diff --git "a/llmga-gemma-pretrain/trainer_state.json" "b/llmga-gemma-pretrain/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/llmga-gemma-pretrain/trainer_state.json"
@@ -0,0 +1,15290 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998853342506593,
+  "eval_steps": 500,
+  "global_step": 2180,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 849.1024478728034,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 18.3289,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 874.7083282050851,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 18.609,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 577.5007959626174,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 16.2579,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 77.17921992387498,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 11.7081,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 879.0917750778681,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 20.838,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 622.479079866714,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 17.078,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 56.873322945544615,
+      "learning_rate": 0.00010606060606060606,
+      "loss": 10.8345,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 430.8110186434264,
+      "learning_rate": 0.00012121212121212122,
+      "loss": 13.7737,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 180.78356622240403,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 10.7919,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 54.67610190532367,
+      "learning_rate": 0.00015151515151515152,
+      "loss": 9.9228,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 69.95753147133752,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 10.643,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 225.54667945369687,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 10.9514,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.39540445295766,
+      "learning_rate": 0.00019696969696969695,
+      "loss": 9.5177,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.643204407976825,
+      "learning_rate": 0.00021212121212121213,
+      "loss": 8.8856,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.09060097286181,
+      "learning_rate": 0.00022727272727272727,
+      "loss": 8.5516,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.519151740669482,
+      "learning_rate": 0.00024242424242424245,
+      "loss": 8.2104,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.65359102290558,
+      "learning_rate": 0.00025757575757575756,
+      "loss": 7.7657,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.089602348249024,
+      "learning_rate": 0.00027272727272727274,
+      "loss": 7.4234,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.463215551057672,
+      "learning_rate": 0.0002878787878787879,
+      "loss": 7.0094,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.306838019581889,
+      "learning_rate": 0.00030303030303030303,
+      "loss": 7.0393,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.315250557903088,
+      "learning_rate": 0.0003181818181818182,
+      "loss": 6.9513,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.71936934420649,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 6.8552,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.277864533683237,
+      "learning_rate": 0.0003484848484848485,
+      "loss": 6.7618,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.645144795586457,
+      "learning_rate": 0.00036363636363636367,
+      "loss": 6.6753,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.535921470660035,
+      "learning_rate": 0.0003787878787878788,
+      "loss": 6.5648,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.533627093732919,
+      "learning_rate": 0.0003939393939393939,
+      "loss": 6.5194,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.261539400917472,
+      "learning_rate": 0.00040909090909090913,
+      "loss": 6.3567,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.554365081742184,
+      "learning_rate": 0.00042424242424242425,
+      "loss": 6.2715,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.320050948899247,
+      "learning_rate": 0.0004393939393939394,
+      "loss": 6.4037,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.943786199426835,
+      "learning_rate": 0.00045454545454545455,
+      "loss": 6.221,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.517228496117076,
+      "learning_rate": 0.0004696969696969697,
+      "loss": 6.2147,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.378510508395376,
+      "learning_rate": 0.0004848484848484849,
+      "loss": 6.2062,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.225822882353064,
+      "learning_rate": 0.0005,
+      "loss": 6.145,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7497805170794294,
+      "learning_rate": 0.0005151515151515151,
+      "loss": 6.0276,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4258401957564253,
+      "learning_rate": 0.0005303030303030302,
+      "loss": 6.0233,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8769824011117278,
+      "learning_rate": 0.0005454545454545455,
+      "loss": 5.7637,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.011438505442924,
+      "learning_rate": 0.0005606060606060606,
+      "loss": 5.9306,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.318926378689546,
+      "learning_rate": 0.0005757575757575758,
+      "loss": 5.7862,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.116660408767277,
+      "learning_rate": 0.0005909090909090909,
+      "loss": 5.838,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7653603571299916,
+      "learning_rate": 0.0006060606060606061,
+      "loss": 5.823,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7521085804466625,
+      "learning_rate": 0.0006212121212121212,
+      "loss": 5.6889,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0900722738499677,
+      "learning_rate": 0.0006363636363636364,
+      "loss": 5.6978,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9641485894426558,
+      "learning_rate": 0.0006515151515151515,
+      "loss": 5.7009,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.629410976950863,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 5.6662,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.723263809246651,
+      "learning_rate": 0.0006818181818181818,
+      "loss": 5.6326,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1621092877964325,
+      "learning_rate": 0.000696969696969697,
+      "loss": 5.5781,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9185216839407433,
+      "learning_rate": 0.0007121212121212122,
+      "loss": 5.553,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5619740662622428,
+      "learning_rate": 0.0007272727272727273,
+      "loss": 5.532,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6187264976166547,
+      "learning_rate": 0.0007424242424242425,
+      "loss": 5.6368,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4871321000850637,
+      "learning_rate": 0.0007575757575757576,
+      "loss": 5.5825,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4504384224224192,
+      "learning_rate": 0.0007727272727272727,
+      "loss": 5.5478,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3584396761662416,
+      "learning_rate": 0.0007878787878787878,
+      "loss": 5.6419,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4555920173953485,
+      "learning_rate": 0.000803030303030303,
+      "loss": 5.6553,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2745455868266466,
+      "learning_rate": 0.0008181818181818183,
+      "loss": 5.4872,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9268541429021937,
+      "learning_rate": 0.0008333333333333334,
+      "loss": 5.4964,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.678676333835535,
+      "learning_rate": 0.0008484848484848485,
+      "loss": 5.6003,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2634035801406691,
+      "learning_rate": 0.0008636363636363636,
+      "loss": 5.4726,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1101341601737222,
+      "learning_rate": 0.0008787878787878789,
+      "loss": 5.5956,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7327890047410668,
+      "learning_rate": 0.000893939393939394,
+      "loss": 5.4358,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0353656064747123,
+      "learning_rate": 0.0009090909090909091,
+      "loss": 5.3685,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6956602799998404,
+      "learning_rate": 0.0009242424242424242,
+      "loss": 5.2759,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.925344217037855,
+      "learning_rate": 0.0009393939393939394,
+      "loss": 5.37,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1406901439461792,
+      "learning_rate": 0.0009545454545454546,
+      "loss": 5.3397,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.87041274245238,
+      "learning_rate": 0.0009696969696969698,
+      "loss": 5.5086,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8173588568653534,
+      "learning_rate": 0.000984848484848485,
+      "loss": 5.286,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2766226350077292,
+      "learning_rate": 0.001,
+      "loss": 5.3558,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1691001197730198,
+      "learning_rate": 0.0009999994478847943,
+      "loss": 5.3125,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.538954507738684,
+      "learning_rate": 0.0009999977915403962,
+      "loss": 5.1995,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.228975866540403,
+      "learning_rate": 0.0009999950309704639,
+      "loss": 5.2818,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.373756244631725,
+      "learning_rate": 0.000999991166181094,
+      "loss": 5.2648,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9231975920847155,
+      "learning_rate": 0.0009999861971808216,
+      "loss": 5.2977,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1543541806631286,
+      "learning_rate": 0.0009999801239806208,
+      "loss": 5.2243,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.024579256794025,
+      "learning_rate": 0.0009999729465939035,
+      "loss": 5.1599,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8183788303244683,
+      "learning_rate": 0.0009999646650365212,
+      "loss": 5.22,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.188004843433445,
+      "learning_rate": 0.0009999552793267634,
+      "loss": 5.2939,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9943579317472722,
+      "learning_rate": 0.0009999447894853577,
+      "loss": 5.1432,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6903347071923294,
+      "learning_rate": 0.0009999331955354708,
+      "loss": 5.1021,
+      "step": 77
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2088117435392876,
+      "learning_rate": 0.0009999204975027073,
+      "loss": 5.1208,
+      "step": 78
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2188626719917552,
+      "learning_rate": 0.0009999066954151103,
+      "loss": 5.009,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5773432289609617,
+      "learning_rate": 0.0009998917893031614,
+      "loss": 5.1039,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5084595807507444,
+      "learning_rate": 0.0009998757791997801,
+      "loss": 5.0651,
+      "step": 81
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0969318016827792,
+      "learning_rate": 0.0009998586651403238,
+      "loss": 5.0618,
+      "step": 82
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3263310403774387,
+      "learning_rate": 0.0009998404471625885,
+      "loss": 4.9805,
+      "step": 83
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.241901364335578,
+      "learning_rate": 0.0009998211253068078,
+      "loss": 4.9813,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1502069680012563,
+      "learning_rate": 0.0009998006996156535,
+      "loss": 5.1047,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0589035755771035,
+      "learning_rate": 0.0009997791701342347,
+      "loss": 5.0637,
+      "step": 86
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2540261551743412,
+      "learning_rate": 0.0009997565369100983,
+      "loss": 4.8899,
+      "step": 87
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.757291275479657,
+      "learning_rate": 0.0009997327999932291,
+      "loss": 5.0356,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.477423930941434,
+      "learning_rate": 0.000999707959436049,
+      "loss": 4.9875,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.047850625550379,
+      "learning_rate": 0.0009996820152934176,
+      "loss": 5.0057,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7441190325771136,
+      "learning_rate": 0.000999654967622631,
+      "loss": 4.857,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4817740318350183,
+      "learning_rate": 0.0009996268164834238,
+      "loss": 4.8841,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1407886724788174,
+      "learning_rate": 0.000999597561937966,
+      "loss": 4.9254,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1895800867488613,
+      "learning_rate": 0.0009995672040508656,
+      "loss": 4.9416,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2297028873087126,
+      "learning_rate": 0.0009995357428891662,
+      "loss": 4.9365,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2804652615717091,
+      "learning_rate": 0.0009995031785223491,
+      "loss": 4.8698,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0145299026468768,
+      "learning_rate": 0.000999469511022331,
+      "loss": 4.9388,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0585243798582193,
+      "learning_rate": 0.0009994347404634657,
+      "loss": 4.8988,
+      "step": 98
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0054120822651518,
+      "learning_rate": 0.0009993988669225423,
+      "loss": 4.8344,
+      "step": 99
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0389841011356726,
+      "learning_rate": 0.000999361890478786,
+      "loss": 4.8608,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1315122044996662,
+      "learning_rate": 0.0009993238112138583,
+      "loss": 4.8501,
+      "step": 101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.850393693890169,
+      "learning_rate": 0.0009992846292118554,
+      "loss": 4.6994,
+      "step": 102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0581450846525704,
+      "learning_rate": 0.000999244344559309,
+      "loss": 4.7566,
+      "step": 103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1181644117075173,
+      "learning_rate": 0.0009992029573451869,
+      "loss": 4.8249,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9377343295485343,
+      "learning_rate": 0.0009991604676608905,
+      "loss": 4.7603,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8894525529834808,
+      "learning_rate": 0.0009991168756002568,
+      "loss": 4.8682,
+      "step": 106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9889698079435518,
+      "learning_rate": 0.0009990721812595574,
+      "loss": 4.6711,
+      "step": 107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9934405609840105,
+      "learning_rate": 0.0009990263847374976,
+      "loss": 4.6894,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9516162374487156,
+      "learning_rate": 0.0009989794861352173,
+      "loss": 4.8183,
+      "step": 109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9106124787767145,
+      "learning_rate": 0.0009989314855562905,
+      "loss": 4.817,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9045317593099511,
+      "learning_rate": 0.0009988823831067245,
+      "loss": 4.8271,
+      "step": 111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9081211379439946,
+      "learning_rate": 0.0009988321788949597,
+      "loss": 4.7967,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7924892168804776,
+      "learning_rate": 0.0009987808730318709,
+      "loss": 4.6874,
+      "step": 113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9975130725197418,
+      "learning_rate": 0.0009987284656307644,
+      "loss": 4.8049,
+      "step": 114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9746799521039936,
+      "learning_rate": 0.0009986749568073802,
+      "loss": 4.719,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1001866630230974,
+      "learning_rate": 0.0009986203466798905,
+      "loss": 4.751,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8885060682552715,
+      "learning_rate": 0.0009985646353688996,
+      "loss": 4.8331,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9424427796124706,
+      "learning_rate": 0.0009985078229974437,
+      "loss": 4.7505,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9008672315532076,
+      "learning_rate": 0.0009984499096909905,
+      "loss": 4.6989,
+      "step": 119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9151299144798222,
+      "learning_rate": 0.0009983908955774397,
+      "loss": 4.7293,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.040480252453662,
+      "learning_rate": 0.0009983307807871211,
+      "loss": 4.7155,
+      "step": 121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.053781196088565,
+      "learning_rate": 0.0009982695654527965,
+      "loss": 4.6135,
+      "step": 122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0126029698018915,
+      "learning_rate": 0.0009982072497096571,
+      "loss": 4.6883,
+      "step": 123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9794811850867537,
+      "learning_rate": 0.000998143833695325,
+      "loss": 4.6696,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2852834335892849,
+      "learning_rate": 0.0009980793175498517,
+      "loss": 4.6805,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1699283178708422,
+      "learning_rate": 0.000998013701415719,
+      "loss": 4.6936,
+      "step": 126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0423622223659692,
+      "learning_rate": 0.0009979469854378372,
+      "loss": 4.5901,
+      "step": 127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0279536338860986,
+      "learning_rate": 0.000997879169763546,
+      "loss": 4.5731,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4499032408022174,
+      "learning_rate": 0.000997810254542614,
+      "loss": 4.5979,
+      "step": 129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0358179844570643,
+      "learning_rate": 0.0009977402399272374,
+      "loss": 4.5378,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.272480306983464,
+      "learning_rate": 0.0009976691260720407,
+      "loss": 4.489,
+      "step": 131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1306980452381632,
+      "learning_rate": 0.0009975969131340763,
+      "loss": 4.6576,
+      "step": 132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2985805049048957,
+      "learning_rate": 0.0009975236012728236,
+      "loss": 4.5474,
+      "step": 133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9405842230613763,
+      "learning_rate": 0.0009974491906501886,
+      "loss": 4.5585,
+      "step": 134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3959403480970891,
+      "learning_rate": 0.0009973736814305049,
+      "loss": 4.6042,
+      "step": 135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.483966614562128,
+      "learning_rate": 0.0009972970737805312,
+      "loss": 4.6063,
+      "step": 136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0922368693420674,
+      "learning_rate": 0.0009972193678694525,
+      "loss": 4.5495,
+      "step": 137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6522766749406093,
+      "learning_rate": 0.0009971405638688794,
+      "loss": 4.617,
+      "step": 138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9498160471179888,
+      "learning_rate": 0.0009970606619528475,
+      "loss": 4.5567,
+      "step": 139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0451387764959665,
+      "learning_rate": 0.000996979662297817,
+      "loss": 4.5459,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1227901333856147,
+      "learning_rate": 0.0009968975650826721,
+      "loss": 4.4808,
+      "step": 141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.107510153497332,
+      "learning_rate": 0.000996814370488722,
+      "loss": 4.5773,
+      "step": 142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3962485042202044,
+      "learning_rate": 0.000996730078699698,
+      "loss": 4.5589,
+      "step": 143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1628758840604438,
+      "learning_rate": 0.0009966446899017558,
+      "loss": 4.5339,
+      "step": 144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9905075650292238,
+      "learning_rate": 0.0009965582042834728,
+      "loss": 4.626,
+      "step": 145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6415131265736471,
+      "learning_rate": 0.0009964706220358492,
+      "loss": 4.5089,
+      "step": 146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1279896547870332,
+      "learning_rate": 0.000996381943352307,
+      "loss": 4.5199,
+      "step": 147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1844584007464076,
+      "learning_rate": 0.0009962921684286896,
+      "loss": 4.54,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2939612620815695,
+      "learning_rate": 0.0009962012974632614,
+      "loss": 4.5973,
+      "step": 149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.280493954403419,
+      "learning_rate": 0.0009961093306567075,
+      "loss": 4.5232,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4497574484282498,
+      "learning_rate": 0.0009960162682121328,
+      "loss": 4.5419,
+      "step": 151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2581104307991586,
+      "learning_rate": 0.0009959221103350623,
+      "loss": 4.4479,
+      "step": 152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0719580333605256,
+      "learning_rate": 0.0009958268572334394,
+      "loss": 4.3741,
+      "step": 153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.079922413400243,
+      "learning_rate": 0.0009957305091176274,
+      "loss": 4.4375,
+      "step": 154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2382259617301634,
+      "learning_rate": 0.0009956330662004075,
+      "loss": 4.4919,
+      "step": 155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.462006477093048,
+      "learning_rate": 0.0009955345286969779,
+      "loss": 4.4083,
+      "step": 156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1455254382207671,
+      "learning_rate": 0.0009954348968249551,
+      "loss": 4.4066,
+      "step": 157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0968911580631677,
+      "learning_rate": 0.0009953341708043724,
+      "loss": 4.429,
+      "step": 158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0098433069894681,
+      "learning_rate": 0.0009952323508576793,
+      "loss": 4.4493,
+      "step": 159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0951065204690482,
+      "learning_rate": 0.0009951294372097406,
+      "loss": 4.3567,
+      "step": 160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1530602297104202,
+      "learning_rate": 0.0009950254300878378,
+      "loss": 4.3826,
+      "step": 161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2188260621111935,
+      "learning_rate": 0.000994920329721666,
+      "loss": 4.2782,
+      "step": 162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1881725228310471,
+      "learning_rate": 0.0009948141363433356,
+      "loss": 4.3798,
+      "step": 163
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6244872333027007,
+      "learning_rate": 0.00099470685018737,
+      "loss": 4.2914,
+      "step": 164
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6194028854294913,
+      "learning_rate": 0.0009945984714907073,
+      "loss": 4.3759,
+      "step": 165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3176587941589923,
+      "learning_rate": 0.000994489000492697,
+      "loss": 4.3565,
+      "step": 166
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0872169626074166,
+      "learning_rate": 0.0009943784374351016,
+      "loss": 4.4252,
+      "step": 167
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2061936735635665,
+      "learning_rate": 0.0009942667825620951,
+      "loss": 4.2637,
+      "step": 168
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.033252711790628,
+      "learning_rate": 0.0009941540361202634,
+      "loss": 4.2384,
+      "step": 169
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.639561165712214,
+      "learning_rate": 0.0009940401983586022,
+      "loss": 4.3303,
+      "step": 170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3339797202408956,
+      "learning_rate": 0.000993925269528518,
+      "loss": 4.4522,
+      "step": 171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1736091811660463,
+      "learning_rate": 0.0009938092498838265,
+      "loss": 4.289,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.46602183506964,
+      "learning_rate": 0.0009936921396807524,
+      "loss": 4.448,
+      "step": 173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5165112752458905,
+      "learning_rate": 0.0009935739391779292,
+      "loss": 4.207,
+      "step": 174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3101548338150177,
+      "learning_rate": 0.000993454648636398,
+      "loss": 4.268,
+      "step": 175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9345224527694576,
+      "learning_rate": 0.0009933342683196074,
+      "loss": 4.2913,
+      "step": 176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8493271780220226,
+      "learning_rate": 0.0009932127984934125,
+      "loss": 4.2377,
+      "step": 177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7951299927707032,
+      "learning_rate": 0.0009930902394260745,
+      "loss": 4.3725,
+      "step": 178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4187768267102416,
+      "learning_rate": 0.0009929665913882607,
+      "loss": 4.2398,
+      "step": 179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6001591177798176,
+      "learning_rate": 0.0009928418546530425,
+      "loss": 4.165,
+      "step": 180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6139307990576726,
+      "learning_rate": 0.0009927160294958964,
+      "loss": 4.3542,
+      "step": 181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8727188615303956,
+      "learning_rate": 0.000992589116194702,
+      "loss": 4.2652,
+      "step": 182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3488720034997872,
+      "learning_rate": 0.000992461115029743,
+      "loss": 4.1771,
+      "step": 183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.746525426378196,
+      "learning_rate": 0.000992332026283704,
+      "loss": 4.2599,
+      "step": 184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.828692508525562,
+      "learning_rate": 0.0009922018502416736,
+      "loss": 4.2044,
+      "step": 185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3858070068185104,
+      "learning_rate": 0.0009920705871911395,
+      "loss": 4.1038,
+      "step": 186
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.288874635701196,
+      "learning_rate": 0.0009919382374219915,
+      "loss": 4.2052,
+      "step": 187
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2639616845580672,
+      "learning_rate": 0.0009918048012265187,
+      "loss": 4.1056,
+      "step": 188
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9009541741918108,
+      "learning_rate": 0.0009916702788994097,
+      "loss": 4.1792,
+      "step": 189
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3725359430098159,
+      "learning_rate": 0.0009915346707377519,
+      "loss": 4.0902,
+      "step": 190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3781179858670678,
+      "learning_rate": 0.0009913979770410305,
+      "loss": 4.2138,
+      "step": 191
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1636115965189597,
+      "learning_rate": 0.0009912601981111285,
+      "loss": 4.2216,
+      "step": 192
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3910179718542435,
+      "learning_rate": 0.0009911213342523248,
+      "loss": 4.2528,
+      "step": 193
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4130822405085453,
+      "learning_rate": 0.000990981385771295,
+      "loss": 4.1545,
+      "step": 194
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7604935147207943,
+      "learning_rate": 0.00099084035297711,
+      "loss": 4.1553,
+      "step": 195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6744151552611846,
+      "learning_rate": 0.000990698236181235,
+      "loss": 4.1124,
+      "step": 196
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8666866392275328,
+      "learning_rate": 0.0009905550356975293,
+      "loss": 4.1129,
+      "step": 197
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.179033373459883,
+      "learning_rate": 0.0009904107518422457,
+      "loss": 4.1344,
+      "step": 198
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6809009235731853,
+      "learning_rate": 0.0009902653849340295,
+      "loss": 4.1471,
+      "step": 199
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5188673066012501,
+      "learning_rate": 0.0009901189352939177,
+      "loss": 4.1238,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4496330581678372,
+      "learning_rate": 0.0009899714032453387,
+      "loss": 3.9746,
+      "step": 201
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4567117230533888,
+      "learning_rate": 0.000989822789114111,
+      "loss": 4.0763,
+      "step": 202
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2923234096695275,
+      "learning_rate": 0.0009896730932284434,
+      "loss": 4.0658,
+      "step": 203
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8180879002089845,
+      "learning_rate": 0.0009895223159189332,
+      "loss": 4.0373,
+      "step": 204
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4737971745540392,
+      "learning_rate": 0.0009893704575185663,
+      "loss": 4.0576,
+      "step": 205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5856786919686807,
+      "learning_rate": 0.000989217518362716,
+      "loss": 4.0988,
+      "step": 206
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.374568700631167,
+      "learning_rate": 0.0009890634987891425,
+      "loss": 4.119,
+      "step": 207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.835770097489722,
+      "learning_rate": 0.0009889083991379917,
+      "loss": 4.1176,
+      "step": 208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3969390460369346,
+      "learning_rate": 0.0009887522197517954,
+      "loss": 4.1087,
+      "step": 209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.119823358473737,
+      "learning_rate": 0.0009885949609754693,
+      "loss": 4.1469,
+      "step": 210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8464124103787336,
+      "learning_rate": 0.000988436623156314,
+      "loss": 4.0829,
+      "step": 211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7514119142291742,
+      "learning_rate": 0.0009882772066440114,
+      "loss": 3.8872,
+      "step": 212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7097888687418081,
+      "learning_rate": 0.0009881167117906276,
+      "loss": 3.8104,
+      "step": 213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5937285427604821,
+      "learning_rate": 0.0009879551389506084,
+      "loss": 4.0306,
+      "step": 214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.068156250611066,
+      "learning_rate": 0.0009877924884807814,
+      "loss": 4.0271,
+      "step": 215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.027108833495585,
+      "learning_rate": 0.000987628760740354,
+      "loss": 4.0386,
+      "step": 216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9412117390958445,
+      "learning_rate": 0.0009874639560909118,
+      "loss": 3.9836,
+      "step": 217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.679969601427816,
+      "learning_rate": 0.0009872980748964202,
+      "loss": 3.9495,
+      "step": 218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7051780790635735,
+      "learning_rate": 0.000987131117523221,
+      "loss": 3.9201,
+      "step": 219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.833483230168737,
+      "learning_rate": 0.000986963084340033,
+      "loss": 3.9782,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6584676765014548,
+      "learning_rate": 0.0009867939757179508,
+      "loss": 3.9242,
+      "step": 221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5311522606721626,
+      "learning_rate": 0.0009866237920304443,
+      "loss": 4.0361,
+      "step": 222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4266111190297386,
+      "learning_rate": 0.0009864525336533577,
+      "loss": 4.0225,
+      "step": 223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1822854136503516,
+      "learning_rate": 0.000986280200964908,
+      "loss": 4.0376,
+      "step": 224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5619851801421305,
+      "learning_rate": 0.0009861067943456856,
+      "loss": 3.9612,
+      "step": 225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5986115118862205,
+      "learning_rate": 0.000985932314178652,
+      "loss": 3.9314,
+      "step": 226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0115209783306858,
+      "learning_rate": 0.00098575676084914,
+      "loss": 3.8205,
+      "step": 227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1987011672759897,
+      "learning_rate": 0.0009855801347448518,
+      "loss": 3.9672,
+      "step": 228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3406276029703155,
+      "learning_rate": 0.0009854024362558596,
+      "loss": 3.9151,
+      "step": 229
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5780176954545144,
+      "learning_rate": 0.0009852236657746035,
+      "loss": 3.8448,
+      "step": 230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1733223862402107,
+      "learning_rate": 0.0009850438236958911,
+      "loss": 3.7943,
+      "step": 231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4661127113722698,
+      "learning_rate": 0.0009848629104168966,
+      "loss": 3.9381,
+      "step": 232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9954272153337405,
+      "learning_rate": 0.00098468092633716,
+      "loss": 3.9646,
+      "step": 233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.418520895954301,
+      "learning_rate": 0.0009844978718585855,
+      "loss": 3.8905,
+      "step": 234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1600234736794968,
+      "learning_rate": 0.0009843137473854423,
+      "loss": 3.8285,
+      "step": 235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7682237162507657,
+      "learning_rate": 0.000984128553324362,
+      "loss": 3.9328,
+      "step": 236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5003684713991836,
+      "learning_rate": 0.0009839422900843383,
+      "loss": 3.9005,
+      "step": 237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4540790018287812,
+      "learning_rate": 0.0009837549580767261,
+      "loss": 3.9722,
+      "step": 238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.823724421480226,
+      "learning_rate": 0.0009835665577152411,
+      "loss": 3.7849,
+      "step": 239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7218920035709293,
+      "learning_rate": 0.000983377089415958,
+      "loss": 3.9542,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4474338982296135,
+      "learning_rate": 0.0009831865535973102,
+      "loss": 3.9128,
+      "step": 241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6268776399040175,
+      "learning_rate": 0.0009829949506800885,
+      "loss": 3.8749,
+      "step": 242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8047724398969327,
+      "learning_rate": 0.0009828022810874405,
+      "loss": 3.7679,
+      "step": 243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.505201835523194,
+      "learning_rate": 0.0009826085452448693,
+      "loss": 3.7249,
+      "step": 244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8028758821828368,
+      "learning_rate": 0.000982413743580233,
+      "loss": 3.8204,
+      "step": 245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.937659400042939,
+      "learning_rate": 0.0009822178765237436,
+      "loss": 3.7842,
+      "step": 246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7915323272935215,
+      "learning_rate": 0.0009820209445079654,
+      "loss": 3.8821,
+      "step": 247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5920628689067065,
+      "learning_rate": 0.0009818229479678158,
+      "loss": 3.8027,
+      "step": 248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5565905910085316,
+      "learning_rate": 0.0009816238873405615,
+      "loss": 3.7449,
+      "step": 249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8537225838842477,
+      "learning_rate": 0.0009814237630658207,
+      "loss": 3.9158,
+      "step": 250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7530398926296,
+      "learning_rate": 0.00098122257558556,
+      "loss": 3.8583,
+      "step": 251
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4214656781902324,
+      "learning_rate": 0.0009810203253440937,
+      "loss": 3.8481,
+      "step": 252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9950489881332862,
+      "learning_rate": 0.0009808170127880837,
+      "loss": 3.9203,
+      "step": 253
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9455792350766699,
+      "learning_rate": 0.000980612638366538,
+      "loss": 3.8967,
+      "step": 254
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7848091541096867,
+      "learning_rate": 0.0009804072025308096,
+      "loss": 3.7581,
+      "step": 255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.819275140567835,
+      "learning_rate": 0.000980200705734595,
+      "loss": 3.8092,
+      "step": 256
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0303161648146766,
+      "learning_rate": 0.0009799931484339344,
+      "loss": 3.7317,
+      "step": 257
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0476907220728853,
+      "learning_rate": 0.0009797845310872103,
+      "loss": 3.7905,
+      "step": 258
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.043122783461441,
+      "learning_rate": 0.0009795748541551457,
+      "loss": 3.6976,
+      "step": 259
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.652137198418318,
+      "learning_rate": 0.000979364118100804,
+      "loss": 3.8263,
+      "step": 260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.129051900784942,
+      "learning_rate": 0.0009791523233895875,
+      "loss": 3.8817,
+      "step": 261
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0094209856949763,
+      "learning_rate": 0.0009789394704892364,
+      "loss": 3.7666,
+      "step": 262
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.962227284446103,
+      "learning_rate": 0.0009787255598698282,
+      "loss": 3.7822,
+      "step": 263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5572518383055178,
+      "learning_rate": 0.0009785105920037758,
+      "loss": 3.887,
+      "step": 264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8160596094330477,
+      "learning_rate": 0.0009782945673658275,
+      "loss": 3.7556,
+      "step": 265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4456344813207074,
+      "learning_rate": 0.0009780774864330654,
+      "loss": 3.7782,
+      "step": 266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.506759487048389,
+      "learning_rate": 0.000977859349684904,
+      "loss": 3.739,
+      "step": 267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8301760760339287,
+      "learning_rate": 0.00097764015760309,
+      "loss": 3.6979,
+      "step": 268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3355738042264673,
+      "learning_rate": 0.0009774199106717004,
+      "loss": 3.7739,
+      "step": 269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9007447506664013,
+      "learning_rate": 0.0009771986093771417,
+      "loss": 3.6904,
+      "step": 270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7252183109830546,
+      "learning_rate": 0.0009769762542081496,
+      "loss": 3.6848,
+      "step": 271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0371431314180177,
+      "learning_rate": 0.000976752845655786,
+      "loss": 3.7592,
+      "step": 272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.430549206086476,
+      "learning_rate": 0.0009765283842134411,
+      "loss": 3.604,
+      "step": 273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6918504078265157,
+      "learning_rate": 0.0009763028703768282,
+      "loss": 3.7253,
+      "step": 274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5130514906277006,
+      "learning_rate": 0.0009760763046439862,
+      "loss": 3.7562,
+      "step": 275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8209842422434852,
+      "learning_rate": 0.0009758486875152766,
+      "loss": 3.753,
+      "step": 276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8279933544732148,
+      "learning_rate": 0.0009756200194933829,
+      "loss": 3.7396,
+      "step": 277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.588875073926259,
+      "learning_rate": 0.0009753903010833094,
+      "loss": 3.7049,
+      "step": 278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8858039015517702,
+      "learning_rate": 0.0009751595327923803,
+      "loss": 3.7594,
+      "step": 279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7239447016921228,
+      "learning_rate": 0.0009749277151302382,
+      "loss": 3.7426,
+      "step": 280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.25545630862283,
+      "learning_rate": 0.0009746948486088435,
+      "loss": 3.7905,
+      "step": 281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5740587795614396,
+      "learning_rate": 0.0009744609337424727,
+      "loss": 3.7264,
+      "step": 282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.254901915668207,
+      "learning_rate": 0.0009742259710477177,
+      "loss": 3.6637,
+      "step": 283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8752997224306358,
+      "learning_rate": 0.0009739899610434841,
+      "loss": 3.8387,
+      "step": 284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3702177584153796,
+      "learning_rate": 0.0009737529042509913,
+      "loss": 3.6517,
+      "step": 285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2289496136370626,
+      "learning_rate": 0.0009735148011937693,
+      "loss": 3.6464,
+      "step": 286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.03192268313918,
+      "learning_rate": 0.00097327565239766,
+      "loss": 3.7662,
+      "step": 287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.0840449782909385,
+      "learning_rate": 0.0009730354583908136,
+      "loss": 3.7294,
+      "step": 288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0158777831599513,
+      "learning_rate": 0.0009727942197036895,
+      "loss": 3.6913,
+      "step": 289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.519583905843892,
+      "learning_rate": 0.0009725519368690539,
+      "loss": 3.7281,
+      "step": 290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.589758959875534,
+      "learning_rate": 0.0009723086104219787,
+      "loss": 3.6596,
+      "step": 291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.800848766795162,
+      "learning_rate": 0.0009720642408998409,
+      "loss": 3.6933,
+      "step": 292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4000715593999975,
+      "learning_rate": 0.0009718188288423211,
+      "loss": 3.6111,
+      "step": 293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0380288212386017,
+      "learning_rate": 0.0009715723747914022,
+      "loss": 3.6653,
+      "step": 294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8807551705519394,
+      "learning_rate": 0.0009713248792913685,
+      "loss": 3.7315,
+      "step": 295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5131582341667553,
+      "learning_rate": 0.0009710763428888037,
+      "loss": 3.7145,
+      "step": 296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2163013768100677,
+      "learning_rate": 0.0009708267661325909,
+      "loss": 3.775,
+      "step": 297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4881105643030779,
+      "learning_rate": 0.0009705761495739107,
+      "loss": 3.7385,
+      "step": 298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.021101979714653,
+      "learning_rate": 0.0009703244937662399,
+      "loss": 3.6086,
+      "step": 299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8072253665741793,
+      "learning_rate": 0.0009700717992653505,
+      "loss": 3.7359,
+      "step": 300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5648788027750626,
+      "learning_rate": 0.0009698180666293083,
+      "loss": 3.7056,
+      "step": 301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9523954820393814,
+      "learning_rate": 0.000969563296418472,
+      "loss": 3.6893,
+      "step": 302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4360571943793963,
+      "learning_rate": 0.0009693074891954914,
+      "loss": 3.6727,
+      "step": 303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.539196388299238,
+      "learning_rate": 0.0009690506455253072,
+      "loss": 3.6571,
+      "step": 304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8340992313666828,
+      "learning_rate": 0.0009687927659751481,
+      "loss": 3.6678,
+      "step": 305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5262972364653904,
+      "learning_rate": 0.0009685338511145312,
+      "loss": 3.6979,
+      "step": 306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7544416393538171,
+      "learning_rate": 0.0009682739015152598,
+      "loss": 3.6094,
+      "step": 307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7248694419726,
+      "learning_rate": 0.0009680129177514226,
+      "loss": 3.592,
+      "step": 308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1748044711117305,
+      "learning_rate": 0.0009677509003993915,
+      "loss": 3.6269,
+      "step": 309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5536884981903807,
+      "learning_rate": 0.0009674878500378221,
+      "loss": 3.706,
+      "step": 310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.444117152658226,
+      "learning_rate": 0.0009672237672476505,
+      "loss": 3.5589,
+      "step": 311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.030378372396853,
+      "learning_rate": 0.0009669586526120935,
+      "loss": 3.5906,
+      "step": 312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2850559114177056,
+      "learning_rate": 0.0009666925067166459,
+      "loss": 3.6716,
+      "step": 313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7681725858640163,
+      "learning_rate": 0.000966425330149081,
+      "loss": 3.6564,
+      "step": 314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.041670501314153,
+      "learning_rate": 0.0009661571234994475,
+      "loss": 3.5473,
+      "step": 315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.370501712168098,
+      "learning_rate": 0.0009658878873600691,
+      "loss": 3.6485,
+      "step": 316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5237967468272475,
+      "learning_rate": 0.0009656176223255438,
+      "loss": 3.6982,
+      "step": 317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.810753700527318,
+      "learning_rate": 0.000965346328992741,
+      "loss": 3.6773,
+      "step": 318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.386155063460771,
+      "learning_rate": 0.0009650740079608014,
+      "loss": 3.6389,
+      "step": 319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3171765931856894,
+      "learning_rate": 0.0009648006598311353,
+      "loss": 3.6296,
+      "step": 320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5775870267920995,
+      "learning_rate": 0.0009645262852074214,
+      "loss": 3.5711,
+      "step": 321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.311794270737017,
+      "learning_rate": 0.0009642508846956053,
+      "loss": 3.6232,
+      "step": 322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7959839173061645,
+      "learning_rate": 0.0009639744589038983,
+      "loss": 3.6163,
+      "step": 323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8615446328913974,
+      "learning_rate": 0.0009636970084427759,
+      "loss": 3.6842,
+      "step": 324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9847611202260347,
+      "learning_rate": 0.0009634185339249766,
+      "loss": 3.6486,
+      "step": 325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9032886739255737,
+      "learning_rate": 0.0009631390359655003,
+      "loss": 3.6372,
+      "step": 326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5689962641105155,
+      "learning_rate": 0.0009628585151816074,
+      "loss": 3.507,
+      "step": 327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4572798772953433,
+      "learning_rate": 0.0009625769721928172,
+      "loss": 3.5708,
+      "step": 328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6472702077520958,
+      "learning_rate": 0.0009622944076209061,
+      "loss": 3.5246,
+      "step": 329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.865868462244012,
+      "learning_rate": 0.0009620108220899071,
+      "loss": 3.6182,
+      "step": 330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7316113536704836,
+      "learning_rate": 0.0009617262162261075,
+      "loss": 3.5954,
+      "step": 331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2078314506262813,
+      "learning_rate": 0.0009614405906580486,
+      "loss": 3.6546,
+      "step": 332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5812821312167562,
+      "learning_rate": 0.000961153946016523,
+      "loss": 3.5582,
+      "step": 333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9913791037723723,
+      "learning_rate": 0.000960866282934574,
+      "loss": 3.4874,
+      "step": 334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7380651624760204,
+      "learning_rate": 0.0009605776020474945,
+      "loss": 3.5457,
+      "step": 335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.601249338831896,
+      "learning_rate": 0.0009602879039928249,
+      "loss": 3.6139,
+      "step": 336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.775884836264968,
+      "learning_rate": 0.0009599971894103521,
+      "loss": 3.5644,
+      "step": 337
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3142369395962814,
+      "learning_rate": 0.0009597054589421077,
+      "loss": 3.5753,
+      "step": 338
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.571719921389793,
+      "learning_rate": 0.0009594127132323669,
+      "loss": 3.5139,
+      "step": 339
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1565252283818004,
+      "learning_rate": 0.0009591189529276474,
+      "loss": 3.65,
+      "step": 340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7081838692732827,
+      "learning_rate": 0.0009588241786767072,
+      "loss": 3.6086,
+      "step": 341
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9586678639587665,
+      "learning_rate": 0.0009585283911305436,
+      "loss": 3.4313,
+      "step": 342
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2776510035476853,
+      "learning_rate": 0.000958231590942392,
+      "loss": 3.5938,
+      "step": 343
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.903579030222948,
+      "learning_rate": 0.0009579337787677238,
+      "loss": 3.5452,
+      "step": 344
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9468224101259228,
+      "learning_rate": 0.0009576349552642456,
+      "loss": 3.7081,
+      "step": 345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5293673880395064,
+      "learning_rate": 0.0009573351210918975,
+      "loss": 3.4599,
+      "step": 346
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.902521509478548,
+      "learning_rate": 0.0009570342769128514,
+      "loss": 3.5568,
+      "step": 347
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4831426018347473,
+      "learning_rate": 0.0009567324233915099,
+      "loss": 3.5397,
+      "step": 348
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8723534020910917,
+      "learning_rate": 0.0009564295611945047,
+      "loss": 3.6145,
+      "step": 349
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.905727375371706,
+      "learning_rate": 0.000956125690990695,
+      "loss": 3.446,
+      "step": 350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.776513193191346,
+      "learning_rate": 0.0009558208134511665,
+      "loss": 3.6854,
+      "step": 351
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1596293399337165,
+      "learning_rate": 0.0009555149292492289,
+      "loss": 3.4363,
+      "step": 352
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.645165846009241,
+      "learning_rate": 0.0009552080390604159,
+      "loss": 3.4653,
+      "step": 353
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.447135186473615,
+      "learning_rate": 0.0009549001435624823,
+      "loss": 3.4984,
+      "step": 354
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3733483207540065,
+      "learning_rate": 0.0009545912434354029,
+      "loss": 3.5004,
+      "step": 355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.062981161435643,
+      "learning_rate": 0.0009542813393613721,
+      "loss": 3.5652,
+      "step": 356
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.323212724109424,
+      "learning_rate": 0.0009539704320248006,
+      "loss": 3.538,
+      "step": 357
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.322462723866227,
+      "learning_rate": 0.0009536585221123151,
+      "loss": 3.3988,
+      "step": 358
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0652298768167228,
+      "learning_rate": 0.0009533456103127565,
+      "loss": 3.719,
+      "step": 359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8542549850214978,
+      "learning_rate": 0.000953031697317178,
+      "loss": 3.5195,
+      "step": 360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7654916751601055,
+      "learning_rate": 0.0009527167838188445,
+      "loss": 3.5898,
+      "step": 361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.363210558320693,
+      "learning_rate": 0.0009524008705132299,
+      "loss": 3.473,
+      "step": 362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7467350627015656,
+      "learning_rate": 0.0009520839580980166,
+      "loss": 3.4507,
+      "step": 363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2525855334131224,
+      "learning_rate": 0.0009517660472730929,
+      "loss": 3.4954,
+      "step": 364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0459484254839873,
+      "learning_rate": 0.0009514471387405526,
+      "loss": 3.5693,
+      "step": 365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.2043465735948145,
+      "learning_rate": 0.0009511272332046926,
+      "loss": 3.464,
+      "step": 366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6445385079729213,
+      "learning_rate": 0.0009508063313720119,
+      "loss": 3.5707,
+      "step": 367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.444705396634754,
+      "learning_rate": 0.0009504844339512095,
+      "loss": 3.5099,
+      "step": 368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6558910958181503,
+      "learning_rate": 0.0009501615416531835,
+      "loss": 3.5841,
+      "step": 369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5725795663573976,
+      "learning_rate": 0.0009498376551910285,
+      "loss": 3.5298,
+      "step": 370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7895016910698225,
+      "learning_rate": 0.0009495127752800352,
+      "loss": 3.5692,
+      "step": 371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9825142665211966,
+      "learning_rate": 0.0009491869026376882,
+      "loss": 3.5662,
+      "step": 372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9253987031590438,
+      "learning_rate": 0.0009488600379836648,
+      "loss": 3.5068,
+      "step": 373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.73198287134028,
+      "learning_rate": 0.0009485321820398321,
+      "loss": 3.4737,
+      "step": 374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8630543018576293,
+      "learning_rate": 0.0009482033355302475,
+      "loss": 3.5384,
+      "step": 375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6579794880138365,
+      "learning_rate": 0.0009478734991811556,
+      "loss": 3.4617,
+      "step": 376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.368215700721825,
+      "learning_rate": 0.0009475426737209871,
+      "loss": 3.7018,
+      "step": 377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.606722649164876,
+      "learning_rate": 0.000947210859880357,
+      "loss": 3.5272,
+      "step": 378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6062969760305512,
+      "learning_rate": 0.0009468780583920631,
+      "loss": 3.4589,
+      "step": 379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.995180433520697,
+      "learning_rate": 0.0009465442699910846,
+      "loss": 3.5278,
+      "step": 380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7996399619987986,
+      "learning_rate": 0.0009462094954145801,
+      "loss": 3.6991,
+      "step": 381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.395979750436394,
+      "learning_rate": 0.0009458737354018859,
+      "loss": 3.4887,
+      "step": 382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7286426031659765,
+      "learning_rate": 0.000945536990694515,
+      "loss": 3.5372,
+      "step": 383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.530814917029733,
+      "learning_rate": 0.0009451992620361551,
+      "loss": 3.5995,
+      "step": 384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.679695584619543,
+      "learning_rate": 0.0009448605501726664,
+      "loss": 3.4579,
+      "step": 385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7474129248144776,
+      "learning_rate": 0.000944520855852081,
+      "loss": 3.5939,
+      "step": 386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.64272080801023,
+      "learning_rate": 0.0009441801798246002,
+      "loss": 3.4266,
+      "step": 387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.05613896405206,
+      "learning_rate": 0.0009438385228425939,
+      "loss": 3.5824,
+      "step": 388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3002083472627626,
+      "learning_rate": 0.0009434958856605982,
+      "loss": 3.6104,
+      "step": 389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.194331205933116,
+      "learning_rate": 0.0009431522690353137,
+      "loss": 3.6211,
+      "step": 390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.691690274827961,
+      "learning_rate": 0.0009428076737256044,
+      "loss": 3.4907,
+      "step": 391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6731792509650225,
+      "learning_rate": 0.0009424621004924954,
+      "loss": 3.5944,
+      "step": 392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6318146551581914,
+      "learning_rate": 0.0009421155500991719,
+      "loss": 3.4372,
+      "step": 393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7760046192881567,
+      "learning_rate": 0.0009417680233109767,
+      "loss": 3.6853,
+      "step": 394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9992977236166725,
+      "learning_rate": 0.000941419520895409,
+      "loss": 3.3896,
+      "step": 395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.726570617509536,
+      "learning_rate": 0.0009410700436221229,
+      "loss": 3.4678,
+      "step": 396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5555208379899448,
+      "learning_rate": 0.0009407195922629252,
+      "loss": 3.4098,
+      "step": 397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.823777951397034,
+      "learning_rate": 0.000940368167591774,
+      "loss": 3.338,
+      "step": 398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9465258220967048,
+      "learning_rate": 0.0009400157703847769,
+      "loss": 3.4794,
+      "step": 399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3989130597967026,
+      "learning_rate": 0.0009396624014201895,
+      "loss": 3.4446,
+      "step": 400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5715282960589165,
+      "learning_rate": 0.000939308061478413,
+      "loss": 3.4371,
+      "step": 401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3042110979942683,
+      "learning_rate": 0.0009389527513419935,
+      "loss": 3.5076,
+      "step": 402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6081746165739539,
+      "learning_rate": 0.0009385964717956195,
+      "loss": 3.4799,
+      "step": 403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.669501676585062,
+      "learning_rate": 0.0009382392236261201,
+      "loss": 3.4021,
+      "step": 404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4303959075960422,
+      "learning_rate": 0.0009378810076224644,
+      "loss": 3.3759,
+      "step": 405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.576416725638094,
+      "learning_rate": 0.0009375218245757582,
+      "loss": 3.444,
+      "step": 406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.935907079674561,
+      "learning_rate": 0.0009371616752792432,
+      "loss": 3.3947,
+      "step": 407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4561885982475906,
+      "learning_rate": 0.0009368005605282949,
+      "loss": 3.4526,
+      "step": 408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7202209415554994,
+      "learning_rate": 0.0009364384811204212,
+      "loss": 3.5111,
+      "step": 409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6758862323952628,
+      "learning_rate": 0.00093607543785526,
+      "loss": 3.5464,
+      "step": 410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6099235431228816,
+      "learning_rate": 0.0009357114315345787,
+      "loss": 3.4522,
+      "step": 411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0881738891082535,
+      "learning_rate": 0.0009353464629622705,
+      "loss": 3.4902,
+      "step": 412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0599948997662434,
+      "learning_rate": 0.0009349805329443544,
+      "loss": 3.4213,
+      "step": 413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7344341132374546,
+      "learning_rate": 0.0009346136422889724,
+      "loss": 3.416,
+      "step": 414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7109549636909225,
+      "learning_rate": 0.0009342457918063882,
+      "loss": 3.468,
+      "step": 415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.731527476257493,
+      "learning_rate": 0.0009338769823089853,
+      "loss": 3.4005,
+      "step": 416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5564932093988855,
+      "learning_rate": 0.0009335072146112648,
+      "loss": 3.4531,
+      "step": 417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8576589207571153,
+      "learning_rate": 0.0009331364895298444,
+      "loss": 3.5917,
+      "step": 418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.770879766916332,
+      "learning_rate": 0.0009327648078834559,
+      "loss": 3.3388,
+      "step": 419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8124718580716064,
+      "learning_rate": 0.0009323921704929434,
+      "loss": 3.4826,
+      "step": 420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0634580228380703,
+      "learning_rate": 0.0009320185781812623,
+      "loss": 3.4785,
+      "step": 421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8784036644581337,
+      "learning_rate": 0.0009316440317734762,
+      "loss": 3.3707,
+      "step": 422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6687892110181501,
+      "learning_rate": 0.0009312685320967565,
+      "loss": 3.4043,
+      "step": 423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7705816669080519,
+      "learning_rate": 0.0009308920799803793,
+      "loss": 3.5596,
+      "step": 424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5903059770978016,
+      "learning_rate": 0.0009305146762557246,
+      "loss": 3.3598,
+      "step": 425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0737811500232133,
+      "learning_rate": 0.0009301363217562736,
+      "loss": 3.4015,
+      "step": 426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5043397306248527,
+      "learning_rate": 0.0009297570173176074,
+      "loss": 3.4656,
+      "step": 427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8237923881453229,
+      "learning_rate": 0.000929376763777405,
+      "loss": 3.5637,
+      "step": 428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.649098712523963,
+      "learning_rate": 0.0009289955619754413,
+      "loss": 3.3972,
+      "step": 429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4465461726064968,
+      "learning_rate": 0.0009286134127535859,
+      "loss": 3.5047,
+      "step": 430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6576976327388024,
+      "learning_rate": 0.0009282303169558,
+      "loss": 3.356,
+      "step": 431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9801314023750924,
+      "learning_rate": 0.0009278462754281359,
+      "loss": 3.3969,
+      "step": 432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.880680977994443,
+      "learning_rate": 0.0009274612890187342,
+      "loss": 3.4335,
+      "step": 433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.684435933347235,
+      "learning_rate": 0.0009270753585778222,
+      "loss": 3.4583,
+      "step": 434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7365229633066825,
+      "learning_rate": 0.0009266884849577124,
+      "loss": 3.3409,
+      "step": 435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.385031340182239,
+      "learning_rate": 0.0009263006690127998,
+      "loss": 3.4214,
+      "step": 436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8128207633491114,
+      "learning_rate": 0.0009259119115995609,
+      "loss": 3.4349,
+      "step": 437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1185640465297286,
+      "learning_rate": 0.0009255222135765511,
+      "loss": 3.506,
+      "step": 438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7369966072756575,
+      "learning_rate": 0.0009251315758044032,
+      "loss": 3.5326,
+      "step": 439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.277366829006584,
+      "learning_rate": 0.0009247399991458255,
+      "loss": 3.474,
+      "step": 440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8120670216799835,
+      "learning_rate": 0.0009243474844655996,
+      "loss": 3.4433,
+      "step": 441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.901458389663372,
+      "learning_rate": 0.0009239540326305791,
+      "loss": 3.3993,
+      "step": 442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9663373856195,
+      "learning_rate": 0.0009235596445096864,
+      "loss": 3.5039,
+      "step": 443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0818799744402567,
+      "learning_rate": 0.0009231643209739127,
+      "loss": 3.5387,
+      "step": 444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.127543290742337,
+      "learning_rate": 0.0009227680628963145,
+      "loss": 3.4809,
+      "step": 445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.245303571818071,
+      "learning_rate": 0.000922370871152012,
+      "loss": 3.4785,
+      "step": 446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.95926956292938,
+      "learning_rate": 0.0009219727466181877,
+      "loss": 3.4772,
+      "step": 447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6174311987257208,
+      "learning_rate": 0.0009215736901740841,
+      "loss": 3.4694,
+      "step": 448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4581576732257604,
+      "learning_rate": 0.0009211737027010016,
+      "loss": 3.6154,
+      "step": 449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.526454551098847,
+      "learning_rate": 0.0009207727850822971,
+      "loss": 3.3546,
+      "step": 450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7520609076781084,
+      "learning_rate": 0.0009203709382033814,
+      "loss": 3.4754,
+      "step": 451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2591339907058434,
+      "learning_rate": 0.0009199681629517173,
+      "loss": 3.4977,
+      "step": 452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6234327700136917,
+      "learning_rate": 0.0009195644602168184,
+      "loss": 3.4331,
+      "step": 453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3407939838780325,
+      "learning_rate": 0.0009191598308902464,
+      "loss": 3.3638,
+      "step": 454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.630067229877852,
+      "learning_rate": 0.0009187542758656091,
+      "loss": 3.4307,
+      "step": 455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2002288835796837,
+      "learning_rate": 0.0009183477960385591,
+      "loss": 3.3241,
+      "step": 456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.707296414713657,
+      "learning_rate": 0.0009179403923067912,
+      "loss": 3.4136,
+      "step": 457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9109288072338313,
+      "learning_rate": 0.0009175320655700406,
+      "loss": 3.3404,
+      "step": 458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6820695108884152,
+      "learning_rate": 0.0009171228167300805,
+      "loss": 3.4075,
+      "step": 459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8319297512491712,
+      "learning_rate": 0.0009167126466907215,
+      "loss": 3.4339,
+      "step": 460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7678304014193786,
+      "learning_rate": 0.0009163015563578074,
+      "loss": 3.4466,
+      "step": 461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5724630986655956,
+      "learning_rate": 0.0009158895466392158,
+      "loss": 3.492,
+      "step": 462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.921178377962709,
+      "learning_rate": 0.0009154766184448535,
+      "loss": 3.3428,
+      "step": 463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7400922659208098,
+      "learning_rate": 0.0009150627726866568,
+      "loss": 3.3615,
+      "step": 464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.661187780578636,
+      "learning_rate": 0.000914648010278587,
+      "loss": 3.3258,
+      "step": 465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.62586576884231,
+      "learning_rate": 0.0009142323321366315,
+      "loss": 3.4417,
+      "step": 466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4314119354340953,
+      "learning_rate": 0.0009138157391787986,
+      "loss": 3.4901,
+      "step": 467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0224138669542877,
+      "learning_rate": 0.0009133982323251177,
+      "loss": 3.2735,
+      "step": 468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6925384146419682,
+      "learning_rate": 0.0009129798124976365,
+      "loss": 3.5049,
+      "step": 469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.674486989595892,
+      "learning_rate": 0.0009125604806204187,
+      "loss": 3.3607,
+      "step": 470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8518739064657557,
+      "learning_rate": 0.0009121402376195421,
+      "loss": 3.2419,
+      "step": 471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7328067175202182,
+      "learning_rate": 0.0009117190844230972,
+      "loss": 3.453,
+      "step": 472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.96966136859594,
+      "learning_rate": 0.0009112970219611841,
+      "loss": 3.304,
+      "step": 473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7985657125785213,
+      "learning_rate": 0.0009108740511659115,
+      "loss": 3.4432,
+      "step": 474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3223605657850603,
+      "learning_rate": 0.0009104501729713935,
+      "loss": 3.3415,
+      "step": 475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8486279434291994,
+      "learning_rate": 0.0009100253883137488,
+      "loss": 3.3076,
+      "step": 476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.219321369812922,
+      "learning_rate": 0.0009095996981310974,
+      "loss": 3.4052,
+      "step": 477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.776250841446645,
+      "learning_rate": 0.0009091731033635596,
+      "loss": 3.3758,
+      "step": 478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2141015445149375,
+      "learning_rate": 0.0009087456049532529,
+      "loss": 3.3954,
+      "step": 479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8554374905009505,
+      "learning_rate": 0.0009083172038442914,
+      "loss": 3.4415,
+      "step": 480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5351297877494183,
+      "learning_rate": 0.0009078879009827817,
+      "loss": 3.3278,
+      "step": 481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8409082360534232,
+      "learning_rate": 0.0009074576973168223,
+      "loss": 3.3779,
+      "step": 482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7310698124009687,
+      "learning_rate": 0.0009070265937965015,
+      "loss": 3.4068,
+      "step": 483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4787944863607425,
+      "learning_rate": 0.0009065945913738942,
+      "loss": 3.4092,
+      "step": 484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8313488883989413,
+      "learning_rate": 0.0009061616910030609,
+      "loss": 3.3577,
+      "step": 485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7348537843964962,
+      "learning_rate": 0.0009057278936400453,
+      "loss": 3.2885,
+      "step": 486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7119651834201104,
+      "learning_rate": 0.0009052932002428715,
+      "loss": 3.4091,
+      "step": 487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.272349035042926,
+      "learning_rate": 0.0009048576117715435,
+      "loss": 3.4104,
+      "step": 488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2498018370365873,
+      "learning_rate": 0.0009044211291880407,
+      "loss": 3.3229,
+      "step": 489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.824106150939754,
+      "learning_rate": 0.000903983753456318,
+      "loss": 3.3189,
+      "step": 490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.609935440142885,
+      "learning_rate": 0.0009035454855423026,
+      "loss": 3.4166,
+      "step": 491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1141254972422283,
+      "learning_rate": 0.0009031063264138922,
+      "loss": 3.3913,
+      "step": 492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7380660853385321,
+      "learning_rate": 0.0009026662770409522,
+      "loss": 3.3438,
+      "step": 493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1811362036142588,
+      "learning_rate": 0.0009022253383953147,
+      "loss": 3.45,
+      "step": 494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.591462683814806,
+      "learning_rate": 0.0009017835114507753,
+      "loss": 3.3614,
+      "step": 495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9425986004292877,
+      "learning_rate": 0.0009013407971830914,
+      "loss": 3.3491,
+      "step": 496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5630557926322657,
+      "learning_rate": 0.0009008971965699801,
+      "loss": 3.3664,
+      "step": 497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0788868059004635,
+      "learning_rate": 0.0009004527105911163,
+      "loss": 3.405,
+      "step": 498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6266886367789748,
+      "learning_rate": 0.0009000073402281295,
+      "loss": 3.358,
+      "step": 499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3024002943298507,
+      "learning_rate": 0.0008995610864646028,
+      "loss": 3.3413,
+      "step": 500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9408551085458854,
+      "learning_rate": 0.0008991139502860703,
+      "loss": 3.3105,
+      "step": 501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8979616484252637,
+      "learning_rate": 0.0008986659326800146,
+      "loss": 3.2568,
+      "step": 502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6938110041869125,
+      "learning_rate": 0.0008982170346358651,
+      "loss": 3.411,
+      "step": 503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9963465475241668,
+      "learning_rate": 0.0008977672571449956,
+      "loss": 3.2782,
+      "step": 504
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6627237232397292,
+      "learning_rate": 0.0008973166012007217,
+      "loss": 3.4484,
+      "step": 505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3203543409766016,
+      "learning_rate": 0.0008968650677982998,
+      "loss": 3.3438,
+      "step": 506
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.758497109336238,
+      "learning_rate": 0.0008964126579349236,
+      "loss": 3.3084,
+      "step": 507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1973061917691656,
+      "learning_rate": 0.0008959593726097226,
+      "loss": 3.2597,
+      "step": 508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5940537867073092,
+      "learning_rate": 0.0008955052128237596,
+      "loss": 3.3105,
+      "step": 509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.195055168391578,
+      "learning_rate": 0.0008950501795800288,
+      "loss": 3.4197,
+      "step": 510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0418348261677286,
+      "learning_rate": 0.0008945942738834532,
+      "loss": 3.2564,
+      "step": 511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4280819467601507,
+      "learning_rate": 0.0008941374967408826,
+      "loss": 3.338,
+      "step": 512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2888548243564,
+      "learning_rate": 0.0008936798491610916,
+      "loss": 3.3466,
+      "step": 513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8295871252618472,
+      "learning_rate": 0.0008932213321547768,
+      "loss": 3.3985,
+      "step": 514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6997592803909591,
+      "learning_rate": 0.0008927619467345554,
+      "loss": 3.326,
+      "step": 515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2495292711728596,
+      "learning_rate": 0.0008923016939149615,
+      "loss": 3.2952,
+      "step": 516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8388708751299068,
+      "learning_rate": 0.0008918405747124458,
+      "loss": 3.2362,
+      "step": 517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.781614116964968,
+      "learning_rate": 0.0008913785901453721,
+      "loss": 3.376,
+      "step": 518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6270969308575836,
+      "learning_rate": 0.000890915741234015,
+      "loss": 3.3344,
+      "step": 519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0925814208208435,
+      "learning_rate": 0.0008904520290005582,
+      "loss": 3.4346,
+      "step": 520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8397152383033297,
+      "learning_rate": 0.000889987454469092,
+      "loss": 3.4677,
+      "step": 521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9564533708950318,
+      "learning_rate": 0.0008895220186656111,
+      "loss": 3.257,
+      "step": 522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8116627974008217,
+      "learning_rate": 0.0008890557226180122,
+      "loss": 3.3259,
+      "step": 523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6874684079076412,
+      "learning_rate": 0.0008885885673560921,
+      "loss": 3.3317,
+      "step": 524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6524033709996893,
+      "learning_rate": 0.0008881205539115444,
+      "loss": 3.4599,
+      "step": 525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.185942515422613,
+      "learning_rate": 0.0008876516833179589,
+      "loss": 3.3128,
+      "step": 526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0921994030954054,
+      "learning_rate": 0.0008871819566108177,
+      "loss": 3.2384,
+      "step": 527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7822835935721109,
+      "learning_rate": 0.000886711374827494,
+      "loss": 3.3213,
+      "step": 528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4185008413029958,
+      "learning_rate": 0.0008862399390072491,
+      "loss": 3.3381,
+      "step": 529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7512388894927207,
+      "learning_rate": 0.0008857676501912305,
+      "loss": 3.2829,
+      "step": 530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4232963640970344,
+      "learning_rate": 0.0008852945094224697,
+      "loss": 3.3689,
+      "step": 531
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9024385337608616,
+      "learning_rate": 0.0008848205177458795,
+      "loss": 3.341,
+      "step": 532
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0304521109574383,
+      "learning_rate": 0.0008843456762082518,
+      "loss": 3.3124,
+      "step": 533
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.663749541693311,
+      "learning_rate": 0.0008838699858582557,
+      "loss": 3.3696,
+      "step": 534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0181175850990134,
+      "learning_rate": 0.0008833934477464347,
+      "loss": 3.4151,
+      "step": 535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.117002381369704,
+      "learning_rate": 0.0008829160629252045,
+      "loss": 3.329,
+      "step": 536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.863025243224994,
+      "learning_rate": 0.0008824378324488509,
+      "loss": 3.3564,
+      "step": 537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3140113319528286,
+      "learning_rate": 0.0008819587573735268,
+      "loss": 3.4483,
+      "step": 538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.861241616789604,
+      "learning_rate": 0.0008814788387572513,
+      "loss": 3.2453,
+      "step": 539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9386193096067776,
+      "learning_rate": 0.0008809980776599053,
+      "loss": 3.3605,
+      "step": 540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7805069391035158,
+      "learning_rate": 0.0008805164751432312,
+      "loss": 3.224,
+      "step": 541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.870288981489457,
+      "learning_rate": 0.0008800340322708292,
+      "loss": 3.4167,
+      "step": 542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8453224513220046,
+      "learning_rate": 0.0008795507501081555,
+      "loss": 3.1834,
+      "step": 543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8352834463790855,
+      "learning_rate": 0.0008790666297225196,
+      "loss": 3.1546,
+      "step": 544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2383655760986514,
+      "learning_rate": 0.0008785816721830829,
+      "loss": 3.3233,
+      "step": 545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6772643971288024,
+      "learning_rate": 0.0008780958785608546,
+      "loss": 3.4182,
+      "step": 546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3837529690021717,
+      "learning_rate": 0.0008776092499286912,
+      "loss": 3.2168,
+      "step": 547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7728586505749226,
+      "learning_rate": 0.0008771217873612929,
+      "loss": 3.2463,
+      "step": 548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8613987334801412,
+      "learning_rate": 0.0008766334919352017,
+      "loss": 3.4169,
+      "step": 549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4710610344954709,
+      "learning_rate": 0.0008761443647287987,
+      "loss": 3.2841,
+      "step": 550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0371229058722533,
+      "learning_rate": 0.0008756544068223026,
+      "loss": 3.3636,
+      "step": 551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7865249072822424,
+      "learning_rate": 0.0008751636192977659,
+      "loss": 3.3086,
+      "step": 552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.88333169691346,
+      "learning_rate": 0.0008746720032390737,
+      "loss": 3.3316,
+      "step": 553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4512278068120044,
+      "learning_rate": 0.0008741795597319408,
+      "loss": 3.2887,
+      "step": 554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6615086210003165,
+      "learning_rate": 0.0008736862898639095,
+      "loss": 3.4159,
+      "step": 555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7486853471781831,
+      "learning_rate": 0.0008731921947243468,
+      "loss": 3.3083,
+      "step": 556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.550367158434715,
+      "learning_rate": 0.0008726972754044427,
+      "loss": 3.3366,
+      "step": 557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7867429950362206,
+      "learning_rate": 0.0008722015329972069,
+      "loss": 3.3256,
+      "step": 558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7039659731653147,
+      "learning_rate": 0.0008717049685974672,
+      "loss": 3.3435,
+      "step": 559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.134878554851738,
+      "learning_rate": 0.0008712075833018665,
+      "loss": 3.3222,
+      "step": 560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8039013776264234,
+      "learning_rate": 0.0008707093782088608,
+      "loss": 3.3,
+      "step": 561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.3290639415624215,
+      "learning_rate": 0.0008702103544187167,
+      "loss": 3.1758,
+      "step": 562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8035191181213226,
+      "learning_rate": 0.0008697105130335085,
+      "loss": 3.3053,
+      "step": 563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3638673326044772,
+      "learning_rate": 0.0008692098551571164,
+      "loss": 3.4292,
+      "step": 564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.376026297976182,
+      "learning_rate": 0.0008687083818952235,
+      "loss": 3.4196,
+      "step": 565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.613149360688895,
+      "learning_rate": 0.0008682060943553143,
+      "loss": 3.326,
+      "step": 566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6365627898460071,
+      "learning_rate": 0.0008677029936466707,
+      "loss": 3.2567,
+      "step": 567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8557870980695061,
+      "learning_rate": 0.0008671990808803711,
+      "loss": 3.4336,
+      "step": 568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.33864457210145,
+      "learning_rate": 0.0008666943571692871,
+      "loss": 3.3482,
+      "step": 569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.548689034074765,
+      "learning_rate": 0.0008661888236280813,
+      "loss": 3.3593,
+      "step": 570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8438784463105635,
+      "learning_rate": 0.0008656824813732045,
+      "loss": 3.2068,
+      "step": 571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.730792380037019,
+      "learning_rate": 0.000865175331522894,
+      "loss": 3.3034,
+      "step": 572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2491821385618223,
+      "learning_rate": 0.0008646673751971703,
+      "loss": 3.2503,
+      "step": 573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5740825514983365,
+      "learning_rate": 0.000864158613517835,
+      "loss": 3.2725,
+      "step": 574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.185589956120587,
+      "learning_rate": 0.0008636490476084681,
+      "loss": 3.3181,
+      "step": 575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.634248626680023,
+      "learning_rate": 0.0008631386785944264,
+      "loss": 3.2235,
+      "step": 576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2924411955336925,
+      "learning_rate": 0.0008626275076028397,
+      "loss": 3.4162,
+      "step": 577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7567491349967659,
+      "learning_rate": 0.0008621155357626091,
+      "loss": 3.3291,
+      "step": 578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5615222203844277,
+      "learning_rate": 0.0008616027642044042,
+      "loss": 3.3117,
+      "step": 579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6778790408202329,
+      "learning_rate": 0.000861089194060661,
+      "loss": 3.357,
+      "step": 580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8946696872637532,
+      "learning_rate": 0.000860574826465579,
+      "loss": 3.2355,
+      "step": 581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6273163155038646,
+      "learning_rate": 0.0008600596625551191,
+      "loss": 3.3791,
+      "step": 582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4952230585247044,
+      "learning_rate": 0.0008595437034670006,
+      "loss": 3.2865,
+      "step": 583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4709272213824849,
+      "learning_rate": 0.0008590269503406985,
+      "loss": 3.232,
+      "step": 584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7827498808437716,
+      "learning_rate": 0.0008585094043174423,
+      "loss": 3.3423,
+      "step": 585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5851314887881403,
+      "learning_rate": 0.0008579910665402118,
+      "loss": 3.3411,
+      "step": 586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5855188261813333,
+      "learning_rate": 0.000857471938153736,
+      "loss": 3.4136,
+      "step": 587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.678464953847863,
+      "learning_rate": 0.0008569520203044892,
+      "loss": 3.2553,
+      "step": 588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.59374643213086,
+      "learning_rate": 0.0008564313141406901,
+      "loss": 3.2333,
+      "step": 589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8954803411062628,
+      "learning_rate": 0.0008559098208122973,
+      "loss": 3.1869,
+      "step": 590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9611709634653063,
+      "learning_rate": 0.0008553875414710089,
+      "loss": 3.3059,
+      "step": 591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.02833831469526,
+      "learning_rate": 0.0008548644772702579,
+      "loss": 3.2143,
+      "step": 592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.031217093657953,
+      "learning_rate": 0.0008543406293652116,
+      "loss": 3.3228,
+      "step": 593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0091486035320276,
+      "learning_rate": 0.0008538159989127671,
+      "loss": 3.3931,
+      "step": 594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.529633956266805,
+      "learning_rate": 0.0008532905870715505,
+      "loss": 3.2764,
+      "step": 595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.858018711935005,
+      "learning_rate": 0.0008527643950019131,
+      "loss": 3.3451,
+      "step": 596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3075214232841805,
+      "learning_rate": 0.0008522374238659296,
+      "loss": 3.2885,
+      "step": 597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8467457681123634,
+      "learning_rate": 0.0008517096748273951,
+      "loss": 3.2178,
+      "step": 598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3201022265076268,
+      "learning_rate": 0.0008511811490518227,
+      "loss": 3.3459,
+      "step": 599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6570543791880168,
+      "learning_rate": 0.0008506518477064405,
+      "loss": 3.3686,
+      "step": 600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7320253754317436,
+      "learning_rate": 0.0008501217719601903,
+      "loss": 3.2651,
+      "step": 601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7356300822211161,
+      "learning_rate": 0.0008495909229837233,
+      "loss": 3.4141,
+      "step": 602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2907365140058658,
+      "learning_rate": 0.000849059301949399,
+      "loss": 3.1965,
+      "step": 603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7315803349844119,
+      "learning_rate": 0.0008485269100312812,
+      "loss": 3.2143,
+      "step": 604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.936930620964027,
+      "learning_rate": 0.0008479937484051368,
+      "loss": 3.3501,
+      "step": 605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8913047484801562,
+      "learning_rate": 0.0008474598182484323,
+      "loss": 3.3652,
+      "step": 606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5355041001379583,
+      "learning_rate": 0.0008469251207403317,
+      "loss": 3.2047,
+      "step": 607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6764167614756063,
+      "learning_rate": 0.0008463896570616934,
+      "loss": 3.2914,
+      "step": 608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5291054515699833,
+      "learning_rate": 0.0008458534283950678,
+      "loss": 3.3395,
+      "step": 609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9184746558407748,
+      "learning_rate": 0.0008453164359246952,
+      "loss": 3.2896,
+      "step": 610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4944574068675767,
+      "learning_rate": 0.0008447786808365022,
+      "loss": 3.2726,
+      "step": 611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4689886000708652,
+      "learning_rate": 0.0008442401643181,
+      "loss": 3.3845,
+      "step": 612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5385297471812438,
+      "learning_rate": 0.0008437008875587811,
+      "loss": 3.2654,
+      "step": 613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9604500907643785,
+      "learning_rate": 0.0008431608517495171,
+      "loss": 3.2073,
+      "step": 614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5668938585863246,
+      "learning_rate": 0.0008426200580829561,
+      "loss": 3.2007,
+      "step": 615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9012734605208454,
+      "learning_rate": 0.0008420785077534195,
+      "loss": 3.2283,
+      "step": 616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6270365599128624,
+      "learning_rate": 0.0008415362019569001,
+      "loss": 3.3658,
+      "step": 617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0824453101345703,
+      "learning_rate": 0.0008409931418910591,
+      "loss": 3.2513,
+      "step": 618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.557332063817956,
+      "learning_rate": 0.0008404493287552232,
+      "loss": 3.2874,
+      "step": 619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.119438314477117,
+      "learning_rate": 0.0008399047637503825,
+      "loss": 3.2559,
+      "step": 620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.513885439772425,
+      "learning_rate": 0.0008393594480791875,
+      "loss": 3.2442,
+      "step": 621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.132035909354857,
+      "learning_rate": 0.0008388133829459463,
+      "loss": 3.321,
+      "step": 622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.784783485577915,
+      "learning_rate": 0.0008382665695566227,
+      "loss": 3.3442,
+      "step": 623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8788302785292716,
+      "learning_rate": 0.0008377190091188324,
+      "loss": 3.1885,
+      "step": 624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5178194304415435,
+      "learning_rate": 0.0008371707028418413,
+      "loss": 3.2908,
+      "step": 625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.774439470622937,
+      "learning_rate": 0.0008366216519365621,
+      "loss": 3.3697,
+      "step": 626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5926630739731682,
+      "learning_rate": 0.0008360718576155525,
+      "loss": 3.2184,
+      "step": 627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8879180995201594,
+      "learning_rate": 0.0008355213210930118,
+      "loss": 3.2693,
+      "step": 628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.734027998042461,
+      "learning_rate": 0.0008349700435847778,
+      "loss": 3.328,
+      "step": 629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6089493506575585,
+      "learning_rate": 0.0008344180263083256,
+      "loss": 3.3093,
+      "step": 630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9012205475341957,
+      "learning_rate": 0.000833865270482764,
+      "loss": 3.1791,
+      "step": 631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7167795784242552,
+      "learning_rate": 0.0008333117773288324,
+      "loss": 3.1349,
+      "step": 632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5225098312867988,
+      "learning_rate": 0.0008327575480688985,
+      "loss": 3.1746,
+      "step": 633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4541304445349534,
+      "learning_rate": 0.000832202583926956,
+      "loss": 3.2395,
+      "step": 634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.501094025951985,
+      "learning_rate": 0.0008316468861286217,
+      "loss": 3.2419,
+      "step": 635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.584852677052243,
+      "learning_rate": 0.0008310904559011323,
+      "loss": 3.2189,
+      "step": 636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6953742978001891,
+      "learning_rate": 0.0008305332944733419,
+      "loss": 3.2605,
+      "step": 637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5150863277629674,
+      "learning_rate": 0.0008299754030757202,
+      "loss": 3.2322,
+      "step": 638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9411671267670116,
+      "learning_rate": 0.0008294167829403481,
+      "loss": 3.3244,
+      "step": 639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.366225187442535,
+      "learning_rate": 0.0008288574353009164,
+      "loss": 3.2694,
+      "step": 640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4690552559356713,
+      "learning_rate": 0.0008282973613927225,
+      "loss": 3.2528,
+      "step": 641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9084292847323951,
+      "learning_rate": 0.0008277365624526675,
+      "loss": 3.2738,
+      "step": 642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6921728769728186,
+      "learning_rate": 0.0008271750397192541,
+      "loss": 3.1199,
+      "step": 643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6365330530082103,
+      "learning_rate": 0.0008266127944325832,
+      "loss": 3.2161,
+      "step": 644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4821829854373865,
+      "learning_rate": 0.0008260498278343513,
+      "loss": 3.2194,
+      "step": 645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2077887381837242,
+      "learning_rate": 0.0008254861411678485,
+      "loss": 3.2157,
+      "step": 646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.218017572475962,
+      "learning_rate": 0.0008249217356779544,
+      "loss": 3.2348,
+      "step": 647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7056540316586297,
+      "learning_rate": 0.0008243566126111363,
+      "loss": 3.1797,
+      "step": 648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7753730742496463,
+      "learning_rate": 0.0008237907732154466,
+      "loss": 3.2712,
+      "step": 649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8498605437313873,
+      "learning_rate": 0.0008232242187405194,
+      "loss": 3.2276,
+      "step": 650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.625675930926069,
+      "learning_rate": 0.000822656950437568,
+      "loss": 3.2881,
+      "step": 651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2570009302715484,
+      "learning_rate": 0.0008220889695593823,
+      "loss": 3.2903,
+      "step": 652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.186257060413251,
+      "learning_rate": 0.0008215202773603259,
+      "loss": 3.2817,
+      "step": 653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5746107112952663,
+      "learning_rate": 0.0008209508750963328,
+      "loss": 3.1865,
+      "step": 654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9149625008871842,
+      "learning_rate": 0.0008203807640249062,
+      "loss": 3.2574,
+      "step": 655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6050674213773084,
+      "learning_rate": 0.0008198099454051136,
+      "loss": 3.3231,
+      "step": 656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.667855666901752,
+      "learning_rate": 0.0008192384204975857,
+      "loss": 3.2268,
+      "step": 657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5122845842026393,
+      "learning_rate": 0.000818666190564513,
+      "loss": 3.2615,
+      "step": 658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.572438910147217,
+      "learning_rate": 0.0008180932568696426,
+      "loss": 3.1187,
+      "step": 659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9191530583228213,
+      "learning_rate": 0.0008175196206782764,
+      "loss": 3.2416,
+      "step": 660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5189775851859841,
+      "learning_rate": 0.0008169452832572675,
+      "loss": 3.1698,
+      "step": 661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5547057914520774,
+      "learning_rate": 0.0008163702458750173,
+      "loss": 3.3375,
+      "step": 662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4324753988768066,
+      "learning_rate": 0.0008157945098014734,
+      "loss": 3.2925,
+      "step": 663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6481647655269158,
+      "learning_rate": 0.0008152180763081267,
+      "loss": 3.2841,
+      "step": 664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4547746922567637,
+      "learning_rate": 0.0008146409466680076,
+      "loss": 3.2777,
+      "step": 665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9039716751476037,
+      "learning_rate": 0.0008140631221556845,
+      "loss": 3.2677,
+      "step": 666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9652240777327876,
+      "learning_rate": 0.0008134846040472599,
+      "loss": 3.3126,
+      "step": 667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3155298029478684,
+      "learning_rate": 0.0008129053936203688,
+      "loss": 3.3557,
+      "step": 668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7857987920606522,
+      "learning_rate": 0.0008123254921541745,
+      "loss": 3.2684,
+      "step": 669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7247279100938093,
+      "learning_rate": 0.0008117449009293668,
+      "loss": 3.2268,
+      "step": 670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.658142300366259,
+      "learning_rate": 0.0008111636212281586,
+      "loss": 3.2669,
+      "step": 671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7449411517622884,
+      "learning_rate": 0.0008105816543342833,
+      "loss": 3.2614,
+      "step": 672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7655773336304372,
+      "learning_rate": 0.0008099990015329919,
+      "loss": 3.2546,
+      "step": 673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1078083764044595,
+      "learning_rate": 0.0008094156641110504,
+      "loss": 3.2361,
+      "step": 674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.60525240333491,
+      "learning_rate": 0.0008088316433567369,
+      "loss": 3.3166,
+      "step": 675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8106558271549442,
+      "learning_rate": 0.0008082469405598378,
+      "loss": 3.3211,
+      "step": 676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9484061229681748,
+      "learning_rate": 0.0008076615570116468,
+      "loss": 3.2326,
+      "step": 677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6219908421959777,
+      "learning_rate": 0.0008070754940049603,
+      "loss": 3.2618,
+      "step": 678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5475953180602613,
+      "learning_rate": 0.0008064887528340756,
+      "loss": 3.1618,
+      "step": 679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.20893414541729,
+      "learning_rate": 0.0008059013347947874,
+      "loss": 3.2445,
+      "step": 680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3917020194020395,
+      "learning_rate": 0.0008053132411843857,
+      "loss": 3.216,
+      "step": 681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6855781155795742,
+      "learning_rate": 0.0008047244733016521,
+      "loss": 3.2361,
+      "step": 682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.002077066474074,
+      "learning_rate": 0.0008041350324468573,
+      "loss": 3.1587,
+      "step": 683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0910455289122436,
+      "learning_rate": 0.0008035449199217583,
+      "loss": 3.2212,
+      "step": 684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6755627122099985,
+      "learning_rate": 0.0008029541370295957,
+      "loss": 3.3213,
+      "step": 685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6928380931849847,
+      "learning_rate": 0.0008023626850750903,
+      "loss": 3.2462,
+      "step": 686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0378230576532266,
+      "learning_rate": 0.0008017705653644406,
+      "loss": 3.265,
+      "step": 687
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.157296478120413,
+      "learning_rate": 0.0008011777792053195,
+      "loss": 3.1538,
+      "step": 688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8281523042513712,
+      "learning_rate": 0.0008005843279068725,
+      "loss": 3.2552,
+      "step": 689
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.090635657186393,
+      "learning_rate": 0.000799990212779713,
+      "loss": 3.2403,
+      "step": 690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3673169019920453,
+      "learning_rate": 0.0007993954351359214,
+      "loss": 3.2366,
+      "step": 691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8079639963908294,
+      "learning_rate": 0.0007987999962890406,
+      "loss": 3.3953,
+      "step": 692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.025920396574933,
+      "learning_rate": 0.0007982038975540742,
+      "loss": 3.1545,
+      "step": 693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6163936940434052,
+      "learning_rate": 0.0007976071402474826,
+      "loss": 3.2087,
+      "step": 694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.088965932299249,
+      "learning_rate": 0.0007970097256871811,
+      "loss": 3.3125,
+      "step": 695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.577412773323062,
+      "learning_rate": 0.0007964116551925364,
+      "loss": 3.2191,
+      "step": 696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5281456507288242,
+      "learning_rate": 0.0007958129300843637,
+      "loss": 3.0747,
+      "step": 697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8386835458916255,
+      "learning_rate": 0.0007952135516849239,
+      "loss": 3.0903,
+      "step": 698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6710332406295618,
+      "learning_rate": 0.0007946135213179207,
+      "loss": 3.3258,
+      "step": 699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8381029706644612,
+      "learning_rate": 0.0007940128403084977,
+      "loss": 3.3274,
+      "step": 700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6444167200313866,
+      "learning_rate": 0.0007934115099832355,
+      "loss": 3.2321,
+      "step": 701
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5865564476523442,
+      "learning_rate": 0.0007928095316701483,
+      "loss": 3.204,
+      "step": 702
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4085903740741983,
+      "learning_rate": 0.0007922069066986819,
+      "loss": 3.2365,
+      "step": 703
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7161612897559329,
+      "learning_rate": 0.0007916036363997097,
+      "loss": 3.1803,
+      "step": 704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4576485600964313,
+      "learning_rate": 0.0007909997221055308,
+      "loss": 3.2443,
+      "step": 705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.06651719331994,
+      "learning_rate": 0.0007903951651498658,
+      "loss": 3.1647,
+      "step": 706
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7096757727245953,
+      "learning_rate": 0.0007897899668678557,
+      "loss": 3.195,
+      "step": 707
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7683778684369704,
+      "learning_rate": 0.0007891841285960566,
+      "loss": 3.2613,
+      "step": 708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8010771472402651,
+      "learning_rate": 0.0007885776516724388,
+      "loss": 3.1374,
+      "step": 709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.951405188215397,
+      "learning_rate": 0.0007879705374363831,
+      "loss": 3.1955,
+      "step": 710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6745342350624652,
+      "learning_rate": 0.000787362787228677,
+      "loss": 3.2199,
+      "step": 711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9214864180582365,
+      "learning_rate": 0.0007867544023915134,
+      "loss": 3.2867,
+      "step": 712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9106943510732517,
+      "learning_rate": 0.0007861453842684861,
+      "loss": 3.154,
+      "step": 713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0352394642324056,
+      "learning_rate": 0.0007855357342045882,
+      "loss": 3.1987,
+      "step": 714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6026867651507097,
+      "learning_rate": 0.0007849254535462074,
+      "loss": 3.1834,
+      "step": 715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1832723711084974,
+      "learning_rate": 0.0007843145436411252,
+      "loss": 3.2072,
+      "step": 716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8756942125036153,
+      "learning_rate": 0.0007837030058385117,
+      "loss": 3.2,
+      "step": 717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9912745319099179,
+      "learning_rate": 0.0007830908414889246,
+      "loss": 3.3468,
+      "step": 718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3397929586206634,
+      "learning_rate": 0.0007824780519443046,
+      "loss": 3.1953,
+      "step": 719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7393798894606853,
+      "learning_rate": 0.0007818646385579735,
+      "loss": 3.1519,
+      "step": 720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.512863573733873,
+      "learning_rate": 0.0007812506026846307,
+      "loss": 3.2201,
+      "step": 721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6544149459777076,
+      "learning_rate": 0.0007806359456803504,
+      "loss": 3.2748,
+      "step": 722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0318783725445577,
+      "learning_rate": 0.0007800206689025785,
+      "loss": 3.2332,
+      "step": 723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4574612874830737,
+      "learning_rate": 0.0007794047737101297,
+      "loss": 3.2183,
+      "step": 724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9163628482455417,
+      "learning_rate": 0.0007787882614631843,
+      "loss": 3.2221,
+      "step": 725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8179379659308197,
+      "learning_rate": 0.0007781711335232856,
+      "loss": 3.2157,
+      "step": 726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7927994173637265,
+      "learning_rate": 0.0007775533912533363,
+      "loss": 3.2988,
+      "step": 727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5163647733683836,
+      "learning_rate": 0.0007769350360175962,
+      "loss": 3.3804,
+      "step": 728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4697130889826326,
+      "learning_rate": 0.0007763160691816784,
+      "loss": 3.2476,
+      "step": 729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.878575399499073,
+      "learning_rate": 0.000775696492112547,
+      "loss": 3.1393,
+      "step": 730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6266011995414054,
+      "learning_rate": 0.0007750763061785137,
+      "loss": 3.1769,
+      "step": 731
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6516890588901725,
+      "learning_rate": 0.000774455512749235,
+      "loss": 3.204,
+      "step": 732
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.586379755541604,
+      "learning_rate": 0.0007738341131957085,
+      "loss": 3.1408,
+      "step": 733
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9226582939479724,
+      "learning_rate": 0.000773212108890271,
+      "loss": 3.1321,
+      "step": 734
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6962646168686792,
+      "learning_rate": 0.0007725895012065947,
+      "loss": 3.0934,
+      "step": 735
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5755694696212907,
+      "learning_rate": 0.0007719662915196844,
+      "loss": 3.1296,
+      "step": 736
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8406356088854496,
+      "learning_rate": 0.0007713424812058736,
+      "loss": 3.2235,
+      "step": 737
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4592940285045928,
+      "learning_rate": 0.0007707180716428237,
+      "loss": 3.3698,
+      "step": 738
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6882198573196878,
+      "learning_rate": 0.0007700930642095184,
+      "loss": 3.235,
+      "step": 739
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6256710878360054,
+      "learning_rate": 0.0007694674602862621,
+      "loss": 3.1936,
+      "step": 740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4300788778406077,
+      "learning_rate": 0.0007688412612546769,
+      "loss": 3.1979,
+      "step": 741
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9974377083572334,
+      "learning_rate": 0.0007682144684976983,
+      "loss": 3.0689,
+      "step": 742
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7054462066273737,
+      "learning_rate": 0.0007675870833995739,
+      "loss": 3.1627,
+      "step": 743
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6260377336471654,
+      "learning_rate": 0.0007669591073458592,
+      "loss": 3.3312,
+      "step": 744
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7265083112119854,
+      "learning_rate": 0.0007663305417234146,
+      "loss": 3.1772,
+      "step": 745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9619677503696025,
+      "learning_rate": 0.0007657013879204022,
+      "loss": 3.1702,
+      "step": 746
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.411620179950431,
+      "learning_rate": 0.0007650716473262842,
+      "loss": 3.0489,
+      "step": 747
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5092407213949726,
+      "learning_rate": 0.0007644413213318177,
+      "loss": 3.1417,
+      "step": 748
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6654110672605138,
+      "learning_rate": 0.0007638104113290531,
+      "loss": 3.1357,
+      "step": 749
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6005785138564161,
+      "learning_rate": 0.0007631789187113303,
+      "loss": 3.1938,
+      "step": 750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.556438698024105,
+      "learning_rate": 0.000762546844873276,
+      "loss": 3.1223,
+      "step": 751
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6659487571491298,
+      "learning_rate": 0.0007619141912108007,
+      "loss": 3.3353,
+      "step": 752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.102816381124882,
+      "learning_rate": 0.000761280959121095,
+      "loss": 3.0863,
+      "step": 753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6263920595211936,
+      "learning_rate": 0.0007606471500026273,
+      "loss": 3.2202,
+      "step": 754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8187866726733068,
+      "learning_rate": 0.0007600127652551401,
+      "loss": 3.0783,
+      "step": 755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5721795629578137,
+      "learning_rate": 0.0007593778062796472,
+      "loss": 3.1889,
+      "step": 756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7149815564292814,
+      "learning_rate": 0.000758742274478431,
+      "loss": 3.3861,
+      "step": 757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6874558502229173,
+      "learning_rate": 0.0007581061712550381,
+      "loss": 3.2173,
+      "step": 758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1261167488888506,
+      "learning_rate": 0.0007574694980142779,
+      "loss": 3.1617,
+      "step": 759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4305797515028313,
+      "learning_rate": 0.0007568322561622183,
+      "loss": 3.138,
+      "step": 760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3129366821629165,
+      "learning_rate": 0.0007561944471061826,
+      "loss": 3.2028,
+      "step": 761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.712508040049101,
+      "learning_rate": 0.0007555560722547475,
+      "loss": 3.0381,
+      "step": 762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3179660423582655,
+      "learning_rate": 0.0007549171330177387,
+      "loss": 3.3383,
+      "step": 763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0021985370394653,
+      "learning_rate": 0.0007542776308062285,
+      "loss": 3.1707,
+      "step": 764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3161006077704913,
+      "learning_rate": 0.0007536375670325325,
+      "loss": 3.1827,
+      "step": 765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0148454233049806,
+      "learning_rate": 0.0007529969431102063,
+      "loss": 3.2658,
+      "step": 766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.075062658278405,
+      "learning_rate": 0.000752355760454043,
+      "loss": 3.2421,
+      "step": 767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.323218710096702,
+      "learning_rate": 0.0007517140204800693,
+      "loss": 3.2062,
+      "step": 768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6807239807123646,
+      "learning_rate": 0.0007510717246055425,
+      "loss": 3.2061,
+      "step": 769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0255943237685536,
+      "learning_rate": 0.0007504288742489482,
+      "loss": 3.2271,
+      "step": 770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2665670251579866,
+      "learning_rate": 0.0007497854708299963,
+      "loss": 3.1349,
+      "step": 771
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1243190329433546,
+      "learning_rate": 0.0007491415157696178,
+      "loss": 3.1308,
+      "step": 772
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.332683133182367,
+      "learning_rate": 0.0007484970104899623,
+      "loss": 3.151,
+      "step": 773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.651005462122057,
+      "learning_rate": 0.0007478519564143945,
+      "loss": 3.2394,
+      "step": 774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5373902039508047,
+      "learning_rate": 0.000747206354967491,
+      "loss": 3.2163,
+      "step": 775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7381798652498675,
+      "learning_rate": 0.0007465602075750373,
+      "loss": 3.1161,
+      "step": 776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5631235907759242,
+      "learning_rate": 0.0007459135156640247,
+      "loss": 3.246,
+      "step": 777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.521095952362014,
+      "learning_rate": 0.0007452662806626468,
+      "loss": 3.1417,
+      "step": 778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7606683880589544,
+      "learning_rate": 0.0007446185040002967,
+      "loss": 3.0674,
+      "step": 779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7407251679192068,
+      "learning_rate": 0.0007439701871075642,
+      "loss": 3.2889,
+      "step": 780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7505544192978821,
+      "learning_rate": 0.0007433213314162313,
+      "loss": 3.1224,
+      "step": 781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.47913580077128,
+      "learning_rate": 0.0007426719383592705,
+      "loss": 3.1218,
+      "step": 782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6410772041931971,
+      "learning_rate": 0.000742022009370841,
+      "loss": 3.0781,
+      "step": 783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5089566442658133,
+      "learning_rate": 0.0007413715458862855,
+      "loss": 3.1643,
+      "step": 784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4076095791246044,
+      "learning_rate": 0.0007407205493421272,
+      "loss": 3.1746,
+      "step": 785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3926091624563461,
+      "learning_rate": 0.0007400690211760661,
+      "loss": 3.0795,
+      "step": 786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.868482364668368,
+      "learning_rate": 0.0007394169628269771,
+      "loss": 3.0074,
+      "step": 787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6311128031982587,
+      "learning_rate": 0.0007387643757349051,
+      "loss": 3.0572,
+      "step": 788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.913701513748886,
+      "learning_rate": 0.0007381112613410635,
+      "loss": 3.1157,
+      "step": 789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8722807020423806,
+      "learning_rate": 0.0007374576210878298,
+      "loss": 3.2709,
+      "step": 790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.655327290623419,
+      "learning_rate": 0.0007368034564187425,
+      "loss": 3.1216,
+      "step": 791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4507540500984517,
+      "learning_rate": 0.0007361487687784989,
+      "loss": 3.1682,
+      "step": 792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.899845121320468,
+      "learning_rate": 0.0007354935596129513,
+      "loss": 3.1984,
+      "step": 793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6489314881090773,
+      "learning_rate": 0.000734837830369103,
+      "loss": 3.142,
+      "step": 794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8403989004377288,
+      "learning_rate": 0.0007341815824951066,
+      "loss": 3.0636,
+      "step": 795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9162957346685565,
+      "learning_rate": 0.0007335248174402597,
+      "loss": 3.1162,
+      "step": 796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.569858367691908,
+      "learning_rate": 0.0007328675366550023,
+      "loss": 3.0428,
+      "step": 797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4931999338315192,
+      "learning_rate": 0.0007322097415909134,
+      "loss": 3.1993,
+      "step": 798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9860394006884592,
+      "learning_rate": 0.0007315514337007071,
+      "loss": 3.0724,
+      "step": 799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7278402279167804,
+      "learning_rate": 0.0007308926144382312,
+      "loss": 3.1784,
+      "step": 800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7214341597506415,
+      "learning_rate": 0.0007302332852584619,
+      "loss": 3.2013,
+      "step": 801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.450293527541703,
+      "learning_rate": 0.0007295734476175018,
+      "loss": 3.1306,
+      "step": 802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4929611021318305,
+      "learning_rate": 0.0007289131029725768,
+      "loss": 3.157,
+      "step": 803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.354889165200498,
+      "learning_rate": 0.0007282522527820319,
+      "loss": 3.0947,
+      "step": 804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.683017971380692,
+      "learning_rate": 0.000727590898505329,
+      "loss": 3.0944,
+      "step": 805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8678719814946176,
+      "learning_rate": 0.0007269290416030429,
+      "loss": 3.1058,
+      "step": 806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6141259375651669,
+      "learning_rate": 0.000726266683536859,
+      "loss": 3.1105,
+      "step": 807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.113676715476671,
+      "learning_rate": 0.0007256038257695687,
+      "loss": 3.1189,
+      "step": 808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5359819737571316,
+      "learning_rate": 0.0007249404697650678,
+      "loss": 3.1186,
+      "step": 809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9645979193038337,
+      "learning_rate": 0.0007242766169883518,
+      "loss": 3.2304,
+      "step": 810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9511568090318838,
+      "learning_rate": 0.0007236122689055138,
+      "loss": 3.2126,
+      "step": 811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6893886023916138,
+      "learning_rate": 0.0007229474269837401,
+      "loss": 3.2523,
+      "step": 812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8273029211219423,
+      "learning_rate": 0.0007222820926913085,
+      "loss": 3.0867,
+      "step": 813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8811231121528642,
+      "learning_rate": 0.0007216162674975833,
+      "loss": 3.2349,
+      "step": 814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3205803892513632,
+      "learning_rate": 0.0007209499528730138,
+      "loss": 3.1508,
+      "step": 815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9234703308090473,
+      "learning_rate": 0.0007202831502891294,
+      "loss": 3.0858,
+      "step": 816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.727929060292976,
+      "learning_rate": 0.0007196158612185375,
+      "loss": 3.2121,
+      "step": 817
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.725379772522385,
+      "learning_rate": 0.0007189480871349201,
+      "loss": 3.12,
+      "step": 818
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7202311465482414,
+      "learning_rate": 0.0007182798295130299,
+      "loss": 3.092,
+      "step": 819
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9479313497985353,
+      "learning_rate": 0.0007176110898286878,
+      "loss": 3.1203,
+      "step": 820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.691145081758446,
+      "learning_rate": 0.0007169418695587791,
+      "loss": 3.0941,
+      "step": 821
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7293499367792555,
+      "learning_rate": 0.0007162721701812506,
+      "loss": 3.1087,
+      "step": 822
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.494160183158333,
+      "learning_rate": 0.0007156019931751072,
+      "loss": 2.99,
+      "step": 823
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5784100303938573,
+      "learning_rate": 0.0007149313400204082,
+      "loss": 3.1292,
+      "step": 824
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.615411957145368,
+      "learning_rate": 0.0007142602121982653,
+      "loss": 3.1937,
+      "step": 825
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.695059468275302,
+      "learning_rate": 0.0007135886111908379,
+      "loss": 3.1782,
+      "step": 826
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7501834160188223,
+      "learning_rate": 0.0007129165384813303,
+      "loss": 3.1567,
+      "step": 827
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9922249593968737,
+      "learning_rate": 0.0007122439955539888,
+      "loss": 3.2571,
+      "step": 828
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6839999480207588,
+      "learning_rate": 0.0007115709838940983,
+      "loss": 3.2879,
+      "step": 829
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.314396109142333,
+      "learning_rate": 0.0007108975049879785,
+      "loss": 3.1627,
+      "step": 830
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4530195195606805,
+      "learning_rate": 0.0007102235603229814,
+      "loss": 3.1881,
+      "step": 831
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6990194610140126,
+      "learning_rate": 0.000709549151387487,
+      "loss": 3.2108,
+      "step": 832
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7786590279134915,
+      "learning_rate": 0.0007088742796709013,
+      "loss": 3.1522,
+      "step": 833
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2299109153663377,
+      "learning_rate": 0.000708198946663652,
+      "loss": 3.1791,
+      "step": 834
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.961723249979613,
+      "learning_rate": 0.0007075231538571856,
+      "loss": 3.1425,
+      "step": 835
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3144761522834507,
+      "learning_rate": 0.0007068469027439641,
+      "loss": 3.2021,
+      "step": 836
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5319554644192204,
+      "learning_rate": 0.0007061701948174613,
+      "loss": 3.1403,
+      "step": 837
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0200450871503266,
+      "learning_rate": 0.0007054930315721606,
+      "loss": 3.0059,
+      "step": 838
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5717293908755845,
+      "learning_rate": 0.0007048154145035501,
+      "loss": 3.0885,
+      "step": 839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1054133592596465,
+      "learning_rate": 0.0007041373451081207,
+      "loss": 3.1079,
+      "step": 840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5818256296936288,
+      "learning_rate": 0.0007034588248833621,
+      "loss": 3.1456,
+      "step": 841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.140130063920986,
+      "learning_rate": 0.0007027798553277595,
+      "loss": 3.2052,
+      "step": 842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5938038309913194,
+      "learning_rate": 0.0007021004379407909,
+      "loss": 3.197,
+      "step": 843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.54717164790457,
+      "learning_rate": 0.0007014205742229227,
+      "loss": 3.0764,
+      "step": 844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8064533316138534,
+      "learning_rate": 0.0007007402656756072,
+      "loss": 3.0927,
+      "step": 845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6922913237044352,
+      "learning_rate": 0.0007000595138012797,
+      "loss": 3.2371,
+      "step": 846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6995912563505757,
+      "learning_rate": 0.0006993783201033535,
+      "loss": 3.1469,
+      "step": 847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.115797610710769,
+      "learning_rate": 0.0006986966860862182,
+      "loss": 3.1794,
+      "step": 848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2175464886064544,
+      "learning_rate": 0.000698014613255236,
+      "loss": 3.0583,
+      "step": 849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5009707658820797,
+      "learning_rate": 0.0006973321031167382,
+      "loss": 3.0539,
+      "step": 850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.611655013160259,
+      "learning_rate": 0.0006966491571780216,
+      "loss": 3.2693,
+      "step": 851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.671576066184977,
+      "learning_rate": 0.0006959657769473453,
+      "loss": 3.0357,
+      "step": 852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8996039366411634,
+      "learning_rate": 0.000695281963933928,
+      "loss": 3.1822,
+      "step": 853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1909698335652337,
+      "learning_rate": 0.0006945977196479438,
+      "loss": 3.1551,
+      "step": 854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6469937577448077,
+      "learning_rate": 0.0006939130456005196,
+      "loss": 3.2247,
+      "step": 855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.222440492509311,
+      "learning_rate": 0.0006932279433037311,
+      "loss": 3.1467,
+      "step": 856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7020179777112021,
+      "learning_rate": 0.0006925424142705997,
+      "loss": 3.1853,
+      "step": 857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7344365527808279,
+      "learning_rate": 0.0006918564600150896,
+      "loss": 3.1706,
+      "step": 858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2915749503824463,
+      "learning_rate": 0.0006911700820521042,
+      "loss": 3.1909,
+      "step": 859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5171973417767473,
+      "learning_rate": 0.0006904832818974818,
+      "loss": 3.1013,
+      "step": 860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.335634544970706,
+      "learning_rate": 0.0006897960610679939,
+      "loss": 3.2227,
+      "step": 861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.671179365255327,
+      "learning_rate": 0.0006891084210813407,
+      "loss": 3.1887,
+      "step": 862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7122844801890789,
+      "learning_rate": 0.0006884203634561483,
+      "loss": 3.13,
+      "step": 863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.125326897029215,
+      "learning_rate": 0.0006877318897119651,
+      "loss": 3.2115,
+      "step": 864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6326034857487297,
+      "learning_rate": 0.0006870430013692579,
+      "loss": 3.1472,
+      "step": 865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2523480564280853,
+      "learning_rate": 0.0006863536999494101,
+      "loss": 3.1074,
+      "step": 866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5676535947764079,
+      "learning_rate": 0.0006856639869747167,
+      "loss": 3.1672,
+      "step": 867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4392167024591687,
+      "learning_rate": 0.0006849738639683818,
+      "loss": 3.146,
+      "step": 868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5301561136740458,
+      "learning_rate": 0.000684283332454515,
+      "loss": 3.1624,
+      "step": 869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.241397181613629,
+      "learning_rate": 0.0006835923939581281,
+      "loss": 3.0948,
+      "step": 870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5374893695614553,
+      "learning_rate": 0.0006829010500051318,
+      "loss": 3.177,
+      "step": 871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0045552507315185,
+      "learning_rate": 0.0006822093021223321,
+      "loss": 3.1742,
+      "step": 872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.016328053106204,
+      "learning_rate": 0.0006815171518374268,
+      "loss": 3.1629,
+      "step": 873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6118614535838869,
+      "learning_rate": 0.0006808246006790031,
+      "loss": 3.0884,
+      "step": 874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9352563515412222,
+      "learning_rate": 0.0006801316501765329,
+      "loss": 3.1638,
+      "step": 875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.507732992864948,
+      "learning_rate": 0.0006794383018603704,
+      "loss": 3.2171,
+      "step": 876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7053826621791635,
+      "learning_rate": 0.0006787445572617481,
+      "loss": 3.1901,
+      "step": 877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6974441072640574,
+      "learning_rate": 0.0006780504179127734,
+      "loss": 3.1393,
+      "step": 878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6305201411754326,
+      "learning_rate": 0.0006773558853464265,
+      "loss": 3.0911,
+      "step": 879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4933171271709322,
+      "learning_rate": 0.000676660961096555,
+      "loss": 3.0986,
+      "step": 880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.945266796007687,
+      "learning_rate": 0.000675965646697872,
+      "loss": 3.2514,
+      "step": 881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.119107579581053,
+      "learning_rate": 0.0006752699436859519,
+      "loss": 3.1317,
+      "step": 882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4381609049088484,
+      "learning_rate": 0.0006745738535972279,
+      "loss": 3.0758,
+      "step": 883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0947640238615595,
+      "learning_rate": 0.0006738773779689874,
+      "loss": 3.161,
+      "step": 884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.550179984749495,
+      "learning_rate": 0.0006731805183393696,
+      "loss": 3.1112,
+      "step": 885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.709025785785192,
+      "learning_rate": 0.0006724832762473618,
+      "loss": 3.0612,
+      "step": 886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9234652471189986,
+      "learning_rate": 0.0006717856532327956,
+      "loss": 3.1251,
+      "step": 887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.721677627894101,
+      "learning_rate": 0.0006710876508363444,
+      "loss": 3.1774,
+      "step": 888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.240685500173532,
+      "learning_rate": 0.0006703892705995189,
+      "loss": 3.2334,
+      "step": 889
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.920671251562104,
+      "learning_rate": 0.0006696905140646647,
+      "loss": 3.2522,
+      "step": 890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.68428397171846,
+      "learning_rate": 0.0006689913827749581,
+      "loss": 3.1461,
+      "step": 891
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4918247941583043,
+      "learning_rate": 0.0006682918782744032,
+      "loss": 3.1372,
+      "step": 892
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6333684087004166,
+      "learning_rate": 0.0006675920021078282,
+      "loss": 3.033,
+      "step": 893
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9805189873163929,
+      "learning_rate": 0.0006668917558208823,
+      "loss": 3.2289,
+      "step": 894
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6319258281375724,
+      "learning_rate": 0.0006661911409600321,
+      "loss": 3.1462,
+      "step": 895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6183120312350097,
+      "learning_rate": 0.0006654901590725577,
+      "loss": 3.1364,
+      "step": 896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9074885807555102,
+      "learning_rate": 0.0006647888117065507,
+      "loss": 3.2026,
+      "step": 897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.659892577934357,
+      "learning_rate": 0.0006640871004109086,
+      "loss": 3.1069,
+      "step": 898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1786669979190987,
+      "learning_rate": 0.000663385026735334,
+      "loss": 2.9962,
+      "step": 899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5170011534376997,
+      "learning_rate": 0.0006626825922303285,
+      "loss": 3.2195,
+      "step": 900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5191128038511814,
+      "learning_rate": 0.0006619797984471915,
+      "loss": 3.1443,
+      "step": 901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6542627279948923,
+      "learning_rate": 0.0006612766469380158,
+      "loss": 3.1775,
+      "step": 902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0371797612885456,
+      "learning_rate": 0.0006605731392556833,
+      "loss": 3.2764,
+      "step": 903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7613978531119094,
+      "learning_rate": 0.0006598692769538637,
+      "loss": 3.1164,
+      "step": 904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5559821311280055,
+      "learning_rate": 0.0006591650615870091,
+      "loss": 3.1462,
+      "step": 905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8181226273009328,
+      "learning_rate": 0.0006584604947103514,
+      "loss": 3.14,
+      "step": 906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7438167381202476,
+      "learning_rate": 0.0006577555778798993,
+      "loss": 3.114,
+      "step": 907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4717798800125022,
+      "learning_rate": 0.0006570503126524336,
+      "loss": 3.096,
+      "step": 908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4213514679430177,
+      "learning_rate": 0.0006563447005855054,
+      "loss": 3.1452,
+      "step": 909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5427617777425804,
+      "learning_rate": 0.000655638743237431,
+      "loss": 3.1416,
+      "step": 910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8891332428432026,
+      "learning_rate": 0.0006549324421672894,
+      "loss": 3.1786,
+      "step": 911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4991314983700288,
+      "learning_rate": 0.0006542257989349194,
+      "loss": 3.0642,
+      "step": 912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.738894542830529,
+      "learning_rate": 0.0006535188151009142,
+      "loss": 3.1494,
+      "step": 913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5242211867828066,
+      "learning_rate": 0.0006528114922266204,
+      "loss": 3.1716,
+      "step": 914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5369598444884751,
+      "learning_rate": 0.0006521038318741327,
+      "loss": 3.1655,
+      "step": 915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7318457702461243,
+      "learning_rate": 0.0006513958356062912,
+      "loss": 3.1392,
+      "step": 916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5963680861544027,
+      "learning_rate": 0.0006506875049866781,
+      "loss": 3.0891,
+      "step": 917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.638510479560546,
+      "learning_rate": 0.0006499788415796137,
+      "loss": 3.1322,
+      "step": 918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.96922010526347,
+      "learning_rate": 0.0006492698469501532,
+      "loss": 3.0388,
+      "step": 919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5683219005595723,
+      "learning_rate": 0.0006485605226640837,
+      "loss": 3.1309,
+      "step": 920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9824509108382127,
+      "learning_rate": 0.00064785087028792,
+      "loss": 3.1795,
+      "step": 921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8473396692225412,
+      "learning_rate": 0.0006471408913889019,
+      "loss": 3.0265,
+      "step": 922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6387952697145323,
+      "learning_rate": 0.0006464305875349892,
+      "loss": 3.0574,
+      "step": 923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9062867721962635,
+      "learning_rate": 0.000645719960294861,
+      "loss": 3.0083,
+      "step": 924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5735644513195686,
+      "learning_rate": 0.0006450090112379092,
+      "loss": 3.083,
+      "step": 925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7523861997174695,
+      "learning_rate": 0.0006442977419342371,
+      "loss": 3.0017,
+      "step": 926
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5351321040768353,
+      "learning_rate": 0.000643586153954655,
+      "loss": 3.1136,
+      "step": 927
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.719082332479609,
+      "learning_rate": 0.0006428742488706772,
+      "loss": 3.0544,
+      "step": 928
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7579345600403649,
+      "learning_rate": 0.0006421620282545182,
+      "loss": 3.0673,
+      "step": 929
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.720079553140734,
+      "learning_rate": 0.0006414494936790892,
+      "loss": 3.1394,
+      "step": 930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1118202830892305,
+      "learning_rate": 0.0006407366467179951,
+      "loss": 3.0982,
+      "step": 931
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1444454535591038,
+      "learning_rate": 0.0006400234889455301,
+      "loss": 3.2082,
+      "step": 932
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9783242716828728,
+      "learning_rate": 0.0006393100219366755,
+      "loss": 3.1895,
+      "step": 933
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.494946305470995,
+      "learning_rate": 0.0006385962472670953,
+      "loss": 3.1251,
+      "step": 934
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5785660797617074,
+      "learning_rate": 0.0006378821665131328,
+      "loss": 3.0398,
+      "step": 935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.884995993758215,
+      "learning_rate": 0.0006371677812518072,
+      "loss": 3.1574,
+      "step": 936
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7705203325294778,
+      "learning_rate": 0.0006364530930608107,
+      "loss": 3.041,
+      "step": 937
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9662112156798564,
+      "learning_rate": 0.0006357381035185038,
+      "loss": 3.0917,
+      "step": 938
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.065390893460396,
+      "learning_rate": 0.0006350228142039131,
+      "loss": 3.1093,
+      "step": 939
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6306583763461684,
+      "learning_rate": 0.000634307226696727,
+      "loss": 3.0046,
+      "step": 940
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8130980953010158,
+      "learning_rate": 0.0006335913425772926,
+      "loss": 3.1791,
+      "step": 941
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6410035677974113,
+      "learning_rate": 0.0006328751634266117,
+      "loss": 3.0174,
+      "step": 942
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6439640052695121,
+      "learning_rate": 0.0006321586908263382,
+      "loss": 3.0572,
+      "step": 943
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4285895717652652,
+      "learning_rate": 0.0006314419263587732,
+      "loss": 3.1676,
+      "step": 944
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4169738286027103,
+      "learning_rate": 0.0006307248716068637,
+      "loss": 3.0629,
+      "step": 945
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4865580608245732,
+      "learning_rate": 0.0006300075281541964,
+      "loss": 3.0446,
+      "step": 946
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7651335382219715,
+      "learning_rate": 0.0006292898975849966,
+      "loss": 3.2305,
+      "step": 947
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.455113947615389,
+      "learning_rate": 0.000628571981484123,
+      "loss": 3.0558,
+      "step": 948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4843742795025654,
+      "learning_rate": 0.0006278537814370654,
+      "loss": 3.1837,
+      "step": 949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.635140708611304,
+      "learning_rate": 0.0006271352990299406,
+      "loss": 3.1889,
+      "step": 950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.51307415674298,
+      "learning_rate": 0.0006264165358494885,
+      "loss": 3.1108,
+      "step": 951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9528921117650087,
+      "learning_rate": 0.0006256974934830694,
+      "loss": 3.2093,
+      "step": 952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6480265072989657,
+      "learning_rate": 0.0006249781735186606,
+      "loss": 3.17,
+      "step": 953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.592791193496641,
+      "learning_rate": 0.0006242585775448518,
+      "loss": 3.178,
+      "step": 954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5290344128848277,
+      "learning_rate": 0.0006235387071508427,
+      "loss": 3.0545,
+      "step": 955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7210559965244077,
+      "learning_rate": 0.0006228185639264384,
+      "loss": 3.1445,
+      "step": 956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7383195583268876,
+      "learning_rate": 0.0006220981494620475,
+      "loss": 3.1198,
+      "step": 957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.594894219734358,
+      "learning_rate": 0.000621377465348677,
+      "loss": 3.1547,
+      "step": 958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2721387123948587,
+      "learning_rate": 0.0006206565131779293,
+      "loss": 3.0625,
+      "step": 959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4824928556581032,
+      "learning_rate": 0.0006199352945419994,
+      "loss": 3.1238,
+      "step": 960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0462460317578377,
+      "learning_rate": 0.00061921381103367,
+      "loss": 3.1682,
+      "step": 961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6215343589165414,
+      "learning_rate": 0.0006184920642463094,
+      "loss": 3.2052,
+      "step": 962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.981994930642667,
+      "learning_rate": 0.0006177700557738672,
+      "loss": 3.1501,
+      "step": 963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5683578531926254,
+      "learning_rate": 0.0006170477872108706,
+      "loss": 2.9721,
+      "step": 964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.663885628633806,
+      "learning_rate": 0.0006163252601524216,
+      "loss": 3.1027,
+      "step": 965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.959929720419253,
+      "learning_rate": 0.0006156024761941925,
+      "loss": 3.122,
+      "step": 966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3223120127297205,
+      "learning_rate": 0.000614879436932424,
+      "loss": 3.1132,
+      "step": 967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4051183896731336,
+      "learning_rate": 0.0006141561439639196,
+      "loss": 3.0307,
+      "step": 968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.636899807326773,
+      "learning_rate": 0.0006134325988860433,
+      "loss": 3.0695,
+      "step": 969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5255815059056335,
+      "learning_rate": 0.0006127088032967165,
+      "loss": 3.0094,
+      "step": 970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.631453548330054,
+      "learning_rate": 0.0006119847587944131,
+      "loss": 3.0507,
+      "step": 971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.632622304723368,
+      "learning_rate": 0.0006112604669781572,
+      "loss": 3.0651,
+      "step": 972
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3540179822684577,
+      "learning_rate": 0.0006105359294475188,
+      "loss": 3.191,
+      "step": 973
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7232828676923142,
+      "learning_rate": 0.0006098111478026107,
+      "loss": 3.1302,
+      "step": 974
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6504747062127119,
+      "learning_rate": 0.0006090861236440848,
+      "loss": 3.1304,
+      "step": 975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6560376278858822,
+      "learning_rate": 0.0006083608585731282,
+      "loss": 3.0731,
+      "step": 976
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.125335924655279,
+      "learning_rate": 0.0006076353541914609,
+      "loss": 3.0767,
+      "step": 977
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6960124972972803,
+      "learning_rate": 0.0006069096121013307,
+      "loss": 3.1393,
+      "step": 978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7448803200148488,
+      "learning_rate": 0.0006061836339055105,
+      "loss": 3.0839,
+      "step": 979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.823775001237499,
+      "learning_rate": 0.0006054574212072948,
+      "loss": 3.1362,
+      "step": 980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.678772930904242,
+      "learning_rate": 0.0006047309756104958,
+      "loss": 3.0376,
+      "step": 981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.559641643967172,
+      "learning_rate": 0.00060400429871944,
+      "loss": 3.1428,
+      "step": 982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6052175217845739,
+      "learning_rate": 0.0006032773921389654,
+      "loss": 3.0462,
+      "step": 983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7891246881032654,
+      "learning_rate": 0.0006025502574744162,
+      "loss": 3.0993,
+      "step": 984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5283861242088674,
+      "learning_rate": 0.000601822896331641,
+      "loss": 3.0301,
+      "step": 985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7820176671834704,
+      "learning_rate": 0.0006010953103169883,
+      "loss": 2.9828,
+      "step": 986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6249676554868204,
+      "learning_rate": 0.0006003675010373034,
+      "loss": 3.0792,
+      "step": 987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8115016275995162,
+      "learning_rate": 0.0005996394700999246,
+      "loss": 3.1556,
+      "step": 988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5465818939745644,
+      "learning_rate": 0.0005989112191126794,
+      "loss": 3.0141,
+      "step": 989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6464982691357528,
+      "learning_rate": 0.0005981827496838822,
+      "loss": 3.0247,
+      "step": 990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5307343046917956,
+      "learning_rate": 0.0005974540634223286,
+      "loss": 3.0853,
+      "step": 991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4576044959693344,
+      "learning_rate": 0.0005967251619372939,
+      "loss": 3.0957,
+      "step": 992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6853949495434328,
+      "learning_rate": 0.0005959960468385284,
+      "loss": 3.1857,
+      "step": 993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.373781380190026,
+      "learning_rate": 0.0005952667197362542,
+      "loss": 3.072,
+      "step": 994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6188017661833256,
+      "learning_rate": 0.0005945371822411621,
+      "loss": 3.0767,
+      "step": 995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3618284968444687,
+      "learning_rate": 0.0005938074359644063,
+      "loss": 2.9975,
+      "step": 996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5920516106529974,
+      "learning_rate": 0.0005930774825176034,
+      "loss": 3.0952,
+      "step": 997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5500000806719112,
+      "learning_rate": 0.0005923473235128268,
+      "loss": 3.0843,
+      "step": 998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7944877508500965,
+      "learning_rate": 0.0005916169605626042,
+      "loss": 3.0787,
+      "step": 999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3173809165478791,
+      "learning_rate": 0.0005908863952799134,
+      "loss": 3.0289,
+      "step": 1000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6180414779254642,
+      "learning_rate": 0.0005901556292781793,
+      "loss": 3.0748,
+      "step": 1001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5097031879160119,
+      "learning_rate": 0.0005894246641712698,
+      "loss": 3.0799,
+      "step": 1002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4430414773533038,
+      "learning_rate": 0.0005886935015734931,
+      "loss": 3.0999,
+      "step": 1003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4174478971331803,
+      "learning_rate": 0.0005879621430995928,
+      "loss": 2.9652,
+      "step": 1004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4438100822292805,
+      "learning_rate": 0.0005872305903647455,
+      "loss": 3.1088,
+      "step": 1005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7715615914302942,
+      "learning_rate": 0.0005864988449845569,
+      "loss": 3.0441,
+      "step": 1006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5916123171184997,
+      "learning_rate": 0.0005857669085750578,
+      "loss": 3.0716,
+      "step": 1007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7611158850293018,
+      "learning_rate": 0.0005850347827527013,
+      "loss": 3.1794,
+      "step": 1008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.405938268929194,
+      "learning_rate": 0.0005843024691343584,
+      "loss": 2.9934,
+      "step": 1009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7486616137078383,
+      "learning_rate": 0.000583569969337315,
+      "loss": 3.0912,
+      "step": 1010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5709743824771876,
+      "learning_rate": 0.0005828372849792686,
+      "loss": 3.0674,
+      "step": 1011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6795218259166356,
+      "learning_rate": 0.0005821044176783234,
+      "loss": 3.0798,
+      "step": 1012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9889296995871746,
+      "learning_rate": 0.0005813713690529886,
+      "loss": 3.0986,
+      "step": 1013
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3953708361209776,
+      "learning_rate": 0.0005806381407221729,
+      "loss": 3.0715,
+      "step": 1014
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5104829748772381,
+      "learning_rate": 0.0005799047343051826,
+      "loss": 2.9986,
+      "step": 1015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8339895736964142,
+      "learning_rate": 0.0005791711514217171,
+      "loss": 3.1442,
+      "step": 1016
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4666978569221596,
+      "learning_rate": 0.0005784373936918654,
+      "loss": 3.1672,
+      "step": 1017
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.664495164642137,
+      "learning_rate": 0.0005777034627361025,
+      "loss": 3.1268,
+      "step": 1018
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6002886053224428,
+      "learning_rate": 0.0005769693601752864,
+      "loss": 3.0416,
+      "step": 1019
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.821279284455168,
+      "learning_rate": 0.0005762350876306537,
+      "loss": 3.1373,
+      "step": 1020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.474291038435249,
+      "learning_rate": 0.0005755006467238168,
+      "loss": 3.1159,
+      "step": 1021
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5445152792323649,
+      "learning_rate": 0.0005747660390767593,
+      "loss": 3.0323,
+      "step": 1022
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.729812929414656,
+      "learning_rate": 0.0005740312663118338,
+      "loss": 2.9677,
+      "step": 1023
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6811567249783201,
+      "learning_rate": 0.0005732963300517568,
+      "loss": 3.0533,
+      "step": 1024
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0238696858058285,
+      "learning_rate": 0.0005725612319196064,
+      "loss": 3.1486,
+      "step": 1025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.521413342002447,
+      "learning_rate": 0.000571825973538818,
+      "loss": 3.0169,
+      "step": 1026
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4330102908350244,
+      "learning_rate": 0.0005710905565331811,
+      "loss": 3.0009,
+      "step": 1027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.530448086678303,
+      "learning_rate": 0.0005703549825268353,
+      "loss": 3.1431,
+      "step": 1028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5980601810937363,
+      "learning_rate": 0.0005696192531442667,
+      "loss": 3.0197,
+      "step": 1029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.50379172505528,
+      "learning_rate": 0.000568883370010305,
+      "loss": 3.1797,
+      "step": 1030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9114627908293103,
+      "learning_rate": 0.0005681473347501192,
+      "loss": 3.3599,
+      "step": 1031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9884145443802999,
+      "learning_rate": 0.0005674111489892144,
+      "loss": 3.0187,
+      "step": 1032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6010203342765796,
+      "learning_rate": 0.0005666748143534282,
+      "loss": 3.1435,
+      "step": 1033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.049366942440454,
+      "learning_rate": 0.0005659383324689266,
+      "loss": 3.1004,
+      "step": 1034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4928429102724736,
+      "learning_rate": 0.0005652017049622007,
+      "loss": 3.0358,
+      "step": 1035
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7799921475814307,
+      "learning_rate": 0.0005644649334600641,
+      "loss": 3.0611,
+      "step": 1036
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5031946411478314,
+      "learning_rate": 0.0005637280195896474,
+      "loss": 3.12,
+      "step": 1037
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.529574308629453,
+      "learning_rate": 0.0005629909649783961,
+      "loss": 3.115,
+      "step": 1038
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.515826992281273,
+      "learning_rate": 0.0005622537712540664,
+      "loss": 3.2531,
+      "step": 1039
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.510270771059459,
+      "learning_rate": 0.0005615164400447218,
+      "loss": 3.0194,
+      "step": 1040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.676269456356913,
+      "learning_rate": 0.0005607789729787294,
+      "loss": 3.1531,
+      "step": 1041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6816849615405338,
+      "learning_rate": 0.0005600413716847564,
+      "loss": 3.0335,
+      "step": 1042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.797188990035947,
+      "learning_rate": 0.000559303637791766,
+      "loss": 3.0253,
+      "step": 1043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7598841198837727,
+      "learning_rate": 0.0005585657729290151,
+      "loss": 2.9959,
+      "step": 1044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7333653178358168,
+      "learning_rate": 0.000557827778726049,
+      "loss": 3.0496,
+      "step": 1045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.733489768146429,
+      "learning_rate": 0.0005570896568126993,
+      "loss": 3.0292,
+      "step": 1046
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6275088648233207,
+      "learning_rate": 0.0005563514088190788,
+      "loss": 3.042,
+      "step": 1047
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4379147392448015,
+      "learning_rate": 0.0005556130363755798,
+      "loss": 2.9933,
+      "step": 1048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8808732465036653,
+      "learning_rate": 0.0005548745411128688,
+      "loss": 3.0133,
+      "step": 1049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6036772164116793,
+      "learning_rate": 0.0005541359246618835,
+      "loss": 2.956,
+      "step": 1050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1113643687930175,
+      "learning_rate": 0.0005533971886538293,
+      "loss": 3.0923,
+      "step": 1051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.610831377410517,
+      "learning_rate": 0.000552658334720176,
+      "loss": 3.1338,
+      "step": 1052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1562387797596796,
+      "learning_rate": 0.0005519193644926535,
+      "loss": 2.9616,
+      "step": 1053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6397687234314386,
+      "learning_rate": 0.0005511802796032485,
+      "loss": 3.0314,
+      "step": 1054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6581431381678504,
+      "learning_rate": 0.0005504410816842009,
+      "loss": 3.0468,
+      "step": 1055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4484465975015097,
+      "learning_rate": 0.0005497017723680009,
+      "loss": 2.9329,
+      "step": 1056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5771768944761748,
+      "learning_rate": 0.0005489623532873836,
+      "loss": 2.9937,
+      "step": 1057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7644562357353,
+      "learning_rate": 0.0005482228260753273,
+      "loss": 3.0765,
+      "step": 1058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.022210352700627,
+      "learning_rate": 0.0005474831923650488,
+      "loss": 3.0367,
+      "step": 1059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7576307744216477,
+      "learning_rate": 0.00054674345379,
+      "loss": 2.9924,
+      "step": 1060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7882630720184136,
+      "learning_rate": 0.000546003611983865,
+      "loss": 3.0832,
+      "step": 1061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6513630089965947,
+      "learning_rate": 0.0005452636685805552,
+      "loss": 3.0131,
+      "step": 1062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.890105403333004,
+      "learning_rate": 0.0005445236252142066,
+      "loss": 3.2069,
+      "step": 1063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5853479043805085,
+      "learning_rate": 0.000543783483519176,
+      "loss": 3.1166,
+      "step": 1064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9110528729610718,
+      "learning_rate": 0.0005430432451300374,
+      "loss": 3.0564,
+      "step": 1065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.62719239810702,
+      "learning_rate": 0.0005423029116815781,
+      "loss": 2.988,
+      "step": 1066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8775236877784311,
+      "learning_rate": 0.0005415624848087959,
+      "loss": 3.0686,
+      "step": 1067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7305959933411597,
+      "learning_rate": 0.000540821966146894,
+      "loss": 3.2194,
+      "step": 1068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5758028813485496,
+      "learning_rate": 0.0005400813573312793,
+      "loss": 3.0141,
+      "step": 1069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.017917266136407,
+      "learning_rate": 0.0005393406599975572,
+      "loss": 3.0566,
+      "step": 1070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6982718931964749,
+      "learning_rate": 0.0005385998757815287,
+      "loss": 3.0699,
+      "step": 1071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.704767650658607,
+      "learning_rate": 0.0005378590063191867,
+      "loss": 3.0128,
+      "step": 1072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.567632191306601,
+      "learning_rate": 0.0005371180532467124,
+      "loss": 3.0679,
+      "step": 1073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5536589405314187,
+      "learning_rate": 0.000536377018200472,
+      "loss": 3.0486,
+      "step": 1074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9989108926344465,
+      "learning_rate": 0.0005356359028170118,
+      "loss": 3.1187,
+      "step": 1075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9549660125035337,
+      "learning_rate": 0.0005348947087330564,
+      "loss": 3.0829,
+      "step": 1076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5974730691530121,
+      "learning_rate": 0.0005341534375855037,
+      "loss": 3.1013,
+      "step": 1077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7143110237539536,
+      "learning_rate": 0.0005334120910114222,
+      "loss": 3.1159,
+      "step": 1078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6023727649393982,
+      "learning_rate": 0.0005326706706480467,
+      "loss": 3.049,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.543338713225148,
+      "learning_rate": 0.0005319291781327749,
+      "loss": 3.0376,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5682137491435042,
+      "learning_rate": 0.0005311876151031642,
+      "loss": 3.1099,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6347900552912664,
+      "learning_rate": 0.0005304459831969274,
+      "loss": 3.0345,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.844568032830322,
+      "learning_rate": 0.0005297042840519294,
+      "loss": 2.9695,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8684926359284066,
+      "learning_rate": 0.0005289625193061838,
+      "loss": 3.0113,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.541879978138572,
+      "learning_rate": 0.0005282206905978489,
+      "loss": 2.9901,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.483795338242328,
+      "learning_rate": 0.0005274787995652246,
+      "loss": 3.0657,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5485115756630823,
+      "learning_rate": 0.000526736847846748,
+      "loss": 2.9852,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4940518322972816,
+      "learning_rate": 0.0005259948370809901,
+      "loss": 3.124,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7665054252460908,
+      "learning_rate": 0.0005252527689066533,
+      "loss": 3.1042,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.687434024230023,
+      "learning_rate": 0.0005245106449625654,
+      "loss": 3.0004,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7945626846387766,
+      "learning_rate": 0.0005237684668876785,
+      "loss": 3.0434,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7311279901005807,
+      "learning_rate": 0.0005230262363210637,
+      "loss": 3.1468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5564805550527365,
+      "learning_rate": 0.0005222839549019079,
+      "loss": 3.1809,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5679756674435326,
+      "learning_rate": 0.0005215416242695108,
+      "loss": 2.9572,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6416091554322378,
+      "learning_rate": 0.0005207992460632804,
+      "loss": 3.0508,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7264235216587265,
+      "learning_rate": 0.0005200568219227299,
+      "loss": 2.8953,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8789977422277748,
+      "learning_rate": 0.000519314353487474,
+      "loss": 3.0946,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9371897678058017,
+      "learning_rate": 0.0005185718423972251,
+      "loss": 3.0032,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4420060941429707,
+      "learning_rate": 0.0005178292902917898,
+      "loss": 3.0823,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.045578178848679,
+      "learning_rate": 0.0005170866988110656,
+      "loss": 3.0502,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.029378989444008,
+      "learning_rate": 0.0005163440695950362,
+      "loss": 2.9769,
+      "step": 1101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2571799944673256,
+      "learning_rate": 0.0005156014042837695,
+      "loss": 3.0894,
+      "step": 1102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6771210450999576,
+      "learning_rate": 0.0005148587045174128,
+      "loss": 3.1038,
+      "step": 1103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6715665442398364,
+      "learning_rate": 0.0005141159719361891,
+      "loss": 3.0437,
+      "step": 1104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.491821388959583,
+      "learning_rate": 0.0005133732081803945,
+      "loss": 3.017,
+      "step": 1105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4868987802801903,
+      "learning_rate": 0.0005126304148903936,
+      "loss": 2.8768,
+      "step": 1106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7495082719565467,
+      "learning_rate": 0.0005118875937066161,
+      "loss": 3.0833,
+      "step": 1107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6371233260748879,
+      "learning_rate": 0.0005111447462695537,
+      "loss": 3.0977,
+      "step": 1108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.645672539324122,
+      "learning_rate": 0.0005104018742197557,
+      "loss": 3.0169,
+      "step": 1109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3479628803053947,
+      "learning_rate": 0.0005096589791978261,
+      "loss": 2.9704,
+      "step": 1110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6315804151701494,
+      "learning_rate": 0.0005089160628444192,
+      "loss": 3.0905,
+      "step": 1111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.405280201111046,
+      "learning_rate": 0.0005081731268002371,
+      "loss": 3.0804,
+      "step": 1112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5395488316691566,
+      "learning_rate": 0.0005074301727060243,
+      "loss": 2.9697,
+      "step": 1113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6716335219734644,
+      "learning_rate": 0.0005066872022025663,
+      "loss": 2.9523,
+      "step": 1114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4515349752737148,
+      "learning_rate": 0.0005059442169306844,
+      "loss": 3.1051,
+      "step": 1115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.531394074306441,
+      "learning_rate": 0.0005052012185312321,
+      "loss": 3.0116,
+      "step": 1116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6372659490458108,
+      "learning_rate": 0.0005044582086450925,
+      "loss": 3.0477,
+      "step": 1117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5878159758671422,
+      "learning_rate": 0.0005037151889131737,
+      "loss": 3.0607,
+      "step": 1118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5835344868203487,
+      "learning_rate": 0.0005029721609764059,
+      "loss": 2.9643,
+      "step": 1119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6283472496692226,
+      "learning_rate": 0.000502229126475737,
+      "loss": 3.0833,
+      "step": 1120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8811066054378953,
+      "learning_rate": 0.0005014860870521293,
+      "loss": 2.9336,
+      "step": 1121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5582938358216303,
+      "learning_rate": 0.0005007430443465569,
+      "loss": 3.0907,
+      "step": 1122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5369336091305894,
+      "learning_rate": 0.0005,
+      "loss": 3.0583,
+      "step": 1123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7842334085362441,
+      "learning_rate": 0.0004992569556534432,
+      "loss": 3.1529,
+      "step": 1124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7680287250574065,
+      "learning_rate": 0.0004985139129478707,
+      "loss": 3.0098,
+      "step": 1125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4547851974362698,
+      "learning_rate": 0.0004977708735242633,
+      "loss": 3.0157,
+      "step": 1126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4755813135176297,
+      "learning_rate": 0.0004970278390235942,
+      "loss": 3.0078,
+      "step": 1127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.735181138834567,
+      "learning_rate": 0.0004962848110868262,
+      "loss": 2.9656,
+      "step": 1128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.488642179332836,
+      "learning_rate": 0.0004955417913549074,
+      "loss": 3.0486,
+      "step": 1129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.120588433987515,
+      "learning_rate": 0.0004947987814687679,
+      "loss": 3.1748,
+      "step": 1130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1818728106192715,
+      "learning_rate": 0.0004940557830693157,
+      "loss": 2.9357,
+      "step": 1131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4440983457026166,
+      "learning_rate": 0.0004933127977974338,
+      "loss": 2.953,
+      "step": 1132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.919224367829811,
+      "learning_rate": 0.0004925698272939757,
+      "loss": 2.9689,
+      "step": 1133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4096963073328626,
+      "learning_rate": 0.0004918268731997632,
+      "loss": 3.03,
+      "step": 1134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6637093396049603,
+      "learning_rate": 0.0004910839371555809,
+      "loss": 2.9366,
+      "step": 1135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.317346093741528,
+      "learning_rate": 0.0004903410208021739,
+      "loss": 2.9321,
+      "step": 1136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6607256453502948,
+      "learning_rate": 0.0004895981257802443,
+      "loss": 3.0095,
+      "step": 1137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6544388130520578,
+      "learning_rate": 0.0004888552537304463,
+      "loss": 2.9836,
+      "step": 1138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6493646401160458,
+      "learning_rate": 0.00048811240629338394,
+      "loss": 2.9619,
+      "step": 1139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5189343066080112,
+      "learning_rate": 0.00048736958510960663,
+      "loss": 2.9618,
+      "step": 1140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7580444077588147,
+      "learning_rate": 0.00048662679181960564,
+      "loss": 3.0465,
+      "step": 1141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8437702373570144,
+      "learning_rate": 0.00048588402806381094,
+      "loss": 3.0599,
+      "step": 1142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6284216550066812,
+      "learning_rate": 0.0004851412954825874,
+      "loss": 3.01,
+      "step": 1143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.198600527171034,
+      "learning_rate": 0.00048439859571623034,
+      "loss": 3.0224,
+      "step": 1144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.464107229084171,
+      "learning_rate": 0.00048365593040496373,
+      "loss": 3.0229,
+      "step": 1145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.621463143320934,
+      "learning_rate": 0.00048291330118893443,
+      "loss": 2.9357,
+      "step": 1146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.606816732582293,
+      "learning_rate": 0.0004821707097082102,
+      "loss": 2.9572,
+      "step": 1147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7663878269138789,
+      "learning_rate": 0.0004814281576027749,
+      "loss": 3.1086,
+      "step": 1148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.355913234706931,
+      "learning_rate": 0.000480685646512526,
+      "loss": 3.0016,
+      "step": 1149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5211172275739253,
+      "learning_rate": 0.00047994317807727025,
+      "loss": 3.0977,
+      "step": 1150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6495566173133192,
+      "learning_rate": 0.00047920075393671974,
+      "loss": 2.9887,
+      "step": 1151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6860997683939116,
+      "learning_rate": 0.0004784583757304893,
+      "loss": 3.0615,
+      "step": 1152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.556717252061209,
+      "learning_rate": 0.00047771604509809214,
+      "loss": 2.9943,
+      "step": 1153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7491867177374498,
+      "learning_rate": 0.0004769737636789364,
+      "loss": 2.9604,
+      "step": 1154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7797004566252992,
+      "learning_rate": 0.00047623153311232157,
+      "loss": 2.9024,
+      "step": 1155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3869200426238193,
+      "learning_rate": 0.0004754893550374346,
+      "loss": 2.9039,
+      "step": 1156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.907173321142219,
+      "learning_rate": 0.00047474723109334685,
+      "loss": 2.992,
+      "step": 1157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.514488641187699,
+      "learning_rate": 0.00047400516291900993,
+      "loss": 3.021,
+      "step": 1158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6257818855220496,
+      "learning_rate": 0.0004732631521532522,
+      "loss": 3.0884,
+      "step": 1159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7855815609634822,
+      "learning_rate": 0.0004725212004347755,
+      "loss": 2.9267,
+      "step": 1160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.458193308764067,
+      "learning_rate": 0.00047177930940215095,
+      "loss": 2.9718,
+      "step": 1161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5479570889579872,
+      "learning_rate": 0.00047103748069381624,
+      "loss": 2.9869,
+      "step": 1162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5198250072628372,
+      "learning_rate": 0.0004702957159480707,
+      "loss": 3.0798,
+      "step": 1163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4075677725665527,
+      "learning_rate": 0.00046955401680307267,
+      "loss": 3.1509,
+      "step": 1164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6728164256044225,
+      "learning_rate": 0.0004688123848968359,
+      "loss": 3.0287,
+      "step": 1165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.410832730190197,
+      "learning_rate": 0.00046807082186722516,
+      "loss": 3.0103,
+      "step": 1166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.496727823682135,
+      "learning_rate": 0.0004673293293519535,
+      "loss": 2.9473,
+      "step": 1167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6362572836501736,
+      "learning_rate": 0.00046658790898857806,
+      "loss": 3.1274,
+      "step": 1168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4180529041689665,
+      "learning_rate": 0.0004658465624144963,
+      "loss": 2.9134,
+      "step": 1169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6614867073062518,
+      "learning_rate": 0.0004651052912669438,
+      "loss": 3.0455,
+      "step": 1170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4634470110764943,
+      "learning_rate": 0.0004643640971829883,
+      "loss": 2.884,
+      "step": 1171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5404314616954158,
+      "learning_rate": 0.0004636229817995281,
+      "loss": 3.0164,
+      "step": 1172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6047243607738835,
+      "learning_rate": 0.0004628819467532876,
+      "loss": 2.9016,
+      "step": 1173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4775830589302195,
+      "learning_rate": 0.00046214099368081335,
+      "loss": 3.0596,
+      "step": 1174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6278791099587122,
+      "learning_rate": 0.0004614001242184714,
+      "loss": 2.896,
+      "step": 1175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.036904714572199,
+      "learning_rate": 0.000460659340002443,
+      "loss": 3.0239,
+      "step": 1176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5413537311127568,
+      "learning_rate": 0.00045991864266872073,
+      "loss": 2.9587,
+      "step": 1177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0113468432088113,
+      "learning_rate": 0.00045917803385310595,
+      "loss": 3.0564,
+      "step": 1178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.02012514313468,
+      "learning_rate": 0.00045843751519120417,
+      "loss": 3.0323,
+      "step": 1179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.049386544043595,
+      "learning_rate": 0.00045769708831842193,
+      "loss": 3.0427,
+      "step": 1180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1141906074267256,
+      "learning_rate": 0.00045695675486996266,
+      "loss": 3.0087,
+      "step": 1181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5843015219149599,
+      "learning_rate": 0.00045621651648082405,
+      "loss": 2.8822,
+      "step": 1182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1116606968272507,
+      "learning_rate": 0.00045547637478579356,
+      "loss": 2.8727,
+      "step": 1183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5962946132849856,
+      "learning_rate": 0.0004547363314194449,
+      "loss": 3.0659,
+      "step": 1184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5310349912980679,
+      "learning_rate": 0.000453996388016135,
+      "loss": 3.0868,
+      "step": 1185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.820218813878498,
+      "learning_rate": 0.0004532565462099999,
+      "loss": 2.9762,
+      "step": 1186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6115825222831512,
+      "learning_rate": 0.0004525168076349513,
+      "loss": 3.0294,
+      "step": 1187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.728187198730813,
+      "learning_rate": 0.0004517771739246729,
+      "loss": 3.0554,
+      "step": 1188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1984739975284517,
+      "learning_rate": 0.0004510376467126165,
+      "loss": 2.9963,
+      "step": 1189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4987481717699322,
+      "learning_rate": 0.0004502982276319992,
+      "loss": 2.9112,
+      "step": 1190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6088313062522037,
+      "learning_rate": 0.0004495589183157991,
+      "loss": 2.916,
+      "step": 1191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7757728360588503,
+      "learning_rate": 0.0004488197203967517,
+      "loss": 2.9988,
+      "step": 1192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5729560485248986,
+      "learning_rate": 0.0004480806355073467,
+      "loss": 3.0106,
+      "step": 1193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5887834924525823,
+      "learning_rate": 0.000447341665279824,
+      "loss": 2.9749,
+      "step": 1194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5483785908590972,
+      "learning_rate": 0.0004466028113461708,
+      "loss": 2.9645,
+      "step": 1195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7983589689991573,
+      "learning_rate": 0.0004458640753381167,
+      "loss": 3.0036,
+      "step": 1196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.559138476081008,
+      "learning_rate": 0.0004451254588871313,
+      "loss": 3.0643,
+      "step": 1197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7490827549498031,
+      "learning_rate": 0.0004443869636244203,
+      "loss": 2.9993,
+      "step": 1198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7360113178112446,
+      "learning_rate": 0.0004436485911809212,
+      "loss": 2.9288,
+      "step": 1199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5011914495677197,
+      "learning_rate": 0.00044291034318730087,
+      "loss": 2.9985,
+      "step": 1200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5383022534862532,
+      "learning_rate": 0.0004421722212739511,
+      "loss": 3.0061,
+      "step": 1201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.636927057324379,
+      "learning_rate": 0.0004414342270709848,
+      "loss": 2.9521,
+      "step": 1202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6182247495417805,
+      "learning_rate": 0.00044069636220823397,
+      "loss": 2.9825,
+      "step": 1203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.635510514586777,
+      "learning_rate": 0.0004399586283152437,
+      "loss": 3.0663,
+      "step": 1204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5849604088910123,
+      "learning_rate": 0.0004392210270212706,
+      "loss": 3.0214,
+      "step": 1205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5443005375422927,
+      "learning_rate": 0.00043848355995527825,
+      "loss": 2.9564,
+      "step": 1206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6150165149225972,
+      "learning_rate": 0.00043774622874593374,
+      "loss": 2.9766,
+      "step": 1207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.363017519543566,
+      "learning_rate": 0.000437009035021604,
+      "loss": 2.9776,
+      "step": 1208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5139227258822547,
+      "learning_rate": 0.00043627198041035274,
+      "loss": 2.9989,
+      "step": 1209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0382271199612543,
+      "learning_rate": 0.00043553506653993597,
+      "loss": 3.0294,
+      "step": 1210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.189013324857636,
+      "learning_rate": 0.0004347982950377992,
+      "loss": 2.9811,
+      "step": 1211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.621451121098076,
+      "learning_rate": 0.0004340616675310735,
+      "loss": 3.0056,
+      "step": 1212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1564056415907262,
+      "learning_rate": 0.00043332518564657193,
+      "loss": 2.9133,
+      "step": 1213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8635010156079654,
+      "learning_rate": 0.0004325888510107856,
+      "loss": 3.2171,
+      "step": 1214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3751064993462023,
+      "learning_rate": 0.0004318526652498809,
+      "loss": 3.0365,
+      "step": 1215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.755445646159852,
+      "learning_rate": 0.00043111662998969523,
+      "loss": 3.1027,
+      "step": 1216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5116546398054544,
+      "learning_rate": 0.0004303807468557335,
+      "loss": 3.1306,
+      "step": 1217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.214801181871165,
+      "learning_rate": 0.0004296450174731648,
+      "loss": 2.9493,
+      "step": 1218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.559843751673804,
+      "learning_rate": 0.0004289094434668188,
+      "loss": 2.9991,
+      "step": 1219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5561033301023033,
+      "learning_rate": 0.00042817402646118185,
+      "loss": 3.0336,
+      "step": 1220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.466822071470893,
+      "learning_rate": 0.0004274387680803936,
+      "loss": 3.087,
+      "step": 1221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0270625411318655,
+      "learning_rate": 0.00042670366994824327,
+      "loss": 2.9511,
+      "step": 1222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.571792833540858,
+      "learning_rate": 0.0004259687336881663,
+      "loss": 2.9273,
+      "step": 1223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.313258224714201,
+      "learning_rate": 0.0004252339609232408,
+      "loss": 3.0359,
+      "step": 1224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4779593141762266,
+      "learning_rate": 0.0004244993532761834,
+      "loss": 3.0196,
+      "step": 1225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7523876860304117,
+      "learning_rate": 0.00042376491236934634,
+      "loss": 3.0577,
+      "step": 1226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7938024918042585,
+      "learning_rate": 0.0004230306398247136,
+      "loss": 2.84,
+      "step": 1227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4368394776082039,
+      "learning_rate": 0.0004222965372638976,
+      "loss": 3.0424,
+      "step": 1228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.820752044135207,
+      "learning_rate": 0.0004215626063081348,
+      "loss": 3.1341,
+      "step": 1229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.513896054097871,
+      "learning_rate": 0.000420828848578283,
+      "loss": 2.9865,
+      "step": 1230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.523493576649948,
+      "learning_rate": 0.0004200952656948175,
+      "loss": 3.1191,
+      "step": 1231
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5873711238812442,
+      "learning_rate": 0.0004193618592778272,
+      "loss": 3.0334,
+      "step": 1232
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6419885433871533,
+      "learning_rate": 0.0004186286309470116,
+      "loss": 3.0018,
+      "step": 1233
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4165506383468591,
+      "learning_rate": 0.0004178955823216767,
+      "loss": 3.0308,
+      "step": 1234
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2531595646535587,
+      "learning_rate": 0.00041716271502073137,
+      "loss": 3.0897,
+      "step": 1235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6598582383465785,
+      "learning_rate": 0.000416430030662685,
+      "loss": 2.8866,
+      "step": 1236
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.572860841848263,
+      "learning_rate": 0.00041569753086564173,
+      "loss": 2.9265,
+      "step": 1237
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5549817506099015,
+      "learning_rate": 0.0004149652172472988,
+      "loss": 3.0615,
+      "step": 1238
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5429361334226948,
+      "learning_rate": 0.00041423309142494234,
+      "loss": 2.9804,
+      "step": 1239
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.545358959110009,
+      "learning_rate": 0.0004135011550154433,
+      "loss": 2.9774,
+      "step": 1240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5539719074870866,
+      "learning_rate": 0.0004127694096352546,
+      "loss": 3.0414,
+      "step": 1241
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.517086142819113,
+      "learning_rate": 0.00041203785690040743,
+      "loss": 2.9761,
+      "step": 1242
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.499789686284178,
+      "learning_rate": 0.00041130649842650694,
+      "loss": 2.98,
+      "step": 1243
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5788668006056907,
+      "learning_rate": 0.00041057533582873016,
+      "loss": 3.1147,
+      "step": 1244
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.446205727755795,
+      "learning_rate": 0.0004098443707218208,
+      "loss": 3.0966,
+      "step": 1245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5506954703185012,
+      "learning_rate": 0.00040911360472008673,
+      "loss": 3.1149,
+      "step": 1246
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.638591431441541,
+      "learning_rate": 0.0004083830394373959,
+      "loss": 2.9825,
+      "step": 1247
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6011656506655851,
+      "learning_rate": 0.00040765267648717324,
+      "loss": 3.0413,
+      "step": 1248
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.836421468278804,
+      "learning_rate": 0.00040692251748239677,
+      "loss": 2.9357,
+      "step": 1249
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.655146834662203,
+      "learning_rate": 0.00040619256403559383,
+      "loss": 2.8718,
+      "step": 1250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5875293341807613,
+      "learning_rate": 0.000405462817758838,
+      "loss": 2.9972,
+      "step": 1251
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5371012558925563,
+      "learning_rate": 0.0004047332802637457,
+      "loss": 2.9606,
+      "step": 1252
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5253654236132752,
+      "learning_rate": 0.00040400395316147157,
+      "loss": 2.9263,
+      "step": 1253
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.883704005320784,
+      "learning_rate": 0.00040327483806270627,
+      "loss": 3.1026,
+      "step": 1254
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9306523246636023,
+      "learning_rate": 0.0004025459365776715,
+      "loss": 2.9742,
+      "step": 1255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.726222676254312,
+      "learning_rate": 0.00040181725031611794,
+      "loss": 3.0259,
+      "step": 1256
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6403526611017167,
+      "learning_rate": 0.0004010887808873206,
+      "loss": 2.8751,
+      "step": 1257
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4296113006521627,
+      "learning_rate": 0.00040036052990007553,
+      "loss": 2.9387,
+      "step": 1258
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5646632099181292,
+      "learning_rate": 0.0003996324989626967,
+      "loss": 2.9707,
+      "step": 1259
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7321542303206343,
+      "learning_rate": 0.00039890468968301166,
+      "loss": 2.9433,
+      "step": 1260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4857292421694133,
+      "learning_rate": 0.0003981771036683591,
+      "loss": 3.0651,
+      "step": 1261
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5250657822149811,
+      "learning_rate": 0.00039744974252558385,
+      "loss": 2.9565,
+      "step": 1262
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.625665568528434,
+      "learning_rate": 0.00039672260786103463,
+      "loss": 3.0203,
+      "step": 1263
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.659861490683107,
+      "learning_rate": 0.00039599570128055994,
+      "loss": 3.0157,
+      "step": 1264
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5952716206792026,
+      "learning_rate": 0.0003952690243895044,
+      "loss": 2.9629,
+      "step": 1265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6635549516681594,
+      "learning_rate": 0.0003945425787927054,
+      "loss": 3.083,
+      "step": 1266
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7673069802841315,
+      "learning_rate": 0.00039381636609448975,
+      "loss": 2.8468,
+      "step": 1267
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4932682765961776,
+      "learning_rate": 0.0003930903878986693,
+      "loss": 2.9866,
+      "step": 1268
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.716526170608899,
+      "learning_rate": 0.00039236464580853916,
+      "loss": 3.0973,
+      "step": 1269
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5542101920678129,
+      "learning_rate": 0.0003916391414268718,
+      "loss": 2.9661,
+      "step": 1270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5648791898551646,
+      "learning_rate": 0.00039091387635591536,
+      "loss": 3.0613,
+      "step": 1271
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5476762507365098,
+      "learning_rate": 0.0003901888521973894,
+      "loss": 2.8803,
+      "step": 1272
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.661835549080649,
+      "learning_rate": 0.0003894640705524813,
+      "loss": 2.9866,
+      "step": 1273
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6593827070939977,
+      "learning_rate": 0.00038873953302184284,
+      "loss": 2.8358,
+      "step": 1274
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5871060145184717,
+      "learning_rate": 0.000388015241205587,
+      "loss": 2.9009,
+      "step": 1275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.490173165824435,
+      "learning_rate": 0.00038729119670328355,
+      "loss": 2.9798,
+      "step": 1276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4912982935845387,
+      "learning_rate": 0.00038656740111395665,
+      "loss": 3.0522,
+      "step": 1277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6584222508647164,
+      "learning_rate": 0.00038584385603608053,
+      "loss": 3.0475,
+      "step": 1278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6516157378386456,
+      "learning_rate": 0.00038512056306757615,
+      "loss": 3.0463,
+      "step": 1279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.534083974152629,
+      "learning_rate": 0.0003843975238058075,
+      "loss": 2.9437,
+      "step": 1280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6660749506182209,
+      "learning_rate": 0.00038367473984757863,
+      "loss": 2.9927,
+      "step": 1281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4586533971323798,
+      "learning_rate": 0.0003829522127891296,
+      "loss": 3.0398,
+      "step": 1282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8139189336204646,
+      "learning_rate": 0.0003822299442261329,
+      "loss": 2.85,
+      "step": 1283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6586465726775685,
+      "learning_rate": 0.00038150793575369063,
+      "loss": 2.9749,
+      "step": 1284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6334899861704983,
+      "learning_rate": 0.0003807861889663299,
+      "loss": 2.9696,
+      "step": 1285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6631642798338515,
+      "learning_rate": 0.0003800647054580006,
+      "loss": 2.921,
+      "step": 1286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8664620762604884,
+      "learning_rate": 0.00037934348682207064,
+      "loss": 3.0129,
+      "step": 1287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7933699150769544,
+      "learning_rate": 0.00037862253465132306,
+      "loss": 2.9144,
+      "step": 1288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.287265736625869,
+      "learning_rate": 0.00037790185053795245,
+      "loss": 3.0468,
+      "step": 1289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5363067452235106,
+      "learning_rate": 0.0003771814360735616,
+      "loss": 3.0109,
+      "step": 1290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9754591170454587,
+      "learning_rate": 0.00037646129284915755,
+      "loss": 2.986,
+      "step": 1291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0894795022649313,
+      "learning_rate": 0.00037574142245514825,
+      "loss": 2.9644,
+      "step": 1292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5268157208560573,
+      "learning_rate": 0.0003750218264813393,
+      "loss": 3.0396,
+      "step": 1293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7221522490356287,
+      "learning_rate": 0.0003743025065169305,
+      "loss": 2.9825,
+      "step": 1294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8713708663882545,
+      "learning_rate": 0.0003735834641505116,
+      "loss": 3.0782,
+      "step": 1295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5656292438381014,
+      "learning_rate": 0.00037286470097005954,
+      "loss": 2.8478,
+      "step": 1296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.036431621990092,
+      "learning_rate": 0.0003721462185629347,
+      "loss": 2.892,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1440045537351455,
+      "learning_rate": 0.00037142801851587707,
+      "loss": 3.0166,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6676813721419121,
+      "learning_rate": 0.00037071010241500357,
+      "loss": 2.9821,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2298626075202046,
+      "learning_rate": 0.00036999247184580383,
+      "loss": 2.9086,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5885499233956155,
+      "learning_rate": 0.00036927512839313636,
+      "loss": 2.9613,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8897665521195117,
+      "learning_rate": 0.0003685580736412268,
+      "loss": 2.9471,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7685415184926847,
+      "learning_rate": 0.000367841309173662,
+      "loss": 2.87,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.440855612373281,
+      "learning_rate": 0.0003671248365733883,
+      "loss": 2.8656,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7070814556786251,
+      "learning_rate": 0.0003664086574227075,
+      "loss": 2.8701,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6594698215645276,
+      "learning_rate": 0.000365692773303273,
+      "loss": 2.9218,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.456954608244558,
+      "learning_rate": 0.00036497718579608696,
+      "loss": 2.9079,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.543641004888524,
+      "learning_rate": 0.0003642618964814964,
+      "loss": 2.9145,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6385807219146835,
+      "learning_rate": 0.00036354690693918946,
+      "loss": 3.0383,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.363843341801638,
+      "learning_rate": 0.00036283221874819284,
+      "loss": 2.9691,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7778994375954826,
+      "learning_rate": 0.0003621178334868672,
+      "loss": 3.0771,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6116218858580673,
+      "learning_rate": 0.00036140375273290476,
+      "loss": 3.0035,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4842572952635982,
+      "learning_rate": 0.0003606899780633245,
+      "loss": 2.8582,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5426528925715866,
+      "learning_rate": 0.0003599765110544699,
+      "loss": 2.8221,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.63532641531039,
+      "learning_rate": 0.0003592633532820052,
+      "loss": 2.8954,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5644021047339576,
+      "learning_rate": 0.0003585505063209109,
+      "loss": 3.04,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.610797969390034,
+      "learning_rate": 0.00035783797174548194,
+      "loss": 2.9842,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5528087416672856,
+      "learning_rate": 0.00035712575112932277,
+      "loss": 2.8262,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9597800410614374,
+      "learning_rate": 0.000356413846045345,
+      "loss": 2.9859,
+      "step": 1319
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9222960672724123,
+      "learning_rate": 0.000355702258065763,
+      "loss": 2.9636,
+      "step": 1320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8647036890088053,
+      "learning_rate": 0.0003549909887620909,
+      "loss": 2.9303,
+      "step": 1321
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7040987269256869,
+      "learning_rate": 0.00035428003970513914,
+      "loss": 2.9671,
+      "step": 1322
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4619722341828134,
+      "learning_rate": 0.00035356941246501085,
+      "loss": 2.9843,
+      "step": 1323
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9539383623194002,
+      "learning_rate": 0.0003528591086110984,
+      "loss": 2.982,
+      "step": 1324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.528471682237156,
+      "learning_rate": 0.00035214912971208,
+      "loss": 2.9192,
+      "step": 1325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7156291941383408,
+      "learning_rate": 0.0003514394773359163,
+      "loss": 2.7977,
+      "step": 1326
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5332308243919623,
+      "learning_rate": 0.0003507301530498469,
+      "loss": 2.9665,
+      "step": 1327
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.540612198436645,
+      "learning_rate": 0.00035002115842038646,
+      "loss": 2.9738,
+      "step": 1328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4659289146055459,
+      "learning_rate": 0.00034931249501332195,
+      "loss": 2.9281,
+      "step": 1329
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6850936261894143,
+      "learning_rate": 0.00034860416439370885,
+      "loss": 3.0519,
+      "step": 1330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8797337661701188,
+      "learning_rate": 0.0003478961681258674,
+      "loss": 3.0646,
+      "step": 1331
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4648717669206888,
+      "learning_rate": 0.0003471885077733796,
+      "loss": 2.925,
+      "step": 1332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6354797709221596,
+      "learning_rate": 0.0003464811848990859,
+      "loss": 3.0124,
+      "step": 1333
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5551901779814556,
+      "learning_rate": 0.00034577420106508063,
+      "loss": 3.0355,
+      "step": 1334
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5133625026038877,
+      "learning_rate": 0.0003450675578327105,
+      "loss": 2.9584,
+      "step": 1335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6455830168416097,
+      "learning_rate": 0.000344361256762569,
+      "loss": 2.9426,
+      "step": 1336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5014164409602608,
+      "learning_rate": 0.00034365529941449456,
+      "loss": 2.8835,
+      "step": 1337
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6352725536463528,
+      "learning_rate": 0.0003429496873475664,
+      "loss": 2.9254,
+      "step": 1338
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5131511549339582,
+      "learning_rate": 0.0003422444221201009,
+      "loss": 2.8863,
+      "step": 1339
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4432721743941979,
+      "learning_rate": 0.0003415395052896487,
+      "loss": 2.948,
+      "step": 1340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5537354789342548,
+      "learning_rate": 0.0003408349384129912,
+      "loss": 3.0789,
+      "step": 1341
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6114474091637936,
+      "learning_rate": 0.00034013072304613643,
+      "loss": 3.0046,
+      "step": 1342
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5534392450026369,
+      "learning_rate": 0.00033942686074431674,
+      "loss": 2.976,
+      "step": 1343
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3939135222354884,
+      "learning_rate": 0.0003387233530619843,
+      "loss": 2.9227,
+      "step": 1344
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.427272688826071,
+      "learning_rate": 0.0003380202015528084,
+      "loss": 3.0878,
+      "step": 1345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4662350248532068,
+      "learning_rate": 0.0003373174077696715,
+      "loss": 2.9893,
+      "step": 1346
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4835062752038195,
+      "learning_rate": 0.0003366149732646661,
+      "loss": 2.926,
+      "step": 1347
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.084737483749785,
+      "learning_rate": 0.00033591289958909143,
+      "loss": 2.8884,
+      "step": 1348
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.779542080235675,
+      "learning_rate": 0.00033521118829344954,
+      "loss": 2.906,
+      "step": 1349
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7150134923851612,
+      "learning_rate": 0.0003345098409274423,
+      "loss": 3.0901,
+      "step": 1350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3484655062880324,
+      "learning_rate": 0.00033380885903996796,
+      "loss": 2.9031,
+      "step": 1351
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5089062674724676,
+      "learning_rate": 0.00033310824417911766,
+      "loss": 2.9589,
+      "step": 1352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6899640394809492,
+      "learning_rate": 0.00033240799789217184,
+      "loss": 3.0385,
+      "step": 1353
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5780168088710897,
+      "learning_rate": 0.00033170812172559694,
+      "loss": 2.9076,
+      "step": 1354
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5147803672292959,
+      "learning_rate": 0.000331008617225042,
+      "loss": 2.8964,
+      "step": 1355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.121543233288114,
+      "learning_rate": 0.0003303094859353355,
+      "loss": 2.979,
+      "step": 1356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6656366010172299,
+      "learning_rate": 0.0003296107294004812,
+      "loss": 2.8862,
+      "step": 1357
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8767046389451991,
+      "learning_rate": 0.0003289123491636559,
+      "loss": 3.0106,
+      "step": 1358
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.398157974096869,
+      "learning_rate": 0.00032821434676720443,
+      "loss": 3.0338,
+      "step": 1359
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5528537563609524,
+      "learning_rate": 0.00032751672375263836,
+      "loss": 2.9697,
+      "step": 1360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0877681486701634,
+      "learning_rate": 0.0003268194816606305,
+      "loss": 2.9726,
+      "step": 1361
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.710821630701027,
+      "learning_rate": 0.00032612262203101267,
+      "loss": 3.0052,
+      "step": 1362
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.458382348822902,
+      "learning_rate": 0.00032542614640277225,
+      "loss": 2.8586,
+      "step": 1363
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0331627958576415,
+      "learning_rate": 0.0003247300563140481,
+      "loss": 2.8776,
+      "step": 1364
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.143906633779722,
+      "learning_rate": 0.00032403435330212807,
+      "loss": 2.9577,
+      "step": 1365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.629450064586538,
+      "learning_rate": 0.00032333903890344515,
+      "loss": 2.9096,
+      "step": 1366
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5246477729689838,
+      "learning_rate": 0.00032264411465357333,
+      "loss": 2.957,
+      "step": 1367
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0637951937991454,
+      "learning_rate": 0.00032194958208722654,
+      "loss": 2.9039,
+      "step": 1368
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.474378694886027,
+      "learning_rate": 0.00032125544273825204,
+      "loss": 2.8452,
+      "step": 1369
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4930776724703856,
+      "learning_rate": 0.0003205616981396297,
+      "loss": 3.0294,
+      "step": 1370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9301379475938192,
+      "learning_rate": 0.00031986834982346713,
+      "loss": 2.9364,
+      "step": 1371
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.746140145065674,
+      "learning_rate": 0.00031917539932099694,
+      "loss": 2.8717,
+      "step": 1372
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5767281709349625,
+      "learning_rate": 0.00031848284816257336,
+      "loss": 3.0257,
+      "step": 1373
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4435235721039854,
+      "learning_rate": 0.0003177906978776682,
+      "loss": 2.8976,
+      "step": 1374
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5650872205708277,
+      "learning_rate": 0.0003170989499948683,
+      "loss": 3.0314,
+      "step": 1375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.692640348238247,
+      "learning_rate": 0.0003164076060418719,
+      "loss": 3.0032,
+      "step": 1376
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4813268378755282,
+      "learning_rate": 0.000315716667545485,
+      "loss": 2.8071,
+      "step": 1377
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8352471119976326,
+      "learning_rate": 0.00031502613603161836,
+      "loss": 3.0435,
+      "step": 1378
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7266068717679979,
+      "learning_rate": 0.00031433601302528335,
+      "loss": 2.9802,
+      "step": 1379
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.613266958345568,
+      "learning_rate": 0.00031364630005058995,
+      "loss": 2.8918,
+      "step": 1380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8913328037088564,
+      "learning_rate": 0.0003129569986307422,
+      "loss": 3.0171,
+      "step": 1381
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.474795949421643,
+      "learning_rate": 0.00031226811028803515,
+      "loss": 2.9633,
+      "step": 1382
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.474364597263621,
+      "learning_rate": 0.00031157963654385173,
+      "loss": 2.9756,
+      "step": 1383
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5872630914320038,
+      "learning_rate": 0.0003108915789186592,
+      "loss": 2.9909,
+      "step": 1384
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.597351900905089,
+      "learning_rate": 0.00031020393893200604,
+      "loss": 2.9626,
+      "step": 1385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5795043278020011,
+      "learning_rate": 0.00030951671810251823,
+      "loss": 2.9844,
+      "step": 1386
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6023672983895467,
+      "learning_rate": 0.0003088299179478959,
+      "loss": 2.9619,
+      "step": 1387
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.728788232012061,
+      "learning_rate": 0.0003081435399849104,
+      "loss": 3.012,
+      "step": 1388
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6388999334322343,
+      "learning_rate": 0.0003074575857294004,
+      "loss": 2.9612,
+      "step": 1389
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7054539291425803,
+      "learning_rate": 0.0003067720566962691,
+      "loss": 2.9113,
+      "step": 1390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5946881973675533,
+      "learning_rate": 0.0003060869543994806,
+      "loss": 3.0111,
+      "step": 1391
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.579112833044184,
+      "learning_rate": 0.0003054022803520562,
+      "loss": 2.9519,
+      "step": 1392
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8493733018688614,
+      "learning_rate": 0.0003047180360660721,
+      "loss": 2.9692,
+      "step": 1393
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5703491498427762,
+      "learning_rate": 0.00030403422305265475,
+      "loss": 2.8615,
+      "step": 1394
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5359553555036418,
+      "learning_rate": 0.0003033508428219785,
+      "loss": 2.944,
+      "step": 1395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5693689190684206,
+      "learning_rate": 0.00030266789688326184,
+      "loss": 3.0675,
+      "step": 1396
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4972113372051779,
+      "learning_rate": 0.00030198538674476393,
+      "loss": 2.9578,
+      "step": 1397
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4942093248936732,
+      "learning_rate": 0.00030130331391378185,
+      "loss": 2.9145,
+      "step": 1398
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5233922296164328,
+      "learning_rate": 0.0003006216798966468,
+      "loss": 2.9536,
+      "step": 1399
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4624160784368594,
+      "learning_rate": 0.00029994048619872034,
+      "loss": 2.895,
+      "step": 1400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7047015688672538,
+      "learning_rate": 0.0002992597343243927,
+      "loss": 2.9253,
+      "step": 1401
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.632639380465163,
+      "learning_rate": 0.0002985794257770773,
+      "loss": 3.0136,
+      "step": 1402
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7443137682087528,
+      "learning_rate": 0.0002978995620592092,
+      "loss": 3.0467,
+      "step": 1403
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.621061096602618,
+      "learning_rate": 0.0002972201446722405,
+      "loss": 2.9708,
+      "step": 1404
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7682695158903525,
+      "learning_rate": 0.00029654117511663803,
+      "loss": 2.976,
+      "step": 1405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7474219703074103,
+      "learning_rate": 0.0002958626548918795,
+      "loss": 2.8493,
+      "step": 1406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.463330577686259,
+      "learning_rate": 0.00029518458549645014,
+      "loss": 3.0763,
+      "step": 1407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.467232757494785,
+      "learning_rate": 0.00029450696842783954,
+      "loss": 2.8993,
+      "step": 1408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.891397178120462,
+      "learning_rate": 0.00029382980518253865,
+      "loss": 2.951,
+      "step": 1409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7599168039403779,
+      "learning_rate": 0.00029315309725603595,
+      "loss": 3.0634,
+      "step": 1410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5185068157165313,
+      "learning_rate": 0.00029247684614281446,
+      "loss": 2.919,
+      "step": 1411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8253371419114437,
+      "learning_rate": 0.0002918010533363481,
+      "loss": 2.9206,
+      "step": 1412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8422968224779241,
+      "learning_rate": 0.0002911257203290987,
+      "loss": 2.8432,
+      "step": 1413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6496202528819097,
+      "learning_rate": 0.00029045084861251314,
+      "loss": 3.0623,
+      "step": 1414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.012893809877519,
+      "learning_rate": 0.00028977643967701897,
+      "loss": 2.9576,
+      "step": 1415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9168883504738665,
+      "learning_rate": 0.00028910249501202156,
+      "loss": 2.9358,
+      "step": 1416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9423610968115808,
+      "learning_rate": 0.00028842901610590165,
+      "loss": 2.9877,
+      "step": 1417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9506877231917894,
+      "learning_rate": 0.00028775600444601123,
+      "loss": 2.9764,
+      "step": 1418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3334004225026237,
+      "learning_rate": 0.00028708346151866973,
+      "loss": 2.9201,
+      "step": 1419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6154873122732294,
+      "learning_rate": 0.0002864113888091622,
+      "loss": 2.9686,
+      "step": 1420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9663132163825408,
+      "learning_rate": 0.0002857397878017348,
+      "loss": 2.9562,
+      "step": 1421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0519938815976517,
+      "learning_rate": 0.00028506865997959173,
+      "loss": 2.9478,
+      "step": 1422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5177169988656318,
+      "learning_rate": 0.000284398006824893,
+      "loss": 3.0404,
+      "step": 1423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.787514007695418,
+      "learning_rate": 0.00028372782981874963,
+      "loss": 2.9171,
+      "step": 1424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6430796364440303,
+      "learning_rate": 0.00028305813044122096,
+      "loss": 2.8965,
+      "step": 1425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6845035085775544,
+      "learning_rate": 0.0002823889101713122,
+      "loss": 2.993,
+      "step": 1426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4029502775942215,
+      "learning_rate": 0.0002817201704869701,
+      "loss": 3.0604,
+      "step": 1427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6964297995620303,
+      "learning_rate": 0.00028105191286508,
+      "loss": 3.061,
+      "step": 1428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.772940096598246,
+      "learning_rate": 0.00028038413878146245,
+      "loss": 2.9677,
+      "step": 1429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6742471399529648,
+      "learning_rate": 0.00027971684971087073,
+      "loss": 2.8788,
+      "step": 1430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7972572418251187,
+      "learning_rate": 0.00027905004712698643,
+      "loss": 2.8768,
+      "step": 1431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7839159080981906,
+      "learning_rate": 0.0002783837325024167,
+      "loss": 2.9872,
+      "step": 1432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3925395735571318,
+      "learning_rate": 0.00027771790730869153,
+      "loss": 2.9968,
+      "step": 1433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7008308304617028,
+      "learning_rate": 0.0002770525730162599,
+      "loss": 2.9314,
+      "step": 1434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.580406122805253,
+      "learning_rate": 0.00027638773109448645,
+      "loss": 2.9094,
+      "step": 1435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7109514587578447,
+      "learning_rate": 0.00027572338301164824,
+      "loss": 3.0569,
+      "step": 1436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.575838054738935,
+      "learning_rate": 0.0002750595302349324,
+      "loss": 2.88,
+      "step": 1437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.64111912579196,
+      "learning_rate": 0.00027439617423043145,
+      "loss": 2.9151,
+      "step": 1438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5836007325082615,
+      "learning_rate": 0.00027373331646314114,
+      "loss": 2.909,
+      "step": 1439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9386809311449786,
+      "learning_rate": 0.0002730709583969572,
+      "loss": 2.8861,
+      "step": 1440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4726818380073057,
+      "learning_rate": 0.0002724091014946711,
+      "loss": 3.0225,
+      "step": 1441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9487696453081842,
+      "learning_rate": 0.00027174774721796824,
+      "loss": 2.9398,
+      "step": 1442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1348471927367507,
+      "learning_rate": 0.0002710868970274232,
+      "loss": 2.9258,
+      "step": 1443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.48227765469923,
+      "learning_rate": 0.0002704265523824982,
+      "loss": 2.9098,
+      "step": 1444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.812636894124514,
+      "learning_rate": 0.00026976671474153826,
+      "loss": 2.9432,
+      "step": 1445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7325772817498266,
+      "learning_rate": 0.00026910738556176886,
+      "loss": 2.9392,
+      "step": 1446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6434455055710229,
+      "learning_rate": 0.0002684485662992929,
+      "loss": 2.9974,
+      "step": 1447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3998785135292606,
+      "learning_rate": 0.0002677902584090869,
+      "loss": 2.9936,
+      "step": 1448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2857011146047195,
+      "learning_rate": 0.00026713246334499774,
+      "loss": 2.8716,
+      "step": 1449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6837911593026704,
+      "learning_rate": 0.00026647518255974023,
+      "loss": 2.9168,
+      "step": 1450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4851466251187184,
+      "learning_rate": 0.0002658184175048934,
+      "loss": 2.9192,
+      "step": 1451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6057931670418535,
+      "learning_rate": 0.00026516216963089694,
+      "loss": 2.9781,
+      "step": 1452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7460870575144598,
+      "learning_rate": 0.0002645064403870488,
+      "loss": 3.0093,
+      "step": 1453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6472092669898926,
+      "learning_rate": 0.0002638512312215011,
+      "loss": 2.9784,
+      "step": 1454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.463279415653525,
+      "learning_rate": 0.0002631965435812575,
+      "loss": 2.9109,
+      "step": 1455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.764635506677426,
+      "learning_rate": 0.00026254237891217046,
+      "loss": 2.9248,
+      "step": 1456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0090179436877946,
+      "learning_rate": 0.0002618887386589367,
+      "loss": 2.9407,
+      "step": 1457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5349764154897843,
+      "learning_rate": 0.0002612356242650949,
+      "loss": 2.8539,
+      "step": 1458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6188750989189098,
+      "learning_rate": 0.0002605830371730229,
+      "loss": 2.8285,
+      "step": 1459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.631474483328909,
+      "learning_rate": 0.0002599309788239339,
+      "loss": 2.983,
+      "step": 1460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.495424074144567,
+      "learning_rate": 0.00025927945065787306,
+      "loss": 2.9395,
+      "step": 1461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7756282704604038,
+      "learning_rate": 0.0002586284541137145,
+      "loss": 2.9565,
+      "step": 1462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.173073229337383,
+      "learning_rate": 0.00025797799062915905,
+      "loss": 2.964,
+      "step": 1463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6011925175193404,
+      "learning_rate": 0.00025732806164072966,
+      "loss": 2.9618,
+      "step": 1464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.077744695541047,
+      "learning_rate": 0.00025667866858376874,
+      "loss": 3.0181,
+      "step": 1465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.815296462224908,
+      "learning_rate": 0.0002560298128924358,
+      "loss": 2.913,
+      "step": 1466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8858103703148215,
+      "learning_rate": 0.0002553814959997032,
+      "loss": 3.0468,
+      "step": 1467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3420019008631696,
+      "learning_rate": 0.00025473371933735334,
+      "loss": 2.8835,
+      "step": 1468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6778825385242786,
+      "learning_rate": 0.00025408648433597534,
+      "loss": 2.9263,
+      "step": 1469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4118021309265316,
+      "learning_rate": 0.00025343979242496283,
+      "loss": 2.823,
+      "step": 1470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0497359073704535,
+      "learning_rate": 0.00025279364503250925,
+      "loss": 2.9324,
+      "step": 1471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5800702916609484,
+      "learning_rate": 0.0002521480435856056,
+      "loss": 2.9229,
+      "step": 1472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4982796189809016,
+      "learning_rate": 0.0002515029895100378,
+      "loss": 2.8332,
+      "step": 1473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2057197050565596,
+      "learning_rate": 0.0002508584842303822,
+      "loss": 3.0021,
+      "step": 1474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8114638268662162,
+      "learning_rate": 0.0002502145291700038,
+      "loss": 2.8443,
+      "step": 1475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5099089213027126,
+      "learning_rate": 0.0002495711257510517,
+      "loss": 2.936,
+      "step": 1476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.976124566157622,
+      "learning_rate": 0.0002489282753944575,
+      "loss": 2.9514,
+      "step": 1477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6141570147523714,
+      "learning_rate": 0.00024828597951993093,
+      "loss": 2.8578,
+      "step": 1478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6434832273580102,
+      "learning_rate": 0.00024764423954595706,
+      "loss": 2.9379,
+      "step": 1479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5204860209520548,
+      "learning_rate": 0.0002470030568897938,
+      "loss": 2.8826,
+      "step": 1480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5979289471209772,
+      "learning_rate": 0.00024636243296746773,
+      "loss": 2.9543,
+      "step": 1481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.588733270048087,
+      "learning_rate": 0.0002457223691937716,
+      "loss": 2.9174,
+      "step": 1482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5100541831476932,
+      "learning_rate": 0.0002450828669822613,
+      "loss": 2.8684,
+      "step": 1483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5086001868370686,
+      "learning_rate": 0.00024444392774525253,
+      "loss": 2.8994,
+      "step": 1484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.450782648551871,
+      "learning_rate": 0.00024380555289381733,
+      "loss": 3.0069,
+      "step": 1485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4447054638513908,
+      "learning_rate": 0.00024316774383778184,
+      "loss": 2.8727,
+      "step": 1486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4898841541000822,
+      "learning_rate": 0.0002425305019857222,
+      "loss": 2.8744,
+      "step": 1487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6993199675387893,
+      "learning_rate": 0.00024189382874496184,
+      "loss": 3.0014,
+      "step": 1488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.650859869626695,
+      "learning_rate": 0.00024125772552156916,
+      "loss": 3.0137,
+      "step": 1489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4260351249485441,
+      "learning_rate": 0.00024062219372035292,
+      "loss": 2.9275,
+      "step": 1490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4587359361733256,
+      "learning_rate": 0.00023998723474486007,
+      "loss": 2.9233,
+      "step": 1491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5967972434990227,
+      "learning_rate": 0.00023935284999737272,
+      "loss": 2.9578,
+      "step": 1492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5140567377845602,
+      "learning_rate": 0.00023871904087890505,
+      "loss": 2.9578,
+      "step": 1493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4890697612198782,
+      "learning_rate": 0.00023808580878919945,
+      "loss": 2.9735,
+      "step": 1494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.517471960823647,
+      "learning_rate": 0.00023745315512672398,
+      "loss": 2.9365,
+      "step": 1495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4814341341178476,
+      "learning_rate": 0.0002368210812886698,
+      "loss": 2.8808,
+      "step": 1496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.450001509621225,
+      "learning_rate": 0.0002361895886709471,
+      "loss": 3.0162,
+      "step": 1497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.651193932871942,
+      "learning_rate": 0.0002355586786681823,
+      "loss": 2.9472,
+      "step": 1498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5500637448824848,
+      "learning_rate": 0.00023492835267371575,
+      "loss": 2.8926,
+      "step": 1499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.569200226775873,
+      "learning_rate": 0.0002342986120795978,
+      "loss": 2.9428,
+      "step": 1500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.437968889664695,
+      "learning_rate": 0.0002336694582765857,
+      "loss": 2.8881,
+      "step": 1501
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5552014979087858,
+      "learning_rate": 0.00023304089265414085,
+      "loss": 2.9632,
+      "step": 1502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4052475836841227,
+      "learning_rate": 0.00023241291660042613,
+      "loss": 2.9597,
+      "step": 1503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.752966646231565,
+      "learning_rate": 0.00023178553150230186,
+      "loss": 2.9798,
+      "step": 1504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4240897380123736,
+      "learning_rate": 0.00023115873874532324,
+      "loss": 2.8939,
+      "step": 1505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.688427991877061,
+      "learning_rate": 0.00023053253971373796,
+      "loss": 3.0191,
+      "step": 1506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.745064967939218,
+      "learning_rate": 0.00022990693579048166,
+      "loss": 2.8252,
+      "step": 1507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6130176231790239,
+      "learning_rate": 0.00022928192835717644,
+      "loss": 2.9054,
+      "step": 1508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.604152498608473,
+      "learning_rate": 0.00022865751879412634,
+      "loss": 2.9609,
+      "step": 1509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9597197585371093,
+      "learning_rate": 0.00022803370848031585,
+      "loss": 2.9919,
+      "step": 1510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4874356965176743,
+      "learning_rate": 0.00022741049879340542,
+      "loss": 2.9606,
+      "step": 1511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.744576593036428,
+      "learning_rate": 0.00022678789110972897,
+      "loss": 2.7889,
+      "step": 1512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.166068197922419,
+      "learning_rate": 0.00022616588680429155,
+      "loss": 2.9004,
+      "step": 1513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.578302224919738,
+      "learning_rate": 0.00022554448725076526,
+      "loss": 3.0183,
+      "step": 1514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5961367770030592,
+      "learning_rate": 0.0002249236938214863,
+      "loss": 2.9501,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.832155566828616,
+      "learning_rate": 0.00022430350788745296,
+      "loss": 2.9671,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6432574971173606,
+      "learning_rate": 0.00022368393081832166,
+      "loss": 3.0415,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4825851650348905,
+      "learning_rate": 0.00022306496398240383,
+      "loss": 2.9573,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8309804675550556,
+      "learning_rate": 0.00022244660874666373,
+      "loss": 2.8755,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5143518258849638,
+      "learning_rate": 0.00022182886647671452,
+      "loss": 2.9055,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6717227252189408,
+      "learning_rate": 0.0002212117385368157,
+      "loss": 2.8681,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6492430589357554,
+      "learning_rate": 0.00022059522628987038,
+      "loss": 2.8652,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7461307117025051,
+      "learning_rate": 0.00021997933109742162,
+      "loss": 2.9148,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5238307246788927,
+      "learning_rate": 0.00021936405431964969,
+      "loss": 2.9439,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.606684000169697,
+      "learning_rate": 0.00021874939731536926,
+      "loss": 2.8118,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6469402909653346,
+      "learning_rate": 0.00021813536144202656,
+      "loss": 2.8216,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4473613107756589,
+      "learning_rate": 0.00021752194805569553,
+      "loss": 2.8592,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9210256471384464,
+      "learning_rate": 0.0002169091585110754,
+      "loss": 2.8887,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6824266702541943,
+      "learning_rate": 0.00021629699416148828,
+      "loss": 2.8762,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5276484237437646,
+      "learning_rate": 0.000215685456358875,
+      "loss": 3.0055,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.650967430573702,
+      "learning_rate": 0.00021507454645379258,
+      "loss": 2.9529,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.467836797235011,
+      "learning_rate": 0.00021446426579541184,
+      "loss": 2.9541,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5184403000546254,
+      "learning_rate": 0.00021385461573151387,
+      "loss": 2.9063,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4755613221120354,
+      "learning_rate": 0.00021324559760848677,
+      "loss": 2.9334,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8435316952699927,
+      "learning_rate": 0.00021263721277132303,
+      "loss": 2.9096,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5528816049942322,
+      "learning_rate": 0.0002120294625636171,
+      "loss": 2.9324,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8022472109399794,
+      "learning_rate": 0.0002114223483275613,
+      "loss": 2.8272,
+      "step": 1537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6607380397507938,
+      "learning_rate": 0.0002108158714039435,
+      "loss": 2.9731,
+      "step": 1538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6280258291651917,
+      "learning_rate": 0.00021021003313214455,
+      "loss": 3.0223,
+      "step": 1539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4846970176346543,
+      "learning_rate": 0.00020960483485013432,
+      "loss": 2.9022,
+      "step": 1540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.627720605792389,
+      "learning_rate": 0.0002090002778944694,
+      "loss": 2.9267,
+      "step": 1541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4616844668619664,
+      "learning_rate": 0.00020839636360029025,
+      "loss": 2.9249,
+      "step": 1542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.000394979488673,
+      "learning_rate": 0.00020779309330131818,
+      "loss": 2.9365,
+      "step": 1543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.563939551361803,
+      "learning_rate": 0.00020719046832985184,
+      "loss": 2.9283,
+      "step": 1544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.575648692892011,
+      "learning_rate": 0.0002065884900167646,
+      "loss": 2.8973,
+      "step": 1545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7836966390021012,
+      "learning_rate": 0.0002059871596915024,
+      "loss": 2.9627,
+      "step": 1546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5091666644804809,
+      "learning_rate": 0.0002053864786820795,
+      "loss": 3.0173,
+      "step": 1547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5042713191277945,
+      "learning_rate": 0.00020478644831507627,
+      "loss": 2.7886,
+      "step": 1548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3850664676755933,
+      "learning_rate": 0.00020418706991563634,
+      "loss": 2.935,
+      "step": 1549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5453125689425402,
+      "learning_rate": 0.00020358834480746363,
+      "loss": 2.9125,
+      "step": 1550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.709324975485777,
+      "learning_rate": 0.0002029902743128188,
+      "loss": 2.9848,
+      "step": 1551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4631976550816435,
+      "learning_rate": 0.0002023928597525174,
+      "loss": 2.9102,
+      "step": 1552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6145948870160833,
+      "learning_rate": 0.00020179610244592595,
+      "loss": 2.9375,
+      "step": 1553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4586780223654032,
+      "learning_rate": 0.00020120000371095937,
+      "loss": 2.8557,
+      "step": 1554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5939331562291752,
+      "learning_rate": 0.0002006045648640787,
+      "loss": 2.8535,
+      "step": 1555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4288314650899843,
+      "learning_rate": 0.00020000978722028713,
+      "loss": 2.9021,
+      "step": 1556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.910083420323651,
+      "learning_rate": 0.00019941567209312767,
+      "loss": 3.052,
+      "step": 1557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4932168198758526,
+      "learning_rate": 0.00019882222079468036,
+      "loss": 2.9266,
+      "step": 1558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8962489584561792,
+      "learning_rate": 0.0001982294346355595,
+      "loss": 2.8418,
+      "step": 1559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6523693164888842,
+      "learning_rate": 0.00019763731492490976,
+      "loss": 2.9588,
+      "step": 1560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4595715914312297,
+      "learning_rate": 0.00019704586297040422,
+      "loss": 2.8439,
+      "step": 1561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5774161399172488,
+      "learning_rate": 0.0001964550800782417,
+      "loss": 2.9274,
+      "step": 1562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7287881607374254,
+      "learning_rate": 0.00019586496755314288,
+      "loss": 2.8835,
+      "step": 1563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5696691326363457,
+      "learning_rate": 0.00019527552669834798,
+      "loss": 2.8207,
+      "step": 1564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.389051887615987,
+      "learning_rate": 0.0001946867588156142,
+      "loss": 2.8989,
+      "step": 1565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5107964847271647,
+      "learning_rate": 0.00019409866520521258,
+      "loss": 2.8306,
+      "step": 1566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1157399283355014,
+      "learning_rate": 0.00019351124716592455,
+      "loss": 2.9169,
+      "step": 1567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7481227025648727,
+      "learning_rate": 0.0001929245059950397,
+      "loss": 2.8405,
+      "step": 1568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5780464703416155,
+      "learning_rate": 0.0001923384429883533,
+      "loss": 2.9657,
+      "step": 1569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7998906124677252,
+      "learning_rate": 0.00019175305944016237,
+      "loss": 2.8179,
+      "step": 1570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5225235916594257,
+      "learning_rate": 0.00019116835664326326,
+      "loss": 2.9237,
+      "step": 1571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.695880977798995,
+      "learning_rate": 0.0001905843358889497,
+      "loss": 2.9431,
+      "step": 1572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8030127938241327,
+      "learning_rate": 0.00019000099846700836,
+      "loss": 2.9851,
+      "step": 1573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.111696332321112,
+      "learning_rate": 0.00018941834566571692,
+      "loss": 2.8494,
+      "step": 1574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4777502826776756,
+      "learning_rate": 0.00018883637877184145,
+      "loss": 2.8052,
+      "step": 1575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5340450874789968,
+      "learning_rate": 0.00018825509907063325,
+      "loss": 2.9251,
+      "step": 1576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6177474500874722,
+      "learning_rate": 0.00018767450784582557,
+      "loss": 2.9518,
+      "step": 1577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1832012551531763,
+      "learning_rate": 0.00018709460637963122,
+      "loss": 2.9006,
+      "step": 1578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.532182723676707,
+      "learning_rate": 0.00018651539595274013,
+      "loss": 2.8299,
+      "step": 1579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7411982492102012,
+      "learning_rate": 0.00018593687784431578,
+      "loss": 2.9114,
+      "step": 1580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.575710696257478,
+      "learning_rate": 0.00018535905333199248,
+      "loss": 2.8119,
+      "step": 1581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5345858883161672,
+      "learning_rate": 0.0001847819236918733,
+      "loss": 2.874,
+      "step": 1582
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7507909903485182,
+      "learning_rate": 0.00018420549019852655,
+      "loss": 2.9765,
+      "step": 1583
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.614853851938068,
+      "learning_rate": 0.00018362975412498266,
+      "loss": 2.9586,
+      "step": 1584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7071210344986194,
+      "learning_rate": 0.00018305471674273261,
+      "loss": 2.8786,
+      "step": 1585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.523657578901562,
+      "learning_rate": 0.0001824803793217237,
+      "loss": 2.8041,
+      "step": 1586
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.728578902872943,
+      "learning_rate": 0.00018190674313035737,
+      "loss": 2.9156,
+      "step": 1587
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6171334879023633,
+      "learning_rate": 0.00018133380943548716,
+      "loss": 2.8948,
+      "step": 1588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.32734164031763,
+      "learning_rate": 0.00018076157950241452,
+      "loss": 2.9072,
+      "step": 1589
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.437750324550463,
+      "learning_rate": 0.00018019005459488652,
+      "loss": 2.9262,
+      "step": 1590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0576421421126807,
+      "learning_rate": 0.00017961923597509388,
+      "loss": 3.0327,
+      "step": 1591
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7597478662949821,
+      "learning_rate": 0.00017904912490366722,
+      "loss": 2.9316,
+      "step": 1592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4829153680635956,
+      "learning_rate": 0.00017847972263967433,
+      "loss": 3.0064,
+      "step": 1593
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5011810954688871,
+      "learning_rate": 0.0001779110304406177,
+      "loss": 2.9352,
+      "step": 1594
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7309444313142408,
+      "learning_rate": 0.000177343049562432,
+      "loss": 2.8138,
+      "step": 1595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.861315362155302,
+      "learning_rate": 0.0001767757812594807,
+      "loss": 2.9014,
+      "step": 1596
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5750567234687645,
+      "learning_rate": 0.0001762092267845534,
+      "loss": 2.8707,
+      "step": 1597
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6217684758113966,
+      "learning_rate": 0.00017564338738886365,
+      "loss": 2.8933,
+      "step": 1598
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7280969565460895,
+      "learning_rate": 0.0001750782643220457,
+      "loss": 3.0143,
+      "step": 1599
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5857575509252646,
+      "learning_rate": 0.00017451385883215166,
+      "loss": 3.0419,
+      "step": 1600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6638601047633876,
+      "learning_rate": 0.00017395017216564863,
+      "loss": 2.789,
+      "step": 1601
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3841660994377334,
+      "learning_rate": 0.00017338720556741687,
+      "loss": 2.915,
+      "step": 1602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7239488480074179,
+      "learning_rate": 0.00017282496028074606,
+      "loss": 3.0139,
+      "step": 1603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.586758145914941,
+      "learning_rate": 0.00017226343754733254,
+      "loss": 2.8973,
+      "step": 1604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3966059783732871,
+      "learning_rate": 0.00017170263860727769,
+      "loss": 2.9135,
+      "step": 1605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5532350617147523,
+      "learning_rate": 0.0001711425646990838,
+      "loss": 2.8298,
+      "step": 1606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4590370493934266,
+      "learning_rate": 0.00017058321705965202,
+      "loss": 2.9054,
+      "step": 1607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6112303189506407,
+      "learning_rate": 0.0001700245969242798,
+      "loss": 2.9216,
+      "step": 1608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5009635877418905,
+      "learning_rate": 0.00016946670552665804,
+      "loss": 2.8847,
+      "step": 1609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6609446837121848,
+      "learning_rate": 0.00016890954409886795,
+      "loss": 2.8736,
+      "step": 1610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.519152944517092,
+      "learning_rate": 0.00016835311387137836,
+      "loss": 2.8795,
+      "step": 1611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5909878593499307,
+      "learning_rate": 0.0001677974160730441,
+      "loss": 2.9307,
+      "step": 1612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5276057630215325,
+      "learning_rate": 0.00016724245193110176,
+      "loss": 2.9395,
+      "step": 1613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7070727050989027,
+      "learning_rate": 0.00016668822267116784,
+      "loss": 2.9137,
+      "step": 1614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5827009712747229,
+      "learning_rate": 0.00016613472951723597,
+      "loss": 2.899,
+      "step": 1615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.47808418583521,
+      "learning_rate": 0.00016558197369167434,
+      "loss": 2.8818,
+      "step": 1616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7738888909045838,
+      "learning_rate": 0.00016502995641522216,
+      "loss": 2.9539,
+      "step": 1617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5086267462756844,
+      "learning_rate": 0.00016447867890698843,
+      "loss": 2.9097,
+      "step": 1618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.643142335490357,
+      "learning_rate": 0.00016392814238444753,
+      "loss": 2.9378,
+      "step": 1619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.559455600665628,
+      "learning_rate": 0.00016337834806343782,
+      "loss": 2.9154,
+      "step": 1620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6358421942624877,
+      "learning_rate": 0.0001628292971581588,
+      "loss": 2.8887,
+      "step": 1621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6446261862061582,
+      "learning_rate": 0.00016228099088116772,
+      "loss": 2.9826,
+      "step": 1622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5204528148467293,
+      "learning_rate": 0.00016173343044337734,
+      "loss": 2.9719,
+      "step": 1623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.433630866245793,
+      "learning_rate": 0.00016118661705405356,
+      "loss": 2.8579,
+      "step": 1624
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6375924863070221,
+      "learning_rate": 0.00016064055192081255,
+      "loss": 2.9417,
+      "step": 1625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.679973279026685,
+      "learning_rate": 0.00016009523624961757,
+      "loss": 2.8753,
+      "step": 1626
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.483922092710041,
+      "learning_rate": 0.00015955067124477678,
+      "loss": 2.8357,
+      "step": 1627
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5670847713738632,
+      "learning_rate": 0.000159006858108941,
+      "loss": 2.9255,
+      "step": 1628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5438434444322824,
+      "learning_rate": 0.00015846379804310002,
+      "loss": 2.847,
+      "step": 1629
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.519205816778686,
+      "learning_rate": 0.00015792149224658054,
+      "loss": 2.7868,
+      "step": 1630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5657500381818303,
+      "learning_rate": 0.00015737994191704385,
+      "loss": 2.9517,
+      "step": 1631
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4748546401143636,
+      "learning_rate": 0.0001568391482504829,
+      "loss": 2.8237,
+      "step": 1632
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6265381687914875,
+      "learning_rate": 0.00015629911244121903,
+      "loss": 2.8586,
+      "step": 1633
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7721597604931751,
+      "learning_rate": 0.0001557598356819,
+      "loss": 3.0244,
+      "step": 1634
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.545613028964107,
+      "learning_rate": 0.00015522131916349786,
+      "loss": 2.8882,
+      "step": 1635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6877204259992578,
+      "learning_rate": 0.00015468356407530493,
+      "loss": 2.9623,
+      "step": 1636
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.798094377562649,
+      "learning_rate": 0.00015414657160493217,
+      "loss": 2.9509,
+      "step": 1637
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6644244676181663,
+      "learning_rate": 0.00015361034293830673,
+      "loss": 2.8928,
+      "step": 1638
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4621880593737615,
+      "learning_rate": 0.00015307487925966844,
+      "loss": 2.9008,
+      "step": 1639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.358337526288481,
+      "learning_rate": 0.00015254018175156776,
+      "loss": 2.9265,
+      "step": 1640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4840529064077292,
+      "learning_rate": 0.0001520062515948632,
+      "loss": 2.971,
+      "step": 1641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5610152977931466,
+      "learning_rate": 0.0001514730899687189,
+      "loss": 2.802,
+      "step": 1642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6841632970142366,
+      "learning_rate": 0.00015094069805060122,
+      "loss": 2.8862,
+      "step": 1643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6855813547037304,
+      "learning_rate": 0.00015040907701627666,
+      "loss": 2.8757,
+      "step": 1644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8803335736362743,
+      "learning_rate": 0.00014987822803980976,
+      "loss": 2.7767,
+      "step": 1645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3900473671360671,
+      "learning_rate": 0.00014934815229355965,
+      "loss": 2.9385,
+      "step": 1646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6285585480877767,
+      "learning_rate": 0.00014881885094817748,
+      "loss": 2.9041,
+      "step": 1647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5059411693406595,
+      "learning_rate": 0.00014829032517260488,
+      "loss": 2.9726,
+      "step": 1648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8445417361533605,
+      "learning_rate": 0.0001477625761340704,
+      "loss": 2.9602,
+      "step": 1649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4517945894574233,
+      "learning_rate": 0.0001472356049980868,
+      "loss": 2.921,
+      "step": 1650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4265980897705626,
+      "learning_rate": 0.00014670941292844954,
+      "loss": 2.875,
+      "step": 1651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5326766220182233,
+      "learning_rate": 0.00014618400108723295,
+      "loss": 2.8267,
+      "step": 1652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5068664458405445,
+      "learning_rate": 0.00014565937063478862,
+      "loss": 2.7897,
+      "step": 1653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.58435539745169,
+      "learning_rate": 0.00014513552272974207,
+      "loss": 2.8901,
+      "step": 1654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5665403237314484,
+      "learning_rate": 0.0001446124585289913,
+      "loss": 2.8437,
+      "step": 1655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.586843478226266,
+      "learning_rate": 0.00014409017918770266,
+      "loss": 2.8545,
+      "step": 1656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7496592215711018,
+      "learning_rate": 0.00014356868585930994,
+      "loss": 2.9694,
+      "step": 1657
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3959892592018024,
+      "learning_rate": 0.00014304797969551077,
+      "loss": 2.8241,
+      "step": 1658
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6125474060568383,
+      "learning_rate": 0.00014252806184626417,
+      "loss": 3.1002,
+      "step": 1659
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5450364026306962,
+      "learning_rate": 0.00014200893345978817,
+      "loss": 2.992,
+      "step": 1660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5224512872551674,
+      "learning_rate": 0.00014149059568255778,
+      "loss": 2.9441,
+      "step": 1661
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5244098666709878,
+      "learning_rate": 0.00014097304965930157,
+      "loss": 2.8768,
+      "step": 1662
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4027130207903098,
+      "learning_rate": 0.00014045629653299953,
+      "loss": 3.0016,
+      "step": 1663
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7625494342293935,
+      "learning_rate": 0.00013994033744488076,
+      "loss": 2.9396,
+      "step": 1664
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.683836826538522,
+      "learning_rate": 0.00013942517353442092,
+      "loss": 2.8211,
+      "step": 1665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5839915076311049,
+      "learning_rate": 0.0001389108059393391,
+      "loss": 2.9157,
+      "step": 1666
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6747591079529167,
+      "learning_rate": 0.00013839723579559581,
+      "loss": 2.9708,
+      "step": 1667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.41812529521965,
+      "learning_rate": 0.00013788446423739103,
+      "loss": 2.8943,
+      "step": 1668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.580058782406527,
+      "learning_rate": 0.00013737249239716042,
+      "loss": 2.9412,
+      "step": 1669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5134875188574084,
+      "learning_rate": 0.00013686132140557355,
+      "loss": 2.9541,
+      "step": 1670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3745378742669694,
+      "learning_rate": 0.00013635095239153188,
+      "loss": 2.9256,
+      "step": 1671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.487168866451424,
+      "learning_rate": 0.00013584138648216527,
+      "loss": 2.9117,
+      "step": 1672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7386917322542212,
+      "learning_rate": 0.0001353326248028298,
+      "loss": 2.9746,
+      "step": 1673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5601361451958473,
+      "learning_rate": 0.00013482466847710594,
+      "loss": 2.9341,
+      "step": 1674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5061068243284967,
+      "learning_rate": 0.00013431751862679554,
+      "loss": 2.9249,
+      "step": 1675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.513311050371285,
+      "learning_rate": 0.00013381117637191887,
+      "loss": 2.7799,
+      "step": 1676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5111148602147533,
+      "learning_rate": 0.00013330564283071293,
+      "loss": 2.8092,
+      "step": 1677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6107214341801916,
+      "learning_rate": 0.000132800919119629,
+      "loss": 2.8755,
+      "step": 1678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.509441281914715,
+      "learning_rate": 0.00013229700635332948,
+      "loss": 2.8738,
+      "step": 1679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6536222625564403,
+      "learning_rate": 0.00013179390564468585,
+      "loss": 2.9368,
+      "step": 1680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.511809290433128,
+      "learning_rate": 0.00013129161810477641,
+      "loss": 2.9024,
+      "step": 1681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5402014064134297,
+      "learning_rate": 0.0001307901448428837,
+      "loss": 2.8821,
+      "step": 1682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5663699439805387,
+      "learning_rate": 0.0001302894869664916,
+      "loss": 2.9421,
+      "step": 1683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.576602894458502,
+      "learning_rate": 0.00012978964558128336,
+      "loss": 2.8685,
+      "step": 1684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5108807180744428,
+      "learning_rate": 0.00012929062179113925,
+      "loss": 2.8077,
+      "step": 1685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4045332348248678,
+      "learning_rate": 0.00012879241669813368,
+      "loss": 2.8314,
+      "step": 1686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7210562594102605,
+      "learning_rate": 0.00012829503140253295,
+      "loss": 3.0356,
+      "step": 1687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4940881488895021,
+      "learning_rate": 0.0001277984670027933,
+      "loss": 2.9574,
+      "step": 1688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.635521301159955,
+      "learning_rate": 0.00012730272459555737,
+      "loss": 2.9551,
+      "step": 1689
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3472973823519743,
+      "learning_rate": 0.00012680780527565312,
+      "loss": 2.8083,
+      "step": 1690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5619240834880797,
+      "learning_rate": 0.0001263137101360905,
+      "loss": 2.7986,
+      "step": 1691
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.529823045395325,
+      "learning_rate": 0.00012582044026805922,
+      "loss": 2.8773,
+      "step": 1692
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6592609098529698,
+      "learning_rate": 0.00012532799676092627,
+      "loss": 2.8762,
+      "step": 1693
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.419089692060868,
+      "learning_rate": 0.00012483638070223414,
+      "loss": 2.9733,
+      "step": 1694
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6183634427750329,
+      "learning_rate": 0.00012434559317769752,
+      "loss": 2.9078,
+      "step": 1695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.516269599222086,
+      "learning_rate": 0.0001238556352712012,
+      "loss": 2.8798,
+      "step": 1696
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.597625436196857,
+      "learning_rate": 0.00012336650806479827,
+      "loss": 2.9309,
+      "step": 1697
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.478555242564991,
+      "learning_rate": 0.00012287821263870708,
+      "loss": 2.9219,
+      "step": 1698
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4110940840225143,
+      "learning_rate": 0.00012239075007130885,
+      "loss": 2.9342,
+      "step": 1699
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.921132590916206,
+      "learning_rate": 0.00012190412143914536,
+      "loss": 2.7877,
+      "step": 1700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5330436055352636,
+      "learning_rate": 0.0001214183278169172,
+      "loss": 2.7852,
+      "step": 1701
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.539547454393412,
+      "learning_rate": 0.00012093337027748042,
+      "loss": 3.0024,
+      "step": 1702
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6121432009990502,
+      "learning_rate": 0.00012044924989184459,
+      "loss": 2.8892,
+      "step": 1703
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5427779147110452,
+      "learning_rate": 0.0001199659677291709,
+      "loss": 2.8706,
+      "step": 1704
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5452542751091647,
+      "learning_rate": 0.00011948352485676895,
+      "loss": 2.8838,
+      "step": 1705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.662179469165797,
+      "learning_rate": 0.00011900192234009477,
+      "loss": 2.9066,
+      "step": 1706
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6371713518535134,
+      "learning_rate": 0.00011852116124274875,
+      "loss": 2.8313,
+      "step": 1707
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4986332253826837,
+      "learning_rate": 0.00011804124262647314,
+      "loss": 2.8912,
+      "step": 1708
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7907160981281278,
+      "learning_rate": 0.00011756216755114929,
+      "loss": 2.815,
+      "step": 1709
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4984846145176856,
+      "learning_rate": 0.00011708393707479548,
+      "loss": 3.0168,
+      "step": 1710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4514266005027268,
+      "learning_rate": 0.00011660655225356531,
+      "loss": 2.8368,
+      "step": 1711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3877687990874763,
+      "learning_rate": 0.0001161300141417444,
+      "loss": 2.8961,
+      "step": 1712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8380167903296862,
+      "learning_rate": 0.00011565432379174823,
+      "loss": 2.9601,
+      "step": 1713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5126775314023262,
+      "learning_rate": 0.00011517948225412056,
+      "loss": 2.9045,
+      "step": 1714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.527232630994529,
+      "learning_rate": 0.00011470549057753032,
+      "loss": 2.9585,
+      "step": 1715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5798740086888097,
+      "learning_rate": 0.00011423234980876957,
+      "loss": 2.9607,
+      "step": 1716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4461310825360298,
+      "learning_rate": 0.00011376006099275099,
+      "loss": 2.8605,
+      "step": 1717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.499385623515957,
+      "learning_rate": 0.00011328862517250609,
+      "loss": 2.9519,
+      "step": 1718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9586181778001877,
+      "learning_rate": 0.00011281804338918239,
+      "loss": 2.8806,
+      "step": 1719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.709820096488973,
+      "learning_rate": 0.00011234831668204115,
+      "loss": 2.8707,
+      "step": 1720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6331323045655075,
+      "learning_rate": 0.00011187944608845569,
+      "loss": 2.872,
+      "step": 1721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8863662970276862,
+      "learning_rate": 0.00011141143264390801,
+      "loss": 2.8942,
+      "step": 1722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.445568514081014,
+      "learning_rate": 0.0001109442773819877,
+      "loss": 2.8552,
+      "step": 1723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.592066482222681,
+      "learning_rate": 0.0001104779813343889,
+      "loss": 2.8929,
+      "step": 1724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5415136031664014,
+      "learning_rate": 0.00011001254553090812,
+      "loss": 2.9044,
+      "step": 1725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6010988799443686,
+      "learning_rate": 0.00010954797099944186,
+      "loss": 2.9785,
+      "step": 1726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.548698993031613,
+      "learning_rate": 0.0001090842587659851,
+      "loss": 2.8443,
+      "step": 1727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6164203368317187,
+      "learning_rate": 0.00010862140985462804,
+      "loss": 2.9396,
+      "step": 1728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9225816880260593,
+      "learning_rate": 0.00010815942528755418,
+      "loss": 2.8465,
+      "step": 1729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.695252599829415,
+      "learning_rate": 0.00010769830608503844,
+      "loss": 2.9456,
+      "step": 1730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5419947754298884,
+      "learning_rate": 0.00010723805326544473,
+      "loss": 2.8972,
+      "step": 1731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6999706416241236,
+      "learning_rate": 0.00010677866784522316,
+      "loss": 2.937,
+      "step": 1732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.87491038759697,
+      "learning_rate": 0.00010632015083890839,
+      "loss": 2.8191,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8629223548362634,
+      "learning_rate": 0.00010586250325911745,
+      "loss": 3.0387,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5335312884310106,
+      "learning_rate": 0.00010540572611654697,
+      "loss": 2.9246,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.752769651710078,
+      "learning_rate": 0.00010494982041997126,
+      "loss": 2.9615,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1028067970488884,
+      "learning_rate": 0.0001044947871762405,
+      "loss": 2.9539,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6767359413207843,
+      "learning_rate": 0.00010404062739027753,
+      "loss": 2.9569,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.583568345217933,
+      "learning_rate": 0.00010358734206507641,
+      "loss": 2.9653,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.514755687555455,
+      "learning_rate": 0.00010313493220170017,
+      "loss": 2.8421,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6080656389959587,
+      "learning_rate": 0.00010268339879927836,
+      "loss": 2.8244,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.614028777725661,
+      "learning_rate": 0.00010223274285500466,
+      "loss": 2.8707,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8260915989587525,
+      "learning_rate": 0.00010178296536413495,
+      "loss": 2.9182,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6750020476693457,
+      "learning_rate": 0.00010133406731998546,
+      "loss": 2.8792,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5352557401713172,
+      "learning_rate": 0.00010088604971392979,
+      "loss": 2.8707,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7243450429274425,
+      "learning_rate": 0.0001004389135353972,
+      "loss": 2.8954,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7043794793147335,
+      "learning_rate": 9.999265977187049e-05,
+      "loss": 2.9321,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6714793126213972,
+      "learning_rate": 9.95472894088838e-05,
+      "loss": 2.9256,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5300341619993458,
+      "learning_rate": 9.910280343001993e-05,
+      "loss": 2.8286,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3575088995880071,
+      "learning_rate": 9.865920281690866e-05,
+      "loss": 2.862,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5453272322241085,
+      "learning_rate": 9.821648854922482e-05,
+      "loss": 2.8941,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.750629370964988,
+      "learning_rate": 9.77746616046854e-05,
+      "loss": 2.8958,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4757326829297304,
+      "learning_rate": 9.733372295904774e-05,
+      "loss": 3.0011,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6956442097159874,
+      "learning_rate": 9.68936735861079e-05,
+      "loss": 2.9447,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.580154417548426,
+      "learning_rate": 9.645451445769737e-05,
+      "loss": 3.089,
+      "step": 1755
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6635383634424994,
+      "learning_rate": 9.601624654368196e-05,
+      "loss": 2.8492,
+      "step": 1756
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5738361122052533,
+      "learning_rate": 9.557887081195938e-05,
+      "loss": 2.8009,
+      "step": 1757
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8646360887831783,
+      "learning_rate": 9.514238822845667e-05,
+      "loss": 3.0324,
+      "step": 1758
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.586247013285444,
+      "learning_rate": 9.470679975712837e-05,
+      "loss": 2.8533,
+      "step": 1759
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.613815399473487,
+      "learning_rate": 9.427210635995481e-05,
+      "loss": 2.8503,
+      "step": 1760
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7137368654002059,
+      "learning_rate": 9.383830899693923e-05,
+      "loss": 2.867,
+      "step": 1761
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5334860183348806,
+      "learning_rate": 9.340540862610591e-05,
+      "loss": 2.8407,
+      "step": 1762
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.601787332688766,
+      "learning_rate": 9.297340620349854e-05,
+      "loss": 2.9719,
+      "step": 1763
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5250007401888728,
+      "learning_rate": 9.25423026831777e-05,
+      "loss": 2.9208,
+      "step": 1764
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3900540110200232,
+      "learning_rate": 9.211209901721846e-05,
+      "loss": 2.8442,
+      "step": 1765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6844990110018245,
+      "learning_rate": 9.168279615570863e-05,
+      "loss": 2.8239,
+      "step": 1766
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4587256157588673,
+      "learning_rate": 9.125439504674699e-05,
+      "loss": 2.9497,
+      "step": 1767
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3959046856506658,
+      "learning_rate": 9.082689663644057e-05,
+      "loss": 2.8657,
+      "step": 1768
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5826517891517,
+      "learning_rate": 9.040030186890264e-05,
+      "loss": 2.8274,
+      "step": 1769
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4781139148736993,
+      "learning_rate": 8.997461168625138e-05,
+      "loss": 2.9306,
+      "step": 1770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6190752880979076,
+      "learning_rate": 8.954982702860664e-05,
+      "loss": 2.9977,
+      "step": 1771
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4755709140197775,
+      "learning_rate": 8.912594883408865e-05,
+      "loss": 2.8362,
+      "step": 1772
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5890228290030042,
+      "learning_rate": 8.870297803881589e-05,
+      "loss": 2.9632,
+      "step": 1773
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6943884902662936,
+      "learning_rate": 8.828091557690287e-05,
+      "loss": 2.876,
+      "step": 1774
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4856510369826468,
+      "learning_rate": 8.785976238045801e-05,
+      "loss": 2.8697,
+      "step": 1775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5433221480253105,
+      "learning_rate": 8.743951937958144e-05,
+      "loss": 2.7984,
+      "step": 1776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5224457354670955,
+      "learning_rate": 8.702018750236357e-05,
+      "loss": 2.9615,
+      "step": 1777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5597501652803716,
+      "learning_rate": 8.660176767488237e-05,
+      "loss": 2.931,
+      "step": 1778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5722093616683257,
+      "learning_rate": 8.618426082120146e-05,
+      "loss": 2.8732,
+      "step": 1779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.51975880899728,
+      "learning_rate": 8.576766786336854e-05,
+      "loss": 2.8118,
+      "step": 1780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.512074761380704,
+      "learning_rate": 8.535198972141294e-05,
+      "loss": 2.8316,
+      "step": 1781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4959274026481508,
+      "learning_rate": 8.493722731334347e-05,
+      "loss": 2.8902,
+      "step": 1782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6555416336207913,
+      "learning_rate": 8.452338155514644e-05,
+      "loss": 2.9818,
+      "step": 1783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5901052154532949,
+      "learning_rate": 8.411045336078426e-05,
+      "loss": 2.9397,
+      "step": 1784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5764190413410712,
+      "learning_rate": 8.369844364219264e-05,
+      "loss": 2.9789,
+      "step": 1785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5951642266867696,
+      "learning_rate": 8.328735330927873e-05,
+      "loss": 2.9014,
+      "step": 1786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4422853720723146,
+      "learning_rate": 8.287718326991961e-05,
+      "loss": 2.8608,
+      "step": 1787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4343996398897358,
+      "learning_rate": 8.246793442995954e-05,
+      "loss": 2.9919,
+      "step": 1788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5862216339296258,
+      "learning_rate": 8.205960769320875e-05,
+      "loss": 2.898,
+      "step": 1789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4328021282398504,
+      "learning_rate": 8.165220396144085e-05,
+      "loss": 2.8186,
+      "step": 1790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.916942656429615,
+      "learning_rate": 8.12457241343909e-05,
+      "loss": 3.0028,
+      "step": 1791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4553769353501413,
+      "learning_rate": 8.084016910975367e-05,
+      "loss": 2.8874,
+      "step": 1792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4799805568806625,
+      "learning_rate": 8.043553978318169e-05,
+      "loss": 2.8345,
+      "step": 1793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5837950018161324,
+      "learning_rate": 8.003183704828281e-05,
+      "loss": 2.8932,
+      "step": 1794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5443321586115653,
+      "learning_rate": 7.962906179661872e-05,
+      "loss": 2.927,
+      "step": 1795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5221482226062086,
+      "learning_rate": 7.922721491770296e-05,
+      "loss": 2.881,
+      "step": 1796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7582933354007764,
+      "learning_rate": 7.882629729899832e-05,
+      "loss": 2.9287,
+      "step": 1797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5589560707685162,
+      "learning_rate": 7.842630982591598e-05,
+      "loss": 2.8217,
+      "step": 1798
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.604478320749652,
+      "learning_rate": 7.802725338181232e-05,
+      "loss": 2.8382,
+      "step": 1799
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.86252010602163,
+      "learning_rate": 7.762912884798812e-05,
+      "loss": 2.911,
+      "step": 1800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6849111218115989,
+      "learning_rate": 7.723193710368564e-05,
+      "loss": 2.9376,
+      "step": 1801
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4083371334004482,
+      "learning_rate": 7.683567902608729e-05,
+      "loss": 2.7512,
+      "step": 1802
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4270790964847921,
+      "learning_rate": 7.644035549031364e-05,
+      "loss": 2.9882,
+      "step": 1803
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4720457988638307,
+      "learning_rate": 7.604596736942115e-05,
+      "loss": 2.8556,
+      "step": 1804
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6295745081018895,
+      "learning_rate": 7.56525155344004e-05,
+      "loss": 2.8981,
+      "step": 1805
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.482797567407719,
+      "learning_rate": 7.52600008541745e-05,
+      "loss": 2.8622,
+      "step": 1806
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5472292629387696,
+      "learning_rate": 7.486842419559681e-05,
+      "loss": 2.9227,
+      "step": 1807
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.499925742939567,
+      "learning_rate": 7.447778642344898e-05,
+      "loss": 2.8359,
+      "step": 1808
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6402100687282821,
+      "learning_rate": 7.408808840043912e-05,
+      "loss": 2.8942,
+      "step": 1809
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.522510041279581,
+      "learning_rate": 7.369933098720021e-05,
+      "loss": 2.8899,
+      "step": 1810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4620778048315748,
+      "learning_rate": 7.331151504228767e-05,
+      "loss": 2.8613,
+      "step": 1811
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.498677822478166,
+      "learning_rate": 7.292464142217775e-05,
+      "loss": 2.8878,
+      "step": 1812
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.535613202371982,
+      "learning_rate": 7.25387109812658e-05,
+      "loss": 2.8274,
+      "step": 1813
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4909453128339663,
+      "learning_rate": 7.215372457186415e-05,
+      "loss": 2.8555,
+      "step": 1814
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5201709298426886,
+      "learning_rate": 7.176968304420007e-05,
+      "loss": 2.9086,
+      "step": 1815
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.656264295835627,
+      "learning_rate": 7.138658724641417e-05,
+      "loss": 2.8703,
+      "step": 1816
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.59601993654129,
+      "learning_rate": 7.10044380245587e-05,
+      "loss": 2.8112,
+      "step": 1817
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7250745754176193,
+      "learning_rate": 7.062323622259515e-05,
+      "loss": 2.8181,
+      "step": 1818
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4668213758759916,
+      "learning_rate": 7.024298268239265e-05,
+      "loss": 2.8121,
+      "step": 1819
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4748749603744966,
+      "learning_rate": 6.986367824372647e-05,
+      "loss": 2.8769,
+      "step": 1820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.494189055651729,
+      "learning_rate": 6.948532374427541e-05,
+      "loss": 2.9045,
+      "step": 1821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5153048261701998,
+      "learning_rate": 6.910792001962063e-05,
+      "loss": 2.9405,
+      "step": 1822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7762403318426678,
+      "learning_rate": 6.873146790324358e-05,
+      "loss": 2.9055,
+      "step": 1823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6758371577436613,
+      "learning_rate": 6.83559682265239e-05,
+      "loss": 2.9042,
+      "step": 1824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4786961518349315,
+      "learning_rate": 6.798142181873784e-05,
+      "loss": 2.8226,
+      "step": 1825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.629376345340511,
+      "learning_rate": 6.760782950705662e-05,
+      "loss": 2.9135,
+      "step": 1826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4432958040428752,
+      "learning_rate": 6.723519211654422e-05,
+      "loss": 2.8737,
+      "step": 1827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6741254058441817,
+      "learning_rate": 6.686351047015554e-05,
+      "loss": 2.8852,
+      "step": 1828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5973863740397851,
+      "learning_rate": 6.649278538873515e-05,
+      "loss": 2.762,
+      "step": 1829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.752194994052514,
+      "learning_rate": 6.612301769101465e-05,
+      "loss": 2.9257,
+      "step": 1830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6075070947743555,
+      "learning_rate": 6.575420819361177e-05,
+      "loss": 3.033,
+      "step": 1831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5822134977214013,
+      "learning_rate": 6.538635771102757e-05,
+      "loss": 2.7738,
+      "step": 1832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.556733784583965,
+      "learning_rate": 6.501946705564566e-05,
+      "loss": 2.976,
+      "step": 1833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7001810851066954,
+      "learning_rate": 6.465353703772959e-05,
+      "loss": 2.7998,
+      "step": 1834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6492738212971598,
+      "learning_rate": 6.428856846542136e-05,
+      "loss": 2.9037,
+      "step": 1835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5094249447202255,
+      "learning_rate": 6.392456214473996e-05,
+      "loss": 2.8655,
+      "step": 1836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4686601097895335,
+      "learning_rate": 6.3561518879579e-05,
+      "loss": 2.8018,
+      "step": 1837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6531922117325164,
+      "learning_rate": 6.31994394717052e-05,
+      "loss": 2.8126,
+      "step": 1838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5296319490719477,
+      "learning_rate": 6.283832472075685e-05,
+      "loss": 2.8901,
+      "step": 1839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.84921694223044,
+      "learning_rate": 6.247817542424178e-05,
+      "loss": 2.9112,
+      "step": 1840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4810860342576364,
+      "learning_rate": 6.211899237753559e-05,
+      "loss": 2.8711,
+      "step": 1841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5411639902023009,
+      "learning_rate": 6.176077637387984e-05,
+      "loss": 2.8673,
+      "step": 1842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.557524200649555,
+      "learning_rate": 6.140352820438066e-05,
+      "loss": 2.7863,
+      "step": 1843
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4569316982027463,
+      "learning_rate": 6.104724865800665e-05,
+      "loss": 2.8116,
+      "step": 1844
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5223772558892077,
+      "learning_rate": 6.069193852158711e-05,
+      "loss": 2.8607,
+      "step": 1845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.517418969764908,
+      "learning_rate": 6.0337598579810584e-05,
+      "loss": 2.8424,
+      "step": 1846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.476092469710869,
+      "learning_rate": 5.9984229615223096e-05,
+      "loss": 2.8002,
+      "step": 1847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4433105903349785,
+      "learning_rate": 5.963183240822606e-05,
+      "loss": 2.8389,
+      "step": 1848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6227524752055447,
+      "learning_rate": 5.9280407737074825e-05,
+      "loss": 2.8017,
+      "step": 1849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6486016784230564,
+      "learning_rate": 5.8929956377877125e-05,
+      "loss": 2.8678,
+      "step": 1850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7070901937002474,
+      "learning_rate": 5.8580479104591075e-05,
+      "loss": 2.8817,
+      "step": 1851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.50319688780993,
+      "learning_rate": 5.823197668902341e-05,
+      "loss": 2.8681,
+      "step": 1852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5415952198177345,
+      "learning_rate": 5.78844499008282e-05,
+      "loss": 2.8225,
+      "step": 1853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4689243324365235,
+      "learning_rate": 5.753789950750454e-05,
+      "loss": 2.8637,
+      "step": 1854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5731608084272992,
+      "learning_rate": 5.719232627439558e-05,
+      "loss": 2.8949,
+      "step": 1855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8251546849420626,
+      "learning_rate": 5.6847730964686315e-05,
+      "loss": 2.7629,
+      "step": 1856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4153011515317018,
+      "learning_rate": 5.650411433940189e-05,
+      "loss": 2.9016,
+      "step": 1857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6414133740035526,
+      "learning_rate": 5.61614771574061e-05,
+      "loss": 2.8179,
+      "step": 1858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4252898078090077,
+      "learning_rate": 5.581982017539988e-05,
+      "loss": 2.7409,
+      "step": 1859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.639276747299036,
+      "learning_rate": 5.5479144147919216e-05,
+      "loss": 2.8191,
+      "step": 1860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.541247695456947,
+      "learning_rate": 5.51394498273336e-05,
+      "loss": 2.8638,
+      "step": 1861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.53610585145542,
+      "learning_rate": 5.480073796384494e-05,
+      "loss": 3.003,
+      "step": 1862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8138312126123657,
+      "learning_rate": 5.446300930548492e-05,
+      "loss": 3.0234,
+      "step": 1863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4772925447902598,
+      "learning_rate": 5.412626459811415e-05,
+      "loss": 2.7702,
+      "step": 1864
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7512370125433259,
+      "learning_rate": 5.3790504585419954e-05,
+      "loss": 2.8575,
+      "step": 1865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5368114321265638,
+      "learning_rate": 5.345573000891541e-05,
+      "loss": 2.7417,
+      "step": 1866
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6230339112708894,
+      "learning_rate": 5.312194160793693e-05,
+      "loss": 2.8127,
+      "step": 1867
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4613110317547824,
+      "learning_rate": 5.278914011964303e-05,
+      "loss": 2.9001,
+      "step": 1868
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5618183234348069,
+      "learning_rate": 5.2457326279013006e-05,
+      "loss": 2.9038,
+      "step": 1869
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.413146477749218,
+      "learning_rate": 5.2126500818844514e-05,
+      "loss": 3.0586,
+      "step": 1870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4989489102698013,
+      "learning_rate": 5.1796664469752566e-05,
+      "loss": 2.9331,
+      "step": 1871
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4093705683081048,
+      "learning_rate": 5.1467817960167975e-05,
+      "loss": 2.8642,
+      "step": 1872
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4148995528934127,
+      "learning_rate": 5.113996201633536e-05,
+      "loss": 2.9434,
+      "step": 1873
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4801907749319303,
+      "learning_rate": 5.0813097362311765e-05,
+      "loss": 2.8668,
+      "step": 1874
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4634891540791295,
+      "learning_rate": 5.048722471996475e-05,
+      "loss": 2.7984,
+      "step": 1875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7444522716248454,
+      "learning_rate": 5.016234480897158e-05,
+      "loss": 2.9055,
+      "step": 1876
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.577587245942577,
+      "learning_rate": 4.9838458346816664e-05,
+      "loss": 2.8187,
+      "step": 1877
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3948486769620996,
+      "learning_rate": 4.9515566048790485e-05,
+      "loss": 2.8439,
+      "step": 1878
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6049171128436577,
+      "learning_rate": 4.9193668627988074e-05,
+      "loss": 2.8086,
+      "step": 1879
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3448340431392578,
+      "learning_rate": 4.887276679530744e-05,
+      "loss": 2.8375,
+      "step": 1880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4165474865728367,
+      "learning_rate": 4.855286125944752e-05,
+      "loss": 2.8546,
+      "step": 1881
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4808983108650835,
+      "learning_rate": 4.8233952726907224e-05,
+      "loss": 2.9539,
+      "step": 1882
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7205035863913096,
+      "learning_rate": 4.7916041901983565e-05,
+      "loss": 2.899,
+      "step": 1883
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5726145137994478,
+      "learning_rate": 4.7599129486770145e-05,
+      "loss": 2.8537,
+      "step": 1884
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6294004036349703,
+      "learning_rate": 4.728321618115555e-05,
+      "loss": 2.823,
+      "step": 1885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7174207138593391,
+      "learning_rate": 4.696830268282204e-05,
+      "loss": 2.8472,
+      "step": 1886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4887318047028357,
+      "learning_rate": 4.665438968724361e-05,
+      "loss": 2.9178,
+      "step": 1887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.762551472676163,
+      "learning_rate": 4.634147788768489e-05,
+      "loss": 2.8688,
+      "step": 1888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.568662368646745,
+      "learning_rate": 4.6029567975199414e-05,
+      "loss": 2.963,
+      "step": 1889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.627716399498809,
+      "learning_rate": 4.571866063862795e-05,
+      "loss": 2.8779,
+      "step": 1890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6283660983674033,
+      "learning_rate": 4.540875656459703e-05,
+      "loss": 3.0344,
+      "step": 1891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.433438226855833,
+      "learning_rate": 4.509985643751785e-05,
+      "loss": 2.8518,
+      "step": 1892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6750444348323519,
+      "learning_rate": 4.479196093958421e-05,
+      "loss": 2.8586,
+      "step": 1893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8632111929048816,
+      "learning_rate": 4.4485070750771187e-05,
+      "loss": 2.9819,
+      "step": 1894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.754871370851778,
+      "learning_rate": 4.417918654883363e-05,
+      "loss": 2.8565,
+      "step": 1895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7795978374284736,
+      "learning_rate": 4.3874309009305e-05,
+      "loss": 2.8643,
+      "step": 1896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.689065454302286,
+      "learning_rate": 4.357043880549538e-05,
+      "loss": 3.0176,
+      "step": 1897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.569837511850935,
+      "learning_rate": 4.326757660849012e-05,
+      "loss": 2.9608,
+      "step": 1898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.512823988192231,
+      "learning_rate": 4.2965723087148635e-05,
+      "loss": 2.9018,
+      "step": 1899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.674394260776728,
+      "learning_rate": 4.266487890810256e-05,
+      "loss": 2.837,
+      "step": 1900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.651298085123302,
+      "learning_rate": 4.2365044735754365e-05,
+      "loss": 2.7942,
+      "step": 1901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7036648094975293,
+      "learning_rate": 4.2066221232276266e-05,
+      "loss": 2.8339,
+      "step": 1902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4514830661132014,
+      "learning_rate": 4.176840905760815e-05,
+      "loss": 2.8008,
+      "step": 1903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5864857015502518,
+      "learning_rate": 4.1471608869456443e-05,
+      "loss": 2.7599,
+      "step": 1904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4303070669298903,
+      "learning_rate": 4.117582132329284e-05,
+      "loss": 2.9616,
+      "step": 1905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4500413183702718,
+      "learning_rate": 4.088104707235263e-05,
+      "loss": 2.8833,
+      "step": 1906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.502405318668058,
+      "learning_rate": 4.058728676763313e-05,
+      "loss": 2.9672,
+      "step": 1907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5564876329609305,
+      "learning_rate": 4.0294541057892375e-05,
+      "loss": 2.8747,
+      "step": 1908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4639933889950005,
+      "learning_rate": 4.000281058964794e-05,
+      "loss": 2.8516,
+      "step": 1909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.499868969921477,
+      "learning_rate": 3.971209600717507e-05,
+      "loss": 2.7715,
+      "step": 1910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5111586814901425,
+      "learning_rate": 3.9422397952505465e-05,
+      "loss": 2.7541,
+      "step": 1911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5944294250275566,
+      "learning_rate": 3.913371706542596e-05,
+      "loss": 2.9425,
+      "step": 1912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5580275423783787,
+      "learning_rate": 3.884605398347707e-05,
+      "loss": 2.7524,
+      "step": 1913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.683905020645839,
+      "learning_rate": 3.8559409341951456e-05,
+      "loss": 2.9666,
+      "step": 1914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6377212468045117,
+      "learning_rate": 3.8273783773892404e-05,
+      "loss": 2.9577,
+      "step": 1915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5323690846454083,
+      "learning_rate": 3.798917791009293e-05,
+      "loss": 2.8446,
+      "step": 1916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4311362470289313,
+      "learning_rate": 3.770559237909393e-05,
+      "loss": 2.8404,
+      "step": 1917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5328544761255127,
+      "learning_rate": 3.742302780718288e-05,
+      "loss": 2.8188,
+      "step": 1918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4663454718230928,
+      "learning_rate": 3.7141484818392635e-05,
+      "loss": 2.7942,
+      "step": 1919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6400701608398311,
+      "learning_rate": 3.686096403449973e-05,
+      "loss": 2.8692,
+      "step": 1920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5102343229227375,
+      "learning_rate": 3.658146607502344e-05,
+      "loss": 2.8692,
+      "step": 1921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4036744441877635,
+      "learning_rate": 3.630299155722411e-05,
+      "loss": 2.9044,
+      "step": 1922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4959073188378265,
+      "learning_rate": 3.6025541096101676e-05,
+      "loss": 2.8899,
+      "step": 1923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5376951244806973,
+      "learning_rate": 3.574911530439473e-05,
+      "loss": 2.9399,
+      "step": 1924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4364576078878546,
+      "learning_rate": 3.5473714792578606e-05,
+      "loss": 2.939,
+      "step": 1925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5555068532045278,
+      "learning_rate": 3.519934016886478e-05,
+      "loss": 2.9731,
+      "step": 1926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5503484858841499,
+      "learning_rate": 3.4925992039198776e-05,
+      "loss": 2.9722,
+      "step": 1927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.571635720170499,
+      "learning_rate": 3.465367100725908e-05,
+      "loss": 2.925,
+      "step": 1928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5269373291040813,
+      "learning_rate": 3.438237767445618e-05,
+      "loss": 2.7766,
+      "step": 1929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5464844702339466,
+      "learning_rate": 3.4112112639930804e-05,
+      "loss": 2.9233,
+      "step": 1930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7451535915348053,
+      "learning_rate": 3.3842876500552564e-05,
+      "loss": 2.8883,
+      "step": 1931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4665119144304581,
+      "learning_rate": 3.357466985091906e-05,
+      "loss": 2.9739,
+      "step": 1932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.384489384330484,
+      "learning_rate": 3.330749328335414e-05,
+      "loss": 2.8155,
+      "step": 1933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6795591162836185,
+      "learning_rate": 3.304134738790659e-05,
+      "loss": 2.8918,
+      "step": 1934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4402140556573797,
+      "learning_rate": 3.277623275234953e-05,
+      "loss": 2.834,
+      "step": 1935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4529663494166023,
+      "learning_rate": 3.2512149962177994e-05,
+      "loss": 2.9157,
+      "step": 1936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.590567862847257,
+      "learning_rate": 3.224909960060851e-05,
+      "loss": 2.8381,
+      "step": 1937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4714345724483815,
+      "learning_rate": 3.198708224857755e-05,
+      "loss": 2.8784,
+      "step": 1938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4726849731668494,
+      "learning_rate": 3.172609848474023e-05,
+      "loss": 2.8212,
+      "step": 1939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4060337482477416,
+      "learning_rate": 3.1466148885468895e-05,
+      "loss": 2.8775,
+      "step": 1940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4875144255791442,
+      "learning_rate": 3.120723402485198e-05,
+      "loss": 2.9733,
+      "step": 1941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4970631190152253,
+      "learning_rate": 3.094935447469294e-05,
+      "loss": 2.8211,
+      "step": 1942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5005998991667864,
+      "learning_rate": 3.069251080450863e-05,
+      "loss": 2.862,
+      "step": 1943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.464965848145526,
+      "learning_rate": 3.0436703581528113e-05,
+      "loss": 2.8855,
+      "step": 1944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5335386551872414,
+      "learning_rate": 3.0181933370691694e-05,
+      "loss": 2.9302,
+      "step": 1945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.484008739331667,
+      "learning_rate": 2.9928200734649523e-05,
+      "loss": 2.8072,
+      "step": 1946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5358441924359179,
+      "learning_rate": 2.9675506233760142e-05,
+      "loss": 2.8687,
+      "step": 1947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5973523042482358,
+      "learning_rate": 2.942385042608925e-05,
+      "loss": 2.8793,
+      "step": 1948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.582223534483645,
+      "learning_rate": 2.9173233867409054e-05,
+      "loss": 2.9423,
+      "step": 1949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4678066550787998,
+      "learning_rate": 2.892365711119638e-05,
+      "loss": 2.9109,
+      "step": 1950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4347146695874915,
+      "learning_rate": 2.8675120708631596e-05,
+      "loss": 2.8819,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6025566848049118,
+      "learning_rate": 2.8427625208597764e-05,
+      "loss": 2.8904,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4909591674742582,
+      "learning_rate": 2.8181171157678874e-05,
+      "loss": 2.8805,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.47832464221613,
+      "learning_rate": 2.7935759100159053e-05,
+      "loss": 2.8914,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6972196717084973,
+      "learning_rate": 2.7691389578021365e-05,
+      "loss": 2.9087,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5757128569203995,
+      "learning_rate": 2.7448063130946223e-05,
+      "loss": 2.8744,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5538214364837772,
+      "learning_rate": 2.7205780296310544e-05,
+      "loss": 2.8828,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4731770562511992,
+      "learning_rate": 2.6964541609186378e-05,
+      "loss": 2.8521,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4665879583042831,
+      "learning_rate": 2.6724347602340104e-05,
+      "loss": 2.7618,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4815390477016337,
+      "learning_rate": 2.6485198806230682e-05,
+      "loss": 2.8734,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4349096382687316,
+      "learning_rate": 2.6247095749008797e-05,
+      "loss": 2.9209,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7833695769984879,
+      "learning_rate": 2.6010038956515826e-05,
+      "loss": 2.9418,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6296732463605148,
+      "learning_rate": 2.5774028952282423e-05,
+      "loss": 2.8611,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5831547685757865,
+      "learning_rate": 2.5539066257527277e-05,
+      "loss": 2.9252,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.469656350932715,
+      "learning_rate": 2.530515139115652e-05,
+      "loss": 2.8771,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4787897090913968,
+      "learning_rate": 2.5072284869761874e-05,
+      "loss": 2.887,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.484189165325405,
+      "learning_rate": 2.4840467207619786e-05,
+      "loss": 2.8504,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5294827175806416,
+      "learning_rate": 2.460969891669068e-05,
+      "loss": 2.9081,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6027837273608372,
+      "learning_rate": 2.4379980506617272e-05,
+      "loss": 2.8572,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5673656830612,
+      "learning_rate": 2.4151312484723464e-05,
+      "loss": 2.8468,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4731562037011865,
+      "learning_rate": 2.3923695356013798e-05,
+      "loss": 2.8148,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5578472190648955,
+      "learning_rate": 2.3697129623171833e-05,
+      "loss": 2.8224,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4660743943581256,
+      "learning_rate": 2.3471615786559042e-05,
+      "loss": 2.9389,
+      "step": 1973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5079763677086122,
+      "learning_rate": 2.3247154344213818e-05,
+      "loss": 2.8485,
+      "step": 1974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5041664589213408,
+      "learning_rate": 2.3023745791850625e-05,
+      "loss": 2.8479,
+      "step": 1975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5604549841682265,
+      "learning_rate": 2.2801390622858354e-05,
+      "loss": 2.9428,
+      "step": 1976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4072160826426907,
+      "learning_rate": 2.2580089328299746e-05,
+      "loss": 2.8331,
+      "step": 1977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6438604681318603,
+      "learning_rate": 2.235984239690997e-05,
+      "loss": 2.9018,
+      "step": 1978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4320500088021482,
+      "learning_rate": 2.2140650315095934e-05,
+      "loss": 2.7994,
+      "step": 1979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.534064340755125,
+      "learning_rate": 2.192251356693459e-05,
+      "loss": 2.9454,
+      "step": 1980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4072471487692266,
+      "learning_rate": 2.170543263417246e-05,
+      "loss": 2.9149,
+      "step": 1981
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4874578196742538,
+      "learning_rate": 2.1489407996224286e-05,
+      "loss": 2.9319,
+      "step": 1982
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8273672893919206,
+      "learning_rate": 2.127444013017199e-05,
+      "loss": 2.9181,
+      "step": 1983
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3942046266177364,
+      "learning_rate": 2.1060529510763648e-05,
+      "loss": 2.8285,
+      "step": 1984
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4930103108346384,
+      "learning_rate": 2.084767661041259e-05,
+      "loss": 2.9399,
+      "step": 1985
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.614343291849434,
+      "learning_rate": 2.063588189919596e-05,
+      "loss": 2.8734,
+      "step": 1986
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8127584809648278,
+      "learning_rate": 2.0425145844854275e-05,
+      "loss": 2.9724,
+      "step": 1987
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8037086954329054,
+      "learning_rate": 2.0215468912789693e-05,
+      "loss": 2.8721,
+      "step": 1988
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7497334435742615,
+      "learning_rate": 2.0006851566065575e-05,
+      "loss": 2.9059,
+      "step": 1989
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7026835928865063,
+      "learning_rate": 1.9799294265405166e-05,
+      "loss": 2.7976,
+      "step": 1990
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5717010758013226,
+      "learning_rate": 1.9592797469190572e-05,
+      "loss": 2.9775,
+      "step": 1991
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5991540332869905,
+      "learning_rate": 1.938736163346194e-05,
+      "loss": 2.8874,
+      "step": 1992
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5637072959436045,
+      "learning_rate": 1.9182987211916246e-05,
+      "loss": 2.8109,
+      "step": 1993
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5745896865194786,
+      "learning_rate": 1.8979674655906332e-05,
+      "loss": 2.9609,
+      "step": 1994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5539240823916454,
+      "learning_rate": 1.8777424414440024e-05,
+      "loss": 2.828,
+      "step": 1995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4025417579911514,
+      "learning_rate": 1.8576236934179202e-05,
+      "loss": 2.9362,
+      "step": 1996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5050863593089898,
+      "learning_rate": 1.8376112659438393e-05,
+      "loss": 2.9596,
+      "step": 1997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4799545771805533,
+      "learning_rate": 1.8177052032184282e-05,
+      "loss": 2.8515,
+      "step": 1998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6370075649842235,
+      "learning_rate": 1.7979055492034435e-05,
+      "loss": 2.9223,
+      "step": 1999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7332140982857032,
+      "learning_rate": 1.7782123476256407e-05,
+      "loss": 2.9252,
+      "step": 2000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5388126458604086,
+      "learning_rate": 1.7586256419766965e-05,
+      "loss": 2.9439,
+      "step": 2001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4693043476586793,
+      "learning_rate": 1.7391454755130766e-05,
+      "loss": 2.7673,
+      "step": 2002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5355137938875654,
+      "learning_rate": 1.7197718912559557e-05,
+      "loss": 2.8348,
+      "step": 2003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4118626295927916,
+      "learning_rate": 1.700504931991148e-05,
+      "loss": 2.8659,
+      "step": 2004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5051451372854154,
+      "learning_rate": 1.681344640268978e-05,
+      "loss": 2.8737,
+      "step": 2005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.47400337301359,
+      "learning_rate": 1.6622910584041974e-05,
+      "loss": 2.9754,
+      "step": 2006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5012463101060307,
+      "learning_rate": 1.6433442284758903e-05,
+      "loss": 2.8863,
+      "step": 2007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4198639382717881,
+      "learning_rate": 1.624504192327392e-05,
+      "loss": 2.7836,
+      "step": 2008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5954261264842526,
+      "learning_rate": 1.6057709915661856e-05,
+      "loss": 2.8596,
+      "step": 2009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6481036916953042,
+      "learning_rate": 1.5871446675638057e-05,
+      "loss": 2.8829,
+      "step": 2010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5339498346494753,
+      "learning_rate": 1.5686252614557638e-05,
+      "loss": 2.7937,
+      "step": 2011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.393557889800209,
+      "learning_rate": 1.5502128141414497e-05,
+      "loss": 2.9023,
+      "step": 2012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.88986773515513,
+      "learning_rate": 1.5319073662840188e-05,
+      "loss": 2.7743,
+      "step": 2013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5966387010924286,
+      "learning_rate": 1.5137089583103391e-05,
+      "loss": 2.9486,
+      "step": 2014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3994100632865987,
+      "learning_rate": 1.4956176304108893e-05,
+      "loss": 2.9123,
+      "step": 2015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4978715254602573,
+      "learning_rate": 1.4776334225396481e-05,
+      "loss": 2.9195,
+      "step": 2016
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4385628066120812,
+      "learning_rate": 1.4597563744140397e-05,
+      "loss": 2.8311,
+      "step": 2017
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5471527050455438,
+      "learning_rate": 1.4419865255148269e-05,
+      "loss": 2.9322,
+      "step": 2018
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8327583444194242,
+      "learning_rate": 1.4243239150860122e-05,
+      "loss": 2.7309,
+      "step": 2019
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8850396573787256,
+      "learning_rate": 1.4067685821347932e-05,
+      "loss": 2.8314,
+      "step": 2020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4936714332986873,
+      "learning_rate": 1.389320565431429e-05,
+      "loss": 2.8937,
+      "step": 2021
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4956226151967322,
+      "learning_rate": 1.3719799035091851e-05,
+      "loss": 2.7761,
+      "step": 2022
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6033053339187568,
+      "learning_rate": 1.3547466346642278e-05,
+      "loss": 2.854,
+      "step": 2023
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4776702601067804,
+      "learning_rate": 1.3376207969555577e-05,
+      "loss": 2.8748,
+      "step": 2024
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6799425520501161,
+      "learning_rate": 1.32060242820492e-05,
+      "loss": 2.8671,
+      "step": 2025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5885598073189569,
+      "learning_rate": 1.3036915659967118e-05,
+      "loss": 2.8139,
+      "step": 2026
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4171086120023548,
+      "learning_rate": 1.2868882476779087e-05,
+      "loss": 2.8508,
+      "step": 2027
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5168505214152104,
+      "learning_rate": 1.2701925103579815e-05,
+      "loss": 2.8751,
+      "step": 2028
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4437626793992913,
+      "learning_rate": 1.2536043909088191e-05,
+      "loss": 2.9438,
+      "step": 2029
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.814281030396273,
+      "learning_rate": 1.2371239259646228e-05,
+      "loss": 2.9143,
+      "step": 2030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6619546796392395,
+      "learning_rate": 1.2207511519218672e-05,
+      "loss": 2.9273,
+      "step": 2031
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8887156476744498,
+      "learning_rate": 1.2044861049391676e-05,
+      "loss": 2.8618,
+      "step": 2032
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4588444357105845,
+      "learning_rate": 1.1883288209372512e-05,
+      "loss": 2.8062,
+      "step": 2033
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.578319729400997,
+      "learning_rate": 1.1722793355988471e-05,
+      "loss": 2.9831,
+      "step": 2034
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5420847883700262,
+      "learning_rate": 1.1563376843686135e-05,
+      "loss": 2.9085,
+      "step": 2035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5679244156550947,
+      "learning_rate": 1.140503902453055e-05,
+      "loss": 2.7953,
+      "step": 2036
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8364615041958108,
+      "learning_rate": 1.1247780248204665e-05,
+      "loss": 2.8381,
+      "step": 2037
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4897926606454344,
+      "learning_rate": 1.1091600862008333e-05,
+      "loss": 2.8884,
+      "step": 2038
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5848765946736403,
+      "learning_rate": 1.0936501210857652e-05,
+      "loss": 2.8648,
+      "step": 2039
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6175298730592484,
+      "learning_rate": 1.0782481637284013e-05,
+      "loss": 2.9331,
+      "step": 2040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4917936375091316,
+      "learning_rate": 1.0629542481433663e-05,
+      "loss": 2.8692,
+      "step": 2041
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.436200854069479,
+      "learning_rate": 1.0477684081066751e-05,
+      "loss": 2.9199,
+      "step": 2042
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7438399043286865,
+      "learning_rate": 1.0326906771556566e-05,
+      "loss": 2.9482,
+      "step": 2043
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.435118602016525,
+      "learning_rate": 1.017721088588891e-05,
+      "loss": 2.8844,
+      "step": 2044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4950453406702178,
+      "learning_rate": 1.0028596754661334e-05,
+      "loss": 2.8055,
+      "step": 2045
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5750797839944721,
+      "learning_rate": 9.881064706082298e-06,
+      "loss": 2.845,
+      "step": 2046
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7260328791785635,
+      "learning_rate": 9.734615065970454e-06,
+      "loss": 2.7853,
+      "step": 2047
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.600488879331733,
+      "learning_rate": 9.58924815775425e-06,
+      "loss": 2.911,
+      "step": 2048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4719363724111145,
+      "learning_rate": 9.444964302470715e-06,
+      "loss": 2.8467,
+      "step": 2049
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6641142813161849,
+      "learning_rate": 9.301763818765018e-06,
+      "loss": 2.8522,
+      "step": 2050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5297354713655371,
+      "learning_rate": 9.15964702288996e-06,
+      "loss": 2.9144,
+      "step": 2051
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6465906424541557,
+      "learning_rate": 9.018614228704925e-06,
+      "loss": 2.7477,
+      "step": 2052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4029411802728182,
+      "learning_rate": 8.878665747675152e-06,
+      "loss": 2.8064,
+      "step": 2053
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4713897874510444,
+      "learning_rate": 8.739801888871469e-06,
+      "loss": 2.8162,
+      "step": 2054
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3862184812122866,
+      "learning_rate": 8.602022958969336e-06,
+      "loss": 2.8841,
+      "step": 2055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3425302286773029,
+      "learning_rate": 8.465329262248078e-06,
+      "loss": 2.8259,
+      "step": 2056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.481598435000286,
+      "learning_rate": 8.32972110059027e-06,
+      "loss": 2.8877,
+      "step": 2057
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4197574402815962,
+      "learning_rate": 8.195198773481406e-06,
+      "loss": 2.9348,
+      "step": 2058
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7481587577918951,
+      "learning_rate": 8.061762578008613e-06,
+      "loss": 2.925,
+      "step": 2059
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4797379243106081,
+      "learning_rate": 7.929412808860559e-06,
+      "loss": 2.8116,
+      "step": 2060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6512173363065439,
+      "learning_rate": 7.79814975832649e-06,
+      "loss": 2.9112,
+      "step": 2061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6550133641278555,
+      "learning_rate": 7.667973716295851e-06,
+      "loss": 2.9183,
+      "step": 2062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3795522268718496,
+      "learning_rate": 7.5388849702571205e-06,
+      "loss": 2.8346,
+      "step": 2063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.491178519659264,
+      "learning_rate": 7.4108838052979185e-06,
+      "loss": 2.8363,
+      "step": 2064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4271778359926224,
+      "learning_rate": 7.283970504103732e-06,
+      "loss": 2.7577,
+      "step": 2065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4932681485251926,
+      "learning_rate": 7.1581453469575785e-06,
+      "loss": 2.7733,
+      "step": 2066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.553765796592862,
+      "learning_rate": 7.033408611739456e-06,
+      "loss": 2.8634,
+      "step": 2067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5887782617601862,
+      "learning_rate": 6.909760573925561e-06,
+      "loss": 2.9104,
+      "step": 2068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.560370878264193,
+      "learning_rate": 6.787201506587626e-06,
+      "loss": 2.8573,
+      "step": 2069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5179770132768409,
+      "learning_rate": 6.66573168039264e-06,
+      "loss": 2.9291,
+      "step": 2070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.473788361267198,
+      "learning_rate": 6.545351363601959e-06,
+      "loss": 2.9331,
+      "step": 2071
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.511427342947344,
+      "learning_rate": 6.426060822070812e-06,
+      "loss": 2.8179,
+      "step": 2072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.479821614829401,
+      "learning_rate": 6.3078603192475716e-06,
+      "loss": 2.9935,
+      "step": 2073
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5714443834065885,
+      "learning_rate": 6.1907501161735934e-06,
+      "loss": 2.8862,
+      "step": 2074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.618277096754612,
+      "learning_rate": 6.074730471482049e-06,
+      "loss": 2.8478,
+      "step": 2075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4486934298729226,
+      "learning_rate": 5.959801641397755e-06,
+      "loss": 2.8357,
+      "step": 2076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5097582741040247,
+      "learning_rate": 5.845963879736627e-06,
+      "loss": 2.7522,
+      "step": 2077
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5372374982287638,
+      "learning_rate": 5.733217437904892e-06,
+      "loss": 2.8364,
+      "step": 2078
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.524030997769299,
+      "learning_rate": 5.621562564898597e-06,
+      "loss": 2.9527,
+      "step": 2079
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4651782369634005,
+      "learning_rate": 5.51099950730316e-06,
+      "loss": 2.8722,
+      "step": 2080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5372217167793154,
+      "learning_rate": 5.401528509292763e-06,
+      "loss": 2.9598,
+      "step": 2081
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4960171509913391,
+      "learning_rate": 5.2931498126298495e-06,
+      "loss": 2.9297,
+      "step": 2082
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5128630932867422,
+      "learning_rate": 5.1858636566645135e-06,
+      "loss": 2.8304,
+      "step": 2083
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.604338333171454,
+      "learning_rate": 5.0796702783340035e-06,
+      "loss": 2.7825,
+      "step": 2084
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5131272598202807,
+      "learning_rate": 4.97456991216233e-06,
+      "loss": 2.803,
+      "step": 2085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.660000980221689,
+      "learning_rate": 4.870562790259325e-06,
+      "loss": 2.8959,
+      "step": 2086
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5885865533449666,
+      "learning_rate": 4.7676491423208625e-06,
+      "loss": 2.8071,
+      "step": 2087
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5040679344943872,
+      "learning_rate": 4.66582919562758e-06,
+      "loss": 2.8891,
+      "step": 2088
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5658836249701653,
+      "learning_rate": 4.5651031750448825e-06,
+      "loss": 2.9697,
+      "step": 2089
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4210956827895254,
+      "learning_rate": 4.465471303022217e-06,
+      "loss": 2.7792,
+      "step": 2090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.436851691896154,
+      "learning_rate": 4.366933799592743e-06,
+      "loss": 2.9187,
+      "step": 2091
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5112751874572803,
+      "learning_rate": 4.269490882372551e-06,
+      "loss": 2.8447,
+      "step": 2092
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4293935828070912,
+      "learning_rate": 4.1731427665606115e-06,
+      "loss": 2.8681,
+      "step": 2093
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5166164993582885,
+      "learning_rate": 4.077889664937884e-06,
+      "loss": 2.8435,
+      "step": 2094
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.437581977955914,
+      "learning_rate": 3.983731787867207e-06,
+      "loss": 2.756,
+      "step": 2095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6880841284557726,
+      "learning_rate": 3.890669343292464e-06,
+      "loss": 2.8628,
+      "step": 2096
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.448538016950513,
+      "learning_rate": 3.7987025367384743e-06,
+      "loss": 2.918,
+      "step": 2097
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5265747055521879,
+      "learning_rate": 3.707831571310327e-06,
+      "loss": 2.8671,
+      "step": 2098
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3846339627503044,
+      "learning_rate": 3.6180566476929912e-06,
+      "loss": 2.8209,
+      "step": 2099
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3486140408773666,
+      "learning_rate": 3.529377964150815e-06,
+      "loss": 2.8314,
+      "step": 2100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4442875050263744,
+      "learning_rate": 3.441795716527307e-06,
+      "loss": 2.9235,
+      "step": 2101
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8003966359703822,
+      "learning_rate": 3.355310098244302e-06,
+      "loss": 2.9187,
+      "step": 2102
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.550264867272245,
+      "learning_rate": 3.269921300301959e-06,
+      "loss": 2.8438,
+      "step": 2103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.442644788224796,
+      "learning_rate": 3.1856295112780988e-06,
+      "loss": 2.8719,
+      "step": 2104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4437986486097043,
+      "learning_rate": 3.102434917327812e-06,
+      "loss": 2.9403,
+      "step": 2105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4977996226139574,
+      "learning_rate": 3.0203377021831292e-06,
+      "loss": 2.9147,
+      "step": 2106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.479035525193425,
+      "learning_rate": 2.939338047152573e-06,
+      "loss": 2.7613,
+      "step": 2107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.487584478944744,
+      "learning_rate": 2.8594361311206073e-06,
+      "loss": 2.8024,
+      "step": 2108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.405440574505884,
+      "learning_rate": 2.7806321305475225e-06,
+      "loss": 2.8191,
+      "step": 2109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4683203952694448,
+      "learning_rate": 2.7029262194688818e-06,
+      "loss": 2.7873,
+      "step": 2110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5206005184996711,
+      "learning_rate": 2.626318569495134e-06,
+      "loss": 2.852,
+      "step": 2111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5338481062188865,
+      "learning_rate": 2.550809349811334e-06,
+      "loss": 2.8735,
+      "step": 2112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.463823059198892,
+      "learning_rate": 2.476398727176532e-06,
+      "loss": 2.8412,
+      "step": 2113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4923051774213183,
+      "learning_rate": 2.4030868659237204e-06,
+      "loss": 2.9158,
+      "step": 2114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.461734723295787,
+      "learning_rate": 2.3308739279593317e-06,
+      "loss": 2.8733,
+      "step": 2115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4811685403745933,
+      "learning_rate": 2.2597600727626845e-06,
+      "loss": 2.8369,
+      "step": 2116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4690245644078024,
+      "learning_rate": 2.1897454573860387e-06,
+      "loss": 2.8863,
+      "step": 2117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4439631572852036,
+      "learning_rate": 2.1208302364538746e-06,
+      "loss": 2.7857,
+      "step": 2118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4504883799468544,
+      "learning_rate": 2.0530145621627804e-06,
+      "loss": 2.8481,
+      "step": 2119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5712846977520416,
+      "learning_rate": 1.9862985842810653e-06,
+      "loss": 2.9003,
+      "step": 2120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4618001006324173,
+      "learning_rate": 1.920682450148259e-06,
+      "loss": 2.9018,
+      "step": 2121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.44200793447349,
+      "learning_rate": 1.856166304675111e-06,
+      "loss": 2.9493,
+      "step": 2122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5588902481169093,
+      "learning_rate": 1.792750290342926e-06,
+      "loss": 2.862,
+      "step": 2123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4422826940992963,
+      "learning_rate": 1.7304345472035632e-06,
+      "loss": 2.8537,
+      "step": 2124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4151754371460405,
+      "learning_rate": 1.6692192128788253e-06,
+      "loss": 2.9006,
+      "step": 2125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.476028730444538,
+      "learning_rate": 1.6091044225604035e-06,
+      "loss": 2.8356,
+      "step": 2126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4983546242197796,
+      "learning_rate": 1.5500903090094888e-06,
+      "loss": 2.8427,
+      "step": 2127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4813560149575726,
+      "learning_rate": 1.492177002556383e-06,
+      "loss": 2.856,
+      "step": 2128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.634182601651669,
+      "learning_rate": 1.4353646311004443e-06,
+      "loss": 2.805,
+      "step": 2129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4930452452819154,
+      "learning_rate": 1.3796533201094752e-06,
+      "loss": 2.8237,
+      "step": 2130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5137816665481156,
+      "learning_rate": 1.3250431926197793e-06,
+      "loss": 2.9058,
+      "step": 2131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4520051467762336,
+      "learning_rate": 1.2715343692356607e-06,
+      "loss": 2.9287,
+      "step": 2132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.428375214692525,
+      "learning_rate": 1.2191269681292582e-06,
+      "loss": 2.9125,
+      "step": 2133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5745104543064323,
+      "learning_rate": 1.1678211050402676e-06,
+      "loss": 2.9409,
+      "step": 2134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4779719142655328,
+      "learning_rate": 1.117616893275719e-06,
+      "loss": 2.9152,
+      "step": 2135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.408013776591615,
+      "learning_rate": 1.068514443709534e-06,
+      "loss": 2.7898,
+      "step": 2136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4896026925751995,
+      "learning_rate": 1.0205138647826905e-06,
+      "loss": 2.838,
+      "step": 2137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.588359791836602,
+      "learning_rate": 9.73615262502503e-07,
+      "loss": 2.9336,
+      "step": 2138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4092096302161734,
+      "learning_rate": 9.278187404426763e-07,
+      "loss": 2.9494,
+      "step": 2139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5171615535631873,
+      "learning_rate": 8.831243997431404e-07,
+      "loss": 2.9468,
+      "step": 2140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4902087158541766,
+      "learning_rate": 8.395323391094944e-07,
+      "loss": 2.8707,
+      "step": 2141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6089502941755045,
+      "learning_rate": 7.970426548131183e-07,
+      "loss": 2.8516,
+      "step": 2142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4652701502600067,
+      "learning_rate": 7.556554406908389e-07,
+      "loss": 2.7547,
+      "step": 2143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7666860159180373,
+      "learning_rate": 7.153707881446536e-07,
+      "loss": 2.9179,
+      "step": 2144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5241929339734375,
+      "learning_rate": 6.761887861417293e-07,
+      "loss": 2.8455,
+      "step": 2145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4792571324403863,
+      "learning_rate": 6.381095212139032e-07,
+      "loss": 2.7573,
+      "step": 2146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4904795297151339,
+      "learning_rate": 6.011330774577384e-07,
+      "loss": 2.8708,
+      "step": 2147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5429832752666304,
+      "learning_rate": 5.652595365343016e-07,
+      "loss": 2.8457,
+      "step": 2148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.512697187082373,
+      "learning_rate": 5.304889776688859e-07,
+      "loss": 2.9566,
+      "step": 2149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4164801983368362,
+      "learning_rate": 4.968214776508994e-07,
+      "loss": 2.8511,
+      "step": 2150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4372895358872773,
+      "learning_rate": 4.6425711083375454e-07,
+      "loss": 2.8223,
+      "step": 2151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.461005951519057,
+      "learning_rate": 4.3279594913447906e-07,
+      "loss": 2.8496,
+      "step": 2152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4842996707075866,
+      "learning_rate": 4.02438062033883e-07,
+      "loss": 2.7794,
+      "step": 2153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5132708745253738,
+      "learning_rate": 3.7318351657616987e-07,
+      "loss": 2.8632,
+      "step": 2154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4461415333412584,
+      "learning_rate": 3.4503237736882573e-07,
+      "loss": 2.8995,
+      "step": 2155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5163029207171845,
+      "learning_rate": 3.179847065825081e-07,
+      "loss": 2.8205,
+      "step": 2156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.477348603008028,
+      "learning_rate": 2.9204056395104594e-07,
+      "loss": 2.832,
+      "step": 2157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5989019391969521,
+      "learning_rate": 2.672000067709956e-07,
+      "loss": 2.7874,
+      "step": 2158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4883205932371564,
+      "learning_rate": 2.4346308990175204e-07,
+      "loss": 2.8548,
+      "step": 2159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4558890244291165,
+      "learning_rate": 2.208298657653818e-07,
+      "loss": 2.8863,
+      "step": 2160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5205397869093804,
+      "learning_rate": 1.9930038434645692e-07,
+      "loss": 2.8617,
+      "step": 2161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6158444752737502,
+      "learning_rate": 1.7887469319205484e-07,
+      "loss": 2.859,
+      "step": 2162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.48694798147771,
+      "learning_rate": 1.5955283741142523e-07,
+      "loss": 2.908,
+      "step": 2163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4668354199254765,
+      "learning_rate": 1.4133485967615655e-07,
+      "loss": 2.8105,
+      "step": 2164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.591717419275098,
+      "learning_rate": 1.2422080021995407e-07,
+      "loss": 2.9679,
+      "step": 2165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5638799798970995,
+      "learning_rate": 1.082106968385288e-07,
+      "loss": 2.8005,
+      "step": 2166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4497120087338233,
+      "learning_rate": 9.330458488959748e-08,
+      "loss": 2.9536,
+      "step": 2167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5329852679877578,
+      "learning_rate": 7.950249729271608e-08,
+      "loss": 2.8174,
+      "step": 2168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6244331251030222,
+      "learning_rate": 6.680446452922429e-08,
+      "loss": 2.8932,
+      "step": 2169
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5140721728212518,
+      "learning_rate": 5.521051464230098e-08,
+      "loss": 2.8334,
+      "step": 2170
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6405194772419227,
+      "learning_rate": 4.4720673236631206e-08,
+      "loss": 2.9845,
+      "step": 2171
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5907867078086944,
+      "learning_rate": 3.53349634786837e-08,
+      "loss": 2.7952,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.44673487807436,
+      "learning_rate": 2.7053406096433365e-08,
+      "loss": 2.964,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4715881735561782,
+      "learning_rate": 1.987601937930572e-08,
+      "loss": 2.8704,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4083924002869224,
+      "learning_rate": 1.3802819178398984e-08,
+      "loss": 2.8862,
+      "step": 2175
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.636655594988435,
+      "learning_rate": 8.833818906039959e-09,
+      "loss": 2.7875,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4534222518680877,
+      "learning_rate": 4.969029536061598e-09,
+      "loss": 2.8833,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4654478883568558,
+      "learning_rate": 2.2084596038030037e-09,
+      "loss": 2.8866,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.52024890564369,
+      "learning_rate": 5.521152057763601e-10,
+      "loss": 2.964,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4923017296492271,
+      "learning_rate": 0.0,
+      "loss": 2.8003,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0,
+      "step": 2180,
+      "total_flos": 280522689478656.0,
+      "train_loss": 3.333065954698335,
+      "train_runtime": 6070.2935,
+      "train_samples_per_second": 91.944,
+      "train_steps_per_second": 0.359
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2180,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 24000,
+  "total_flos": 280522689478656.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}