028809,
+      "learning_rate": 1.9906017780649267e-07,
+      "loss": 0.3562,
+      "step": 16331
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5545585769788984,
+      "learning_rate": 1.9869089291653544e-07,
+      "loss": 0.3261,
+      "step": 16332
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24578812393967553,
+      "learning_rate": 1.9832194744549117e-07,
+      "loss": 0.1533,
+      "step": 16333
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3471473882109775,
+      "learning_rate": 1.9795334140613741e-07,
+      "loss": 0.2771,
+      "step": 16334
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7065238619042722,
+      "learning_rate": 1.9758507481123734e-07,
+      "loss": 0.3106,
+      "step": 16335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37428051274745233,
+      "learning_rate": 1.9721714767354516e-07,
+      "loss": 0.2519,
+      "step": 16336
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8941617265103616,
+      "learning_rate": 1.968495600058018e-07,
+      "loss": 0.5307,
+      "step": 16337
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31408047918318077,
+      "learning_rate": 1.9648231182073484e-07,
+      "loss": 0.2839,
+      "step": 16338
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21954413848341273,
+      "learning_rate": 1.9611540313106526e-07,
+      "loss": 0.1522,
+      "step": 16339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4172099353996386,
+      "learning_rate": 1.957488339494973e-07,
+      "loss": 0.4308,
+      "step": 16340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4576229631078692,
+      "learning_rate": 1.9538260428872636e-07,
+      "loss": 0.2894,
+      "step": 16341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29422998452537247,
+      "learning_rate": 1.9501671416143342e-07,
+      "loss": 0.2752,
+      "step": 16342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7309287759530217,
+      "learning_rate": 1.9465116358029057e-07,
+      "loss": 0.3446,
+      "step": 16343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3900915110844801,
+      "learning_rate": 1.9428595255795657e-07,
+      "loss": 0.2888,
+      "step": 16344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37353761465583235,
+      "learning_rate": 1.9392108110707686e-07,
+      "loss": 0.2302,
+      "step": 16345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3062970278982408,
+      "learning_rate": 1.935565492402891e-07,
+      "loss": 0.2056,
+      "step": 16346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5981284968410989,
+      "learning_rate": 1.9319235697021766e-07,
+      "loss": 0.3264,
+      "step": 16347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35572231889161077,
+      "learning_rate": 1.9282850430947242e-07,
+      "loss": 0.3058,
+      "step": 16348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3250188011219933,
+      "learning_rate": 1.9246499127065333e-07,
+      "loss": 0.2664,
+      "step": 16349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5737774291728073,
+      "learning_rate": 1.9210181786635028e-07,
+      "loss": 0.2784,
+      "step": 16350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4012088600693107,
+      "learning_rate": 1.9173898410913995e-07,
+      "loss": 0.2982,
+      "step": 16351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29848629529359066,
+      "learning_rate": 1.9137649001158665e-07,
+      "loss": 0.1829,
+      "step": 16352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48598014526128647,
+      "learning_rate": 1.9101433558624483e-07,
+      "loss": 0.2018,
+      "step": 16353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.26365262037215137,
+      "learning_rate": 1.9065252084565222e-07,
+      "loss": 0.2652,
+      "step": 16354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2940730599519867,
+      "learning_rate": 1.9029104580234325e-07,
+      "loss": 0.7013,
+      "step": 16355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6116318818103755,
+      "learning_rate": 1.8992991046883236e-07,
+      "loss": 0.1767,
+      "step": 16356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23131950704158802,
+      "learning_rate": 1.895691148576273e-07,
+      "loss": 0.1767,
+      "step": 16357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3293918268569444,
+      "learning_rate": 1.8920865898122143e-07,
+      "loss": 0.2735,
+      "step": 16358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4511372616577314,
+      "learning_rate": 1.888485428520992e-07,
+      "loss": 0.2337,
+      "step": 16359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3753456647512689,
+      "learning_rate": 1.884887664827284e-07,
+      "loss": 0.2896,
+      "step": 16360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44961231456689443,
+      "learning_rate": 1.881293298855713e-07,
+      "loss": 0.3533,
+      "step": 16361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28662578208318146,
+      "learning_rate": 1.877702330730724e-07,
+      "loss": 0.1942,
+      "step": 16362
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6551363416683533,
+      "learning_rate": 1.874114760576684e-07,
+      "loss": 0.3072,
+      "step": 16363
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32419575634159736,
+      "learning_rate": 1.870530588517827e-07,
+      "loss": 0.1952,
+      "step": 16364
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3382957834583971,
+      "learning_rate": 1.8669498146782871e-07,
+      "loss": 0.2804,
+      "step": 16365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27619581453105596,
+      "learning_rate": 1.863372439182054e-07,
+      "loss": 0.2148,
+      "step": 16366
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2113862017013661,
+      "learning_rate": 1.8597984621530063e-07,
+      "loss": 0.8154,
+      "step": 16367
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.391102015454029,
+      "learning_rate": 1.8562278837149228e-07,
+      "loss": 0.4795,
+      "step": 16368
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3298554751561711,
+      "learning_rate": 1.852660703991438e-07,
+      "loss": 0.1875,
+      "step": 16369
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33732172554386936,
+      "learning_rate": 1.8490969231061085e-07,
+      "loss": 0.3013,
+      "step": 16370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41691582029462937,
+      "learning_rate": 1.8455365411823134e-07,
+      "loss": 0.2349,
+      "step": 16371
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23903592330484194,
+      "learning_rate": 1.8419795583433763e-07,
+      "loss": 0.1662,
+      "step": 16372
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48747978730964747,
+      "learning_rate": 1.8384259747124766e-07,
+      "loss": 0.3685,
+      "step": 16373
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4383352697210021,
+      "learning_rate": 1.834875790412649e-07,
+      "loss": 0.3244,
+      "step": 16374
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34083610981257056,
+      "learning_rate": 1.831329005566851e-07,
+      "loss": 0.2185,
+      "step": 16375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6216981447352833,
+      "learning_rate": 1.8277856202979282e-07,
+      "loss": 0.3541,
+      "step": 16376
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2519742262206242,
+      "learning_rate": 1.8242456347285498e-07,
+      "loss": 0.2207,
+      "step": 16377
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27485795073992864,
+      "learning_rate": 1.8207090489813284e-07,
+      "loss": 0.2107,
+      "step": 16378
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9112149768074835,
+      "learning_rate": 1.8171758631787327e-07,
+      "loss": 0.5171,
+      "step": 16379
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6028629951517451,
+      "learning_rate": 1.8136460774431097e-07,
+      "loss": 0.2255,
+      "step": 16380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3830929246588872,
+      "learning_rate": 1.8101196918967056e-07,
+      "loss": 0.2811,
+      "step": 16381
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3110335698157431,
+      "learning_rate": 1.806596706661634e-07,
+      "loss": 0.2559,
+      "step": 16382
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3532742177417911,
+      "learning_rate": 1.8030771218598863e-07,
+      "loss": 0.2153,
+      "step": 16383
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27118052800532905,
+      "learning_rate": 1.799560937613365e-07,
+      "loss": 0.2216,
+      "step": 16384
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30009922060906546,
+      "learning_rate": 1.7960481540438278e-07,
+      "loss": 0.259,
+      "step": 16385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2439511677777768,
+      "learning_rate": 1.7925387712729113e-07,
+      "loss": 0.3727,
+      "step": 16386
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3074681980467877,
+      "learning_rate": 1.7890327894221515e-07,
+      "loss": 0.2405,
+      "step": 16387
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6218933087042393,
+      "learning_rate": 1.7855302086129734e-07,
+      "loss": 0.2886,
+      "step": 16388
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25243688522148056,
+      "learning_rate": 1.7820310289666577e-07,
+      "loss": 0.2225,
+      "step": 16389
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24248323345775227,
+      "learning_rate": 1.7785352506043852e-07,
+      "loss": 0.2009,
+      "step": 16390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.140361489090267,
+      "learning_rate": 1.7750428736472146e-07,
+      "loss": 0.7097,
+      "step": 16391
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6494164060996802,
+      "learning_rate": 1.7715538982160717e-07,
+      "loss": 0.2132,
+      "step": 16392
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29027986290742447,
+      "learning_rate": 1.7680683244318154e-07,
+      "loss": 0.274,
+      "step": 16393
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4739590651850567,
+      "learning_rate": 1.7645861524151152e-07,
+      "loss": 0.3406,
+      "step": 16394
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.47333802545971043,
+      "learning_rate": 1.7611073822865753e-07,
+      "loss": 0.1103,
+      "step": 16395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34722614276277375,
+      "learning_rate": 1.7576320141666548e-07,
+      "loss": 0.297,
+      "step": 16396
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4741094317552018,
+      "learning_rate": 1.7541600481757238e-07,
+      "loss": 0.3714,
+      "step": 16397
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30253332304639474,
+      "learning_rate": 1.7506914844340084e-07,
+      "loss": 0.2009,
+      "step": 16398
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3499583597663478,
+      "learning_rate": 1.7472263230616126e-07,
+      "loss": 0.2693,
+      "step": 16399
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9313377074422858,
+      "learning_rate": 1.7437645641785404e-07,
+      "loss": 0.4164,
+      "step": 16400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3084187512136342,
+      "learning_rate": 1.7403062079046851e-07,
+      "loss": 0.2116,
+      "step": 16401
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3905940549961325,
+      "learning_rate": 1.736851254359795e-07,
+      "loss": 0.2868,
+      "step": 16402
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45116394117225134,
+      "learning_rate": 1.7333997036635296e-07,
+      "loss": 0.3161,
+      "step": 16403
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.587337210292463,
+      "learning_rate": 1.7299515559354052e-07,
+      "loss": 0.3472,
+      "step": 16404
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24303637781952572,
+      "learning_rate": 1.7265068112948257e-07,
+      "loss": 0.1978,
+      "step": 16405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3114290454839184,
+      "learning_rate": 1.7230654698610848e-07,
+      "loss": 0.2325,
+      "step": 16406
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4056712180482611,
+      "learning_rate": 1.7196275317533761e-07,
+      "loss": 0.4493,
+      "step": 16407
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28533591224264054,
+      "learning_rate": 1.7161929970907266e-07,
+      "loss": 0.2084,
+      "step": 16408
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33232194989980035,
+      "learning_rate": 1.7127618659920963e-07,
+      "loss": 0.2976,
+      "step": 16409
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5558040330434648,
+      "learning_rate": 1.7093341385762907e-07,
+      "loss": 0.3583,
+      "step": 16410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1846507440261721,
+      "learning_rate": 1.7059098149620257e-07,
+      "loss": 0.1403,
+      "step": 16411
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3482449752952768,
+      "learning_rate": 1.702488895267862e-07,
+      "loss": 0.5575,
+      "step": 16412
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3510275776236248,
+      "learning_rate": 1.6990713796122938e-07,
+      "loss": 0.2993,
+      "step": 16413
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2897476651472436,
+      "learning_rate": 1.6956572681136485e-07,
+      "loss": 0.1903,
+      "step": 16414
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6730125448322015,
+      "learning_rate": 1.692246560890176e-07,
+      "loss": 0.4091,
+      "step": 16415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43855648011694554,
+      "learning_rate": 1.688839258059971e-07,
+      "loss": 0.3141,
+      "step": 16416
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4039676932669491,
+      "learning_rate": 1.6854353597410278e-07,
+      "loss": 0.2686,
+      "step": 16417
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24169214358315877,
+      "learning_rate": 1.68203486605123e-07,
+      "loss": 0.1656,
+      "step": 16418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4027029186189741,
+      "learning_rate": 1.6786377771083496e-07,
+      "loss": 0.5126,
+      "step": 16419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4016732571923031,
+      "learning_rate": 1.675244093030015e-07,
+      "loss": 0.272,
+      "step": 16420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31401939566076453,
+      "learning_rate": 1.6718538139337325e-07,
+      "loss": 0.251,
+      "step": 16421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7274054388947614,
+      "learning_rate": 1.6684669399369412e-07,
+      "loss": 0.4075,
+      "step": 16422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3134544640656414,
+      "learning_rate": 1.6650834711569031e-07,
+      "loss": 0.2514,
+      "step": 16423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29983200999934284,
+      "learning_rate": 1.661703407710802e-07,
+      "loss": 0.0921,
+      "step": 16424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3718216113469689,
+      "learning_rate": 1.6583267497156663e-07,
+      "loss": 0.3093,
+      "step": 16425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3162630195231657,
+      "learning_rate": 1.6549534972884584e-07,
+      "loss": 0.2512,
+      "step": 16426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5407154342108497,
+      "learning_rate": 1.6515836505459848e-07,
+      "loss": 0.3433,
+      "step": 16427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4685866469053301,
+      "learning_rate": 1.648217209604941e-07,
+      "loss": 0.2603,
+      "step": 16428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2709135275382767,
+      "learning_rate": 1.6448541745819113e-07,
+      "loss": 0.2503,
+      "step": 16429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3388604124213988,
+      "learning_rate": 1.6414945455933363e-07,
+      "loss": 0.1982,
+      "step": 16430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9826337080521574,
+      "learning_rate": 1.638138322755578e-07,
+      "loss": 0.2629,
+      "step": 16431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.300363016379023,
+      "learning_rate": 1.634785506184866e-07,
+      "loss": 0.2463,
+      "step": 16432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34111656134447776,
+      "learning_rate": 1.6314360959973075e-07,
+      "loss": 0.2959,
+      "step": 16433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0121583179292197,
+      "learning_rate": 1.628090092308876e-07,
+      "loss": 0.3706,
+      "step": 16434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39314249015309166,
+      "learning_rate": 1.6247474952354568e-07,
+      "loss": 0.2884,
+      "step": 16435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2459622684578627,
+      "learning_rate": 1.621408304892802e-07,
+      "loss": 0.19,
+      "step": 16436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3447675748059979,
+      "learning_rate": 1.618072521396552e-07,
+      "loss": 0.247,
+      "step": 16437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41316622660519126,
+      "learning_rate": 1.6147401448622145e-07,
+      "loss": 0.2924,
+      "step": 16438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44203133836751707,
+      "learning_rate": 1.6114111754051976e-07,
+      "loss": 0.3195,
+      "step": 16439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48436498902588176,
+      "learning_rate": 1.6080856131407862e-07,
+      "loss": 0.3431,
+      "step": 16440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28460937853923546,
+      "learning_rate": 1.6047634581841331e-07,
+      "loss": 0.184,
+      "step": 16441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2880830569609324,
+      "learning_rate": 1.6014447106502907e-07,
+      "loss": 0.1768,
+      "step": 16442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8736808553813734,
+      "learning_rate": 1.5981293706541888e-07,
+      "loss": 0.3771,
+      "step": 16443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31317366638812105,
+      "learning_rate": 1.5948174383106362e-07,
+      "loss": 0.2167,
+      "step": 16444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3277336661706098,
+      "learning_rate": 1.5915089137343186e-07,
+      "loss": 0.3097,
+      "step": 16445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1496737135542834,
+      "learning_rate": 1.5882037970398111e-07,
+      "loss": 0.7183,
+      "step": 16446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29250473166722574,
+      "learning_rate": 1.584902088341589e-07,
+      "loss": 0.1838,
+      "step": 16447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2878214529095352,
+      "learning_rate": 1.5816037877539715e-07,
+      "loss": 0.1679,
+      "step": 16448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3309627514911893,
+      "learning_rate": 1.5783088953911784e-07,
+      "loss": 0.3004,
+      "step": 16449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33326136046621196,
+      "learning_rate": 1.5750174113673077e-07,
+      "loss": 0.2167,
+      "step": 16450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.179074426417154,
+      "learning_rate": 1.5717293357963682e-07,
+      "loss": 0.5454,
+      "step": 16451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4564239880365591,
+      "learning_rate": 1.5684446687922017e-07,
+      "loss": 0.3404,
+      "step": 16452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30674601902753773,
+      "learning_rate": 1.565163410468562e-07,
+      "loss": 0.2332,
+      "step": 16453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2190549081107368,
+      "learning_rate": 1.5618855609390803e-07,
+      "loss": 0.1242,
+      "step": 16454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5729376141354351,
+      "learning_rate": 1.558611120317266e-07,
+      "loss": 0.3633,
+      "step": 16455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5545582014596254,
+      "learning_rate": 1.5553400887165172e-07,
+      "loss": 0.2793,
+      "step": 16456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2458146065487911,
+      "learning_rate": 1.5520724662501207e-07,
+      "loss": 0.2315,
+      "step": 16457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1168303612442845,
+      "learning_rate": 1.5488082530312087e-07,
+      "loss": 0.694,
+      "step": 16458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6203300463305141,
+      "learning_rate": 1.545547449172835e-07,
+      "loss": 0.2942,
+      "step": 16459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24751408093210903,
+      "learning_rate": 1.5422900547879206e-07,
+      "loss": 0.199,
+      "step": 16460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4335318922233136,
+      "learning_rate": 1.5390360699892636e-07,
+      "loss": 0.3313,
+      "step": 16461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3373096852746386,
+      "learning_rate": 1.5357854948895634e-07,
+      "loss": 0.2321,
+      "step": 16462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3630705794262418,
+      "learning_rate": 1.532538329601363e-07,
+      "loss": 0.2461,
+      "step": 16463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5126707395414409,
+      "learning_rate": 1.529294574237139e-07,
+      "loss": 0.3202,
+      "step": 16464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2925570376583863,
+      "learning_rate": 1.5260542289092016e-07,
+      "loss": 0.2425,
+      "step": 16465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5729303620205985,
+      "learning_rate": 1.5228172937297837e-07,
+      "loss": 0.3226,
+      "step": 16466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4405506344298435,
+      "learning_rate": 1.5195837688109506e-07,
+      "loss": 0.2103,
+      "step": 16467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23129493919039548,
+      "learning_rate": 1.5163536542647018e-07,
+      "loss": 0.2064,
+      "step": 16468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34775353831476136,
+      "learning_rate": 1.5131269502029034e-07,
+      "loss": 0.3053,
+      "step": 16469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0191758063869858,
+      "learning_rate": 1.509903656737277e-07,
+      "loss": 0.5071,
+      "step": 16470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6469909048210354,
+      "learning_rate": 1.506683773979445e-07,
+      "loss": 0.352,
+      "step": 16471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3790611624365381,
+      "learning_rate": 1.5034673020409173e-07,
+      "loss": 0.2984,
+      "step": 16472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2816475223008202,
+      "learning_rate": 1.5002542410330946e-07,
+      "loss": 0.2091,
+      "step": 16473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33164768844534576,
+      "learning_rate": 1.4970445910672205e-07,
+      "loss": 0.1769,
+      "step": 16474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.400650506425723,
+      "learning_rate": 1.493838352254462e-07,
+      "loss": 0.3036,
+      "step": 16475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32037790280791784,
+      "learning_rate": 1.4906355247058412e-07,
+      "loss": 0.2392,
+      "step": 16476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.605568545808694,
+      "learning_rate": 1.487436108532292e-07,
+      "loss": 0.33,
+      "step": 16477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3674500331935406,
+      "learning_rate": 1.4842401038445808e-07,
+      "loss": 0.2867,
+      "step": 16478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9639296926238876,
+      "learning_rate": 1.4810475107533973e-07,
+      "loss": 0.3768,
+      "step": 16479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2325175694217112,
+      "learning_rate": 1.4778583293692972e-07,
+      "loss": 0.1823,
+      "step": 16480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30507855821640345,
+      "learning_rate": 1.4746725598027367e-07,
+      "loss": 0.257,
+      "step": 16481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3321349423875684,
+      "learning_rate": 1.4714902021640277e-07,
+      "loss": 0.7486,
+      "step": 16482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5040720847135953,
+      "learning_rate": 1.4683112565633706e-07,
+      "loss": 0.2263,
+      "step": 16483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3693885098403536,
+      "learning_rate": 1.4651357231108555e-07,
+      "loss": 0.2884,
+      "step": 16484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3716837935292826,
+      "learning_rate": 1.4619636019164608e-07,
+      "loss": 0.3051,
+      "step": 16485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20352716304565405,
+      "learning_rate": 1.458794893090032e-07,
+      "loss": 0.1185,
+      "step": 16486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5404248538254562,
+      "learning_rate": 1.4556295967412925e-07,
+      "loss": 0.3645,
+      "step": 16487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35677414223681986,
+      "learning_rate": 1.4524677129798547e-07,
+      "loss": 0.2902,
+      "step": 16488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5106964245865684,
+      "learning_rate": 1.449309241915231e-07,
+      "loss": 0.1938,
+      "step": 16489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3935425552716978,
+      "learning_rate": 1.4461541836568004e-07,
+      "loss": 0.3145,
+      "step": 16490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3403643170390191,
+      "learning_rate": 1.443002538313798e-07,
+      "loss": 0.5119,
+      "step": 16491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26232504943660767,
+      "learning_rate": 1.4398543059953918e-07,
+      "loss": 0.2427,
+      "step": 16492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30415871484094553,
+      "learning_rate": 1.4367094868105725e-07,
+      "loss": 0.2046,
+      "step": 16493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42062338088496104,
+      "learning_rate": 1.433568080868286e-07,
+      "loss": 0.2729,
+      "step": 16494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5711244914388044,
+      "learning_rate": 1.4304300882772903e-07,
+      "loss": 0.3911,
+      "step": 16495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2384222357994241,
+      "learning_rate": 1.4272955091462648e-07,
+      "loss": 0.2252,
+      "step": 16496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3215017477278463,
+      "learning_rate": 1.424164343583745e-07,
+      "loss": 0.532,
+      "step": 16497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0583273806987223,
+      "learning_rate": 1.4210365916981882e-07,
+      "loss": 0.4026,
+      "step": 16498
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3118227358818172,
+      "learning_rate": 1.4179122535978862e-07,
+      "loss": 0.2325,
+      "step": 16499
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34214996786542395,
+      "learning_rate": 1.414791329391052e-07,
+      "loss": 0.3061,
+      "step": 16500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3931429662229714,
+      "learning_rate": 1.4116738191857437e-07,
+      "loss": 0.3039,
+      "step": 16501
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23757841266563745,
+      "learning_rate": 1.4085597230899418e-07,
+      "loss": 0.167,
+      "step": 16502
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2952326611305722,
+      "learning_rate": 1.4054490412114817e-07,
+      "loss": 0.5202,
+      "step": 16503
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3303112645977888,
+      "learning_rate": 1.402341773658078e-07,
+      "loss": 0.2998,
+      "step": 16504
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3496116138968531,
+      "learning_rate": 1.3992379205373219e-07,
+      "loss": 0.2641,
+      "step": 16505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6284477927092317,
+      "learning_rate": 1.3961374819567386e-07,
+      "loss": 0.3239,
+      "step": 16506
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29321421158262606,
+      "learning_rate": 1.3930404580236646e-07,
+      "loss": 0.1722,
+      "step": 16507
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27799789142741405,
+      "learning_rate": 1.3899468488453583e-07,
+      "loss": 0.2652,
+      "step": 16508
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5390190495988705,
+      "learning_rate": 1.3868566545289563e-07,
+      "loss": 0.2118,
+      "step": 16509
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2106732546022587,
+      "learning_rate": 1.383769875181462e-07,
+      "loss": 0.6369,
+      "step": 16510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3303768375574665,
+      "learning_rate": 1.38068651090979e-07,
+      "loss": 0.2693,
+      "step": 16511
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31858858994348516,
+      "learning_rate": 1.377606561820699e-07,
+      "loss": 0.2569,
+      "step": 16512
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41073201497419076,
+      "learning_rate": 1.3745300280208373e-07,
+      "loss": 0.249,
+      "step": 16513
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25752711484513774,
+      "learning_rate": 1.371456909616764e-07,
+      "loss": 0.2332,
+      "step": 16514
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4245672657375734,
+      "learning_rate": 1.3683872067149052e-07,
+      "loss": 0.6246,
+      "step": 16515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27505237674289085,
+      "learning_rate": 1.3653209194215534e-07,
+      "loss": 0.2159,
+      "step": 16516
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3313341959115871,
+      "learning_rate": 1.3622580478428903e-07,
+      "loss": 0.2806,
+      "step": 16517
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6390544637914571,
+      "learning_rate": 1.3591985920849981e-07,
+      "loss": 0.3746,
+      "step": 16518
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29075587750528226,
+      "learning_rate": 1.356142552253814e-07,
+      "loss": 0.1736,
+      "step": 16519
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28201683673325467,
+      "learning_rate": 1.3530899284551756e-07,
+      "loss": 0.2628,
+      "step": 16520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5231549591416089,
+      "learning_rate": 1.3500407207947875e-07,
+      "loss": 0.2276,
+      "step": 16521
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6591201808581352,
+      "learning_rate": 1.3469949293782426e-07,
+      "loss": 0.2343,
+      "step": 16522
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35442038050860036,
+      "learning_rate": 1.3439525543110232e-07,
+      "loss": 0.285,
+      "step": 16523
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3256582404882992,
+      "learning_rate": 1.3409135956984897e-07,
+      "loss": 0.2858,
+      "step": 16524
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4867984784902485,
+      "learning_rate": 1.337878053645869e-07,
+      "loss": 0.1892,
+      "step": 16525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2675448451742374,
+      "learning_rate": 1.334845928258288e-07,
+      "loss": 0.2195,
+      "step": 16526
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4860536461056225,
+      "learning_rate": 1.331817219640752e-07,
+      "loss": 0.2422,
+      "step": 16527
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3300567767788356,
+      "learning_rate": 1.3287919278981544e-07,
+      "loss": 0.2799,
+      "step": 16528
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33493831219224574,
+      "learning_rate": 1.3257700531352334e-07,
+      "loss": 0.2111,
+      "step": 16529
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7971942715761426,
+      "learning_rate": 1.3227515954566506e-07,
+      "loss": 0.4725,
+      "step": 16530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2101032035959423,
+      "learning_rate": 1.319736554966955e-07,
+      "loss": 0.4656,
+      "step": 16531
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2161492687029256,
+      "learning_rate": 1.31672493177053e-07,
+      "loss": 0.2144,
+      "step": 16532
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2706126200232461,
+      "learning_rate": 1.3137167259716698e-07,
+      "loss": 0.1613,
+      "step": 16533
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7201379209477112,
+      "learning_rate": 1.310711937674569e-07,
+      "loss": 0.3627,
+      "step": 16534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3188833536266433,
+      "learning_rate": 1.3077105669832556e-07,
+      "loss": 0.2127,
+      "step": 16535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.353800166876077,
+      "learning_rate": 1.3047126140016907e-07,
+      "loss": 0.29,
+      "step": 16536
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.152362615048209,
+      "learning_rate": 1.3017180788336804e-07,
+      "loss": 0.5908,
+      "step": 16537
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2201443708188749,
+      "learning_rate": 1.29872696158293e-07,
+      "loss": 0.1572,
+      "step": 16538
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39768952646586436,
+      "learning_rate": 1.295739262353013e-07,
+      "loss": 0.2506,
+      "step": 16539
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3711468804434103,
+      "learning_rate": 1.2927549812474128e-07,
+      "loss": 0.3222,
+      "step": 16540
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3875931478557323,
+      "learning_rate": 1.2897741183694578e-07,
+      "loss": 0.2718,
+      "step": 16541
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5040208709722376,
+      "learning_rate": 1.286796673822388e-07,
+      "loss": 0.2931,
+      "step": 16542
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3644096645649208,
+      "learning_rate": 1.2838226477092875e-07,
+      "loss": 0.2835,
+      "step": 16543
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.43199632651889797,
+      "learning_rate": 1.2808520401331737e-07,
+      "loss": 0.2786,
+      "step": 16544
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26276909694452727,
+      "learning_rate": 1.27788485119692e-07,
+      "loss": 0.142,
+      "step": 16545
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6481758566901482,
+      "learning_rate": 1.2749210810032664e-07,
+      "loss": 0.3816,
+      "step": 16546
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3026318614423293,
+      "learning_rate": 1.2719607296548309e-07,
+      "loss": 0.2441,
+      "step": 16547
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31330158406573605,
+      "learning_rate": 1.2690037972541646e-07,
+      "loss": 0.2622,
+      "step": 16548
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5618122703655006,
+      "learning_rate": 1.2660502839036526e-07,
+      "loss": 0.5628,
+      "step": 16549
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3455198798238632,
+      "learning_rate": 1.2631001897055683e-07,
+      "loss": 0.2721,
+      "step": 16550
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2731384236231438,
+      "learning_rate": 1.2601535147620746e-07,
+      "loss": 0.1375,
+      "step": 16551
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3394524321219831,
+      "learning_rate": 1.2572102591752234e-07,
+      "loss": 0.3059,
+      "step": 16552
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30354042132773396,
+      "learning_rate": 1.2542704230469326e-07,
+      "loss": 0.2326,
+      "step": 16553
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1255505639202197,
+      "learning_rate": 1.2513340064790102e-07,
+      "loss": 0.6941,
+      "step": 16554
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35766936306315206,
+      "learning_rate": 1.2484010095731414e-07,
+      "loss": 0.246,
+      "step": 16555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3323040325151218,
+      "learning_rate": 1.2454714324309115e-07,
+      "loss": 0.2591,
+      "step": 16556
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7689015974335848,
+      "learning_rate": 1.2425452751537503e-07,
+      "loss": 0.3713,
+      "step": 16557
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2691954444647513,
+      "learning_rate": 1.2396225378430105e-07,
+      "loss": 0.144,
+      "step": 16558
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.38664284669524407,
+      "learning_rate": 1.2367032205998775e-07,
+      "loss": 0.2492,
+      "step": 16559
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30289171963027917,
+      "learning_rate": 1.2337873235254704e-07,
+      "loss": 0.3089,
+      "step": 16560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3377802262789582,
+      "learning_rate": 1.2308748467207753e-07,
+      "loss": 0.2227,
+      "step": 16561
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39685089389976136,
+      "learning_rate": 1.2279657902866226e-07,
+      "loss": 0.2741,
+      "step": 16562
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34991767076091934,
+      "learning_rate": 1.225060154323776e-07,
+      "loss": 0.2956,
+      "step": 16563
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2420095759472581,
+      "learning_rate": 1.222157938932833e-07,
+      "loss": 0.1584,
+      "step": 16564
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4161152844228987,
+      "learning_rate": 1.219259144214324e-07,
+      "loss": 0.2804,
+      "step": 16565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5728617107352475,
+      "learning_rate": 1.216363770268625e-07,
+      "loss": 0.4119,
+      "step": 16566
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5569364520388921,
+      "learning_rate": 1.2134718171960103e-07,
+      "loss": 0.3633,
+      "step": 16567
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2639185558147655,
+      "learning_rate": 1.2105832850966004e-07,
+      "loss": 0.2085,
+      "step": 16568
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.59735708477971,
+      "learning_rate": 1.2076981740704485e-07,
+      "loss": 0.3744,
+      "step": 16569
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34765113855029134,
+      "learning_rate": 1.2048164842174636e-07,
+      "loss": 0.1818,
+      "step": 16570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2991838323871734,
+      "learning_rate": 1.2019382156374326e-07,
+      "loss": 0.1983,
+      "step": 16571
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33619733070672253,
+      "learning_rate": 1.1990633684300424e-07,
+      "loss": 0.2832,
+      "step": 16572
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9724457561401293,
+      "learning_rate": 1.1961919426948244e-07,
+      "loss": 0.4762,
+      "step": 16573
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3870489682298207,
+      "learning_rate": 1.1933239385312324e-07,
+      "loss": 0.2104,
+      "step": 16574
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.361921769610525,
+      "learning_rate": 1.190459356038598e-07,
+      "loss": 0.2939,
+      "step": 16575
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3518873026409385,
+      "learning_rate": 1.1875981953160975e-07,
+      "loss": 0.2099,
+      "step": 16576
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3527276444123807,
+      "learning_rate": 1.1847404564628185e-07,
+      "loss": 0.1475,
+      "step": 16577
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2743531371236152,
+      "learning_rate": 1.181886139577737e-07,
+      "loss": 0.2313,
+      "step": 16578
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.338837035760285,
+      "learning_rate": 1.1790352447596853e-07,
+      "loss": 0.3039,
+      "step": 16579
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5475170217274716,
+      "learning_rate": 1.1761877721073845e-07,
+      "loss": 0.3508,
+      "step": 16580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5025763106213428,
+      "learning_rate": 1.1733437217194665e-07,
+      "loss": 0.2398,
+      "step": 16581
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.535478845267447,
+      "learning_rate": 1.1705030936943973e-07,
+      "loss": 0.2167,
+      "step": 16582
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3968889128451416,
+      "learning_rate": 1.1676658881305647e-07,
+      "loss": 0.304,
+      "step": 16583
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21367273624373442,
+      "learning_rate": 1.1648321051262012e-07,
+      "loss": 0.196,
+      "step": 16584
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.80451182074773,
+      "learning_rate": 1.1620017447794507e-07,
+      "loss": 0.3895,
+      "step": 16585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5701986470439201,
+      "learning_rate": 1.1591748071883458e-07,
+      "loss": 0.2978,
+      "step": 16586
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2685745996616732,
+      "learning_rate": 1.1563512924507525e-07,
+      "loss": 0.246,
+      "step": 16587
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4199611521261792,
+      "learning_rate": 1.1535312006644706e-07,
+      "loss": 0.7062,
+      "step": 16588
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3948249665352689,
+      "learning_rate": 1.150714531927144e-07,
+      "loss": 0.2906,
+      "step": 16589
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5276002741172402,
+      "learning_rate": 1.1479012863363282e-07,
+      "loss": 0.3783,
+      "step": 16590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24658030549592613,
+      "learning_rate": 1.1450914639894451e-07,
+      "loss": 0.2207,
+      "step": 16591
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3155271224796519,
+      "learning_rate": 1.1422850649837836e-07,
+      "loss": 0.223,
+      "step": 16592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5925497152046855,
+      "learning_rate": 1.1394820894165437e-07,
+      "loss": 0.2979,
+      "step": 16593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5226704293130526,
+      "learning_rate": 1.1366825373847923e-07,
+      "loss": 0.2853,
+      "step": 16594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4706632168946928,
+      "learning_rate": 1.1338864089854629e-07,
+      "loss": 0.2925,
+      "step": 16595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35243983811365676,
+      "learning_rate": 1.1310937043154113e-07,
+      "loss": 0.2916,
+      "step": 16596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31547774494065683,
+      "learning_rate": 1.128304423471327e-07,
+      "loss": 0.1765,
+      "step": 16597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.640983379760982,
+      "learning_rate": 1.1255185665497992e-07,
+      "loss": 0.3632,
+      "step": 16598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2510122641943343,
+      "learning_rate": 1.1227361336473175e-07,
+      "loss": 0.256,
+      "step": 16599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4540579733239993,
+      "learning_rate": 1.119957124860238e-07,
+      "loss": 0.203,
+      "step": 16600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6368166691950049,
+      "learning_rate": 1.1171815402847841e-07,
+      "loss": 0.3969,
+      "step": 16601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.351165725596696,
+      "learning_rate": 1.1144093800170786e-07,
+      "loss": 0.2781,
+      "step": 16602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3683204475698124,
+      "learning_rate": 1.1116406441531335e-07,
+      "loss": 0.3086,
+      "step": 16603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23192559345215039,
+      "learning_rate": 1.1088753327888169e-07,
+      "loss": 0.1139,
+      "step": 16604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4094259485578723,
+      "learning_rate": 1.1061134460198964e-07,
+      "loss": 0.2794,
+      "step": 16605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2833862226613364,
+      "learning_rate": 1.1033549839420066e-07,
+      "loss": 0.5491,
+      "step": 16606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2600204249198676,
+      "learning_rate": 1.1005999466506822e-07,
+      "loss": 0.2268,
+      "step": 16607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36192005945864036,
+      "learning_rate": 1.0978483342413359e-07,
+      "loss": 0.288,
+      "step": 16608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5971870477119776,
+      "learning_rate": 1.0951001468092471e-07,
+      "loss": 0.3491,
+      "step": 16609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1664000066752861,
+      "learning_rate": 1.0923553844495838e-07,
+      "loss": 0.1133,
+      "step": 16610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3779109104685889,
+      "learning_rate": 1.0896140472574035e-07,
+      "loss": 0.3052,
+      "step": 16611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.48651148623474383,
+      "learning_rate": 1.0868761353276414e-07,
+      "loss": 0.3653,
+      "step": 16612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6419661588127807,
+      "learning_rate": 1.0841416487550994e-07,
+      "loss": 0.2488,
+      "step": 16613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3649414029689163,
+      "learning_rate": 1.0814105876344794e-07,
+      "loss": 0.2928,
+      "step": 16614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3596194680368247,
+      "learning_rate": 1.0786829520603503e-07,
+      "loss": 0.3083,
+      "step": 16615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2692613514757282,
+      "learning_rate": 1.0759587421271811e-07,
+      "loss": 0.1606,
+      "step": 16616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3453726228128851,
+      "learning_rate": 1.0732379579293184e-07,
+      "loss": 0.1923,
+      "step": 16617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3122016638571212,
+      "learning_rate": 1.0705205995609536e-07,
+      "loss": 0.5729,
+      "step": 16618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3154318714815696,
+      "learning_rate": 1.0678066671162113e-07,
+      "loss": 0.2759,
+      "step": 16619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3255968830256943,
+      "learning_rate": 1.0650961606890719e-07,
+      "loss": 0.2382,
+      "step": 16620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2859265206571604,
+      "learning_rate": 1.062389080373405e-07,
+      "loss": 0.8102,
+      "step": 16621
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26955193137320504,
+      "learning_rate": 1.0596854262629352e-07,
+      "loss": 0.1498,
+      "step": 16622
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24722787460287485,
+      "learning_rate": 1.0569851984513102e-07,
+      "loss": 0.2281,
+      "step": 16623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3302779187313543,
+      "learning_rate": 1.0542883970320328e-07,
+      "loss": 0.5446,
+      "step": 16624
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5332660462662445,
+      "learning_rate": 1.051595022098506e-07,
+      "loss": 0.3458,
+      "step": 16625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30330182429363667,
+      "learning_rate": 1.0489050737439777e-07,
+      "loss": 0.2059,
+      "step": 16626
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36778833648956744,
+      "learning_rate": 1.0462185520616064e-07,
+      "loss": 0.315,
+      "step": 16627
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23797657274293577,
+      "learning_rate": 1.0435354571444401e-07,
+      "loss": 0.1594,
+      "step": 16628
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3561924831436635,
+      "learning_rate": 1.0408557890853821e-07,
+      "loss": 0.2595,
+      "step": 16629
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4752465406071092,
+      "learning_rate": 1.038179547977236e-07,
+      "loss": 0.2584,
+      "step": 16630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.39518269810907725,
+      "learning_rate": 1.0355067339126723e-07,
+      "loss": 0.2972,
+      "step": 16631
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36175612579907546,
+      "learning_rate": 1.0328373469842502e-07,
+      "loss": 0.2588,
+      "step": 16632
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.571137981694321,
+      "learning_rate": 1.0301713872844288e-07,
+      "loss": 0.414,
+      "step": 16633
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3156274111668739,
+      "learning_rate": 1.0275088549055123e-07,
+      "loss": 0.226,
+      "step": 16634
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2669295187752056,
+      "learning_rate": 1.0248497499396936e-07,
+      "loss": 0.2455,
+      "step": 16635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4475781213109659,
+      "learning_rate": 1.022194072479088e-07,
+      "loss": 0.1764,
+      "step": 16636
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7397982641900223,
+      "learning_rate": 1.019541822615644e-07,
+      "loss": 0.4622,
+      "step": 16637
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32650303175664785,
+      "learning_rate": 1.0168930004412103e-07,
+      "loss": 0.2467,
+      "step": 16638
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3796386468131196,
+      "learning_rate": 1.0142476060475137e-07,
+      "loss": 0.2462,
+      "step": 16639
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8287190883309027,
+      "learning_rate": 1.0116056395261586e-07,
+      "loss": 0.327,
+      "step": 16640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20625582020929892,
+      "learning_rate": 1.0089671009686497e-07,
+      "loss": 0.2011,
+      "step": 16641
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5948962808042882,
+      "learning_rate": 1.0063319904663471e-07,
+      "loss": 0.4045,
+      "step": 16642
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29778674786226755,
+      "learning_rate": 1.0037003081105223e-07,
+      "loss": 0.2352,
+      "step": 16643
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33424304060632115,
+      "learning_rate": 1.0010720539922914e-07,
+      "loss": 0.2542,
+      "step": 16644
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1934592174173186,
+      "learning_rate": 9.984472282026814e-08,
+      "loss": 0.702,
+      "step": 16645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3274171866966553,
+      "learning_rate": 9.958258308325975e-08,
+      "loss": 0.2261,
+      "step": 16646
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34650531219482306,
+      "learning_rate": 9.932078619727892e-08,
+      "loss": 0.2777,
+      "step": 16647
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3254717650580907,
+      "learning_rate": 9.905933217139397e-08,
+      "loss": 0.2203,
+      "step": 16648
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7048051219031087,
+      "learning_rate": 9.879822101465874e-08,
+      "loss": 0.2505,
+      "step": 16649
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3501335317692874,
+      "learning_rate": 9.853745273611604e-08,
+      "loss": 0.275,
+      "step": 16650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34539266905883126,
+      "learning_rate": 9.827702734479528e-08,
+      "loss": 0.3104,
+      "step": 16651
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0335167832865215,
+      "learning_rate": 9.801694484971369e-08,
+      "loss": 0.2343,
+      "step": 16652
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30600946235260734,
+      "learning_rate": 9.775720525988076e-08,
+      "loss": 0.2483,
+      "step": 16653
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3128501072923363,
+      "learning_rate": 9.749780858429036e-08,
+      "loss": 0.2964,
+      "step": 16654
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2661670102671037,
+      "learning_rate": 9.723875483192536e-08,
+      "loss": 0.1817,
+      "step": 16655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2967207607658545,
+      "learning_rate": 9.69800440117552e-08,
+      "loss": 0.2061,
+      "step": 16656
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3288708256421893,
+      "learning_rate": 9.672167613274053e-08,
+      "loss": 0.7043,
+      "step": 16657
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3621252628942739,
+      "learning_rate": 9.64636512038286e-08,
+      "loss": 0.2824,
+      "step": 16658
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2996790863119093,
+      "learning_rate": 9.620596923395454e-08,
+      "loss": 0.2111,
+      "step": 16659
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7797861094274411,
+      "learning_rate": 9.594863023204226e-08,
+      "loss": 0.3865,
+      "step": 16660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24464123847918023,
+      "learning_rate": 9.569163420700245e-08,
+      "loss": 0.183,
+      "step": 16661
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3100921626514089,
+      "learning_rate": 9.543498116773576e-08,
+      "loss": 0.1861,
+      "step": 16662
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3610536933026674,
+      "learning_rate": 9.51786711231295e-08,
+      "loss": 0.2764,
+      "step": 16663
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6986584506879514,
+      "learning_rate": 9.492270408206106e-08,
+      "loss": 0.3745,
+      "step": 16664
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2895372923549709,
+      "learning_rate": 9.46670800533922e-08,
+      "loss": 0.1944,
+      "step": 16665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3473184880519653,
+      "learning_rate": 9.441179904597697e-08,
+      "loss": 0.3016,
+      "step": 16666
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38263612931781815,
+      "learning_rate": 9.415686106865496e-08,
+      "loss": 0.2469,
+      "step": 16667
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4296040570362124,
+      "learning_rate": 9.390226613025466e-08,
+      "loss": 0.248,
+      "step": 16668
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31216699027746436,
+      "learning_rate": 9.364801423959235e-08,
+      "loss": 0.1625,
+      "step": 16669
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33552034711426854,
+      "learning_rate": 9.339410540547433e-08,
+      "loss": 0.2803,
+      "step": 16670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3310185221024679,
+      "learning_rate": 9.314053963669245e-08,
+      "loss": 0.254,
+      "step": 16671
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7250290339843236,
+      "learning_rate": 9.288731694202747e-08,
+      "loss": 0.2567,
+      "step": 16672
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2242252500615893,
+      "learning_rate": 9.263443733024791e-08,
+      "loss": 0.7467,
+      "step": 16673
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32309648425980414,
+      "learning_rate": 9.238190081011345e-08,
+      "loss": 0.2512,
+      "step": 16674
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19987086897763234,
+      "learning_rate": 9.212970739036709e-08,
+      "loss": 0.1648,
+      "step": 16675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5338994500229073,
+      "learning_rate": 9.187785707974183e-08,
+      "loss": 0.3529,
+      "step": 16676
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32266698545901074,
+      "learning_rate": 9.162634988696184e-08,
+      "loss": 0.2438,
+      "step": 16677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5167703108082058,
+      "learning_rate": 9.137518582073345e-08,
+      "loss": 0.3471,
+      "step": 16678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4186625502370127,
+      "learning_rate": 9.112436488975751e-08,
+      "loss": 0.2912,
+      "step": 16679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3821134635943973,
+      "learning_rate": 9.087388710271927e-08,
+      "loss": 0.2757,
+      "step": 16680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25700002896802115,
+      "learning_rate": 9.06237524682918e-08,
+      "loss": 0.1689,
+      "step": 16681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3561604016730922,
+      "learning_rate": 9.037396099513707e-08,
+      "loss": 0.2703,
+      "step": 16682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.39261394030257535,
+      "learning_rate": 9.012451269190592e-08,
+      "loss": 0.2591,
+      "step": 16683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6363852734792065,
+      "learning_rate": 8.987540756723811e-08,
+      "loss": 0.3443,
+      "step": 16684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3103970018735047,
+      "learning_rate": 8.962664562975676e-08,
+      "loss": 0.3247,
+      "step": 16685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3028875377252202,
+      "learning_rate": 8.93782268880794e-08,
+      "loss": 0.2357,
+      "step": 16686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2752372436372155,
+      "learning_rate": 8.913015135080805e-08,
+      "loss": 0.2525,
+      "step": 16687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36823697113929454,
+      "learning_rate": 8.888241902653361e-08,
+      "loss": 0.1942,
+      "step": 16688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4023075320401017,
+      "learning_rate": 8.863502992383477e-08,
+      "loss": 0.2551,
+      "step": 16689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2965441139243796,
+      "learning_rate": 8.838798405127802e-08,
+      "loss": 0.2671,
+      "step": 16690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.325159274198456,
+      "learning_rate": 8.814128141741984e-08,
+      "loss": 0.6663,
+      "step": 16691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2863213468271409,
+      "learning_rate": 8.78949220308023e-08,
+      "loss": 0.2085,
+      "step": 16692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7635493682524934,
+      "learning_rate": 8.764890589995745e-08,
+      "loss": 0.4148,
+      "step": 16693
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2738614036932938,
+      "learning_rate": 8.740323303340514e-08,
+      "loss": 0.2462,
+      "step": 16694
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26102284400214015,
+      "learning_rate": 8.715790343965192e-08,
+      "loss": 0.1567,
+      "step": 16695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.418513087327854,
+      "learning_rate": 8.691291712719541e-08,
+      "loss": 0.4353,
+      "step": 16696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1335322604001175,
+      "learning_rate": 8.666827410451772e-08,
+      "loss": 0.7633,
+      "step": 16697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24218728721153193,
+      "learning_rate": 8.642397438008987e-08,
+      "loss": 0.2166,
+      "step": 16698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44698841899876174,
+      "learning_rate": 8.618001796237507e-08,
+      "loss": 0.3161,
+      "step": 16699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8319313930088971,
+      "learning_rate": 8.593640485981991e-08,
+      "loss": 0.194,
+      "step": 16700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3124725689478206,
+      "learning_rate": 8.569313508086097e-08,
+      "loss": 0.0728,
+      "step": 16701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2663921373594587,
+      "learning_rate": 8.545020863392262e-08,
+      "loss": 0.2698,
+      "step": 16702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0964726846573836,
+      "learning_rate": 8.520762552741591e-08,
+      "loss": 0.5473,
+      "step": 16703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5929084586700467,
+      "learning_rate": 8.496538576974412e-08,
+      "loss": 0.3214,
+      "step": 16704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3420457220417289,
+      "learning_rate": 8.472348936929387e-08,
+      "loss": 0.2463,
+      "step": 16705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33912810289817497,
+      "learning_rate": 8.448193633444291e-08,
+      "loss": 0.3064,
+      "step": 16706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2234517371912271,
+      "learning_rate": 8.424072667355565e-08,
+      "loss": 0.1544,
+      "step": 16707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44621753014878657,
+      "learning_rate": 8.399986039498653e-08,
+      "loss": 0.2155,
+      "step": 16708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8496786623009368,
+      "learning_rate": 8.375933750707554e-08,
+      "loss": 0.4511,
+      "step": 16709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2951581272073628,
+      "learning_rate": 8.35191580181527e-08,
+      "loss": 0.2546,
+      "step": 16710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3656795630471628,
+      "learning_rate": 8.327932193653355e-08,
+      "loss": 0.2506,
+      "step": 16711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5578558576279988,
+      "learning_rate": 8.30398292705259e-08,
+      "loss": 0.2439,
+      "step": 16712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27556519383523753,
+      "learning_rate": 8.280068002842312e-08,
+      "loss": 0.2176,
+      "step": 16713
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2691015371715588,
+      "learning_rate": 8.256187421850636e-08,
+      "loss": 0.2203,
+      "step": 16714
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6831366435714303,
+      "learning_rate": 8.232341184904458e-08,
+      "loss": 0.3823,
+      "step": 16715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.39566843622236897,
+      "learning_rate": 8.20852929282967e-08,
+      "loss": 0.2655,
+      "step": 16716
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5305390025366589,
+      "learning_rate": 8.184751746450947e-08,
+      "loss": 0.3602,
+      "step": 16717
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3804937384566211,
+      "learning_rate": 8.16100854659152e-08,
+      "loss": 0.243,
+      "step": 16718
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24366733203716592,
+      "learning_rate": 8.13729969407373e-08,
+      "loss": 0.1461,
+      "step": 16719
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3626781267030705,
+      "learning_rate": 8.113625189718588e-08,
+      "loss": 0.2923,
+      "step": 16720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3382050876755194,
+      "learning_rate": 8.089985034346104e-08,
+      "loss": 0.2521,
+      "step": 16721
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6097586807838828,
+      "learning_rate": 8.066379228774624e-08,
+      "loss": 0.3351,
+      "step": 16722
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34159479554399735,
+      "learning_rate": 8.042807773821826e-08,
+      "loss": 0.285,
+      "step": 16723
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4091610780001418,
+      "learning_rate": 8.019270670303946e-08,
+      "loss": 0.2663,
+      "step": 16724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2633520677569486,
+      "learning_rate": 7.995767919036002e-08,
+      "loss": 0.1632,
+      "step": 16725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25363763611530493,
+      "learning_rate": 7.972299520832005e-08,
+      "loss": 0.2677,
+      "step": 16726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0371752079589232,
+      "learning_rate": 7.948865476504641e-08,
+      "loss": 0.2723,
+      "step": 16727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5501126198181006,
+      "learning_rate": 7.925465786865372e-08,
+      "loss": 0.3554,
+      "step": 16728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3783143100194757,
+      "learning_rate": 7.902100452724548e-08,
+      "loss": 0.3148,
+      "step": 16729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35404309098921694,
+      "learning_rate": 7.878769474891413e-08,
+      "loss": 0.2972,
+      "step": 16730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19637934444938035,
+      "learning_rate": 7.855472854173763e-08,
+      "loss": 0.0779,
+      "step": 16731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3869329443831731,
+      "learning_rate": 7.832210591378398e-08,
+      "loss": 0.2814,
+      "step": 16732
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42170939268755797,
+      "learning_rate": 7.808982687311006e-08,
+      "loss": 0.3055,
+      "step": 16733
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2777667725765214,
+      "learning_rate": 7.785789142775834e-08,
+      "loss": 0.2124,
+      "step": 16734
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5345760328975254,
+      "learning_rate": 7.762629958576129e-08,
+      "loss": 0.3851,
+      "step": 16735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.316711270975337,
+      "learning_rate": 7.739505135513803e-08,
+      "loss": 0.541,
+      "step": 16736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2516530707494128,
+      "learning_rate": 7.716414674389771e-08,
+      "loss": 0.148,
+      "step": 16737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23169324602476324,
+      "learning_rate": 7.693358576003617e-08,
+      "loss": 0.214,
+      "step": 16738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5344331125889435,
+      "learning_rate": 7.670336841153925e-08,
+      "loss": 0.3644,
+      "step": 16739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4752035059563894,
+      "learning_rate": 7.64734947063761e-08,
+      "loss": 0.2514,
+      "step": 16740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36429165610906206,
+      "learning_rate": 7.624396465251038e-08,
+      "loss": 0.2732,
+      "step": 16741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3708596666694886,
+      "learning_rate": 7.601477825788905e-08,
+      "loss": 0.2794,
+      "step": 16742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3845077104005944,
+      "learning_rate": 7.57859355304491e-08,
+      "loss": 0.1939,
+      "step": 16743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3137435856663455,
+      "learning_rate": 7.55574364781153e-08,
+      "loss": 0.2363,
+      "step": 16744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33829037017062646,
+      "learning_rate": 7.532928110880133e-08,
+      "loss": 0.2301,
+      "step": 16745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40890395308043054,
+      "learning_rate": 7.510146943040641e-08,
+      "loss": 0.3042,
+      "step": 16746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3051775333200117,
+      "learning_rate": 7.487400145082203e-08,
+      "loss": 0.2107,
+      "step": 16747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3662893373185263,
+      "learning_rate": 7.464687717792407e-08,
+      "loss": 0.4789,
+      "step": 16748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33082857403563304,
+      "learning_rate": 7.442009661957738e-08,
+      "loss": 0.2837,
+      "step": 16749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33747596972844696,
+      "learning_rate": 7.41936597836368e-08,
+      "loss": 0.2082,
+      "step": 16750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3298350815450901,
+      "learning_rate": 7.396756667794158e-08,
+      "loss": 0.2119,
+      "step": 16751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7593553931238317,
+      "learning_rate": 7.374181731032326e-08,
+      "loss": 0.5444,
+      "step": 16752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3203570525125074,
+      "learning_rate": 7.35164116885989e-08,
+      "loss": 0.2598,
+      "step": 16753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33154457148997163,
+      "learning_rate": 7.329134982057562e-08,
+      "loss": 0.2206,
+      "step": 16754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7366957717865998,
+      "learning_rate": 7.306663171404494e-08,
+      "loss": 0.3779,
+      "step": 16755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30293447315905003,
+      "learning_rate": 7.284225737678952e-08,
+      "loss": 0.2432,
+      "step": 16756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25522354037367,
+      "learning_rate": 7.261822681657982e-08,
+      "loss": 0.1827,
+      "step": 16757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.388424240051691,
+      "learning_rate": 7.239454004117519e-08,
+      "loss": 0.8719,
+      "step": 16758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31357229248114443,
+      "learning_rate": 7.217119705831943e-08,
+      "loss": 0.2698,
+      "step": 16759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8776215112811757,
+      "learning_rate": 7.19481978757497e-08,
+      "loss": 0.2509,
+      "step": 16760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3567502082271883,
+      "learning_rate": 7.172554250118535e-08,
+      "loss": 0.3052,
+      "step": 16761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3012814796438905,
+      "learning_rate": 7.150323094233912e-08,
+      "loss": 0.2416,
+      "step": 16762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44014096773481015,
+      "learning_rate": 7.128126320690931e-08,
+      "loss": 0.0926,
+      "step": 16763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7301742900613,
+      "learning_rate": 7.105963930258308e-08,
+      "loss": 0.4903,
+      "step": 16764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22739964947646552,
+      "learning_rate": 7.083835923703319e-08,
+      "loss": 0.2198,
+      "step": 16765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4968594490313015,
+      "learning_rate": 7.061742301792462e-08,
+      "loss": 0.3535,
+      "step": 16766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49300324716515354,
+      "learning_rate": 7.039683065290792e-08,
+      "loss": 0.2496,
+      "step": 16767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3440836293024113,
+      "learning_rate": 7.017658214962142e-08,
+      "loss": 0.2844,
+      "step": 16768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7243014548865391,
+      "learning_rate": 6.995667751569346e-08,
+      "loss": 0.3251,
+      "step": 16769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3702799611297618,
+      "learning_rate": 6.973711675873795e-08,
+      "loss": 0.2672,
+      "step": 16770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37034677737138777,
+      "learning_rate": 6.951789988635992e-08,
+      "loss": 0.2754,
+      "step": 16771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33444108241821474,
+      "learning_rate": 6.929902690614998e-08,
+      "loss": 0.2079,
+      "step": 16772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3614151509687135,
+      "learning_rate": 6.90804978256876e-08,
+      "loss": 0.2578,
+      "step": 16773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34113505165220437,
+      "learning_rate": 6.886231265254007e-08,
+      "loss": 0.2359,
+      "step": 16774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2018390028919796,
+      "learning_rate": 6.864447139426356e-08,
+      "loss": 0.5018,
+      "step": 16775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3092635740058215,
+      "learning_rate": 6.842697405840204e-08,
+      "loss": 0.1559,
+      "step": 16776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2583079412979954,
+      "learning_rate": 6.820982065248837e-08,
+      "loss": 0.252,
+      "step": 16777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3110707555011408,
+      "learning_rate": 6.79930111840399e-08,
+      "loss": 0.2555,
+      "step": 16778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9741701424650024,
+      "learning_rate": 6.777654566056724e-08,
+      "loss": 0.3839,
+      "step": 16779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3323721668501855,
+      "learning_rate": 6.756042408956554e-08,
+      "loss": 0.1876,
+      "step": 16780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5357257838410326,
+      "learning_rate": 6.73446464785199e-08,
+      "loss": 0.3502,
+      "step": 16781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44370665817522104,
+      "learning_rate": 6.712921283490103e-08,
+      "loss": 0.3634,
+      "step": 16782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28311071631336937,
+      "learning_rate": 6.691412316617075e-08,
+      "loss": 0.1912,
+      "step": 16783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5741547571259257,
+      "learning_rate": 6.66993774797775e-08,
+      "loss": 0.3838,
+      "step": 16784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2227072118964059,
+      "learning_rate": 6.64849757831576e-08,
+      "loss": 0.205,
+      "step": 16785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2907201308397229,
+      "learning_rate": 6.627091808373509e-08,
+      "loss": 0.1812,
+      "step": 16786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.299735719613476,
+      "learning_rate": 6.605720438892515e-08,
+      "loss": 0.7159,
+      "step": 16787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4745022016836694,
+      "learning_rate": 6.584383470612631e-08,
+      "loss": 0.3508,
+      "step": 16788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2982418095480476,
+      "learning_rate": 6.563080904272712e-08,
+      "loss": 0.1854,
+      "step": 16789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3813855226753747,
+      "learning_rate": 6.54181274061072e-08,
+      "loss": 0.3044,
+      "step": 16790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32301668216479656,
+      "learning_rate": 6.520578980362957e-08,
+      "loss": 0.1707,
+      "step": 16791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41005198828464845,
+      "learning_rate": 6.499379624264834e-08,
+      "loss": 0.2734,
+      "step": 16792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.277449872251276,
+      "learning_rate": 6.478214673050542e-08,
+      "loss": 0.2324,
+      "step": 16793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2249416877156012,
+      "learning_rate": 6.45708412745294e-08,
+      "loss": 0.7376,
+      "step": 16794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3198672617571951,
+      "learning_rate": 6.435987988203662e-08,
+      "loss": 0.241,
+      "step": 16795
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6394970517054889,
+      "learning_rate": 6.414926256033461e-08,
+      "loss": 0.2757,
+      "step": 16796
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22601691033016935,
+      "learning_rate": 6.393898931671749e-08,
+      "loss": 0.2039,
+      "step": 16797
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33681906182926935,
+      "learning_rate": 6.372906015846502e-08,
+      "loss": 0.2788,
+      "step": 16798
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.303947428546528,
+      "learning_rate": 6.351947509284695e-08,
+      "loss": 0.2012,
+      "step": 16799
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.44996544450292303,
+      "learning_rate": 6.331023412712411e-08,
+      "loss": 0.3338,
+      "step": 16800
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2537708506596748,
+      "learning_rate": 6.310133726853962e-08,
+      "loss": 0.23,
+      "step": 16801
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8024519107979677,
+      "learning_rate": 6.289278452432768e-08,
+      "loss": 0.3105,
+      "step": 16802
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3602472486183106,
+      "learning_rate": 6.268457590171251e-08,
+      "loss": 0.1993,
+      "step": 16803
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5559272095999094,
+      "learning_rate": 6.247671140790279e-08,
+      "loss": 0.2386,
+      "step": 16804
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2690640064334697,
+      "learning_rate": 6.226919105009721e-08,
+      "loss": 0.269,
+      "step": 16805
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5439058823170019,
+      "learning_rate": 6.206201483548224e-08,
+      "loss": 0.3134,
+      "step": 16806
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4256086949463779,
+      "learning_rate": 6.185518277123215e-08,
+      "loss": 0.2782,
+      "step": 16807
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5052015564502045,
+      "learning_rate": 6.164869486451008e-08,
+      "loss": 0.3196,
+      "step": 16808
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20772748215582018,
+      "learning_rate": 6.144255112246589e-08,
+      "loss": 0.1723,
+      "step": 16809
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40896993870319204,
+      "learning_rate": 6.12367515522394e-08,
+      "loss": 0.3089,
+      "step": 16810
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5635204637734937,
+      "learning_rate": 6.103129616095605e-08,
+      "loss": 0.3505,
+      "step": 16811
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4005554484110214,
+      "learning_rate": 6.082618495573234e-08,
+      "loss": 0.2544,
+      "step": 16812
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33308458781620476,
+      "learning_rate": 6.062141794366927e-08,
+      "loss": 0.2608,
+      "step": 16813
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.66920581948784,
+      "learning_rate": 6.041699513186005e-08,
+      "loss": 0.4106,
+      "step": 16814
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32580508384371565,
+      "learning_rate": 6.021291652738348e-08,
+      "loss": 0.09,
+      "step": 16815
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3156954468518769,
+      "learning_rate": 6.0009182137305e-08,
+      "loss": 0.241,
+      "step": 16816
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35874233754083895,
+      "learning_rate": 5.98057919686823e-08,
+      "loss": 0.3018,
+      "step": 16817
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5415758946024039,
+      "learning_rate": 5.9602746028556425e-08,
+      "loss": 0.3908,
+      "step": 16818
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32112130011980533,
+      "learning_rate": 5.9400044323960625e-08,
+      "loss": 0.2137,
+      "step": 16819
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3192494544105202,
+      "learning_rate": 5.919768686191263e-08,
+      "loss": 0.6982,
+      "step": 16820
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27660356577726986,
+      "learning_rate": 5.8995673649422383e-08,
+      "loss": 0.2368,
+      "step": 16821
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31310707140945643,
+      "learning_rate": 5.879400469348429e-08,
+      "loss": 0.0847,
+      "step": 16822
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38135073290776306,
+      "learning_rate": 5.859268000108276e-08,
+      "loss": 0.3013,
+      "step": 16823
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3411340438737624,
+      "learning_rate": 5.8391699579188885e-08,
+      "loss": 0.2999,
+      "step": 16824
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3468521220123775,
+      "learning_rate": 5.819106343476266e-08,
+      "loss": 0.1525,
+      "step": 16825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38139159913608844,
+      "learning_rate": 5.799077157475297e-08,
+      "loss": 0.3019,
+      "step": 16826
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35172703587555404,
+      "learning_rate": 5.779082400609426e-08,
+      "loss": 0.1948,
+      "step": 16827
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3715864195679574,
+      "learning_rate": 5.7591220735712105e-08,
+      "loss": 0.1732,
+      "step": 16828
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2882289558296544,
+      "learning_rate": 5.7391961770519865e-08,
+      "loss": 0.3003,
+      "step": 16829
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8134034249660722,
+      "learning_rate": 5.7193047117415356e-08,
+      "loss": 0.4185,
+      "step": 16830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5514544419384692,
+      "learning_rate": 5.699447678328751e-08,
+      "loss": 0.373,
+      "step": 16831
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29257468097948075,
+      "learning_rate": 5.6796250775014164e-08,
+      "loss": 0.2282,
+      "step": 16832
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5000178068678629,
+      "learning_rate": 5.6598369099458705e-08,
+      "loss": 0.2971,
+      "step": 16833
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22716384655094343,
+      "learning_rate": 5.640083176347455e-08,
+      "loss": 0.1603,
+      "step": 16834
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3518311840836283,
+      "learning_rate": 5.620363877390178e-08,
+      "loss": 0.2388,
+      "step": 16835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45494879435889785,
+      "learning_rate": 5.600679013756938e-08,
+      "loss": 0.3139,
+      "step": 16836
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31631780442761426,
+      "learning_rate": 5.581028586129411e-08,
+      "loss": 0.2677,
+      "step": 16837
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5723125886588851,
+      "learning_rate": 5.561412595188165e-08,
+      "loss": 0.2285,
+      "step": 16838
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9015022205745714,
+      "learning_rate": 5.541831041612322e-08,
+      "loss": 0.4787,
+      "step": 16839
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37034387207268293,
+      "learning_rate": 5.5222839260802294e-08,
+      "loss": 0.2729,
+      "step": 16840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30520368304807893,
+      "learning_rate": 5.502771249268568e-08,
+      "loss": 0.2811,
+      "step": 16841
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2898764190409881,
+      "learning_rate": 5.483293011853241e-08,
+      "loss": 0.145,
+      "step": 16842
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.756039605785083,
+      "learning_rate": 5.4638492145087096e-08,
+      "loss": 0.3867,
+      "step": 16843
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3807809778880975,
+      "learning_rate": 5.4444398579083235e-08,
+      "loss": 0.2543,
+      "step": 16844
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37486671787998804,
+      "learning_rate": 5.425064942724212e-08,
+      "loss": 0.2385,
+      "step": 16845
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.596115630845532,
+      "learning_rate": 5.4057244696272826e-08,
+      "loss": 0.3478,
+      "step": 16846
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36027385165231446,
+      "learning_rate": 5.386418439287444e-08,
+      "loss": 0.2733,
+      "step": 16847
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26856822778458994,
+      "learning_rate": 5.3671468523731617e-08,
+      "loss": 0.1785,
+      "step": 16848
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37989515893189935,
+      "learning_rate": 5.3479097095516795e-08,
+      "loss": 0.3023,
+      "step": 16849
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2918242952616663,
+      "learning_rate": 5.328707011489465e-08,
+      "loss": 0.2222,
+      "step": 16850
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7682960082736956,
+      "learning_rate": 5.3095387588512074e-08,
+      "loss": 0.2587,
+      "step": 16851
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36395943813381715,
+      "learning_rate": 5.2904049523009315e-08,
+      "loss": 0.2663,
+      "step": 16852
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38743575817296055,
+      "learning_rate": 5.271305592501108e-08,
+      "loss": 0.3176,
+      "step": 16853
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.417320125421086,
+      "learning_rate": 5.252240680113319e-08,
+      "loss": 0.2276,
+      "step": 16854
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22973767123986893,
+      "learning_rate": 5.233210215797591e-08,
+      "loss": 0.104,
+      "step": 16855
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3841946691455176,
+      "learning_rate": 5.2142142002129524e-08,
+      "loss": 0.3001,
+      "step": 16856
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34093585258614106,
+      "learning_rate": 5.195252634017434e-08,
+      "loss": 0.2781,
+      "step": 16857
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5523198990332471,
+      "learning_rate": 5.1763255178673974e-08,
+      "loss": 0.2075,
+      "step": 16858
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4296501087072229,
+      "learning_rate": 5.1574328524184316e-08,
+      "loss": 0.2917,
+      "step": 16859
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47266839506660946,
+      "learning_rate": 5.1385746383249e-08,
+      "loss": 0.32,
+      "step": 16860
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21272369188801774,
+      "learning_rate": 5.1197508762397265e-08,
+      "loss": 0.1139,
+      "step": 16861
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2604253939433585,
+      "learning_rate": 5.1009615668147217e-08,
+      "loss": 0.2097,
+      "step": 16862
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.610360799421692,
+      "learning_rate": 5.082206710700699e-08,
+      "loss": 0.3483,
+      "step": 16863
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.295175344003875,
+      "learning_rate": 5.063486308547028e-08,
+      "loss": 0.2115,
+      "step": 16864
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33680233097645323,
+      "learning_rate": 5.044800361002078e-08,
+      "loss": 0.2811,
+      "step": 16865
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.372302629109178,
+      "learning_rate": 5.0261488687128876e-08,
+      "loss": 0.62,
+      "step": 16866
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24720787872927993,
+      "learning_rate": 5.007531832325385e-08,
+      "loss": 0.1813,
+      "step": 16867
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2469605731267985,
+      "learning_rate": 4.9889492524842766e-08,
+      "loss": 0.2309,
+      "step": 16868
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5687789725130987,
+      "learning_rate": 4.970401129833047e-08,
+      "loss": 0.3839,
+      "step": 16869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7360556561940557,
+      "learning_rate": 4.9518874650139604e-08,
+      "loss": 0.3072,
+      "step": 16870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36413256453148496,
+      "learning_rate": 4.933408258668393e-08,
+      "loss": 0.2074,
+      "step": 16871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33474022926915215,
+      "learning_rate": 4.914963511436055e-08,
+      "loss": 0.2959,
+      "step": 16872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3531737978180413,
+      "learning_rate": 4.896553223955658e-08,
+      "loss": 0.2193,
+      "step": 16873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3508589055917248,
+      "learning_rate": 4.878177396864914e-08,
+      "loss": 0.2221,
+      "step": 16874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.42703325762760735,
+      "learning_rate": 4.859836030800091e-08,
+      "loss": 0.261,
+      "step": 16875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29443074994484736,
+      "learning_rate": 4.8415291263962383e-08,
+      "loss": 0.2574,
+      "step": 16876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37489421083434077,
+      "learning_rate": 4.823256684287625e-08,
+      "loss": 0.2146,
+      "step": 16877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1846274794014118,
+      "learning_rate": 4.805018705106745e-08,
+      "loss": 0.6589,
+      "step": 16878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9644830863353985,
+      "learning_rate": 4.7868151894852054e-08,
+      "loss": 0.3864,
+      "step": 16879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24407539568102385,
+      "learning_rate": 4.768646138053501e-08,
+      "loss": 0.2382,
+      "step": 16880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2693149381967699,
+      "learning_rate": 4.750511551440906e-08,
+      "loss": 0.189,
+      "step": 16881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.318880587886877,
+      "learning_rate": 4.732411430275141e-08,
+      "loss": 0.455,
+      "step": 16882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32987152777098505,
+      "learning_rate": 4.71434577518326e-08,
+      "loss": 0.2373,
+      "step": 16883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33653567396866907,
+      "learning_rate": 4.696314586790762e-08,
+      "loss": 0.2356,
+      "step": 16884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6980998648832283,
+      "learning_rate": 4.6783178657221486e-08,
+      "loss": 0.4677,
+      "step": 16885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32680210012665406,
+      "learning_rate": 4.6603556126004756e-08,
+      "loss": 0.2597,
+      "step": 16886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23594603800560673,
+      "learning_rate": 4.642427828047913e-08,
+      "loss": 0.1121,
+      "step": 16887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32282382268927773,
+      "learning_rate": 4.624534512685297e-08,
+      "loss": 0.2613,
+      "step": 16888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3497080476951133,
+      "learning_rate": 4.6066756671322434e-08,
+      "loss": 0.2655,
+      "step": 16889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9983668851762743,
+      "learning_rate": 4.588851292007257e-08,
+      "loss": 0.5257,
+      "step": 16890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40613672708852766,
+      "learning_rate": 4.5710613879275115e-08,
+      "loss": 0.2886,
+      "step": 16891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3492455486029241,
+      "learning_rate": 4.5533059555090684e-08,
+      "loss": 0.2554,
+      "step": 16892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4983795945378701,
+      "learning_rate": 4.535584995366882e-08,
+      "loss": 0.3268,
+      "step": 16893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2967937991637615,
+      "learning_rate": 4.517898508114571e-08,
+      "loss": 0.0822,
+      "step": 16894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3401773758082715,
+      "learning_rate": 4.500246494364535e-08,
+      "loss": 0.2568,
+      "step": 16895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33076498271613985,
+      "learning_rate": 4.482628954728285e-08,
+      "loss": 0.3311,
+      "step": 16896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6425786521864342,
+      "learning_rate": 4.465045889815778e-08,
+      "loss": 0.3003,
+      "step": 16897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2972202986795993,
+      "learning_rate": 4.447497300235859e-08,
+      "loss": 0.2615,
+      "step": 16898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3100186032119291,
+      "learning_rate": 4.4299831865962653e-08,
+      "loss": 0.1789,
+      "step": 16899
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3280509719318347,
+      "learning_rate": 4.412503549503622e-08,
+      "loss": 0.2357,
+      "step": 16900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33145619269969046,
+      "learning_rate": 4.3950583895631116e-08,
+      "loss": 0.2536,
+      "step": 16901
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7463227424917745,
+      "learning_rate": 4.377647707379029e-08,
+      "loss": 0.5095,
+      "step": 16902
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31520197157597846,
+      "learning_rate": 4.360271503554114e-08,
+      "loss": 0.2509,
+      "step": 16903
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33216346525271806,
+      "learning_rate": 4.342929778690108e-08,
+      "loss": 0.2673,
+      "step": 16904
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4729941097580517,
+      "learning_rate": 4.325622533387752e-08,
+      "loss": 0.5392,
+      "step": 16905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3713416552428508,
+      "learning_rate": 4.308349768246234e-08,
+      "loss": 0.1466,
+      "step": 16906
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28593857290381,
+      "learning_rate": 4.291111483863741e-08,
+      "loss": 0.1936,
+      "step": 16907
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34487575236955365,
+      "learning_rate": 4.273907680837241e-08,
+      "loss": 0.2875,
+      "step": 16908
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5930762811522095,
+      "learning_rate": 4.2567383597624804e-08,
+      "loss": 0.394,
+      "step": 16909
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2870976878198615,
+      "learning_rate": 4.239603521234092e-08,
+      "loss": 0.188,
+      "step": 16910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5038426691479727,
+      "learning_rate": 4.2225031658453816e-08,
+      "loss": 0.6532,
+      "step": 16911
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22444525476756616,
+      "learning_rate": 4.205437294188541e-08,
+      "loss": 0.2015,
+      "step": 16912
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2935048318220947,
+      "learning_rate": 4.1884059068546534e-08,
+      "loss": 0.1828,
+      "step": 16913
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5488045078725402,
+      "learning_rate": 4.171409004433358e-08,
+      "loss": 0.3798,
+      "step": 16914
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.887381775060886,
+      "learning_rate": 4.154446587513406e-08,
+      "loss": 0.5585,
+      "step": 16915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2382508451881509,
+      "learning_rate": 4.137518656682216e-08,
+      "loss": 0.2414,
+      "step": 16916
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6288484829732698,
+      "learning_rate": 4.120625212525875e-08,
+      "loss": 0.2158,
+      "step": 16917
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2514810105291575,
+      "learning_rate": 4.10376625562936e-08,
+      "loss": 0.1322,
+      "step": 16918
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4334570249701634,
+      "learning_rate": 4.086941786576759e-08,
+      "loss": 0.2788,
+      "step": 16919
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2745662783710371,
+      "learning_rate": 4.070151805950384e-08,
+      "loss": 0.2448,
+      "step": 16920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8311620037180578,
+      "learning_rate": 4.053396314331881e-08,
+      "loss": 0.5275,
+      "step": 16921
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3124622685734896,
+      "learning_rate": 4.0366753123014526e-08,
+      "loss": 0.2508,
+      "step": 16922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9206862580118762,
+      "learning_rate": 4.0199888004381907e-08,
+      "loss": 0.2277,
+      "step": 16923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22491161238978494,
+      "learning_rate": 4.003336779319855e-08,
+      "loss": 0.2061,
+      "step": 16924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3165803242891882,
+      "learning_rate": 3.9867192495230965e-08,
+      "loss": 0.2662,
+      "step": 16925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6313708675545323,
+      "learning_rate": 3.970136211623343e-08,
+      "loss": 0.2733,
+      "step": 16926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3125624883391193,
+      "learning_rate": 3.9535876661951356e-08,
+      "loss": 0.3,
+      "step": 16927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5088215981290167,
+      "learning_rate": 3.937073613811237e-08,
+      "loss": 0.2573,
+      "step": 16928
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.422677352010954,
+      "learning_rate": 3.920594055043636e-08,
+      "loss": 0.2661,
+      "step": 16929
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32130140929879897,
+      "learning_rate": 3.9041489904629857e-08,
+      "loss": 0.1425,
+      "step": 16930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4239268555816805,
+      "learning_rate": 3.8877384206389426e-08,
+      "loss": 0.3111,
+      "step": 16931
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29273897539680954,
+      "learning_rate": 3.8713623461396066e-08,
+      "loss": 0.2897,
+      "step": 16932
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7208118477900014,
+      "learning_rate": 3.855020767532191e-08,
+      "loss": 0.3823,
+      "step": 16933
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.39365836561310663,
+      "learning_rate": 3.8387136853825776e-08,
+      "loss": 0.2538,
+      "step": 16934
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3864142535331856,
+      "learning_rate": 3.822441100255425e-08,
+      "loss": 0.2928,
+      "step": 16935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4594334507085826,
+      "learning_rate": 3.806203012714394e-08,
+      "loss": 0.2365,
+      "step": 16936
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36622375232538923,
+      "learning_rate": 3.7899994233216996e-08,
+      "loss": 0.2491,
+      "step": 16937
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2574689612287729,
+      "learning_rate": 3.77383033263834e-08,
+      "loss": 0.2046,
+      "step": 16938
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32583886214228214,
+      "learning_rate": 3.757695741224532e-08,
+      "loss": 0.2641,
+      "step": 16939
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30582437989842254,
+      "learning_rate": 3.7415956496388295e-08,
+      "loss": 0.2625,
+      "step": 16940
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.272482518357978,
+      "learning_rate": 3.7255300584388976e-08,
+      "loss": 0.6096,
+      "step": 16941
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6476407086373608,
+      "learning_rate": 3.709498968180958e-08,
+      "loss": 0.371,
+      "step": 16942
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24076540832073126,
+      "learning_rate": 3.693502379420233e-08,
+      "loss": 0.1992,
+      "step": 16943
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47951685652698056,
+      "learning_rate": 3.677540292710724e-08,
+      "loss": 0.3247,
+      "step": 16944
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.434123742526139,
+      "learning_rate": 3.6616127086051e-08,
+      "loss": 0.3129,
+      "step": 16945
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1810864281589373,
+      "learning_rate": 3.64571962765492e-08,
+      "loss": 0.0715,
+      "step": 16946
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32436565764474184,
+      "learning_rate": 3.629861050410743e-08,
+      "loss": 0.2812,
+      "step": 16947
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4181731715703347,
+      "learning_rate": 3.6140369774215755e-08,
+      "loss": 0.3126,
+      "step": 16948
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5866495465077649,
+      "learning_rate": 3.5982474092355334e-08,
+      "loss": 0.1783,
+      "step": 16949
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28246328115634156,
+      "learning_rate": 3.58249234639918e-08,
+      "loss": 0.2231,
+      "step": 16950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3389229225390919,
+      "learning_rate": 3.566771789458412e-08,
+      "loss": 0.3261,
+      "step": 16951
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21439535606761287,
+      "learning_rate": 3.55108573895746e-08,
+      "loss": 0.1327,
+      "step": 16952
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5498107837423801,
+      "learning_rate": 3.535434195439558e-08,
+      "loss": 0.3353,
+      "step": 16953
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6198576731338329,
+      "learning_rate": 3.5198171594467145e-08,
+      "loss": 0.3519,
+      "step": 16954
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38369011378607276,
+      "learning_rate": 3.504234631519721e-08,
+      "loss": 0.2421,
+      "step": 16955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2903915426775922,
+      "learning_rate": 3.4886866121982555e-08,
+      "loss": 0.2236,
+      "step": 16956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.078379182005405,
+      "learning_rate": 3.473173102020666e-08,
+      "loss": 0.7698,
+      "step": 16957
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2756133154465786,
+      "learning_rate": 3.4576941015243003e-08,
+      "loss": 0.1562,
+      "step": 16958
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27127846778563536,
+      "learning_rate": 3.4422496112451745e-08,
+      "loss": 0.2316,
+      "step": 16959
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40386928409495443,
+      "learning_rate": 3.426839631718082e-08,
+      "loss": 0.3061,
+      "step": 16960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4168805787576852,
+      "learning_rate": 3.411464163476597e-08,
+      "loss": 0.2353,
+      "step": 16961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47659370916054206,
+      "learning_rate": 3.3961232070532927e-08,
+      "loss": 0.2327,
+      "step": 16962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34162007685684154,
+      "learning_rate": 3.380816762979411e-08,
+      "loss": 0.3236,
+      "step": 16963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2637327032439912,
+      "learning_rate": 3.3655448317849725e-08,
+      "loss": 0.1839,
+      "step": 16964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4139189954002851,
+      "learning_rate": 3.350307413998888e-08,
+      "loss": 0.2518,
+      "step": 16965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6670530591028055,
+      "learning_rate": 3.335104510148734e-08,
+      "loss": 0.3724,
+      "step": 16966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31514853852025554,
+      "learning_rate": 3.3199361207610916e-08,
+      "loss": 0.2825,
+      "step": 16967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.39693340246090936,
+      "learning_rate": 3.304802246361205e-08,
+      "loss": 0.2747,
+      "step": 16968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.937981789113555,
+      "learning_rate": 3.2897028874731006e-08,
+      "loss": 0.5085,
+      "step": 16969
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35797398099669914,
+      "learning_rate": 3.274638044619805e-08,
+      "loss": 0.2483,
+      "step": 16970
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24761901524686186,
+      "learning_rate": 3.2596077183228993e-08,
+      "loss": 0.2691,
+      "step": 16971
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4298641626817817,
+      "learning_rate": 3.244611909102857e-08,
+      "loss": 0.1868,
+      "step": 16972
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6401200224266734,
+      "learning_rate": 3.229650617479152e-08,
+      "loss": 0.3064,
+      "step": 16973
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3937630678490877,
+      "learning_rate": 3.2147238439697026e-08,
+      "loss": 0.3115,
+      "step": 16974
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31684257741207245,
+      "learning_rate": 3.19983158909154e-08,
+      "loss": 0.248,
+      "step": 16975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.47739044032650685,
+      "learning_rate": 3.1849738533603625e-08,
+      "loss": 0.273,
+      "step": 16976
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4075103274369792,
+      "learning_rate": 3.1701506372906476e-08,
+      "loss": 0.2934,
+      "step": 16977
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3212448915171606,
+      "learning_rate": 3.155361941395763e-08,
+      "loss": 0.1268,
+      "step": 16978
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32874130733070184,
+      "learning_rate": 3.140607766187853e-08,
+      "loss": 0.251,
+      "step": 16979
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39753759795242666,
+      "learning_rate": 3.125888112177733e-08,
+      "loss": 0.3094,
+      "step": 16980
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8441004038029325,
+      "learning_rate": 3.1112029798753274e-08,
+      "loss": 0.5548,
+      "step": 16981
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5623939560608282,
+      "learning_rate": 3.096552369789119e-08,
+      "loss": 0.2477,
+      "step": 16982
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24903041506543536,
+      "learning_rate": 3.081936282426368e-08,
+      "loss": 0.2619,
+      "step": 16983
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29800042452218295,
+      "learning_rate": 3.067354718293336e-08,
+      "loss": 0.1459,
+      "step": 16984
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8593406624322196,
+      "learning_rate": 3.0528076778949536e-08,
+      "loss": 0.1092,
+      "step": 16985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34340164329049183,
+      "learning_rate": 3.038295161734928e-08,
+      "loss": 0.2913,
+      "step": 16986
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32860953048223024,
+      "learning_rate": 3.023817170315857e-08,
+      "loss": 0.2968,
+      "step": 16987
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6455934770032342,
+      "learning_rate": 3.0093737041392293e-08,
+      "loss": 0.2445,
+      "step": 16988
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4043626940495485,
+      "learning_rate": 2.994964763704977e-08,
+      "loss": 0.2695,
+      "step": 16989
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22943688435845902,
+      "learning_rate": 2.980590349512258e-08,
+      "loss": 0.16,
+      "step": 16990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3037975887387826,
+      "learning_rate": 2.966250462058895e-08,
+      "loss": 0.2288,
+      "step": 16991
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3512971318276826,
+      "learning_rate": 2.9519451018413804e-08,
+      "loss": 0.2918,
+      "step": 16992
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9975814522598434,
+      "learning_rate": 2.9376742693550954e-08,
+      "loss": 0.4212,
+      "step": 16993
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7888704691744978,
+      "learning_rate": 2.9234379650943113e-08,
+      "loss": 0.3391,
+      "step": 16994
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2782399609130223,
+      "learning_rate": 2.9092361895519673e-08,
+      "loss": 0.2218,
+      "step": 16995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22388047250656798,
+      "learning_rate": 2.895068943219892e-08,
+      "loss": 0.1858,
+      "step": 16996
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8291268218917041,
+      "learning_rate": 2.880936226588693e-08,
+      "loss": 0.3896,
+      "step": 16997
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3064483112237091,
+      "learning_rate": 2.866838040147868e-08,
+      "loss": 0.21,
+      "step": 16998
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3857237505293713,
+      "learning_rate": 2.8527743843854704e-08,
+      "loss": 0.309,
+      "step": 16999
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8772816244367653,
+      "learning_rate": 2.8387452597886666e-08,
+      "loss": 0.4426,
+      "step": 17000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3463263519606244,
+      "learning_rate": 2.824750666843179e-08,
+      "loss": 0.1876,
+      "step": 17001
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26138233646250353,
+      "learning_rate": 2.810790606033731e-08,
+      "loss": 0.203,
+      "step": 17002
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5924213734480587,
+      "learning_rate": 2.7968650778438245e-08,
+      "loss": 0.2967,
+      "step": 17003
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3005423451194237,
+      "learning_rate": 2.7829740827555185e-08,
+      "loss": 0.2472,
+      "step": 17004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6290851599106121,
+      "learning_rate": 2.769117621249873e-08,
+      "loss": 0.3087,
+      "step": 17005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5028037464387212,
+      "learning_rate": 2.7552956938068364e-08,
+      "loss": 0.366,
+      "step": 17006
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26322989882195386,
+      "learning_rate": 2.741508300905138e-08,
+      "loss": 0.2482,
+      "step": 17007
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7885337997270865,
+      "learning_rate": 2.727755443021951e-08,
+      "loss": 0.1453,
+      "step": 17008
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3236513169203668,
+      "learning_rate": 2.7140371206337834e-08,
+      "loss": 0.2128,
+      "step": 17009
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.410045538047261,
+      "learning_rate": 2.7003533342156995e-08,
+      "loss": 0.2497,
+      "step": 17010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3294855926484948,
+      "learning_rate": 2.6867040842414316e-08,
+      "loss": 0.2396,
+      "step": 17011
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7913529950119497,
+      "learning_rate": 2.6730893711837124e-08,
+      "loss": 0.4242,
+      "step": 17012
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34653414142176686,
+      "learning_rate": 2.6595091955139428e-08,
+      "loss": 0.2621,
+      "step": 17013
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27650331142096424,
+      "learning_rate": 2.6459635577026355e-08,
+      "loss": 0.1645,
+      "step": 17014
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34663882196846774,
+      "learning_rate": 2.6324524582186374e-08,
+      "loss": 0.2303,
+      "step": 17015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34330890321049773,
+      "learning_rate": 2.6189758975299074e-08,
+      "loss": 0.2632,
+      "step": 17016
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6417232033799707,
+      "learning_rate": 2.6055338761031835e-08,
+      "loss": 0.4148,
+      "step": 17017
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3119913260890965,
+      "learning_rate": 2.592126394403982e-08,
+      "loss": 0.2295,
+      "step": 17018
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.37198388859885884,
+      "learning_rate": 2.5787534528964875e-08,
+      "loss": 0.2445,
+      "step": 17019
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3027242854603744,
+      "learning_rate": 2.5654150520438848e-08,
+      "loss": 0.3769,
+      "step": 17020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3261165338744473,
+      "learning_rate": 2.5521111923080266e-08,
+      "loss": 0.1333,
+      "step": 17021
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3275988014909243,
+      "learning_rate": 2.5388418741497668e-08,
+      "loss": 0.2532,
+      "step": 17022
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3388604127745746,
+      "learning_rate": 2.5256070980284042e-08,
+      "loss": 0.2924,
+      "step": 17023
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4462487671630624,
+      "learning_rate": 2.5124068644024613e-08,
+      "loss": 0.2823,
+      "step": 17024
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3366263902770327,
+      "learning_rate": 2.4992411737289057e-08,
+      "loss": 0.2503,
+      "step": 17025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4836462318839685,
+      "learning_rate": 2.4861100264638172e-08,
+      "loss": 0.6723,
+      "step": 17026
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27144366896564376,
+      "learning_rate": 2.473013423061832e-08,
+      "loss": 0.221,
+      "step": 17027
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31117856512925224,
+      "learning_rate": 2.459951363976476e-08,
+      "loss": 0.2513,
+      "step": 17028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38810367971759135,
+      "learning_rate": 2.4469238496600546e-08,
+      "loss": 0.2505,
+      "step": 17029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33623722779449305,
+      "learning_rate": 2.433930880563762e-08,
+      "loss": 0.2981,
+      "step": 17030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2907484340689656,
+      "learning_rate": 2.4209724571376826e-08,
+      "loss": 0.1823,
+      "step": 17031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.158270296591638,
+      "learning_rate": 2.4080485798302355e-08,
+      "loss": 0.4486,
+      "step": 17032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5911414484895722,
+      "learning_rate": 2.395159249089285e-08,
+      "loss": 0.3196,
+      "step": 17033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2877702121234691,
+      "learning_rate": 2.3823044653610295e-08,
+      "loss": 0.1904,
+      "step": 17034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2822114281047928,
+      "learning_rate": 2.3694842290907792e-08,
+      "loss": 0.2648,
+      "step": 17035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4802254996038075,
+      "learning_rate": 2.356698540722291e-08,
+      "loss": 0.2575,
+      "step": 17036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2828728976084193,
+      "learning_rate": 2.343947400698432e-08,
+      "loss": 0.1845,
+      "step": 17037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3516289049644702,
+      "learning_rate": 2.3312308094607382e-08,
+      "loss": 0.2959,
+      "step": 17038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7073007426790366,
+      "learning_rate": 2.3185487674497463e-08,
+      "loss": 0.3777,
+      "step": 17039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3036204352872146,
+      "learning_rate": 2.3059012751044386e-08,
+      "loss": 0.2045,
+      "step": 17040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0686989334600154,
+      "learning_rate": 2.2932883328629087e-08,
+      "loss": 0.7031,
+      "step": 17041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26671675957112734,
+      "learning_rate": 2.280709941161807e-08,
+      "loss": 0.2414,
+      "step": 17042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3138331144274908,
+      "learning_rate": 2.268166100436897e-08,
+      "loss": 0.2286,
+      "step": 17043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5218556651137621,
+      "learning_rate": 2.2556568111223866e-08,
+      "loss": 0.2401,
+      "step": 17044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7268865897870525,
+      "learning_rate": 2.2431820736517062e-08,
+      "loss": 0.3442,
+      "step": 17045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34937516121279033,
+      "learning_rate": 2.2307418884566225e-08,
+      "loss": 0.2662,
+      "step": 17046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32484541126394684,
+      "learning_rate": 2.2183362559681232e-08,
+      "loss": 0.2435,
+      "step": 17047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2538310862070762,
+      "learning_rate": 2.205965176615643e-08,
+      "loss": 0.1857,
+      "step": 17048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36626610756991873,
+      "learning_rate": 2.1936286508278393e-08,
+      "loss": 0.234,
+      "step": 17049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31560292290674424,
+      "learning_rate": 2.1813266790315922e-08,
+      "loss": 0.2447,
+      "step": 17050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8116602561966607,
+      "learning_rate": 2.169059261653228e-08,
+      "loss": 0.3647,
+      "step": 17051
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.646895185331955,
+      "learning_rate": 2.1568263991174065e-08,
+      "loss": 0.2433,
+      "step": 17052
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3426135098172122,
+      "learning_rate": 2.144628091847678e-08,
+      "loss": 0.2603,
+      "step": 17053
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2182673378928602,
+      "learning_rate": 2.1324643402667045e-08,
+      "loss": 0.2074,
+      "step": 17054
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4041287114922914,
+      "learning_rate": 2.1203351447954824e-08,
+      "loss": 0.2777,
+      "step": 17055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4575676863248586,
+      "learning_rate": 2.10824050585412e-08,
+      "loss": 0.2954,
+      "step": 17056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0410511963285844,
+      "learning_rate": 2.0961804238616156e-08,
+      "loss": 0.2411,
+      "step": 17057
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29720838275806993,
+      "learning_rate": 2.0841548992354132e-08,
+      "loss": 0.2345,
+      "step": 17058
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.37419418678071553,
+      "learning_rate": 2.0721639323919573e-08,
+      "loss": 0.3246,
+      "step": 17059
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.16552952889999026,
+      "learning_rate": 2.0602075237465825e-08,
+      "loss": 0.0844,
+      "step": 17060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41038045330779327,
+      "learning_rate": 2.0482856737132906e-08,
+      "loss": 0.2749,
+      "step": 17061
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29247558336905394,
+      "learning_rate": 2.0363983827049737e-08,
+      "loss": 0.2884,
+      "step": 17062
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2705941693360938,
+      "learning_rate": 2.0245456511333028e-08,
+      "loss": 0.2699,
+      "step": 17063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3362945356885263,
+      "learning_rate": 2.012727479408616e-08,
+      "loss": 0.2474,
+      "step": 17064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.231997007937773,
+      "learning_rate": 2.0009438679403636e-08,
+      "loss": 0.6624,
+      "step": 17065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.329125988206052,
+      "learning_rate": 1.9891948171364417e-08,
+      "loss": 0.2482,
+      "step": 17066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4091946800044302,
+      "learning_rate": 1.9774803274038578e-08,
+      "loss": 0.2754,
+      "step": 17067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35837796725133636,
+      "learning_rate": 1.9658003991480656e-08,
+      "loss": 0.258,
+      "step": 17068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5680477821646825,
+      "learning_rate": 1.9541550327738524e-08,
+      "loss": 0.2052,
+      "step": 17069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2690535902574051,
+      "learning_rate": 1.942544228684229e-08,
+      "loss": 0.2,
+      "step": 17070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3929295717834575,
+      "learning_rate": 1.930967987281429e-08,
+      "loss": 0.3381,
+      "step": 17071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8100108431977137,
+      "learning_rate": 1.9194263089662435e-08,
+      "loss": 0.4147,
+      "step": 17072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4949306445932574,
+      "learning_rate": 1.9079191941384635e-08,
+      "loss": 0.2105,
+      "step": 17073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2799353141961879,
+      "learning_rate": 1.8964466431964367e-08,
+      "loss": 0.2766,
+      "step": 17074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25292916616971384,
+      "learning_rate": 1.8850086565376236e-08,
+      "loss": 0.1642,
+      "step": 17075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43814973227571347,
+      "learning_rate": 1.8736052345580403e-08,
+      "loss": 0.188,
+      "step": 17076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5327994711284099,
+      "learning_rate": 1.862236377652593e-08,
+      "loss": 0.3839,
+      "step": 17077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35915994320120836,
+      "learning_rate": 1.8509020862149673e-08,
+      "loss": 0.3232,
+      "step": 17078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4653115137961894,
+      "learning_rate": 1.8396023606376268e-08,
+      "loss": 0.3291,
+      "step": 17079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29836713476307714,
+      "learning_rate": 1.828337201311925e-08,
+      "loss": 0.1782,
+      "step": 17080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5164581462048005,
+      "learning_rate": 1.817106608628105e-08,
+      "loss": 0.2172,
+      "step": 17081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28972253741358867,
+      "learning_rate": 1.805910582974857e-08,
+      "loss": 0.2673,
+      "step": 17082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3477015924607169,
+      "learning_rate": 1.7947491247399808e-08,
+      "loss": 0.2427,
+      "step": 17083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.577825325414684,
+      "learning_rate": 1.783622234310056e-08,
+      "loss": 0.3721,
+      "step": 17084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41418340683900945,
+      "learning_rate": 1.772529912070442e-08,
+      "loss": 0.2571,
+      "step": 17085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2782321577020592,
+      "learning_rate": 1.7614721584051643e-08,
+      "loss": 0.2388,
+      "step": 17086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3467169897838902,
+      "learning_rate": 1.7504489736971385e-08,
+      "loss": 0.1657,
+      "step": 17087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7927895004364096,
+      "learning_rate": 1.73946035832806e-08,
+      "loss": 0.5113,
+      "step": 17088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29960578060265763,
+      "learning_rate": 1.728506312678624e-08,
+      "loss": 0.2258,
+      "step": 17089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32245881570389173,
+      "learning_rate": 1.7175868371281936e-08,
+      "loss": 0.3108,
+      "step": 17090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.951002243046634,
+      "learning_rate": 1.7067019320546886e-08,
+      "loss": 0.3718,
+      "step": 17091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38421464500733016,
+      "learning_rate": 1.6958515978351407e-08,
+      "loss": 0.2816,
+      "step": 17092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.15888882951589792,
+      "learning_rate": 1.6850358348453612e-08,
+      "loss": 0.0663,
+      "step": 17093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3003963983604446,
+      "learning_rate": 1.6742546434598272e-08,
+      "loss": 0.2845,
+      "step": 17094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36253705809280756,
+      "learning_rate": 1.6635080240520186e-08,
+      "loss": 0.2953,
+      "step": 17095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6824644163387067,
+      "learning_rate": 1.6527959769939705e-08,
+      "loss": 0.2952,
+      "step": 17096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6366164928559752,
+      "learning_rate": 1.6421185026566088e-08,
+      "loss": 0.2909,
+      "step": 17097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24641835079060603,
+      "learning_rate": 1.631475601409749e-08,
+      "loss": 0.2552,
+      "step": 17098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31674059694920764,
+      "learning_rate": 1.6208672736219843e-08,
+      "loss": 0.0891,
+      "step": 17099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5181426442377051,
+      "learning_rate": 1.6102935196606883e-08,
+      "loss": 0.3263,
+      "step": 17100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3799271362409939,
+      "learning_rate": 1.5997543398919013e-08,
+      "loss": 0.2932,
+      "step": 17101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32179637117374854,
+      "learning_rate": 1.5892497346807754e-08,
+      "loss": 0.271,
+      "step": 17102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7109955981462678,
+      "learning_rate": 1.57877970439102e-08,
+      "loss": 0.3308,
+      "step": 17103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3955832021279465,
+      "learning_rate": 1.568344249385123e-08,
+      "loss": 0.2954,
+      "step": 17104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28272887435926286,
+      "learning_rate": 1.5579433700245727e-08,
+      "loss": 0.1893,
+      "step": 17105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3781045982315596,
+      "learning_rate": 1.5475770666694145e-08,
+      "loss": 0.2298,
+      "step": 17106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3939611312355522,
+      "learning_rate": 1.5372453396788057e-08,
+      "loss": 0.284,
+      "step": 17107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7118015447019852,
+      "learning_rate": 1.52694818941046e-08,
+      "loss": 0.439,
+      "step": 17108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4439561205053077,
+      "learning_rate": 1.516685616220981e-08,
+      "loss": 0.1962,
+      "step": 17109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28543838310868797,
+      "learning_rate": 1.506457620465751e-08,
+      "loss": 0.2636,
+      "step": 17110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32304533937000707,
+      "learning_rate": 1.4962642024989316e-08,
+      "loss": 0.1582,
+      "step": 17111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.556431431252915,
+      "learning_rate": 1.4861053626734623e-08,
+      "loss": 0.2584,
+      "step": 17112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3452856886959734,
+      "learning_rate": 1.475981101341284e-08,
+      "loss": 0.2698,
+      "step": 17113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34444279352265655,
+      "learning_rate": 1.4658914188530049e-08,
+      "loss": 0.3137,
+      "step": 17114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0069469226681833,
+      "learning_rate": 1.4558363155579013e-08,
+      "loss": 0.1646,
+      "step": 17115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3590631076036183,
+      "learning_rate": 1.4458157918042503e-08,
+      "loss": 0.2807,
+      "step": 17116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3076290801115484,
+      "learning_rate": 1.4358298479391076e-08,
+      "loss": 0.2447,
+      "step": 17117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4345708272045855,
+      "learning_rate": 1.4258784843081963e-08,
+      "loss": 0.3126,
+      "step": 17118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35796512209441345,
+      "learning_rate": 1.415961701256241e-08,
+      "loss": 0.2191,
+      "step": 17119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43834411557560704,
+      "learning_rate": 1.4060794991265226e-08,
+      "loss": 0.3003,
+      "step": 17120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3023620756441359,
+      "learning_rate": 1.3962318782613226e-08,
+      "loss": 0.267,
+      "step": 17121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33571420647735895,
+      "learning_rate": 1.386418839001702e-08,
+      "loss": 0.1946,
+      "step": 17122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7962831409761743,
+      "learning_rate": 1.3766403816873886e-08,
+      "loss": 0.3344,
+      "step": 17123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40328605944188584,
+      "learning_rate": 1.3668965066571115e-08,
+      "loss": 0.2527,
+      "step": 17124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27251997461785155,
+      "learning_rate": 1.3571872142483789e-08,
+      "loss": 0.1927,
+      "step": 17125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27747022869745464,
+      "learning_rate": 1.3475125047971438e-08,
+      "loss": 0.2665,
+      "step": 17126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2894017763114165,
+      "learning_rate": 1.3378723786386938e-08,
+      "loss": 0.6075,
+      "step": 17127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3370384947172715,
+      "learning_rate": 1.3282668361067619e-08,
+      "loss": 0.185,
+      "step": 17128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33942967898729376,
+      "learning_rate": 1.3186958775339709e-08,
+      "loss": 0.3023,
+      "step": 17129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6884605520324051,
+      "learning_rate": 1.3091595032518333e-08,
+      "loss": 0.3898,
+      "step": 17130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32177530158067874,
+      "learning_rate": 1.2996577135906408e-08,
+      "loss": 0.2731,
+      "step": 17131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22547779418670175,
+      "learning_rate": 1.290190508879241e-08,
+      "loss": 0.1392,
+      "step": 17132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26112428619009537,
+      "learning_rate": 1.280757889445594e-08,
+      "loss": 0.2366,
+      "step": 17133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35491189542978396,
+      "learning_rate": 1.2713598556164386e-08,
+      "loss": 0.2418,
+      "step": 17134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6705417107991852,
+      "learning_rate": 1.2619964077170699e-08,
+      "loss": 0.2883,
+      "step": 17135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7273092453633878,
+      "learning_rate": 1.252667546071784e-08,
+      "loss": 0.3569,
+      "step": 17136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3310531820505807,
+      "learning_rate": 1.2433732710037671e-08,
+      "loss": 0.2706,
+      "step": 17137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23516781376709966,
+      "learning_rate": 1.2341135828347617e-08,
+      "loss": 0.2121,
+      "step": 17138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5182882686249745,
+      "learning_rate": 1.2248884818854001e-08,
+      "loss": 0.2102,
+      "step": 17139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.347744576905726,
+      "learning_rate": 1.2156979684753157e-08,
+      "loss": 0.2669,
+      "step": 17140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30419991686902886,
+      "learning_rate": 1.2065420429225872e-08,
+      "loss": 0.2622,
+      "step": 17141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6698959296242852,
+      "learning_rate": 1.1974207055444054e-08,
+      "loss": 0.6203,
+      "step": 17142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3200165060424897,
+      "learning_rate": 1.1883339566565178e-08,
+      "loss": 0.2531,
+      "step": 17143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.411575213284914,
+      "learning_rate": 1.1792817965736725e-08,
+      "loss": 0.2983,
+      "step": 17144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2582650364241364,
+      "learning_rate": 1.1702642256093965e-08,
+      "loss": 0.1813,
+      "step": 17145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3650810325649037,
+      "learning_rate": 1.1612812440758847e-08,
+      "loss": 0.2605,
+      "step": 17146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5358995044698758,
+      "learning_rate": 1.1523328522843324e-08,
+      "loss": 0.3528,
+      "step": 17147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9446703893782842,
+      "learning_rate": 1.1434190505443809e-08,
+      "loss": 0.384,
+      "step": 17148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2741899067962737,
+      "learning_rate": 1.1345398391650053e-08,
+      "loss": 0.2475,
+      "step": 17149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5004404118735434,
+      "learning_rate": 1.1256952184535153e-08,
+      "loss": 0.3734,
+      "step": 17150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.363040297891496,
+      "learning_rate": 1.1168851887163323e-08,
+      "loss": 0.1338,
+      "step": 17151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3385171036240212,
+      "learning_rate": 1.1081097502584349e-08,
+      "loss": 0.2484,
+      "step": 17152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4682014674094445,
+      "learning_rate": 1.0993689033836907e-08,
+      "loss": 0.3082,
+      "step": 17153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.541802237828164,
+      "learning_rate": 1.090662648394858e-08,
+      "loss": 0.2646,
+      "step": 17154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4216574372674074,
+      "learning_rate": 1.0819909855933618e-08,
+      "loss": 0.2951,
+      "step": 17155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5286671846393574,
+      "learning_rate": 1.073353915279629e-08,
+      "loss": 0.3723,
+      "step": 17156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26883762854101706,
+      "learning_rate": 1.0647514377527535e-08,
+      "loss": 0.2437,
+      "step": 17157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2838856173997619,
+      "learning_rate": 1.0561835533104969e-08,
+      "loss": 0.1802,
+      "step": 17158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48001083283485046,
+      "learning_rate": 1.0476502622496221e-08,
+      "loss": 0.2635,
+      "step": 17159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2070202127543408,
+      "learning_rate": 1.0391515648657813e-08,
+      "loss": 0.6532,
+      "step": 17160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24790652896927554,
+      "learning_rate": 1.0306874614530727e-08,
+      "loss": 0.2034,
+      "step": 17161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5456230843925484,
+      "learning_rate": 1.022257952304706e-08,
+      "loss": 0.3501,
+      "step": 17162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.872555453274828,
+      "learning_rate": 1.0138630377125591e-08,
+      "loss": 0.3599,
+      "step": 17163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30090643775201775,
+      "learning_rate": 1.0055027179675104e-08,
+      "loss": 0.211,
+      "step": 17164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20651735036815919,
+      "learning_rate": 9.971769933587728e-09,
+      "loss": 0.1962,
+      "step": 17165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3557084430328847,
+      "learning_rate": 9.888858641750044e-09,
+      "loss": 0.5679,
+      "step": 17166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4083475818641006,
+      "learning_rate": 9.806293307030868e-09,
+      "loss": 0.2982,
+      "step": 17167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5826905889124464,
+      "learning_rate": 9.724073932289024e-09,
+      "loss": 0.2523,
+      "step": 17168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3194676400384813,
+      "learning_rate": 9.642200520374457e-09,
+      "loss": 0.299,
+      "step": 17169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29693589850133373,
+      "learning_rate": 9.560673074120452e-09,
+      "loss": 0.2682,
+      "step": 17170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.16895972237669396,
+      "learning_rate": 9.47949159635031e-09,
+      "loss": 0.0703,
+      "step": 17171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46335545917394955,
+      "learning_rate": 9.398656089876224e-09,
+      "loss": 0.2993,
+      "step": 17172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29457119427898515,
+      "learning_rate": 9.318166557497066e-09,
+      "loss": 0.2394,
+      "step": 17173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.47416681421393464,
+      "learning_rate": 9.238023001999496e-09,
+      "loss": 0.2596,
+      "step": 17174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6675035464853056,
+      "learning_rate": 9.158225426160183e-09,
+      "loss": 0.3474,
+      "step": 17175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5988924080851362,
+      "learning_rate": 9.078773832742471e-09,
+      "loss": 0.2719,
+      "step": 17176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18081883458352935,
+      "learning_rate": 8.999668224496383e-09,
+      "loss": 0.1884,
+      "step": 17177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1298631304084927,
+      "learning_rate": 8.92090860416195e-09,
+      "loss": 0.555,
+      "step": 17178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40152654309722646,
+      "learning_rate": 8.842494974466986e-09,
+      "loss": 0.2712,
+      "step": 17179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.49817004543825866,
+      "learning_rate": 8.764427338127102e-09,
+      "loss": 0.3426,
+      "step": 17180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31938785931852637,
+      "learning_rate": 8.686705697845688e-09,
+      "loss": 0.2383,
+      "step": 17181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3866142623720773,
+      "learning_rate": 8.609330056313926e-09,
+      "loss": 0.2452,
+      "step": 17182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2502692400459461,
+      "learning_rate": 8.532300416210781e-09,
+      "loss": 0.1758,
+      "step": 17183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5478422286740277,
+      "learning_rate": 8.455616780205234e-09,
+      "loss": 0.243,
+      "step": 17184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3054428534977997,
+      "learning_rate": 8.379279150951824e-09,
+      "loss": 0.259,
+      "step": 17185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.563806009862002,
+      "learning_rate": 8.303287531093996e-09,
+      "loss": 0.3525,
+      "step": 17186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6691683684864551,
+      "learning_rate": 8.227641923264085e-09,
+      "loss": 0.3025,
+      "step": 17187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3912003511948697,
+      "learning_rate": 8.152342330081109e-09,
+      "loss": 0.247,
+      "step": 17188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29904732930023953,
+      "learning_rate": 8.077388754151872e-09,
+      "loss": 0.2665,
+      "step": 17189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28293767680946147,
+      "learning_rate": 8.002781198074294e-09,
+      "loss": 0.1153,
+      "step": 17190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39065053153705637,
+      "learning_rate": 7.928519664430756e-09,
+      "loss": 0.2872,
+      "step": 17191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4524577827612803,
+      "learning_rate": 7.854604155791423e-09,
+      "loss": 0.3092,
+      "step": 17192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3234766790264359,
+      "learning_rate": 7.78103467471869e-09,
+      "loss": 0.301,
+      "step": 17193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5498364437172288,
+      "learning_rate": 7.70781122375941e-09,
+      "loss": 0.0939,
+      "step": 17194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.290298275004318,
+      "learning_rate": 7.634933805448219e-09,
+      "loss": 0.2089,
+      "step": 17195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2698869203749294,
+      "learning_rate": 7.562402422309767e-09,
+      "loss": 0.2444,
+      "step": 17196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5583643820899699,
+      "learning_rate": 7.490217076855377e-09,
+      "loss": 0.1874,
+      "step": 17197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.37483103695911446,
+      "learning_rate": 7.418377771585273e-09,
+      "loss": 0.3031,
+      "step": 17198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2057443310019202,
+      "learning_rate": 7.346884508987462e-09,
+      "loss": 0.5939,
+      "step": 17199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40814755288192517,
+      "learning_rate": 7.275737291536633e-09,
+      "loss": 0.1775,
+      "step": 17200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23334911812197162,
+      "learning_rate": 7.204936121697481e-09,
+      "loss": 0.2408,
+      "step": 17201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.452400964531313,
+      "learning_rate": 7.134481001922488e-09,
+      "loss": 0.253,
+      "step": 17202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3952783924255758,
+      "learning_rate": 7.064371934649706e-09,
+      "loss": 0.2208,
+      "step": 17203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5326550442015726,
+      "learning_rate": 6.9946089223082995e-09,
+      "loss": 0.3627,
+      "step": 17204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34586123136046604,
+      "learning_rate": 6.925191967314115e-09,
+      "loss": 0.3045,
+      "step": 17205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7623488186505916,
+      "learning_rate": 6.856121072070787e-09,
+      "loss": 0.3254,
+      "step": 17206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4536690860312234,
+      "learning_rate": 6.787396238969735e-09,
+      "loss": 0.2152,
+      "step": 17207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23043597811227537,
+      "learning_rate": 6.719017470392386e-09,
+      "loss": 0.2123,
+      "step": 17208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39678585504428626,
+      "learning_rate": 6.650984768704627e-09,
+      "loss": 0.2842,
+      "step": 17209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.484976822562212,
+      "learning_rate": 6.583298136264571e-09,
+      "loss": 0.2722,
+      "step": 17210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1659165788472297,
+      "learning_rate": 6.515957575413678e-09,
+      "loss": 0.5502,
+      "step": 17211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4491331575497026,
+      "learning_rate": 6.448963088486748e-09,
+      "loss": 0.3109,
+      "step": 17212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27854955150557587,
+      "learning_rate": 6.382314677803037e-09,
+      "loss": 0.2077,
+      "step": 17213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.41626719286042163,
+      "learning_rate": 6.316012345668476e-09,
+      "loss": 0.2705,
+      "step": 17214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7680084509491535,
+      "learning_rate": 6.2500560943812295e-09,
+      "loss": 0.3548,
+      "step": 17215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25815630989197186,
+      "learning_rate": 6.184445926225024e-09,
+      "loss": 0.2255,
+      "step": 17216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3151753336653738,
+      "learning_rate": 6.119181843471378e-09,
+      "loss": 0.2446,
+      "step": 17217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9113337178352028,
+      "learning_rate": 6.0542638483818136e-09,
+      "loss": 0.3112,
+      "step": 17218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3686064816471827,
+      "learning_rate": 5.989691943202314e-09,
+      "loss": 0.2553,
+      "step": 17219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4338394855360507,
+      "learning_rate": 5.925466130169977e-09,
+      "loss": 0.2657,
+      "step": 17220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2913724177552018,
+      "learning_rate": 5.86158641150969e-09,
+      "loss": 0.2284,
+      "step": 17221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3690652954085403,
+      "learning_rate": 5.798052789431907e-09,
+      "loss": 0.3086,
+      "step": 17222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44592350230275896,
+      "learning_rate": 5.734865266138201e-09,
+      "loss": 0.1507,
+      "step": 17223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29599152713750027,
+      "learning_rate": 5.67202384381682e-09,
+      "loss": 0.2542,
+      "step": 17224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38827856045711373,
+      "learning_rate": 5.609528524642694e-09,
+      "loss": 0.3015,
+      "step": 17225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5828565839256031,
+      "learning_rate": 5.547379310781864e-09,
+      "loss": 0.2791,
+      "step": 17226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7721637320865353,
+      "learning_rate": 5.485576204383725e-09,
+      "loss": 0.4538,
+      "step": 17227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3448076249812229,
+      "learning_rate": 5.424119207592115e-09,
+      "loss": 0.2812,
+      "step": 17228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21500697134765617,
+      "learning_rate": 5.3630083225331145e-09,
+      "loss": 0.1745,
+      "step": 17229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.028394499800441,
+      "learning_rate": 5.302243551322806e-09,
+      "loss": 0.4091,
+      "step": 17230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3682420380289012,
+      "learning_rate": 5.2418248960661725e-09,
+      "loss": 0.2909,
+      "step": 17231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3277072529659946,
+      "learning_rate": 5.181752358854874e-09,
+      "loss": 0.3007,
+      "step": 17232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4392349799183583,
+      "learning_rate": 5.1220259417705806e-09,
+      "loss": 0.1281,
+      "step": 17233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3499626554649006,
+      "learning_rate": 5.0626456468805265e-09,
+      "loss": 0.2961,
+      "step": 17234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28294895281739385,
+      "learning_rate": 5.003611476240844e-09,
+      "loss": 0.1918,
+      "step": 17235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30877597909823423,
+      "learning_rate": 4.944923431896564e-09,
+      "loss": 0.218,
+      "step": 17236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35228097598873376,
+      "learning_rate": 4.886581515880506e-09,
+      "loss": 0.2528,
+      "step": 17237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8101994452225071,
+      "learning_rate": 4.828585730211055e-09,
+      "loss": 0.3758,
+      "step": 17238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8340505753102018,
+      "learning_rate": 4.770936076898825e-09,
+      "loss": 0.2225,
+      "step": 17239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29161964930278367,
+      "learning_rate": 4.713632557938885e-09,
+      "loss": 0.2682,
+      "step": 17240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3629011642961109,
+      "learning_rate": 4.6566751753163166e-09,
+      "loss": 0.2429,
+      "step": 17241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.455365093725818,
+      "learning_rate": 4.600063931002874e-09,
+      "loss": 0.2451,
+      "step": 17242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.336152325585062,
+      "learning_rate": 4.543798826959211e-09,
+      "loss": 0.2254,
+      "step": 17243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3599515795554048,
+      "learning_rate": 4.4878798651337705e-09,
+      "loss": 0.2767,
+      "step": 17244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9079207839441679,
+      "learning_rate": 4.4323070474638906e-09,
+      "loss": 0.4227,
+      "step": 17245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3120814844657969,
+      "learning_rate": 4.377080375873588e-09,
+      "loss": 0.2139,
+      "step": 17246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2816458631063591,
+      "learning_rate": 4.322199852274667e-09,
+      "loss": 0.2005,
+      "step": 17247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34555457410158374,
+      "learning_rate": 4.267665478567829e-09,
+      "loss": 0.2776,
+      "step": 17248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3508270629408411,
+      "learning_rate": 4.213477256642673e-09,
+      "loss": 0.22,
+      "step": 17249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8936437076660343,
+      "learning_rate": 4.159635188375477e-09,
+      "loss": 0.3986,
+      "step": 17250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2272218348029869,
+      "learning_rate": 4.106139275629195e-09,
+      "loss": 0.6651,
+      "step": 17251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22314666519834417,
+      "learning_rate": 4.0529895202579e-09,
+      "loss": 0.2085,
+      "step": 17252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4471815041901033,
+      "learning_rate": 4.00018592410123e-09,
+      "loss": 0.6434,
+      "step": 17253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3906800807951317,
+      "learning_rate": 3.947728488988833e-09,
+      "loss": 0.2578,
+      "step": 17254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2559969138747373,
+      "learning_rate": 3.895617216735925e-09,
+      "loss": 0.199,
+      "step": 17255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3663610961053545,
+      "learning_rate": 3.843852109148838e-09,
+      "loss": 0.2337,
+      "step": 17256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1749116983977523,
+      "learning_rate": 3.792433168019471e-09,
+      "loss": 0.7044,
+      "step": 17257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32863069988556126,
+      "learning_rate": 3.741360395127513e-09,
+      "loss": 0.2506,
+      "step": 17258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7208685948356803,
+      "learning_rate": 3.6906337922426593e-09,
+      "loss": 0.2665,
+      "step": 17259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3628388138696765,
+      "learning_rate": 3.640253361121282e-09,
+      "loss": 0.309,
+      "step": 17260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.36290377979163024,
+      "learning_rate": 3.590219103508652e-09,
+      "loss": 0.2359,
+      "step": 17261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3023237550567913,
+      "learning_rate": 3.540531021135607e-09,
+      "loss": 0.1374,
+      "step": 17262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3912028186810661,
+      "learning_rate": 3.491189115725213e-09,
+      "loss": 0.3087,
+      "step": 17263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3489042396908763,
+      "learning_rate": 3.4421933889849936e-09,
+      "loss": 0.2758,
+      "step": 17264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3906870321219293,
+      "learning_rate": 3.3935438426113687e-09,
+      "loss": 0.262,
+      "step": 17265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9233473327277593,
+      "learning_rate": 3.3452404782896577e-09,
+      "loss": 0.4265,
+      "step": 17266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2371891988216668,
+      "learning_rate": 3.2972832976918557e-09,
+      "loss": 0.1978,
+      "step": 17267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26164169308525725,
+      "learning_rate": 3.2496723024799672e-09,
+      "loss": 0.2508,
+      "step": 17268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0699215886321223,
+      "learning_rate": 3.2024074943015626e-09,
+      "loss": 0.4488,
+      "step": 17269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3099846689515529,
+      "learning_rate": 3.1554888747942213e-09,
+      "loss": 0.2469,
+      "step": 17270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6191979829959496,
+      "learning_rate": 3.10891644558331e-09,
+      "loss": 0.3591,
+      "step": 17271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34374932351710824,
+      "learning_rate": 3.0626902082797615e-09,
+      "loss": 0.2373,
+      "step": 17272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34319801121283044,
+      "learning_rate": 3.0168101644845183e-09,
+      "loss": 0.2517,
+      "step": 17273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38067607576018514,
+      "learning_rate": 2.9712763157885293e-09,
+      "loss": 0.1608,
+      "step": 17274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3449780354209424,
+      "learning_rate": 2.9260886637672014e-09,
+      "loss": 0.2681,
+      "step": 17275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30939155012806313,
+      "learning_rate": 2.881247209984839e-09,
+      "loss": 0.2485,
+      "step": 17276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6205897551993651,
+      "learning_rate": 2.8367519559957537e-09,
+      "loss": 0.3882,
+      "step": 17277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1477515882330112,
+      "learning_rate": 2.792602903339825e-09,
+      "loss": 0.2529,
+      "step": 17278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38265407718892425,
+      "learning_rate": 2.7488000535458303e-09,
+      "loss": 0.2418,
+      "step": 17279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19879501472534586,
+      "learning_rate": 2.7053434081314447e-09,
+      "loss": 0.2037,
+      "step": 17280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9142794037777149,
+      "learning_rate": 2.6622329686010196e-09,
+      "loss": 0.4647,
+      "step": 17281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28941532117602325,
+      "learning_rate": 2.619468736446695e-09,
+      "loss": 0.1886,
+      "step": 17282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6324264222405012,
+      "learning_rate": 2.5770507131517297e-09,
+      "loss": 0.4013,
+      "step": 17283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.328680654110541,
+      "learning_rate": 2.5349789001827274e-09,
+      "loss": 0.2716,
+      "step": 17284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2842382251427557,
+      "learning_rate": 2.4932532989974113e-09,
+      "loss": 0.1838,
+      "step": 17285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32087338368593543,
+      "learning_rate": 2.4518739110412913e-09,
+      "loss": 0.1578,
+      "step": 17286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46121708923217086,
+      "learning_rate": 2.410840737746556e-09,
+      "loss": 0.3881,
+      "step": 17287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25180801069509645,
+      "learning_rate": 2.37015378053429e-09,
+      "loss": 0.2123,
+      "step": 17288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6193436792258021,
+      "learning_rate": 2.329813040814477e-09,
+      "loss": 0.3735,
+      "step": 17289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.283115078097724,
+      "learning_rate": 2.289818519982667e-09,
+      "loss": 0.4848,
+      "step": 17290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2806743140172014,
+      "learning_rate": 2.2501702194244192e-09,
+      "loss": 0.1837,
+      "step": 17291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28516428759808665,
+      "learning_rate": 2.2108681405141885e-09,
+      "loss": 0.2424,
+      "step": 17292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38473609798441183,
+      "learning_rate": 2.1719122846097783e-09,
+      "loss": 0.2753,
+      "step": 17293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31851548707689087,
+      "learning_rate": 2.1333026530634403e-09,
+      "loss": 0.238,
+      "step": 17294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.181637073888693,
+      "learning_rate": 2.0950392472107726e-09,
+      "loss": 0.3125,
+      "step": 17295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3795736909128083,
+      "learning_rate": 2.0571220683762717e-09,
+      "loss": 0.2708,
+      "step": 17296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6764029907443159,
+      "learning_rate": 2.019551117874441e-09,
+      "loss": 0.3151,
+      "step": 17297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.200451458051388,
+      "learning_rate": 1.9823263970042416e-09,
+      "loss": 0.1499,
+      "step": 17298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34620613205315875,
+      "learning_rate": 1.9454479070579735e-09,
+      "loss": 0.2849,
+      "step": 17299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5328903980935618,
+      "learning_rate": 1.9089156493101722e-09,
+      "loss": 0.3001,
+      "step": 17300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.342820967041687,
+      "learning_rate": 1.8727296250264924e-09,
+      "loss": 0.2557,
+      "step": 17301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1405798965958176,
+      "learning_rate": 1.8368898354603759e-09,
+      "loss": 0.4979,
+      "step": 17302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3883494449017486,
+      "learning_rate": 1.8013962818530516e-09,
+      "loss": 0.273,
+      "step": 17303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21653975646538426,
+      "learning_rate": 1.7662489654324267e-09,
+      "loss": 0.206,
+      "step": 17304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7855936171737142,
+      "learning_rate": 1.7314478874175255e-09,
+      "loss": 0.406,
+      "step": 17305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43869514092740386,
+      "learning_rate": 1.6969930490129406e-09,
+      "loss": 0.2855,
+      "step": 17306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2802090575204732,
+      "learning_rate": 1.662884451411051e-09,
+      "loss": 0.2516,
+      "step": 17307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5347954446359646,
+      "learning_rate": 1.6291220957942443e-09,
+      "loss": 0.23,
+      "step": 17308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43073846683388817,
+      "learning_rate": 1.5957059833293653e-09,
+      "loss": 0.2637,
+      "step": 17309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5787939035509605,
+      "learning_rate": 1.5626361151765967e-09,
+      "loss": 0.3245,
+      "step": 17310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30066704859583493,
+      "learning_rate": 1.5299124924794684e-09,
+      "loss": 0.2534,
+      "step": 17311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3953423827246001,
+      "learning_rate": 1.497535116371518e-09,
+      "loss": 0.2949,
+      "step": 17312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5339570681109478,
+      "learning_rate": 1.4655039879740706e-09,
+      "loss": 0.3517,
+      "step": 17313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2465958574137639,
+      "learning_rate": 1.4338191083962394e-09,
+      "loss": 0.0919,
+      "step": 17314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3375883743148337,
+      "learning_rate": 1.4024804787349244e-09,
+      "loss": 0.2508,
+      "step": 17315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34858518129319876,
+      "learning_rate": 1.371488100075924e-09,
+      "loss": 0.2834,
+      "step": 17316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5732830362171144,
+      "learning_rate": 1.3408419734928235e-09,
+      "loss": 0.2936,
+      "step": 17317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5490046678633443,
+      "learning_rate": 1.3105421000458861e-09,
+      "loss": 0.3244,
+      "step": 17318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25973593846651,
+      "learning_rate": 1.280588480785383e-09,
+      "loss": 0.2386,
+      "step": 17319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40068654789339847,
+      "learning_rate": 1.2509811167482622e-09,
+      "loss": 0.2308,
+      "step": 17320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27862506869394177,
+      "learning_rate": 1.2217200089592596e-09,
+      "loss": 0.0688,
+      "step": 17321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3646002178721917,
+      "learning_rate": 1.192805158432009e-09,
+      "loss": 0.2923,
+      "step": 17322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4507387100999731,
+      "learning_rate": 1.164236566167931e-09,
+      "loss": 0.3267,
+      "step": 17323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2687264725141766,
+      "learning_rate": 1.1360142331562351e-09,
+      "loss": 0.2057,
+      "step": 17324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35005883485614064,
+      "learning_rate": 1.1081381603750275e-09,
+      "loss": 0.2811,
+      "step": 17325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.47338451847770335,
+      "learning_rate": 1.0806083487890917e-09,
+      "loss": 0.2107,
+      "step": 17326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25788576820285475,
+      "learning_rate": 1.0534247993509994e-09,
+      "loss": 0.2088,
+      "step": 17327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5201529560961416,
+      "learning_rate": 1.0265875130033298e-09,
+      "loss": 0.3249,
+      "step": 17328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7291318527937328,
+      "learning_rate": 1.0000964906753396e-09,
+      "loss": 0.3989,
+      "step": 17329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6160151881883443,
+      "learning_rate": 9.739517332829628e-10,
+      "loss": 0.3685,
+      "step": 17330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2861802045422825,
+      "learning_rate": 9.481532417332518e-10,
+      "loss": 0.2201,
+      "step": 17331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33069292196442907,
+      "learning_rate": 9.227010169188256e-10,
+      "loss": 0.1932,
+      "step": 17332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5335904809825592,
+      "learning_rate": 8.975950597212014e-10,
+      "loss": 0.318,
+      "step": 17333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34095909495518917,
+      "learning_rate": 8.728353710107939e-10,
+      "loss": 0.217,
+      "step": 17334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3489415469794577,
+      "learning_rate": 8.484219516435854e-10,
+      "loss": 0.3089,
+      "step": 17335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.77950741238855,
+      "learning_rate": 8.243548024655656e-10,
+      "loss": 0.4937,
+      "step": 17336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3601096759079263,
+      "learning_rate": 8.006339243094019e-10,
+      "loss": 0.2173,
+      "step": 17337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.295337610488898,
+      "learning_rate": 7.772593179977694e-10,
+      "loss": 0.1954,
+      "step": 17338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30542651027800183,
+      "learning_rate": 7.542309843400209e-10,
+      "loss": 0.2761,
+      "step": 17339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31871131623653226,
+      "learning_rate": 7.315489241332963e-10,
+      "loss": 0.2258,
+      "step": 17340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4402612913254227,
+      "learning_rate": 7.092131381625233e-10,
+      "loss": 0.71,
+      "step": 17341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8675495422620723,
+      "learning_rate": 6.872236272026378e-10,
+      "loss": 0.4088,
+      "step": 17342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2562096236525775,
+      "learning_rate": 6.655803920130322e-10,
+      "loss": 0.2502,
+      "step": 17343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3887523928164326,
+      "learning_rate": 6.442834333453274e-10,
+      "loss": 0.1314,
+      "step": 17344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49486610459972546,
+      "learning_rate": 6.233327519356014e-10,
+      "loss": 0.3164,
+      "step": 17345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2790524317225934,
+      "learning_rate": 6.0272834850994e-10,
+      "loss": 0.2222,
+      "step": 17346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3245395596263069,
+      "learning_rate": 5.824702237822167e-10,
+      "loss": 0.2424,
+      "step": 17347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.550353076174964,
+      "learning_rate": 5.62558378452982e-10,
+      "loss": 0.3918,
+      "step": 17348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3946960811855935,
+      "learning_rate": 5.429928132127948e-10,
+      "loss": 0.2718,
+      "step": 17349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.577554881905842,
+      "learning_rate": 5.23773528737781e-10,
+      "loss": 0.289,
+      "step": 17350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26363339789068374,
+      "learning_rate": 5.049005256951845e-10,
+      "loss": 0.2336,
+      "step": 17351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2628831020493132,
+      "learning_rate": 4.863738047378164e-10,
+      "loss": 0.2008,
+      "step": 17352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.316507968789768,
+      "learning_rate": 4.681933665084959e-10,
+      "loss": 0.3222,
+      "step": 17353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1554124886092956,
+      "learning_rate": 4.5035921163449905e-10,
+      "loss": 0.4016,
+      "step": 17354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24058819285647187,
+      "learning_rate": 4.3287134073422e-10,
+      "loss": 0.2455,
+      "step": 17355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6174960471662688,
+      "learning_rate": 4.1572975441384055e-10,
+      "loss": 0.4182,
+      "step": 17356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3324218340581116,
+      "learning_rate": 3.9893445326733003e-10,
+      "loss": 0.1355,
+      "step": 17357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26942764406218134,
+      "learning_rate": 3.824854378753351e-10,
+      "loss": 0.2026,
+      "step": 17358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35914740702980735,
+      "learning_rate": 3.6638270880851034e-10,
+      "loss": 0.2896,
+      "step": 17359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5455096931025434,
+      "learning_rate": 3.5062626662307753e-10,
+      "loss": 0.2895,
+      "step": 17360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3443424930766011,
+      "learning_rate": 3.352161118652664e-10,
+      "loss": 0.2766,
+      "step": 17361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6169748491571364,
+      "learning_rate": 3.2015224506909414e-10,
+      "loss": 0.3999,
+      "step": 17362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22643565427717527,
+      "learning_rate": 3.0543466675636567e-10,
+      "loss": 0.1819,
+      "step": 17363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33718256571038757,
+      "learning_rate": 2.9106337743667336e-10,
+      "loss": 0.2753,
+      "step": 17364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5574590750528694,
+      "learning_rate": 2.7703837760739706e-10,
+      "loss": 0.2991,
+      "step": 17365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36791725805618314,
+      "learning_rate": 2.6335966775370423e-10,
+      "loss": 0.2042,
+      "step": 17366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2976665692719068,
+      "learning_rate": 2.500272483496602e-10,
+      "loss": 0.2583,
+      "step": 17367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.563196407125086,
+      "learning_rate": 2.370411198582279e-10,
+      "loss": 0.3444,
+      "step": 17368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4542464555776609,
+      "learning_rate": 2.2440128272682716e-10,
+      "loss": 0.2276,
+      "step": 17369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2292789630209967,
+      "learning_rate": 2.1210773739510637e-10,
+      "loss": 0.1631,
+      "step": 17370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3201790858158781,
+      "learning_rate": 2.0016048428828095e-10,
+      "loss": 0.2865,
+      "step": 17371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1917003624322648,
+      "learning_rate": 1.8855952381935384e-10,
+      "loss": 0.6258,
+      "step": 17372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.29913451282453496,
+      "learning_rate": 1.7730485639133598e-10,
+      "loss": 0.2036,
+      "step": 17373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4936270409112451,
+      "learning_rate": 1.6639648239280547e-10,
+      "loss": 0.3253,
+      "step": 17374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5255633349642953,
+      "learning_rate": 1.5583440220234835e-10,
+      "loss": 0.3583,
+      "step": 17375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21257105141332128,
+      "learning_rate": 1.4561861618411778e-10,
+      "loss": 0.1564,
+      "step": 17376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4397563294608192,
+      "learning_rate": 1.357491246944953e-10,
+      "loss": 0.3039,
+      "step": 17377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43366888050053,
+      "learning_rate": 1.2622592807320922e-10,
+      "loss": 0.3299,
+      "step": 17378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25224273533898794,
+      "learning_rate": 1.1704902665110596e-10,
+      "loss": 0.2109,
+      "step": 17379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.876683409290362,
+      "learning_rate": 1.082184207445991e-10,
+      "loss": 0.5038,
+      "step": 17380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5709334603042204,
+      "learning_rate": 9.97341106612204e-11,
+      "loss": 0.2467,
+      "step": 17381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3252570517790537,
+      "learning_rate": 9.159609669406876e-11,
+      "loss": 0.2698,
+      "step": 17382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2536394980429637,
+      "learning_rate": 8.380437912514083e-11,
+      "loss": 0.2018,
+      "step": 17383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5485757717623354,
+      "learning_rate": 7.635895822311057e-11,
+      "loss": 0.3742,
+      "step": 17384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3684524592933877,
+      "learning_rate": 6.925983424777016e-11,
+      "loss": 0.2536,
+      "step": 17385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3760098681531618,
+      "learning_rate": 6.250700744336869e-11,
+      "loss": 0.2418,
+      "step": 17386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.44201853065463803,
+      "learning_rate": 5.610047804527341e-11,
+      "loss": 0.2955,
+      "step": 17387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3492915963630672,
+      "learning_rate": 5.00402462733085e-11,
+      "loss": 0.2622,
+      "step": 17388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40702154184454287,
+      "learning_rate": 4.4326312338416333e-11,
+      "loss": 0.1927,
+      "step": 17389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4248298337535507,
+      "learning_rate": 3.895867643932683e-11,
+      "loss": 0.3135,
+      "step": 17390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26443247178119667,
+      "learning_rate": 3.3937338760337e-11,
+      "loss": 0.2393,
+      "step": 17391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.47992056831554575,
+      "learning_rate": 2.92622994768621e-11,
+      "loss": 0.1771,
+      "step": 17392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7800715125968598,
+      "learning_rate": 2.4933558749884456e-11,
+      "loss": 0.4894,
+      "step": 17393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2614907187670138,
+      "learning_rate": 2.0951116729284182e-11,
+      "loss": 0.2234,
+      "step": 17394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33747834456865705,
+      "learning_rate": 1.731497355272893e-11,
+      "loss": 0.319,
+      "step": 17395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5460667339487827,
+      "learning_rate": 1.4025129346784127e-11,
+      "loss": 0.3004,
+      "step": 17396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32489280345231525,
+      "learning_rate": 1.1081584224692521e-11,
+      "loss": 0.2555,
+      "step": 17397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5915103642932011,
+      "learning_rate": 8.484338289704852e-12,
+      "loss": 0.3031,
+      "step": 17398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.37556640988554935,
+      "learning_rate": 6.233391630638963e-12,
+      "loss": 0.2357,
+      "step": 17399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.29506593921851954,
+      "learning_rate": 4.328744325210466e-12,
+      "loss": 0.2447,
+      "step": 17400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7211757381438109,
+      "learning_rate": 2.770396440032741e-12,
+      "loss": 0.4654,
+      "step": 17401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.316378855890181,
+      "learning_rate": 1.5583480295067177e-12,
+      "loss": 0.278,
+      "step": 17402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.38606123702259176,
+      "learning_rate": 6.925991336004245e-13,
+      "loss": 0.2463,
+      "step": 17403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26007754178125586,
+      "learning_rate": 1.7314978451032915e-13,
+      "loss": 0.1748,
+      "step": 17404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5025672310396203,
+      "learning_rate": 0.0,
+      "loss": 0.1774,
+      "step": 17405
+    },
+    {
+      "epoch": 1.0,
+      "step": 17405,
+      "total_flos": 0.0,
+      "train_loss": 0.3205628498764781,
+      "train_runtime": 138539.6755,
+      "train_samples_per_second": 60.365,
+      "train_steps_per_second": 0.126
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17405,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}